Understanding the Scandinavian - S Kasparov

Symbols

Understanding the Scandinavian

Sergey Kasparov

A reliable and challenging system for Black

Contents

Symbols Introduction

!!

brilliant move

!

good move

!?

interesting/probably good

?!

dubious

?

bad move

??

blunder

+–

White is winning

+/–

White is much better

+=

White is slightly better

=

equal position

=+

Black is slightly better

–/+

Black is much better

–+

Black is winning

Ch

Championship

1-0

the game ends in a win for White

1: 3...Qd6 with 4 d4 Nf6 5 Nf3 c6 6 Ne5 2: 3...Qd6: Other Lines 3: 3...Qa5 4: 3...Qd8 5: 2...Qxd5 without 3 Nc3 6: 2...Nf6

½-½ the game ends in a draw

7: Rare Second Moves for White

0-1

the game ends in a win for Black

Conclusion

(n)

nth match game

Exercises

(D)

see next diagram

Answers Index of Variations

Acknowledgements

Index of Main Games

Some of the technical aspects of this work were performed by my wife Tatiana (WIM) and the translation from Russian to English was made by my daughter Eva.

Copyright Information

Introduction

About the Author

I’d like to surprise you with a ‘revolutionary’ claim: the Scandinavian is an improved CaroKann! To be honest, I first read this idea in my childhood in the Russian edition of the Montreal 1979 tournament book.

About Gambit Publications Other Gambit Titles on Kindle

1

1) Panov Attack: 1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4, reaching positions similar to some lines of the Queen’s Gambit.

The game Karpov-Larsen (sensationally won by Black) started with 1 e4 d5. Hort remarked: “Of course, as you are Scandinavian!”. Larsen (who was from Denmark) replied: “Yes, but it is just a good version of the Caro-Kann”.

2) Exchange Variation: 1 e4 c6 2 d4 d5 3 exd5 cxd5 4 Bd3 Nc6 5 c3, a modest but reliable system.

This startling assertion needs to be examined closely. After all, the Caro-Kann is one of the most respected and popular openings...

3) Two Knights Variation: 1 e4 c6 2 Nc3 d5 3 Nf3, a tricky line leading to quirky positions of a very different type from the standard Caro-Kann.

The pawn-structures of these two openings are identical.

4) Advance Variation 1 e4 c6 2 d4 d5 3 e5. Nowadays the positions from the classical line (1 e4 c6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5) are looking quite reliable for Black. That’s why White’s main search for an advantage is currently in the Advance Variation. Now look how much narrower White’s options are in the Scandinavian, with the four Caro-Kann options mentioned above in mind: 1 e4 d5 (D)

Scandinavian Defence (main line with 3...Qd6 and 5...c6): Black has a pawn on c6 and none on the d-file, while White has a pawn on d4 and none on the e-file.

White to play Obviously, lines 1 and 2 don’t apply as Black captures on d5 with the queen instead of a pawn (Black can also choose to meet 2 exd5 with 2...Nf6, although this isn’t my main recommendation in this book).

Caro-Kann Defence (classical 4...Bf5 main line): everything is the same, but in addition the white hpawn is on h5 and the black one on h6 (of course, it is possible for White to refrain from advancing his h-pawn, in which case there are no structural differences at all).

It is unpromising to continue by analogy with line 3: 2 Nc3 can be met by 2...dxe4 3 Nxe4, which is like a ‘classical’ Caro-Kann but with more or less an extra tempo for Black since ...c6 has little purpose here. He can instead play, for example, 3...Nd7 (3...Nc6!?) 4 d4 Ngf6. Later on Black can play ...c5 in one go and equalize. Let’s not forget that the space grab by 2...d4!? looks promising too (in the Caro-Kann version, Black is again hampered by the fact that ...c6 is less useful than White’s developing move Nf3).

The basic ideas of these openings are very similar as well. OK, but what is the basis for the ‘improved’ claim? Firstly, those who play the Caro-Kann have many other possible replies to deal with: 2

That leaves us with option 4, 2 e5, but this is far less challenging than the Advance Caro-Kann. Black can play 2...c5, saving a whole tempo. One might also view this as like an Advance French but with Black able to develop his queen’s bishop actively outside his pawn-chain.

while you are reading this book you already have a stronger chess engine, and when checking the analysis, your assessments may differ. But before rejecting one of my conclusions, be sure to work diligently, as computers don’t necessarily understand all types of positions perfectly. Likewise, everyone has access to game databases. So, if you need further information on a particular line, it will take just a few seconds to find the relevant material. I urge you to focus on games by strong players.

So what other possible advantages might the Scandinavian offer? From a practical viewpoint, it has been studied in less detail than the Caro-Kann and the other major responses to 1 e4. Not only that, but the modern main line of the opening (with 3...Qd6 and 5...c6) is barely covered in older books.

I see the author’s main task as follows: in the most entertaining and comprehensible way (even for relatively weak chess-players) to explain the ideas and plans so as to increase your chances in tournaments and improve your general chess understanding.

For instance, here is a quotation from Opening for White According to Anand 1 e4 Volume 3, published as recently as 2005: “... the variations 3...Qd6 and 3...Qd8 are not the main fare in the Scandinavian Defence ...” Meanwhile, recent practice shows that it is precisely 3...Qd6 which is considered most reliable today. The authors indirectly confirm it on page 336: “The move 5...c6 is more reliable”. Yet they consider just one game with this move. GM Sergei Tiviakov seems to have been the first top-level player to have understood this truth. His main (and during recent years almost only) weapon against 1 e4 is the Scandinavian Defence with 3...Qd6. If at some point I am having doubts about Black’s chances in some line, I will check Sergei’s games, and more often than not find an answer to my problem, or at least a hint in the right direction.

My words don’t claim to be the absolute truth. Quite the contrary; there are no doubt errors in these pages. Feel free to send your suggestions, wishes, criticisms, to [email protected]

Quotes and Differing Views Before we move on to our detailed examination of the Scandinavian, let’s get acquainted with the opinions of some grandmasters about this opening, both favourable and unfavourable. The Elo ratings given are those at the end of 2014.

“I can’t say a lot about the Scandinavian Defence as I haven’t played 1 e4 very much recently. I have to teach my students, though, and it is necessary to know something. There is not much theory but, most likely, one needs very concrete knowledge; otherwise White can fail to get any advantage (if not arrive even in a inferior position).”

Writing Methods and Philosophy I have sought to embrace the full range of strategic and tactical themes in the opening we are studying, while not overburdening your memory with too many lengthy variations. It is a difficult balance to strike, and please note that you shouldn’t feel it necessary to memorize large chunks of the analysis presented here; play through the lines and absorb the main ideas, and only memorize the most critical lines and those paradoxical ideas that it would be unrealistic to uncover at the board.

Zakhartsov (2552)

“My attitude to the Scandinavian Defence is very mixed. On the one hand, my whole chess experience suggests that Black’s play is provocative and must be punished; on the other hand, it is not always clear how to do so! One of the main benefits of the Scandinavian Defence is

Nowadays everyone has a computer, of course. While working on this book I used Houdini as my primary analytical assistant. It is possible that 3

that players often give preparation for this opening a low priority, instead focusing on the Sicilian, French, Caro-Kann, etc.

nuanced game than this simple, if well-intentioned, advice would have us believe. The first question is where to put the queen. 3...Qd6 (Chapters 1 and 2) and 3...Qa5 (Chapter 3) look reasonable, while 3...Qd8 (Chapter 4 ) is a bit passive.

“In fact, it is quite difficult for White to get an advantage. In many lines the positions that arise are quite sharp, and White must be ready to sacrifice to have much hope of an advantage.

3...Qd6 (D)

“Another benefit is the large number of variations at Black’s disposal, starting from the 2nd move (2...Qxd5 and 2...Nf6), and then the choice of squares to put his queen on move 3, and beyond that there are lines with ...a6, ...c6 and ...g6. “The opening is very plucky and it suits the style of chess ‘counter-punchers’.” Tarlev (2561)

“I can say that it’s a terrible and over-adventurous opening. White has a lot of ways to get good positions, which is why I wouldn’t recommend playing this defence.”

White to play This retreat is my main recommendation. It is a multi-functional move. Here the queen prevents Bf4, prepares queenside castling (since d8 is left vacant), and binds together the central foundations (the squares e7, e6, c6, d5, etc.). While the queen can be attacked again by Ne4 or Nb5, these moves may not necessarily prove useful for White.

T.L. Petrosian (2651)

“The Scandinavian Defence is a very interesting opening. Black can either play very solidly with ...c6, which entails positions similar to lines of the Caro-Kann, or act aggressively with ...Nc6 or ...Bg4. It allows any chess-player, regardless of his style, to include this opening in his repertoire.”

Let’s move a little further along the main line: 4 d4 Nf6 5 Nf3 (D)

Kovalenko (2653) 1: 3...Qd6 with 4 d4 Nf6 5 Nf3 c6 6 Ne5

1 e4 d5 2 exd5 This is the most natural reply and clearly the best move. Rare lines like 2 e5?! and 2 Nc3 will be covered in Chapter 7.

Black to play

2...Qxd5 3 Nc3

My main recommendation in this position is the ‘central’ strategy with 5...c6 (Chapter 1 and parts of Chapter 2 ); it is marked by Caro-Kann motifs, and has proved highly reliable in practice.

This is another move that naturally comes to mind. From our chess ‘childhood’, we have been taught that it is a bad idea to bring out the queen in the opening because it will come under attack from the enemy minor pieces and will have to spend further time retreating. However, chess is a more

Kovalenko’s active idea 5...Bg4 is considered in Chapter 2 . The older and somewhat dubious 4

5...a6 also has its own section in Chapter 2 . An option that is quite popular nowadays – the fianchetto with 5...g6 – is discussed in the final section of Chapter 2 . In Chapter 2 we also deal with a variety of rarer options for White on moves 4, 5 and 6 in the lines after 3...Qd6.

a Caro-Kann where Black lacks time to develop his bishop to f5: 5...Qe6 6 Qe2 Nf6 7 f3 (this is the point of White’s strategy: White doesn’t remove the knight but instead consolidates it in the centre) 7...g6 8 g3 Nxe4 9 fxe4 Bg7 10 Bh3 f5 11 Nf3 0-0 (11...c5!?) 12 Ng5 Qd6 13 c3 (13 Bf4!?) 13...h6 14 Bf4 += Bologan-Tiviakov, Šibenik 2009. Not many people have been willing to play this line as Black.

Frequently the plans overlap in these lines (e.g. ...c6 can be played after ...g6, and vice versa), so even if you decide to specialize in a particular line, you may be able to pick up useful ideas by examining material in other variations. I shall try to explain all the possible nuances of nearidentical positions in the clearest way that I can.

5 Nf3 c6 (D) Here we see the first similarities with the CaroKann Defence. Some points of this modest pawn move:

In the current chapter we shall examine White’s most popular reply to 5...c6, namely 6 Ne5.

a) Black places d5 and b5 under control. b) He vacates the c7-square, which will later be used as a more permanent home for the black queen.

Game 1

c) As we shall see further on, in case of a fianchetto with g3 and Bg2 (which is rather popular) the ‘breakwater’ b7-c6 will serve as a good restraint for the g2-bishop.

Swiercz – Tiviakov Wroclaw 2010 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 Other continuations (apart from 3...Qa5 and 3...Qd8, which we discuss in later chapters) do not deserve detailed discussion: 3...Qc6?? 4 Bb5 +–; 3...Qe5+?! 4 Be2 and Nf3 is on the agenda; 3...Qf5?! and 3...Qd7?! block the bishop, and the queen will come under attack from White’s minor pieces. 4 d4 (D)

White to play 6 Ne5 This idea is employed quite often nowadays. Under cover of the powerful outpost White can: 1) Bring the bishop to f4. 2) Consolidate the knight by playing f4. Black to play

3) Attack f7 with Bc4.

4...Nf6

It makes little sense to tolerate the powerful centralized knight, so Black usually tries to exchange it or kick it away. I would like to draw your attention to the importance of exchanges in general. The player who lacks space, as a rule,

A little note on move-order: 4...c6?! gives White the interesting additional possibility of 5 Ne4!?, when the game is starting to look like 5

benefits from simplifications. This natural principle should always be borne in mind. Naturally though, there are exceptions to any rule.

cxd5 (13 Nxg6 Nxf4 14 Nxh8 Nc5 is unclear) 13...Nxe5 14 Bxe5 Qxe5 15 dxc6 Qc7! 16 Qd7+ Qxd7 17 cxd7+ Kd8 =.

6...Nbd7

a2) 10 Bc4 Bxc4 11 Nxc4 Bg7 12 d5 (hurrying to open the d-file before the black king leaves the centre) 12...0-0 (12...Nb6 is even more accurate: 13 dxc6 Qxd2+ 14 Rxd2 bxc6 15 Na5 Nfd5 16 Nxd5 cxd5 = with a perfect position for Black) 13 dxc6 bxc6 14 Rhe1 e6 15 Bd6 Re8 = KanovskyS.Kasparov, Teplice 2013. Black has solved his opening problems. His total control over d5 allows him to view the future with confidence.

Sometimes, possibly trying to avoid opening preparation, Black has used 6...Be6!? (D).

b) 8 Be2 (D) and now:

White to play At first sight this move may seem fanciful. However, this is not the last time we shall meet this bishop move, placing it in front of the e7pawn, and I shall discuss its nuances in more detail at a later point. For now, we shall note that Black places d5 and c4 under control. 7 Bf4 (or 7 f4 g6 8 Be2 Bg7 9 0-0 0-0 10 Be3 Nbd7 11 g4 Nd5 12 Nxd5 Bxd5 13 c4 Be4 = BodnarukTiviakov, St Petersburg 2012) 7...Qd8 and then:

Black to play b1) 8...Nbd7 9 0-0 g6 10 Qd2 Nxe5 11 Bxe5 Bg7 12 Rad1 0-0 13 Bf3 was about equal in EichnerS.Kasparov, Erfurt 2012. b2) 8...g6 and here:

a) 8 Qd2 Nbd7 9 0-0-0 g6 (D) and here:

b21) White enjoyed a certain initiative after 9 Qd2 Nbd7 10 Nf3 Bg7 11 Bh6 Bxh6 12 Qxh6 Nb6 13 0-0 Qd6 14 h3 0-0-0 15 Rad1 Nbd5 16 Nxd5 Nxd5 17 Bc4 in S.Haslinger-Tiviakov, Roosendaal 2012, though Black’s game remains viable. b22) 9 0-0 Bg7 10 Qd2 Nbd7 11 Rad1 0-0 12 h3 Re8 13 Rfe1 Nb6 14 a3 Qc8 15 Bf3 a5 16 a4 Nbd5 is a type of situation we shall see a great many times in this book. The knight moves to the strong point d5 with the intention of simplifying the position. It attacks two enemy pieces simultaneously, so White can’t avoid an exchange. After 17 Nxd5 Bxd5 18 c4 Bxf3 19 Nxf3 b6 20 b3 Ra7 White’s space advantage is not that significant any more since the knight and bishop have left the board, A.Nguyen-Tiviakov, Kuala Lumpur 2012.

White to play a1) In Z.Almasi-Tologontegin, St Petersburg 2012 a leading Hungarian player tried to blast open the centre with 10 d5. However, his low-rated Russian opponent responded in very solid fashion: 10...Bxd5 11 Nxd5 Nxd5 (11...cxd5?? 12 Bb5 Bg7 13 Nxd7 Nxd7 14 Qxd5 +–) 12 c4 Qc7! 13 6

We now return to 6...Nbd7 (D):

Black to play White to play

The play has been very reasonable and natural so far, but now Black must decide which way to jump. But why is that? It seems that the queen must retreat to d8, since surely it shouldn’t stand on d7, in front of the bishop? But this is what we love chess for; it is boundless and diverse.

7 Nc4 Qc7 8 Qf3 8 d5?! is very seldom played, which is not surprising as it doesn’t promise any benefits:

9...Qd7!?

a) Black can attack the audacious pawn immediately by 8...Nb6 9 dxc6 bxc6 10 Qf3 Nbd5 11 Bd2 g6 12 h3 Bg7 13 0-0-0?! 0-0 14 g4 Be6 15 Rg1 Rab8 16 b3 Rfd8 =+ Ristić-Milanović, Kragujevac 2013. White has only gained a headache, as Black exerts pressure on the d- and b-files. The dark squares around the white king have been weakened by b3.

Yes, this move does make sense! The queen attacks the d4-pawn, while keeping an eye on g4. Now White faces a major decision about how to proceed. He can protect the d-pawn, letting the queen move to g4; we shall consider this option in the current game. Kotronias’s idea 10 Nxb6 is covered in Game 2

b) With 8...g6 9 Qd4 Bg7 10 Bf4 Qd8 Black simply ignores the pawn on d5, challenging White to find some way to justify his play. After 11 dxc6 bxc6 12 0-0-0 0-0 13 Qd2 Bb7 14 Bh6 Qc7 15 Bxg7 Kxg7 16 Qe3 Nb6 White has not the slightest advantage, while Black can fight for the initiative thanks to his strong central bastion on d5, L.Dominguez-Ivanchuk, Wijk aan Zee 2010.

. White can also ignore the threat to the pawn and place g4 under control by playing 10 h3 – for this idea of Bologan’s, see Game 3.

8...Nb6

poses an interesting challenge, since (unlike the analogous line with 9...Qd8 – see the notes to Game 6) Black has little choice but to accept the pawn with 10...Qxd4. In this little-explored position, White has enough compensation, but Black has adequate defensive resources; e.g., 11 Rd1 (11 Bd3 g6 12 0-0-0 Qc5) 11...Qb4 12 Be3 (12 Bd2 Bg4) 12...Qxb2 (12...Bg4?! 13 Nxg4 Qxg4 14 Nb5!) 13 Bd4 Qb4! 14 Be2 (14 a3 Qa5 15 Be2 Be6) 14...h5!? with a dynamic balance.

10 Be5 The Polish player doesn’t try to interfere with Black’s plan. 10 Ne5!?

The knight is persistent in the offer to exchange. Note that this idea is a common thread in the whole opening. Indeed, in many openings (and positions), the player who lacks space, as a rule, benefits from simplifications. 9 Bf4 (D)

10...Qg4 (D)

7

“Is it possible for Black to win after the exchange of queens?” I hear you ask. My answer: “Why not? There are other pieces on the board.” Now: b1) 13 Be2 0-0-0 14 0-0-0 h5 (intending ...Ng4) 15 h3 Nfd5 16 Ncxd5 Nxd5 17 c4 Nxe3 18 fxe3 f6 19 Bf4 g5 20 Bh2 g4 and here you can observe active counterplay. I can foresee some sceptical smiles, but let me ask then, can Black win in any other opening if White doesn’t commit serious mistakes?! 21 Bf4 Bg7 22 hxg4 Bxg4 23 Bxg4+ hxg4 24 Kc2 e6 25 Rxh8 Rxh8 26 e4 f5 27 exf5 exf5 with counterplay, Penson-S.Kasparov, Brasschaat 2014.

White to play It is rare to see such a strange thrust by the queen in the opening. Furthermore, it is not part of an attack, but an offer to exchange!

b2) 13 a4 Nbd7 14 f4?! Ng4! (D).

11 Qe3 What if White acquiesces to his opponent’s desire? 11 Qxg4 Bxg4 and then: a) In C.Bauer-R.Ekström, Swiss Team Ch 2009 the French grandmaster directed his knight to the queenside: 12 f3 Be6 13 Na5 (this looks attractive; the knight bothers the pawns on b7 and c6, and Black can’t protect them with his rook since the bishop controls b8) 13...0-0-0 (however, the black king can perform this defensive duty) 14 Rd1 (if there were a knight on d7, then 14 Nxc6 bxc6 15 Ba6+ would mate, but as it is, the king would just run away) 14...Bf5 15 g4 Bg6 16 h4 h5 17 g5 Nfd5 18 Ne4 Ne3 and Black already stood better.

White to play Add this manoeuvre to your chess arsenal – the double attack on e3 and e5; on occasion ...f6 may also follow, to disturb the aggressive white bishop. 15 Nxg4 Bxg4 16 d5 (I have another remark: chess is not draughts, and capturing is not a necessity! That’s why Black often ignores White’s d5 break) 16...Nxe5 17 fxe5 0-0-0 18 dxc6 bxc6 19 a5 Rd4 (19...Be6!?) 20 Be2?! (20 Bd3!?) 20...Bf5 21 Bf3 Kc7 =+ Ganguly-Tiviakov, Khanty-Mansiisk 2007. I guess one doesn’t need a grandmaster title to see that Black has two bishops, while White has a pawn weakness on e5 (and later on, maybe on a5 as well). In a long ending Black managed to realize this slight advantage.

b) 12 Ne3 Be6 (D) gives Black total control over the d5-square, something I will repeat hundreds of times in this book.

We now return to 11 Qe3 (D): White to play

8

13...h5 13...Bf5 is also quite acceptable: 14 Be2 (or 14 f3 Qg6 15 0-0-0 f6 16 Bg3 h5 with normal play) 14...Qg6 (14...Qxg2!?) 15 0-0 h5 (this is actually a standard move in positions like this; Black hinders any ideas of a g4 advance, while making it unappealing for the e5-bishop to retreat to g3, and in the future it may even be possible for Black to make a general advance of his kingside pawns; 15...Bxc2 seemed risky because of White’s lead in development) 16 Ne3 Nxe3 17 fxe3 f6 18 Bf4 (18 Bc7 e6 is about equal) 18...Be4 (18...Bxc2!?) 19 Bf3 Bxf3 20 Rxf3 Qe4 (an important move: the queen occupies a commanding height in the centre, while opening the way for the g-pawn; otherwise White might some day smash open the e-file and the third rank for a rook-lift by an e4 pawn sacrifice) 21 Bc7 Rc8 22 Ba5 g5 (this move is good in all respects; Black prevents Rf4 and e4, while intending ...Bh6 and ...g4) 23 c4 Bh6 24 Re1 g4 25 Rf2 0-0 26 Qc2 Qxc2 with a good game for Black, Boguslavsky-S.Kasparov, Bad Liebenzell 2007.

Black to play 11...Nfd5 Or: a) The exotic 11...Qe6 has also been tried: 12 Nxb6 axb6 13 Be2 Nd7 (13...b5!?) 14 0-0 f6 15 Bc7 Qxe3 16 fxe3 g6 17 a4 (a typical way to restrain the b7- and b6-pawns) 17...Bh6 18 Kf2 00 19 Bc4+ Kg7 20 Rae1 Re8 and now White, by means of 21 g4, paralysed the kingside as well, increasing his advantage, in IordachescuBoguslavsky, Bad Wiessee 2009.

We now return to 13...h5 (D):

b) Another method was used by the author in a rapidplay game against a Lithuanian grandmaster: 11...Nxc4 12 Bxc4 Be6 13 Bxe6 Qxe6 14 0-0 (14 0-0-0!?) 14...Ng4 (seeking beneficial exchanges) 15 Qh3 Qd7 (15...Nxe5 16 Qxe6 Nf3+ 17 gxf3 fxe6 +=) 16 Qg3 h5 17 h3 Nxe5 18 dxe5 Qf5 19 Rad1 (19 e6!? Qxe6 20 Rfe1 += is an interesting file-opening pawn sacrifice) 19...e6 20 Rd3 Rd8 (or 20...h4 21 Qe3 Be7 +=) 21 Rxd8+ Kxd8 22 Re1?! (better is the ‘computer move’ 22 Nb5!, though it would be hard to find when short of time; since 22...cxb5? loses to 23 Qc3, Black has nothing better than 22...Kc8 +=) 22...Kc7 23 Qe3 Bb4 = Rozentalis-S.Kasparov, Palanga rapid 2012.

White to play 14 Be2?!

12 Nxd5 Nxd5 13 Qd2 (D)

14 h3 and 14 Bd3 are better; in the latter case 14...Qxg2?! 15 0-0-0 f6?! 16 h4! fxe5? 17 Nxe5 gives White decisive threats, while 14...f6 15 0-0! Bf5 (only move) 16 h3 Qg6 leads to quieter play. 14...Qxg2 Principled and right! 15 0-0-0 f6 16 h4?? A bluff like this is most unlikely to succeed against Sergei Tiviakov, who simply accepts the sacrificed offering and beats off the attack. The

Black to play 9

fact that he chose 14...Qxg2 showed that he was confident about the defensive capacity of his position.

and plays carefully, it is hard for Black to generate many winning chances. Anyway, after the exchange of queens it is easier for Black to defend.

16...fxe5 17 Nxe5 g6 18 Kb1 Bh6 19 Qe1 Rf8 20 Bd3 Bf5 (D)

In several lines we saw White sacrifice the g2pawn. This typically leads to complications where Black’s chances are not lower. Of course, the exchange of queens on g4 can’t be attractive to a player as White who is truly seeking an advantage. The first measure that comes to mind is keeping the queens on the board with 10 Nxb6 axb6 11 0-0-0, as played in the next game, which the author witnessed one hot Greek summer’s day.

Game 2 White to play Kotronias – Tiviakov Almost all the black pieces are activated, while he has a healthy extra bishop.

Vrachati 2011 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Nc4 Qc7 8 Qf3 Nb6 9 Bf4 Qd7 10 Nxb6!?

21 Rg1 Qh2 22 Bxf5 Rxf5 23 Rxg6 Bf8 Solidly consolidating the king’s residence. 24 Nc4 Rd8

By removing the b6-knight, White reduces Black’s control over d5.

Just not 24...0-0-0??, which loses to 25 Qe6+. I leave the rest without comments.

10...axb6 (D)

25 Qe6 Qxf2 26 a3 Nc7 27 Qe4 Qf4 28 Qe2 Nb5 29 c3 Nd6 30 Nd2 Qxh4 31 Rdg1 Rf2 32 Qd1 Qf4 33 Nb3 Nc4 34 R6g2 Qf5+ 35 Ka1 Rxg2 36 Rxg2 Ne3 0-1

Conclusions White to play

You were probably surprised by the original placement of the queen in front of the bishop (9...Qd7), with the exotic idea of chasing the enemy queen! I have about 30 years of chess experience and I must say that it is rare for an idea like this in the opening to be effective.

11 0-0-0 11 Rd1 avoids placing the king on the queenside, given that Black already has a half-open a-file. 11...e6 12 a3 Nd5 13 Bd2 Nxc3 14 Bxc3 Qd5 (simplest; the exchange of queens guarantees a quiet life in the ending; micro-nuances such as a bad bishop won’t be enough to cause Black serious problems) and now:

In the notes you can see how Black can fight for victory in the positions resulting after an exchange of queens (Ganguly-Tiviakov). At least, the position is very far from a dead draw, and besides, in any opening if White doesn’t take major risks 10

a) 15 Qg3 Qe4+ 16 Be2 Qg6 (seeking an exchange is safer than 16...Qxc2 17 Bf3 Bd7 {17...Kd7!?} 18 Qc7 += or 16...h5?! 17 h4 Rh6? 18 Bd2 +/–) 17 Rd3 (17 Qc7 Qxg2) 17...Qxg3 18 Rxg3 b5 is similar to line ‘b’.

d) In Klek-S.Kasparov, Apolda 2013, I failed to find anything better than 12...e6?! (D) with a draw offer, which the girl sitting in front of me refused...

b) 15 Be2 Qxf3 16 Bxf3 Bd6 17 Bb4 Ke7 (17...Bxb4+ is more accurate: 18 axb4 Ra2 19 Rb1 Ke7 20 Kd2 Bd7 21 Ra1 Rha8 =) 18 Bxd6+ Kxd6 19 0-0 Rd8 20 Rfe1 Kc7 21 g3 Rd6 22 h4 Bd7 23 Re5 f6 and about 40 moves later Dgebuadze-S.Kasparov, Brasschaat 2013 ended in a draw. 11...e6 11...Qd8!? is interesting, though of course it seems odd to move the queen four times just to bring it back home. 12 Qe3 Nd5 13 Nxd5 Qxd5 14 c4 Qa5 15 a3 += Rublevsky-Zolotukhin, Olginka 2011.

White to play Play continued 13 f3 (13 h3 Qh5 14 Be5 +/–) 13...Qh5 14 Be5! +/–. The problem is that the black queen is in an alarming situation. The enemy surrounds it and intends to imprison it by g4. The problems are aggravated by the underdevelopment of the other forces. Fortunately, when facing a real threat of loss I managed to concentrate and find several ‘only’ moves: 14...Nd7! 15 g4! Qh4 16 Bg3 (16 Bc7 h6 +/–) 16...Qd8 17 d5 Bc5 18 Qe2 cxd5 19 Rxd5 0-0 20 Rd3? (I can understand the desire to bring the rook into play via the third rank, but 20 Rd1 is better) 20...b5 (otherwise White has a clear initiative; now the black queen can enter the attack on the queenside) 21 Nxb5 Rxa2 (stronger is 21...Qa5 22 Nc7 {22 Nc3 b5 –/+} 22...Ra7! 23 a3 b5 24 b4 Bxb4 25 axb4 Qxb4 with good compensation) 22 Kb1 Ra6 (22...Qa5 23 Na3 Rxa3 24 bxa3 Bxa3 is unclear) 23 Bc7?! (better is 23 Qd2) 23...Qe7 24 f4?! Nf6 25 Be5 Nd5 26 g5?! Bd7 27 Rh3 Nxf4 –/+.

Black must exercise some caution, as shown by 11...Qg4?! 12 Qe3! (D), when suddenly Black faces problems finding a move.

Black to play White has the straightforward threat of d5, with very unpleasant consequences. Then:

We now return to 11...e6 (D): a) 12...Be6? leads to an immediate loss: 13 d5! Nxd5 14 Nxd5 cxd5 15 f3 Qg6 16 Bb5+ 1-0 Sarić-B.Kovačević, Stari Mikanovci 2009. b) 12...Qe6 13 d5 Qxe3+ 14 Bxe3 b5 15 dxc6 bxc6 16 Bxb5 Bb7 (16...cxb5 17 Nxb5 +–) 17 Bd3 +/– Sipila-Valli, Helsinki 2012. c) The most persistent was 12...Qf5!? 13 d5 Nxd5? (better is 13...e6 14 dxc6 bxc6 15 Qf3 +=) 14 Rxd5! cxd5 15 Nb5 (15 Bd3!) 15...e5 16 Bxe5 Bc5? 17 Bd4+ Kd8 18 Bd3 with a terrible attack, R.Jones-Zeidler, Welsh Ch, Cardiff 2010.

White to play 11

29 Bxh7 Rxa2 30 Bg6+ Kg8 31 Kb1 Ra8 (31...Qa8 32 Qe7 +–) 32 Re1 Rd8 33 Kc1 b5 34 Qe7 +– Solodovnichenko-S.Kasparov, Padua 2012.

12 Be5 Nd5 12...Be7 13 Qg3 +=. 13 Bc4!?

b) In 2015 I played more carefully and managed to equalize: 16...0-0 17 a3 Bd6 18 Qh5 (D).

The queen exchange is unpromising: 13 Nxd5 Qxd5 14 Qxd5 exd5 15 Re1 Be6 = Pavlidis-Stips, Fermo 2009. 13...f6 14 Bf4 Tiviakov gives this move an exclamation mark, but I shall refrain from it, as there is an interesting alternative in 14 Bg3 Bb4 15 Nxd5 exd5 (15...cxd5 16 Bb3 +=) 16 Bd3 (D).

Black to play 18...f5 = (now it is unclear how White should develop his initiative; note that 18...g6? allows 19 Bxg6 hxg6 20 Qxg6+ Kh8 21 Rd3 with an attack) 19 Rde1 (19 Qh4 Be7 =) 19...g6 20 Qh6 f4 21 Bh4 (21 Bxg6?! hxg6 22 Qxg6+ Qg7 23 Qxd6 fxg3 24 hxg3 Bf5 probably even favours Black) 21...Qf7 (the most precise; not 21...Qg7? 22 Qxg7+ Kxg7 23 Be7 Bxe7 24 Rxe7+ Rf7 25 Re8 +=) 22 Bg5 and after 22...Bd7?! I held approximate equality in Carlstedt-S.Kasparov, Århus 2015. However, 22...c5! is much stronger: 23 c3?? cxd4 24 cxd4 Bf5 –+; 23 h4?! c4 24 Be2 c3 =+; or 23 Qh4 Be6 with the initiative.

Black to play This novelty was invented by the ingenious Ukrainian grandmaster Solodovnichenko. My friend had spent about 5 hours searching for a way to give me some trouble in the Scandinavian Defence, and one must admit that he succeeded. Everyone had retreated to b3 so far. Indeed, who is willing to give up the a2-pawn? Then:

We now return to 14 Bf4 (D):

a) 16...Qg4?! 17 Qe3+ Kf7 18 h3 Qe6 19 Qf3 (of course, White avoids the exchange, since his plan is to attack) 19...g6? (Black should try 19...Rxa2 since 20 c3? fails to 20...Bxc3 21 bxc3 Qe7 –+, when it is the white king who risks being mated, so White must settle for 20 Kb1 Ra8 21 c3 +=) 20 c3 Bd6 21 Rde1 Qd7 22 Re3 Bxg3 23 Qxg3 Re8 24 Rhe1 Rxe3 25 Qxe3 Qd8 (surprisingly, there is nothing instantly decisive for White) 26 h4! +/– (after prolonged thought, Solodovnichenko brings in reinforcements) 26...Bd7 27 h5 Qf8 28 Rh1?! (28 hxg6+ hxg6 29 Qg3 Qh6+ 30 Kc2 Re8 31 Rxe8 Bxe8 +/–) 28...g5? (more resilient is 28...Qg7 29 Qf4 Kg8 30 hxg6 hxg6 31 Qd6 +/–)

Black to play 14...Be7 14...Bb4 is slightly more accurate: 15 Nxd5 (15 Ne4 should be met by the unclear 15...0-0 rather than 15...Nxf4 16 Qxf4 0-0 17 a3 Be7 18 Rhe1 +=) 12

and a8-rook in the game, but this can be fixed given a little time.

15...exd5 16 Bb3 0-0 (16...Kd8!? 17 c3 Bd6 +=) 17 c3 (the bishop’s control of e1 was irritating for the white rooks) 17...Be7 18 Rhe1 +=.

20 f3

14...b5!? is also possible.

20 Rd3 Bf8 21 Qh4 Qg4 =; 20 c3?! bxc3 21 Bc2? cxb2+ 22 Kxb2 (22 Kb1 Bf8 –+) 22...Qd8! –+.

15 Rhe1 b5 15...0-0?? still mustn’t be played as it leads to immediate loss after 16 Rxe6 Qxe6 17 Nxd5 +–.

20...b6 Interesting is 20...Bf8!? 21 Qh4, when 21...h5 is unclear. Instead 21...g5 looks like the type of move only a computer would consider, but 22 Qh5+ is no more than unclear.

16 Bb3 b4 Black holds back the c2- and a2-pawns; otherwise White might play c3 and Bc2. Not 16...0-0?? 17 Rxe6 +–, 16...Kd8?! 17 Nxd5 exd5 18 Re3 +/– or 16...Kf7?? 17 Rxe6 +–.

21 g4 (D) After 21 Re3 Bf8 22 Qh4 h5 it’s hard to say whose position is better; ...g5 is threatened.

17 Nxd5 There were also worthy alternatives, such as 17 Na4 Ra5 18 Qh5+ g6 19 Qf3 with the initiative. 17...exd5 18 Qh5+ 18 Rd3 0-0 19 Rde3 Rf7 +=. 18...g6 18...Kd8 19 Re3 +/–. 19 Qh6 Kf7 (D)

Black to play 21...Ra7?! 21...Ba6!? is more attractive, activating the bishop that has been passive so far. 22 Re3 (22 g5 Bc4 23 gxf6 Bxf6 24 Be5 Bxb3 25 Bxf6 Kxf6 26 Qh4+ Kf7 27 Qf4+ =) 22...Bc4 23 Rde1 Bf8 (Black’s defence is based on this retreat; his pieces leave the e-file but they still keep control over the potential entry squares e6, e7 and e8) 24 Qh4 Bxb3 25 axb3 Ra1+ 26 Kd2 Rxe1 27 Rxe1 h5 =. White to play

22 Rd3

Let’s take a closer look at the situation on the board. At a glance, White appears to have an overwhelming position, but the fact that there are no knights on the board makes it harder for White to exploit Black’s apparent weaknesses. Meanwhile, the b4-pawn actually plays an important role as a ‘breakwater’; preventing White from attacking the centre by playing c4. Black doesn’t really have any long-term weaknesses. His main challenge is how to include his c8-bishop

22 Bg3!?. 22...Qd8 23 Rde3 Bf8 24 Qh4 Re7 (D)

13

33...Kg8! would have led to victory: 34 Rd3 h6! 35 Qe4 (35 Bxd5+ Bxd5 36 Rxd5 Qf7 –+) 35...Qf7 36 Rxd5 seems terrible, but the quiet 36...Kh7 –+ leaves White in desperate condition. 34 Ba4 34 Qxh8 cxb3 is unclear. 34...Qxa4 35 Qe6+ 35 h5 is probably more accurate. 35...gxh5 36 Qxh8 gives White the initiative.

White to play Another important defensive resource, not just in this specific position, but in a general sense. Tiviakov trades his passive rook for an active enemy unit, thus easing the tension.

A titanic battle at the seacoast has come to an end.

25 Qg3 Rxe3 26 Rxe3 Qd7

Conclusions

26...c5 can be met by 27 Ba4 += or 27 g5! with a strong attack.

I trust you enjoyed studying not only the main game between Kotronias and Tiviakov, but also the labyrinth of rather complicated variations in the notes. Note that I don’t recommend that you try to memorize them, but rather to understand the ideas; my aim is that by the end of the book you can comprehend and even sense the opening ideas. However, a certain accuracy with the move-order is still advisable. It is possible to get lost in this labyrinth, as you can see in the fragment KlekS.Kasparov. I found myself in an unpleasant situation which frayed my nerves, even though it had a happy ending.

35...Kg7 36 Qe5+ Kg8 37 Qe6+ Kg7 38 h5 gxh5 39 Qf6+ Kg8 40 Qe6+ Kg7 ½-½

27 h4?! 27 Bc7 c5 28 dxc5 bxc5 29 Rd3 Bb7 =; 27 Ba4 +=. 27...c5 28 dxc5 bxc5 29 Rd3 Bb7 = 30 g5 Qe6 Since this could be met by 31 Ba4 Bc6 32 Re3 Qd7 33 Bxc6 Qxc6 34 gxf6 with interesting chances, 30...Be7 might be preferable. 31 Re3 Qd7 (D)

As we saw in Dgebuadze-S.Kasparov, it may not be too difficult to repel White’s assault in a slightly worse ending. Before moving on, I should mention the interesting novelty used against me by Solodovnichenko. He spent 3-5 hours preparing this surprise in the opening. I doubt that most of your opponents will have so much time and diligence. Besides, this novelty did not automatically imply a victory, and careful analysis shows that objectively the situation shouldn’t have been so bad. So, those playing Black can sleep at ease, and focus their energy on solving problems at the board – for the solutions will be there, waiting to be uncovered.

White to play 32 Be5? This looks like a time-trouble mistake, which goes unpunished as Black was also under time-pressure. 32 Rd3 repeats.

In search of an advantage, grandmasters will try all possible methods. In the following game the top Moldavian player found an unconventional idea: on move 10 he preferred to give up the d-

32...fxe5 33 Qxe5 c4? 14

wouldn’t be surprised if Alexandra had prepared everything in advance; the world team championship is a really serious competition, after all) 18 Bxe6 Be7 19 Bd7+ Kf8 20 Qf3 Qxb2 21 00 Qxc2 22 Rfe1? (better is 22 Be6, with unclear play) 22...Bb4 –/+ Yildiz-Kosteniuk, Women’s World Team Ch, Astana 2013.

pawn so as to avoid an exchange of queens by playing 10 h3. Game 3 Bologan – Tiviakov Croatian Team Ch, Šibenik 2010

13 Rd1 (D)

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Nc4 Qc7 8 Qf3 Nb6 9 Bf4 Qd7 10 h3!?(D)

Black to play 13...Qb6

Black to play

In Melia-Mikadze, Georgian Women’s Ch, Anaklia 2011, Black handled the position well: 13...Qc5!? 14 0-0 Be7 15 Rfe1 0-0 16 Re5 (by general considerations it seems that the rook’s penetration to the fifth rank is favourable for White as it can be used for a frontal attack on the king; however, computer analysis confirms that this is not so) 16...Qb6 17 Na4 (the only move, but sufficient) 17...Qa6 18 Nc5 Qb6 19 Na4 Qa6 (a draw by repetition would be a logical end, but White is dreaming of something more) 20 c4 Rd8 21 Rxd8+ Bxd8 22 Qd3 Be7 23 Bc2 b6 (I don’t know how you feel, but to me this looks risky for Black due to his sluggish queenside development; meanwhile, Houdini prefers Black!) 24 Bg5?! (24 Nc3!?) 24...h6 25 Bf4 (25 Bxf6 Bxf6 26 Qh7+ Kf8 –/+ is absolutely safe for Black; the darksquared bishop holds the attackers at bay) 25...Bb7 26 Qg3 Kf8 27 b3 Rd8 –/+. The attack died out and gradually Black managed to realize the extra pawn.

With this quiet preventive move, White takes control of g4, and is ready to give up the d4-pawn. 10...Nxc4 11 Bxc4 Qxd4 12 Bb3 e6 Black has an extra pawn, but he is behind in development and therefore great caution is required. I wouldn’t recommend a hyperactive move like 12...e5. After 13 Bg5 Black feels uneasy. 12...Bf5 is a more principled way to develop. 13 Rd1 and then: a) 13...Qc5 14 0-0 e6 15 Rfe1 Be7 16 Re5 Qb6 (both players have been operating correctly so far) 17 g4 (17 Na4!? keeps the initiative; after 17...Qa6 18 Rde1 Bg6 19 Bd2 he threatens Ra5, winning the queen) 17...Bg6 18 Bxe6!? (this surgical intervention looks effective but Black has enough resources to repel the attack) 18...fxe6 19 Rxe6 Rd8 20 Rde1 (G.Ginsburg-Heinzel, Nuremberg 2011) and after 20...0-0!? 21 Rxe7 Qxb2 22 Rb1 Qxc2 23 Rbxb7 Ne8 it is absolutely unclear who is attacking whom.

14 0-0 Be7 15 Ne4 Nxe4 16 Qxe4 0-0 17 Be5 (D)

b) 13...Qb6 14 Be5 e6 15 g4 Bg6 16 Bxf6 gxf6 17 Qxf6 Rg8 (Black must follow a narrow path, and it looks like the Russian lady knows the way; I 15

solves all the problems. 20 Rxd7 Bxd7 21 Rd1 Be8 22 Rd4 (a transparent hint at a strike on g7) 22...Qc5! (forced but sufficient; the careless 22...Rd8? is punished by 23 Bxg7 Bxg7 24 Rg4 +/–) 23 c3 (23 Bxg7?? Bxg7 24 Rg4 Qf8 –+) 23...Qe7 24 Bd6 Qf6 25 Be5 Qe7 26 Bd6 Qf6 27 Be5 Qg6 (I rejected a draw offer here since I was well ahead on the clock) 28 Rg4 Qb1+ 29 Kh2 g6 30 h4 (30 Qh4 Qf5 31 Rg5 Qd3 =+) and now Black should choose 30...Be7, when 31 Qe3 is unclear (the attack with 31 h5?! doesn’t achieve its aim: 31...Qf5 32 hxg6 hxg6 –/+). Instead I played the ‘natural’ 30...Bg7? 31 Bxg7 Kxg7, when 32 Qe5+! would have won: 32...Kg8 33 Rf4 (it would have been hard to find this quiet move in time-trouble; White’s small army paralyses his opponent) 33...Rd8 34 h5 Qd3 35 h6 Kf8 36 Qg7+ Ke7 37 Qf6+ Kf8 38 Bxe6.

Black to play This is an important moment in the game. As we all know, any pawn move is rather a weighty decision since you can’t move them back. 17...c5?! Black includes the queen in the defence via his third rank, but at the cost of leaving the c8-bishop and a8-rook passive. While this hardly looks like a serious mistake, Tiviakov later played an improvement: 17...Rd8!. Then:

b) 18 Bc4 Rd7! (D).

a) 18 Qg4 Bf8 (avoiding the weakening ...g6) and then: a1) 19 Rfe1 is not very dangerous: 19...Rd7 20 Re3 Qb4! 21 c4 (Filippov-Vakhidov, Agzamov Memorial, Tashkent 2013) 21...Rxd1+ 22 Qxd1 Qc5 and Black restricts the e5-bishop while the other one is blocked by the c4-pawn. White to play a2) 19 Qg3 Before looking through the game further, please try to guess the purpose of this fanciful rook manoeuvre. In an Internet game, one might suppose it was a mouse-slip.

(D) was played in Kanarek-S.Kasparov, Krakow 2012.

All will be revealed as we examine the course of Dzhumaev-Tiviakov, Mashhad 2011: 19 Qg4 (after 19 Bd3 g6 20 Qf4 Qd8 21 Bc3 Bd6 the black queen has made good use of the square left for it by the rook: there is no mate, while the extra pawn is still on hand) 19...Bf8 20 a3 (the rook on d7 is successfully fighting for the only open file, and if it is eliminated by 20 Rxd7 Bxd7 21 Rd1 Be8 =+ the bishop has run from c8 to e8, where it reinforces the castled position and, importantly, allows the unemployed a8-rook to move to d8) 20...Qc5 21 Qe4 g6 22 b4 Qe7 23 h4 (D).

Black to play I recall my Polish opponent thinking for a long time while I walked around and recalled Tiviakov’s notes to his game against Dzhumaev (in line ‘b’). It seems that the threat of Bc7 is extremely unpleasant but the ‘universal’ 19...Rd7 16

This is a common idea: a rook-lift via the third rank with a view to a frontal attack on the king. 19...Rd8 20 Rxd8+ Qxd8 20...Bxd8 is also unsatisfactory: after 21 Qg4 g6 (alas, the a8-rook fails to join the defence: 21...f6 22 Bc3 Ra6 {22...Kf7 23 Qh5+ +–} 23 Bc4 +–) 22 Qf4 Qc6 23 Rf3 Qe8 24 Bc3 Qe7 25 Qe5 f6 26 Qe3 Qf8 27 Bxe6+ Bxe6 28 Qxe6+ Qf7 29 Qe2 b6 30 Re3 +/– Black has weak light squares, an insecure king, and a bad bishop ... and all ‘for nothing’, as the material balance has been restored.

Black to play 23...h5! (both players maintain the tension on the d-file but the onus is on White to make something happen, as he is a pawn down) 24 Rxd7 Bxd7 25 Qf4 Be8 26 Bf6 Qd6 27 Be5 Qe7 28 Bf6 Qd6 29 Be5 Qd8 30 Bf6 Bh6! (Black doesn’t have to repeat moves! You often hear claims that the Scandinavian Defence is passive; I invite you to draw your own conclusions) 31 Qf3 (31 Qe5!? Qb8 32 Qc3 Qf4 gives Black the initiative, while 31 Qxh6 Qxf6 –/+ is definitely unfavourable for White) 31...Qd2 32 Bb3 Qf4 =+. White has run out of threats against the black king, while the extra pawn remains.

21 Rd3 Qf8 (D)

White to play

We now return to 17...c5?! (D):

If Black has not been paralysed yet, it is close. It is unclear how the rook and c8-bishop should be activated. 22 Ba4 22 a4 is also effective. 22...f6 23 Bc7 f5 24 Qc4 Qf6 25 Qb5! The harvest begins; the a5-pawn is indefensible. 25...Ra6?!

White to play

One must agree that there is disharmony in Black’s camp. 25...h6!?.

18 Rfe1 a5 It is difficult to recommend 18...f6?! 19 Bc3 Kh8 (19...Kf7 20 Qxh7 +–) 20 Bxe6 Bxe6 21 Qxe6 Qxe6 22 Rxe6; although the mating attack has been beaten off, material is now equal and White exerts unpleasant pressure. After 18...Rd8 19 Qg4 g6 20 Qf4 (20 h4!?) 20...a5 21 c3 White retains compensation.

26 Bxa5 26 Bd8! is very strong and threatens Qe8+. After 26...Qf8 27 Bxe7 Qxe7 it seems there is nothing terrible since after 28 Qe8+ Qxe8 29 Bxe8 Kf8 30 Rd8 Bd7! the piece doesn’t drop but Black has a lot of other problems: 31 Bxd7+ Ke7 32 Rg8 Kxd7 33 Rxg7+ +/–.

19 Re3

26...e5?? 17

Game 4

After 26...Bd6 27 Bc3 Qe7 28 Qc4 h6 += Black is still suffering but the whole fight lies ahead.

Grishchuk – Tiviakov

27 Rd8+! Bxd8 28 Qe8+ Qf8 29 Bb3+ Be6 30 Bxe6+ Rxe6 31 Qxe6+ Kh8 32 Qc8! (D)

Russian Team Ch, Sochi 2006 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Nc4 Qc7 8 Qf3 Nb6 9 Bf4 Qd8 (D)

Black to play Black’s queenside is falling apart. White to play

32...Be7 33 Qxb7 e4 34 a4 Bf6 35 Bc7 c4 36 a5 1-0

This retreat is clearly the most natural option. But there is also a negative aspect – Black lacks development. The queen returns to its initial square with apparently little to show for its ...Qxd5-d6-c7-d8 itinerary. But there are also some nuances that are in Black’s favour. In positions with this general structure (they are also common in the Caro-Kann) White should try to put pawns on d4 and c4 if he wishes to develop an initiative. This provides control over the centre and brings in possible ideas of a pawn-break by d5. Here, two knights are settled in front of the c2pawn. Meanwhile it is unclear where the f1-bishop should move. Also, the queen being on f3 is not an unequivocal gain, since the d4-pawn is loose and the queen may itself be subject to attack.

Conclusions In our main game, White’s idea proved successful, partly thanks to the element of surprise (as was also the case in my game against Solodovnichenko – see the notes to Game 2). Tiviakov didn’t find the best path at the board and was then outplayed in a difficult position. Nevertheless, later analysis demonstrated Black’s best set-up, featuring the original manoeuvre ...Rfd8-d7!, putting the rook in front of a very passive-looking c8-bishop! Those who followed in Bologan’s footsteps had to struggle against this improvement. In the notes we saw a couple of examples (with both Tiviakov and the author as Black) where White’s attempts to maintain the tension ended dismally. It seems the ball is in White’s court. Either he must abandon the sacrifice or strengthen his play somehow. The former option seems more likely.

10 Be5 White protects d4. The otherwise natural 10 0-0-0? is bad because of 10...Bg4. For 10 Ne5, see Game 6 , where we discuss ideas based on returning the knight to e5 on moves 9 and 10. 10 h3 is discussed in the notes to Game 5.

This brings to an end our examination of the counter-intuitive 9...Qd7. We shall now examine the more obvious queen retreat, 9...Qd8. It is useful to be familiar with this move too, either because of some possible problems after 9...Qd7, or to increase our opponents’ workload by being less predictable in our choice of lines.

10...Bg4 Or: a) 10...Nbd5 18

b21) 13...a6 14 Kb1 Nbd5 15 Ncxd5 Bxd5 16 Qe2 (16 Nxd5 +=) 16...e6 17 f3 b5 (this is more to secure the d5-bishop than an attacking attempt) 18 h4 Be7 19 h5 0-0 (or 19...h6!?; it’s a matter of taste whether to allow White to play h6) 20 Bh4 Nd7 21 Qf2 Nb6 22 Bd3 h6 23 Ng4 Bxh4 24 Rxh4 Kh8 25 Ne5, Sarić-Dražić, Pula 2010. The initiative is definitely on White’s side. He hasn’t advanced his queenside pawns, which gives Black little on which to ‘hook’ his counterplay. Meanwhile, kingside play with f4 and g4-g5 may prove successful.

seeks a quick exchange of queens (which is usually desirable for Black). 11 Nxd5 Qxd5 12 Qe2 Be6 13 c3 Qe4 (13...Ng4!?) 14 Qxe4 Nxe4 15 Bd3 (Gashimov-Epishin, Ourense rapid 2009) and now 15...Nd6 looks solid: 16 Ne3 f6 17 Bf4 Kf7 with acceptable play. b) 10...Be6 attacks the c4-knight at once. 11 Ne3 (D) and then:

b22) 13...Nbd5 must by logical considerations favour White, who has already castled. After 14 Nexd5 (14 Ncxd5 Nxd5 15 c4 Nxe3 16 Qxe3 +=) 14...Nxd5 15 Ne4 += White threatens Ng5, Nc5, c4, etc. b23) 13...g6? (D) was played Hoogeveen 2012.

Black to play b1) 11...Qd7 12 0-0-0 Bg4 (I am not sure that giving up the bishop is a good idea, because even if his attack in the centre fails, White will still have the bishop-pair as a long-term advantage; probably it is more reasonable to capture with the knight: 12...Ng4!? 13 Bd3 Nxe5 14 dxe5 Qc7 15 Bf5, though White retains the initiative) 13 Nxg4 Qxg4 14 Qe3 0-0-0 15 Nb5! (the unsophisticated 15 Be2 also leads to a comfortable advantage: 15...Qf5 16 h3 +/–) 15...cxb5 16 Qc3+ Nc4 17 Bxc4 bxc4 18 Qxc4+ Kd7 19 Qa4+ Kc8 20 Rd3 Ne4 (Naiditsch-Boguslavsky, Deizisau 2009) 21 Re1 Nd6 22 Qxa7 +–.

in

Willemze-S.Kasparov,

White to play It looks like everything is normal. However, the opposition of the rook and the queen on the d-file and Black’s lack of development give White an opportunity for a powerful breakthrough in the centre. However, the game featured a case of mutual blindness, even though we had both plunged into long and diligent thought: 14 Be5? Nbd7 and the fight then swayed back and forth, with the crucial move remaining behind the scenes. The bone-crushing 14 d5! would have more or less won: 14...Nbxd5 (14...cxd5?? 15 Bb5+ Bd7 16 Ncxd5 Nfxd5 17 Nxd5 Bh6+ 18 Kb1 +–; 14...Bd7 15 dxc6 bxc6 16 Ba6 +/– with the idea Bb7) 15 Ncxd5 Nxd5 16 c4 +/–.

b2) 11...Nfd7 12 Bg3 Nf6 13 0-0-0 (D) and here Black can’t equalize but he needn’t be losing either:

b3) 11...Nbd5 (D). Black to play 19

White to play

White to play 12 Nexd5 Bxd5 (capturing with the knight doesn’t equalize: 12...Nxd5 13 Bc4 +=) 13 Nxd5 Qxd5 (White has gained the bishop-pair but here it is more a moral than a real advantage) 14 Qxd5 (it hardly changes the assessment for the exchange to take place on b3: 14 Qb3 Qxb3 15 axb3 e6 16 Be2 Kd7 17 h4 Bd6 18 h5 h6 19 0-0-0 Ke7 = Z.Rahman-Palit, Parsvnath 2011; after the darksquared bishops are exchanged Black has absolutely no problems; moreover, the h5-pawn may turn out weak in the ending) 14...Nxd5 (14...cxd5 looks weaker, although after 15 Bb5+ Nd7 16 c4 Black isn’t much worse: 16...dxc4 17 Rc1 Rc8 +=) 15 Be2 (the central break by 15 c4 Nf6 16 d5?! is not very dangerous: 16...cxd5 17 cxd5 Nxd5 18 Bb5+ Kd8 19 0-0 e6 =; one way or another, the d5-knight must be removed – then White will restore the material balance, but at the cost of simplifying the position) 15...e6 16 0-0 Rd8 (or 16...f6 17 Bg3 h5 18 h4 Kf7 with approximate equality) 17 Rfd1 (17 c4 might appear logical, but White should keep in mind in structures of this type that in the fight for the dark squares Black actually has an extra piece, and putting the pawn on c4 weakens the d4-pawn, which in the future can be subjected to pressure) 17...Nf6?! (17...f6 18 Bg3 Bd6, with an inevitable exchange of bishops, would have fully equalized: 19 Bxd6 Rxd6 =) 18 Rd3! += HamitevichiS.Kasparov, Donetsk 2010.

11 Qg3

We now return to 10...Bg4 (D):

We now return to 11...h5 (D):

For 11 Qf4 see the notes to Game 5. 11...h5 As we have seen previously, this advance often makes sense as: 1) A space grab; 2) Driving the queen back; 3) Preventing a pawn advance by White on the kingside. Some players have tried managing without moving the h-pawn; e.g., 11...Be6 12 Ne3 Nfd5 13 Be2 f6 14 Bf4 g5 (when pushing the pawns to f6 and g5, you should always think about possible checks on the h5-e8 diagonal) 15 Nexd5 (15 Bh5+ Bf7 16 Bxf7+ Kxf7 +=) 15...Nxd5 16 Bd2 Qb6. Black just needs to castle queenside to be happy, but now came 17 Bh5+ Bf7 18 Bxf7+ Kxf7 19 Qd3 e6 (greed is inappropriate: 19...Qxb2 20 Rb1 Qa3 21 Rxb7 +=) 20 Ne4 Be7 21 0-0 += in Iordachescu-Ardelean, Iasi 2014. White will at some point play c4, chasing the knight away from the centre. Black has some problems in the middlegame due to his compromised king position.

20

This standard relieving manoeuvre equalizes. 13...Nfd5 is also possible. 14 Ncxd5 Nxd5 15 Nxd5 Qxd5 16 Bd3 Launching an attack when development is unfinished is a poor idea: 16 c4? Qa5+; 16 Bxg7?? Rg8 –+. 16...f6 This is also a typical move, driving the centralized bishop away from e5, where it protects the d4pawn and exerts pressure on g7.

White to play 12 f3

17 Bb8?!

This is not necessarily the best option; White can also put up a stop sign by 12 h4

Quite an unexpected move! Instead, 17 Bf4 h4 18 Qg6+ Bf7 and 17 Bc7 Bf7 18 Qf2 Qe6+ 19 Be2 g6 are both equal. Swooping in with 17 Bg6+?! is not dangerous because Black safely controls the light-square complex: 17...Bf7 =+.

Nbd5 13 Nxd5 Nxd5 (I don’t see any strong reason to avoid the generally desirable recapture with the pawn: 13...cxd5!? 14 f3 Bd7 15 Ne3 Qa5+ with acceptable play) 14 Qb3 Qd7 (14...b5!?) 15 Ne3 Nxe3 16 Qxe3 0-0-0 17 f3 Bf5 18 0-0-0 f6 19 Bh2, Pavasović-Skoberne, Nova Gorica 2008. After all the knights have been traded the position is absolutely equal. Now the careful 19...g6 would preserve the status quo: 20 Qa3 a6 =.

17...h4! 17...Qa5+ 18 c3 h4 19 Qc7 Qxc7 20 Bxc7 Bf7 =. 18 Qc7 (D)

12 h3 is the subject of Game 5. 12...Be6 (D)

Black to play Such an exotic picture is quite a rare sight. It can be reached by either a very weak or a very strong chess-player. White intends to gobble the b7-pawn and the rook afterwards. However, Black has abundant resources.

White to play 13 Ne3

18...Kf7!

After 13 Nxb6 axb6 the a-file is opened, which is hardly attractive for the white king; with ...b5 on the agenda, where should it castle? White should avoid 13 Bc7?! Qxd4 14 Rd1 Qc5 =+.

Not 18...Qd7?? 19 Bg6+ +–. 19 Qxb7 19 0-0-0? is inappropriate: 19...g6! (19...h3!?) 20 c4 (20 f4 Bh6 21 c4 Qd8 –+) 20...Qd6 21 Qxd6 exd6 22 Bc7 b6 –/+.

13...Nbd5

21

19...c5 20 Qxd5

Conclusions

20 Qb5? c4 –+.

9...Qd8 leads to less critical play than 9...Qd7. On d7 the queen is unstable and must move again to free the bishop. With the queen on d8, Black has more choice; in particular, the bishop can go to e6 or g4, etc. As a rule, a complicated strategic struggle ensues. Note the ‘mutual blindness’ in the fragment Willemze-S.Kasparov ; one rarely sees something like that at the beginning of the game when the players are not so tired yet.

20...Bxd5 White has an extra pawn but Black’s position is preferable. 21 c4 (D) Banal retreats don’t bring equality: 21 Bf4 cxd4 =+; 21 Bc7 cxd4 =+.

The exchange of queens usually makes Black’s life easier (see Hamitevichi-S.Kasparov). The main game itself has many interesting points. Black’s resolute advance by ...h5-h4 catches the eye. In structures of this type it is not so rare for the black king to have a ‘one-room apartment’ on f7. Small, but safe! When the bishop was kicked away from e5, Grishchuk responded with the creative idea of putting his bishop on b8, and the queen was then sent to support it. Even if it wasn’t the soundest of ideas, it’s an impressive feat, don’t you think? Black to play

In the end, White was quite relieved to achieve a draw. For us it is highly reassuring that one of the top Russian players can reach a dubious position as White.

21...Bxf3?! 21...Rxb8 22 cxd5 cxd4 23 0-0-0 is unclear, but 21...Bb7! seems the most accurate: after 22 Bc7 (22 Bf4 cxd4 23 Ke2 e5 gives Black the initiative) 22...cxd4 (22...h3? loses material: 23 d5 hxg2 24 Rg1 e6 25 dxe6+ Kxe6 26 Kf2 +/–) 23 Ke2 e5 24 Ba5 Ke6 Black has the initiative and dominates the centre. The advance of the f-, e- and d-pawns may turn out fatal for White.

Our next game demonstrates Black’s hidden potential in positions with doubled pawns. We shall also discuss such structures in the 3...Qa5 lines (see the first part of Chapter 3, i.e. Games 24 and 25). Game 5

22 Bxa7 Bxg2 23 Rg1 Rxa7 24 Rxg2 cxd4 25 Bg6+ Kg8 26 c5

Shukh – S. Kasparov St Petersburg 2011

White has compensation, as he also would after 26 b4!?.

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Nc4 Qc7 8 Qf3 Nb6 9 Bf4 Qd8 10 Be5

½-½ Please remember that with opposite-coloured bishops the number of pawns often doesn’t play a significant role. Their configuration is more important. In this case the g6-bishop paralyses the black pawns and a number of black pieces too.

Another option is 10 h3!? Be6 = (Black can also transpose to Game 3 by grabbing a pawn with 10...Nxc4 11 Bxc4 Qxd4) 11 Nxb6 axb6 12 Be5 (D), and now:

22

Black to play

Black to play

a) 12...Nd5?!

10...Bg4 11 Qg3

13 Bc4! += f6 (13...Qd7 doesn’t equalize: 14 0-0 Nxc3 15 Bxe6 Qxe6 16 Qxc3 +=) 14 Bg3 Qd7 15 0-0 b5 (15...Nxc3 16 Bxe6 Qxe6 17 Qxc3 Rxa2 18 Qb4 Rxa1 19 Rxa1 +=) 16 Bb3 Nxc3 17 Qxc3 Bd5?! (Black should play actively on the kingside: 17...g6 18 Rfe1 Bxb3 19 axb3 Rxa1 20 Rxa1 Kf7 21 Qc5 h5 22 Qb6 h4 23 Bh2 Rh5 with good counterplay; the bishop secures the king’s safety while the queen and the rook will bother the opponent; now 24 Ra7? doesn’t work due to 24...Qf5 =+) 18 a4! bxa4? (a more careful option is 18...Bxb3 19 Qxb3 bxa4 20 Rxa4 Rxa4 21 Qxa4 b5 +=) 19 Bxa4! Rc8 (19...Rd8!?) 20 Bb3 b5 21 Rfe1 Bxb3 22 Qxb3 Qxd4? (there is no satisfactory defence, so Black might as well take the material) 23 Rad1 +– G.Ginzburg-S.Kasparov, 2nd Bundesliga 2010/11. White’s overwhelming advantage in development results in an irresistible attack.

11 Qf4 Be6 12 Nxb6 Qxb6 13 0-0-0 0-0-0 14 Be2 (D) and now:

Black to play a) 14...Nd7?! is inappropriate as it leads to a loss of control over the centre because of 15 d5 Bxd5 16 Nxd5 cxd5 17 Rxd5 Nxe5 18 Rxd8+ Qxd8 19 Qxe5 +/–. The black king risks getting caught in the draught.

b) 12...b5 looks very natural and interesting. After 13 Bxf6 gxf6 I can’t see anything terrible; 14 d5 Bd7 is unclear.

b) 14...h5 was played in Preotu-Tiviakov, Montreal 2014. An inexperienced player may not get the point of such moves. Whom does Tiviakov intend to assault? However, we have already seen this type of move on the kingside many times. The g4-square is placed under control and the rook, on occasion, is ready to leap to h6, h5, etc. Besides, one can also include the g-pawn in the attack. However, after 15 Kb1 Black went wrong with 15...Ng4? (15...Rg8 is more accurate, intending ...g5 with counterplay) 16 Bxg4 hxg4. At a glance, Black’s position looks quite nice: bishop-pair, both rooks have something to do, and the unpleasant ...f6 is threatened. However, both players missed the effective strike 17 d5!. There is no mate, but White concentrates his efforts on the d-file and Black is powerless to oppose this:

c) 12...Nd7!? is also not bad. After 13 Bg3 (13 Bd3?! Nxe5 14 dxe5 g6 15 Qe3 Bg7 16 f4 0-0 and now 17 0-0 and 17 a4 are both well met by 17...b5) 13...Nf6 it is far from clear what White should play; e.g., 14 Be2 (14 Rd1 b5 with counterplay) 14...g6 (14...Qxd4 15 0-0 Qc5 is unclear) 15 Be5 Bg7 16 a4 0-0 17 0-0 Nd7 18 Bxg7 Kxg7. We now return to 10 Be5 (D):

23

naked eye. After 18...Kf8 19 Bxe6 Re8 20 Na4 Qb4 21 Qb3 Black was left a piece down: 21...Bxf4+ 22 Kb1 Qxb3 23 Bxb3 +–.

17...cxd5 (17...Bxd5? is even worse since after 18 Nxd5 cxd5 19 Rd3 there is no knight on c3, so the rook rushes there, creating a decisive attack: 19...Rh6 20 Qxg4+ Re6 {20...Rd7 21 Rxd5 +–} 21 Rc3+ +–) 18 Rd3 g5 19 Qd2 Rh6 20 Rd1 +/–.

We now return to 14...axb6 (D):

11...h5 12 h3 (D)

White to play 15 Be2 b5

Black to play

15...Bf5!? is interesting: 16 Qd2 e6 17 0-0 b5 18 Bf3 Nd7 =.

12...h4 13 Qe3 In E.Hossain-Tiviakov, Turin Olympiad 2006, White handled the position in a different manner: 13 Qf4 Be6 14 Ne3 Nbd5 15 Nexd5 Nxd5 16 Qd2 Nxc3 17 Qxc3 Bd5 18 Qd2 e6 19 c4 Be4 20 f3. If I were him I would beware of making such moves as now the black h4-pawn unequivocally restrains two others (g2 and h3). An attempt to free them by playing g4 (or g3) will result in the creation of isolated pawns on the h- and f-files. After 20...Bf5 21 a3 Rh5 (I like manoeuvres of this kind; remember that rooks don’t have to stay on the edge of the board!) 22 Rd1 Bd6 23 Qe3 (23 Bxg7? Bg3+ =+) 23...f6 24 Bxd6 Qxd6 = Black didn’t have any problems. However, White managed to hold the draw.

16 Bxf6?! This is an unsuccessful idea. The doubled pawns don’t worsen Black’s position, whereas the bishop-pair will be an appreciable plus. White should play 16 0-0 b4 17 Ne4 Nxe4 18 Qxe4 Qd5 19 Qe3 +=. 16...gxf6 17 0-0 Bh6 18 Qe4 After 18 f4 Rg8 =+ White’s kingside looks vulnerable, especially the dark squares (g3 and f4). 18...f5! (D)

13...Be6 14 Nxb6 axb6 If there is a choice, I generally prefer to capture on b6 with the pawn. The positives are obvious: a) The a8-rook immediately joins the struggle without even having to move. b) An additional resource arises – the advance of the front b-pawn (...b5 and even ...b4), while the other pawn can remain on b7, keeping the black queenside solid.

White to play Obviously, White had not considered this idea, which gives Black an edge.

In Huschenbeth-Dranov, Bonn 2011 Black played less soundly: 14...Qxb6?!

19 Qe5 15 0-0-0 g6 16 Bxf6 exf6 17 Bc4 Bh6 18 f4 +/–. Black’s hardships in the centre are clear to the 24

19 Qxh4?! is even worse: 19...Bg7 20 Qg3 Qxd4 –/+.

30 a3 Bd6 31 Kf1 e5 32 c3 e4 Alas, it is too late. Now it is unrealistic to double rooks on the g-file.

19...Rg8 Houdini recommends 19...0-0 but for a human being it seems too exotic to castle when there is so little pawn-cover.

33 Be2 Ke6 34 Bh5 Bc5 35 Be2 f4 36 f3 e3 When making this move, I realized that White will build a fortress on the light squares, but I didn’t have any worthy alternative. Besides, Black can still hope for a breakthrough on the queenside.

20 d5?! This is an unfounded decision; the situation still doesn’t require such radical measures. 20 Bh5 Kf8 =+; after 20 Rfd1 White stands just a little bit worse.

Weaker is 36...f5?! 37 fxe4 fxe4 38 Bg4+ Kf6 39 Rd7 with counterplay; then it is already Black who must be careful.

20...Bg7 21 Qf4 Bxd5 22 Nxd5 Qxd5

37 Bd3 Bd6 38 Be2 Be5 39 Bd3 Kf6 40 Be2 Ke7 41 Bd3 (D)

White has achieved a position with oppositecoloured bishops, which might be a good idea if we were already in a pure ending. However... 23 Bf3? 23 Rad1 Qe5 24 Qg5 Rh8 =+. 23...Bh6 –/+ 24 Bxd5 Bxf4 25 Bf3 (D)

Black to play 41...Bxc3!? How else can White’s defensive line be broken? 41...Ra5 hardly makes sense, as 42 Ke2 c5 43 Be4 b4 44 cxb4 cxb4 45 Rd5 leads to an exchange of rooks and a draw. Black to play

42 bxc3 Rxa3

Somewhere around here I lost the thread. The ...e5-e4 advance, followed by doubling rooks on the g-file, looks powerful.

At least Black doesn’t have to worry about a draw due to opposite-coloured bishops any more! 43 c4! bxc4

25...e6

43...b4 44 c5 Rc3 45 Ke2 Rxc5 46 Rb1 =.

It seems logical to put the pawns on light squares, but it is not vigorous enough. 25...e5! 26 Rfe1 Rg5 –/+ is better.

44 Bxc4 b5 45 Be2 b4 Surely, it is desirable to keep pawns on light squares but how should it go forward then?

26 Rfd1 Ke7 27 Rd3 Rgd8 28 Rxd8 Rxd8 29 Rd1 Ra8

46 Ke1 (D)

Black must preserve a pair of rooks because a pure opposite-coloured bishop ending is too drawish. 25

Game 6 Dann – S. Kasparov 2nd Bundesliga 2009/10 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Nc4 Qc7 8 Qf3 Nb6 9 Ne5 This is not a repetition of moves, as the queens have moved to f3 and c7 respectively. White’s idea is to avoid the exchange of knights; as I have already mentioned, keeping as many pieces as possible on the board is favourable for the side who has more space, i.e. White.

Black to play 46...c5 There was a much meatier option in 46...Rc3!?, but it probably also misses the target: 47 Rd3 Rxd3 48 Bxd3 Kd6 49 Kd1 Kc5 50 Kc2 Kd4 51 Be2 c5 52 Kb3 f6. Is this zugzwang? Now any bishop move is losing: 53 Bf1?? e2! 54 Bxe2 Ke3 55 Bb5 Kf2 56 Bd7 Kxg2 57 Bg4 f5 –+ and the doubled pawn turns out to be useful! But White can hold the draw by 53 Kc2 c4 54 Kb2 b3 (54...c3+ 55 Kb3 =) 55 Bf1 e2 56 Bxe2 Ke3 57 Bxc4 Kf2 58 Be6 Kxg2 59 Bg4 f5 60 Bxf5 Kxf3 61 Kxb3 Kg3 62 Kc2 f3 63 Kd2 f2 64 Bd3 Kxh3 65 Bf1+ Kg3 66 Ke3 =.

Similar structures arise after 9 Bf4 Qd8 10 Ne5 Be6 (D) (10...Qxd4 transposes to the note to White’s 10th move in Game 1).

47 Rc1! Kd6 48 Bc4 f5 48...Ke5 doesn’t make progress: 49 Bxf7 Kd4 50 Rd1+ (50 Ke2 Rc3 51 Rd1+ Ke5 52 Ke1?! b3 53 Rb1 c4 –/+) 50...Rd3 51 Rc1 Rc3 52 Rd1+ Ke5 53 Bg6 c4 54 Be4 b3 55 Ke2. The rook on c3 is clumsy.

White to play

49 Ke2 Kc6 50 Be6 Kd6 51 Bc4 Kc6 52 Be6 Kd6 53 Bc4 ½-½

With the knight on e5, it looks quite reasonable to develop the bishop on e6. It establishes additional control over d5, while it doesn’t obstruct the king’s bishop, which is going to be fianchettoed. Some typical examples:

Conclusions

a) 11 Rd1

In the notes to this game we have seen the author’s sad experience in the Bundesliga (G.Ginzburg-S.Kasparov) and Tiviakov’s unfounded loss of control over the key d5 point (Preotu-Tiviakov), both the result of inattentive play; don’t repeat our mistakes. In the main game you will have noticed the problems White faces after an unnecessary exchange of bishop for knight. Doubled pawns on both flanks gave me more benefits than negatives, whereas we saw (in Huschenbeth-Dranov) Black getting into serious trouble after avoiding doubled b-pawns.

g6 12 Be2 Bg7 13 0-0 0-0 14 h3 Qc8 (an attempt to keep the game complicated and play for a win; a simpler option is 14...Nfd5 15 Nxd5 Qxd5 =) 15 Ng4 Nxg4 (the right decision) 16 hxg4 Rd8 17 Be5 Bxe5 (the ‘computer-like’ 17...f6 is interesting; e.g., 18 Bf4 f5! 19 gxf5 Bxf5 20 g4 Be6 =) 18 dxe5 Qc7 19 Qe3. Two pairs of minor pieces have left the battlefield and the one open file heralds rook exchanges. Black has solved his opening problems. Rather than 19...Nd5 (as in Lomako-S.Kasparov, Lipetsk 2010), 19...a5 20 f4 Rxd1 21 Rxd1 Rd8 is even more precise: 22 Rxd8+ Qxd8 is equal, while 22 Rf1 preserves the rooks but promises no special benefits: 22...Bc4

And now let’s consider ideas for White based on returning the knight to e5. 26

take White up on this offer. 13...Nfd5 (this leaves the g4-knight out of the picture, as a leap to h6 doesn’t make much sense while there are no other plans in sight; after 13...Nxg4?! 14 hxg4 Nd5 15 Nxd5 Qxd5 16 Qxd5 Bxd5 17 f3 White has a slightly better pawn-configuration) 14 Nxd5 (14 Be5 f6 15 Bg3 {15 Bh2 h5} 15...h5 16 Ne3 Nxc3 17 bxc3 =+) 14...Qxd5 (I imagine you have already become accustomed to this motif; the queen attacks both a2 and f3, forcing the exchange) 15 Qxd5 Nxd5 16 Be5 f6 17 Bh2 h5 (Black gains a tempo thanks to the g4-knight) 18 Ne3 Bh6 19 Re1 and now rather than 19...0-0-0, as I played, 19...Kf7 leaves Black with a certain initiative.

23 f5 Bxe2 24 Qxe2 Nd5 25 fxg6 Qb6+ (25...fxg6 =) 26 Kh2 hxg6 =. b) 11 0-0-0 (D) and here:

b22) 12 h4 Bg7 13 Be2 (D).

Black to play b1) 11...Nbd5 12 Bc4 Nxf4 13 Qxf4 Bxc4 14 Nxc4 e6 15 Ne5 Be7 16 Kb1 0-0 was a rapid simplification in Siugirov-Danin, Belgorod 2010. Black will hardly experience problems, as he has no weaknesses. The Elo-favourite (i.e. White) tried to create an attack but without a ‘hook’ (since the pawns are on h7, g7 and f7) it was all very slow: 17 h4 Nd5 18 Nxd5 exd5 19 h5 Bd6 20 Qg4 Qc8 21 Qf3 Bxe5 22 dxe5 f6! 23 h6 fxe5 24 Qg3 g6 (24...Rf7!?) 25 Qxe5 Rf7 26 f3 Qf5 27 Rhe1 Raf8 28 Qxf5 gxf5 29 Re6 Rf6 30 Re7 R8f7 31 Re8+ Rf8 32 Re7 R8f7 33 Re8+ Rf8 ½-½. Black’s defence was carried out almost ideally and may serve as a model for positions of this kind.

Black to play 13...Nfd5 (it is probably more precise to capture with the other knight, viz. 13...Nbd5!?, leaving its colleague on f6 to consolidate the kingside; after 14 Nxd5 Qxd5 {14...Nxd5!?} 15 c4 Qxf3 16 Bxf3 White’s initiative is within reasonable limits; usually the queens’ disappearance considerably eases Black’s game) 14 Bd2 (14 h5 is hardly dangerous; Black can reply 14...Nxc3 15 Qxc3 Nd5 16 Qd2 Nxf4 17 Qxf4 Qd6 or 14...Nxf4 15 Qxf4 Nd5 16 Nxd5 Qxd5 17 c4 Qd6, when nothing terrible is likely to happen to him) 14...Nxc3 15 Qxc3 Qd5 16 Qa3 0-0 17 Bf4 f6 (Black can try to win a pawn by 17...Qe4 18 Be3 Bxe5 19 dxe5, though after 19...Nd5 20 Bf3 Qxe5 21 Bd4 Qf4+ 22 Rd2 White has compensation; this is hard to assess, but not everyone will be happy to face the h5 advance without a darksquared bishop) 18 Nf3 Nc4 (18...Rfe8!?) 19 Qc3 b5 (this looks reasonable; with opposite-side castling, one should attack as soon as possible) 20 b3 (D).

b2) 11...g6 (D) and then:

White to play b21) The modest 12 h3 Bg7 13 Ng4 was played in T.Pähtz-S.Kasparov, Arco 2011. The idea seems sound, since an exchange on g4 leads to a structural change that augurs well for White. But Black has no need to 27

initiative in Dolzhikova-S.Kasparov, St Petersburg 2013. The game continued 16 Bg2 Nd7 (16...c5! =+) 17 Nxd7 Nxc3+ 18 Qxc3 Qxd7 (better is 18...Rxd7) 19 Qg3 Bd5 20 c3 b5 21 Rhe1 a5 22 Bxd5 cxd5. This is all rather prosaic, but Black did take over the offensive and went on to win. Black had a more incisive option in 14...Nfd5!, activating the so-far sleeping bishop on g7. After 15 Bh2 (or 15 Bc1 c5 –/+) 15...c5! – /+ the white pieces are suddenly skewered on the h8-a1 diagonal. Black to play

b) 11 0-0-0

20...Nb6?! (so far Black’s play has been rather logical, but around here I started to lose the plot; the tough 20...Qe4! is correct; e.g., 21 g3 b4! {distracting the queen from f3} 22 Qe1 {22 Qxb4? Qxe2} 22...Na3 {also equal is 22...Nb6 23 Bd3 Qxe1 24 Rdxe1 Bd5 25 Be4 =} 23 Bd3 Qxe1 {23...Qxf3 24 Qxe6+ +–} 24 Rhxe1 Bg4 25 Be2 {25 Be4 Rac8 =+} 25...Rfd8 =) 21 h5, T.L.Petrosian-S.Kasparov, Bhubaneswar 2011. White has seized the initiative, and while the fight lasted for about 100 moves, it was only White who could hope for winning chances.

(D) is more uncompromising, avoiding spending time playing h3.

We now return to 9 Ne5 (D): Black to play

I must confess that some of the lines here are so intricately intertwined, that I myself confuse them from time to time. However, I remind you that in this case the most important thing is understanding the plans, rather than the moveorders. Our main game shows Black’s resources in this ‘boring’ opening rather nicely.

11...a6 (this preventive move is directed against a possible pawn-break in the centre; e.g., 11...g6 12 d5!? Nfxd5 13 Nxd5 Nxd5 14 Rxd5 Bxd5? {14...g5} 15 Qxd5 cxd5 16 Bb5+ Qc6 17 Nxc6 a6 18 Ba4 +/–) 12 Kb1 g6 13 Be2 Bg7 14 g4 Nbd5 15 Na4?! (I can understand the desire to exploit the weaknesses on b6 and c5, but it is naïve to suppose that Landa hadn’t foreseen it) 15...Nxg4! 16 c4 (not 16 Nxg4? Bxg4 17 Qxg4 Qxg4 18 Bxg4 Nxf4 –/+) 16...Nxf4 17 Qxf4 Nxe5 18 Nb6 Qd8 19 Nxa8 Qxa8 20 dxe5 Qb8 21 h4 and in Sharafiev-Landa, Khanty-Mansiisk 2009 the tactical skirmish had ended in Black’s favour. After the simple 21...Qxe5 22 Qxe5 Bxe5 =+ Black has a comfortable ending where he can fight for victory without any risk. White’s h- and fpawns are weak, while Black’s dark-squared bishop is very strong.

9...Be6 10 Bd3

We now return to 10 Bd3 (D):

Black to play

Or 10 Bf4 Qc8, and now: a) 11 h3 g6 12 0-0-0 Bg7 13 Kb1 0-0 14 g4?! and now 14...Nbd5?! 15 Bc1 Rd8 gave Black the 28

18...f6 19 Bf4 Nxc3 20 bxc3 Bd5 21 Qe3 Re8 22 Nc5 22 Bh6 Qa5 =. 22...Qc8 (D)

Black to play 10...g6 10...Nbd7!? is a more concrete solution: 11 Bf4 Nxe5 12 Bxe5 Qb6 13 0-0 Nd7 14 Ne2 Nxe5 15 dxe5 h5 16 h3 g5 with counterplay, Sax-Brancaleoni, Bratto 2005.

White to play I can’t understand why the computer prefers White’s position here. The bishop-pair and superior pawn-structure speak in Black’s favour.

11 0-0 Bg7 12 Bf4 Qd8 13 Rad1 0-0 14 h3 Nbd5 = 15 Rfe1 Nb4 (D)

23 Qd3 b6 24 Ne4

Based on the point that the bishop is tied to the protection of the c2-pawn. The simple 15...Nxf4 16 Qxf4 Qd6 = is also good.

This is the right decision. The pseudo-active 24 Na6?! is weaker in view of 24...e5 25 Bh2 (25 dxe5 fxe5 26 Bh2 e4 27 Qe2 Re7 28 c4 Bf7 29 Bd6 Rd7 30 c5 b5 leads to trouble with the knight because of 31 Nb4? a5 –+) 25...e4 26 Qe2 Re7 27 Nb4 Qe6 =+. 24...Qf5 25 Bg3 Rad8 (D)

White to play 16 a3 Nxd3 17 Nxd3 Nd5 17...Qxd4 may be objectively stronger but in a team competition I didn’t want to take such a risk. White has sufficient compensation after 18 Be5 Qb6 19 Na4 Qb5 20 Nc3.

White to play Full centralization. Now Houdini agrees that White has problems.

18 Be5 Black is alright in the case of 18 Nxd5 Bxd5 19 Qe2 Re8 20 Nc5 b6 21 Na6, as he has enough resources to repel the aggression: 21...Rc8 22 c4 Be6 23 Nb4 Bd7 24 d5 cxd5 25 Nxd5 e6 26 Ne3 Qe7 =.

26 Re3 Time-trouble. 26 c4 Bf7 27 c5 b5 leaves Black with the initiative. 26...Bh6 27 Re2 Bf7 28 Rde1 Kf8?! 29

The strongest option is 28...Qa5; for example, 29 c4?! f5! (alas, I didn’t find this move; 29...Qa6 misses the target since after 30 Qc3, 30...Qxc4?? doesn’t work due to 31 Nxf6+ +–) 30 Nc3 (30 Nd2? e5 31 c3 exd4 32 Rxe8+ Bxe8 –+) 30...Qxa3 –/+.

have been over and I could have found 42...Bxg2, leading to a win.

29 Qa6 Rd7 30 a4?

Conclusions

38...Ke8 39 Qb6 Bg7 40 Bd6 Bd5 41 Bc5 Qc6 42 Qb8+ 0-1

In time-trouble it would be difficult to venture 30 c4!, when a possible line is 30...Rxd4 (30...Qa5 31 Qxa5 bxa5 32 c3) 31 Qxa7 Qa5 (31...Qd7 32 Qxb6 Rxc4 33 a4 is unclear) 32 Qc7, when Black is only a little better and White has counterplay.

I hope that the reader can appreciate some hidden nuances in these lines. My games with Lomako and T.Pähtz demonstrate the nature of the fight following a queen exchange. Both players can play for a win, even well into the ending. In T.L.Petrosian-S.Kasparov, we saw an interesting struggle with opposite-side castling with the queens still on the board. Sharafiev-Landa featured an instructive exchange sacrifice by Black. Please note an additional resource which I missed (14...Nfd5! and 15...c5!) against Dolzhikova. We shall see a similar motif again in the final section of Chapter 2 (which deals with 5...g6). In Sax-Brancaleoni Black played a sharp advance with ...h5 and ...g5 – quite a rare idea.

30...Red8 31 Kh2 31 a5 Qb5! 32 Qxb5 cxb5 –/+. 31...Qd5 (D)

Our main game took a relatively logical course and in the end the two bishops proved their superiority. White’s broken pawns on the queenside were also a problem for him. White to play As is well-known, the squares g6, g3, b6 and b3 are often poor locations for knights due to their limited space for manoeuvres. A typical way of pushing them aside in such cases is an advance of the rook’s pawn. Is it effective in our opening? In the next game we shall see White making an early advance of his a-pawn. Only the author’s poor play led to a deplorable result for Black; the opening variation itself was hardly to blame.

Black has a large advantage. 32 Nd2 c5 33 Nf3? With the absence of good moves (and time, by the way) one is left only with bad ones. 33...Qa2 –+ Now there is no satisfactory defence against ...Bc4.

Game 7

34 Rxe7

Harutjunyan – S. Kasparov

This is the best practical option. 34 Qb5 Bc4 35 Qc6 Bxe2 36 Rxe2 Qd5 –+ makes matters easier for Black.

Bhubaneswar 2010 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Nc4 Qc7 8 a4!? (D)

34...Rxe7 35 Rxe7 Kxe7 36 Qxa7+ Rd7 37 Qxb6 Qxa4 38 Qxc5+ 38 dxc5! would have given Black the greatest difficulties. After 38...Bd5 39 Nd4 Qa7 40 Qb5 Qb7 41 Qa4 Qa8 42 Qb5 the time-trouble would 30

a) 9...a5!?, preventing White’s advance, seems reasonable. It looks so logical and natural that I don’t understand why other moves have been played more often. Furthermore, the a5-pawn also establishes a strongpoint for the knight on b4. In general, I consider the pawn set-up c6-b7-a5 to be very solid, and I doubt Black would like to castle queenside in any case once White has played a4. This line was confidently played by the Polish veteran in A.Vovk-Pytel, Le Touquet 2009: 10 Qf3 Be6 11 Bf4 Qc8 12 h3 g6 13 Bd3 Bg7 14 0-0 0-0 15 Rfe1 Nbd5 16 Bh2 Nb4 =.

Black to play An interesting idea. White intends to squeeze the opponent by moving the pawn to a5. However, I feel that this idea should not be too dangerous for Black, and that he should continue with his main plan.

b) 9...g6?! seems less attractive, as it allows White to carry out the planned advance, creating more pressure. After 10 a5 (D), the b6-knight starts an unattractive flight in search of some shelter:

8...Nb6!? The knight rushes to b6, before this square is denied to it. 9 a5 After this immediate advance Black can simply exchange on c4, whereupon he will be able to breathe more freely. Although the further advance of the pawn to a6 may be possible, it is unlikely to prove very effective, as the white bishop is not fianchettoed on g2, but stands on c4 where it is subject to the counterblow ...b5.

Black to play b1) 10...Nbd7 is no better than putting the knight on d5. 11 Nc4 (this is the point of the a4-a5 advance; the black knight can’t come back to b6, and actually has little scope; if earlier White’s space advantage was based just on him being the only one with a pawn on a central square, then now he also has a ‘winger’ on a5) 11...Bg7 12 g3 0-0 13 Bf4 e5!?. Black lands a central counterblow before White has finished developing – a reasonable idea, but will it solve the problem of his lack of space? After 14 dxe5 Nh5 (Mihok-G.Szabo, Budapest 2014) 15 Qd4! Re8 16 0-0-0 Nxe5 17 Bxe5 Bxe5 18 Nxe5 Rxe5 19 Qd8+ Qxd8 20 Rxd8+ Kg7 21 Bh3 f5 22 a6 += one doesn’t have to be a grandmaster to understand that Black lacks real counterplay and has only a gloomy defence ahead of him.

Little wonder that White is unhappy with such an outlook, so before pushing the pawn to a5, he has tended to direct the knight back to the centre by 9 Ne5 (D), planning a5.

Black to play Then Black has continuations:

quite

a

b2) 10...Nbd5 11 Nxd5 Nxd5 12 c4 (D). wide

range

of

31

(18...Bxe5?! 19 fxe5 Rxd4 20 Qf3 Nd5 is rotten due to 21 Qxf7 +/–) 19 Ra4! (an excellent plan; although rook-lifts along the third or fourth rank are a standard way to fight for the initiative, they are nevertheless often overlooked, even by experienced players) 19...Kb8 (19...Bxe5? 20 Rc4 +–) 20 Rc4 Qe7 21 Qxd4 Nd5 22 Rc6 +/– Caruana-Granda, Zafra 2009. White has a readymade queenside plan with the c4 advance. d) 9...Be6 (D) has also been tested, developing the bishop before the knight retreats to d7. While this is possible, I don’t really like it.

Black to play 12...Nf6 (entering the enemy camp with 12...Nb4 is no panacea) 13 Be2 a6 14 Bf4 Qd8 15 Bf3 (15 Qa4? Qxd4 =+ 16 Nxc6? Qxf4 –+) 15...Bg7 16 00 0-0 17 Qd2 and it is rather clear, even at a glance, that White stands better. Although in Sangma-S.Kasparov, Mumbai 2010 I managed to equalize by careful defence, I cannot offer a good reason why I allowed a4-a5 and then suffered from a lack of space. The set-up with an a5-pawn versus the c6-b7-a6 structure (one pawn holds back three others!) is favourable for White.

White to play

c) I also don’t like 9...e6 (D).

10 a5 Nbd7 11 Bf4 Nxe5 12 Bxe5 Qd8 (this setup seems more reasonable than ...e6 but only if Black manages to finish his development) 13 a6 b6 14 Be2 Nd5?! (as is sometimes the case in positions of this kind, 14...Nd7!? would make sense, chasing the bishop away from the dominant height; after 15 Bg3 Nf6 16 Bf3 Rc8 17 Be5 Nd7 Black awaits White’s reply with curiosity) 15 0-0 Nxc3 16 bxc3 Qd7 17 c4 (17 Re1 is probably more accurate: 17...f6 18 Bf4 g6 19 Bd3 +/–) 17...f6 18 Bg3 g6 19 Re1 Rd8 20 c3 Bh6 21 Qb3 0-0 22 Rad1 Bf7 23 Bf3 Rfe8 += ParligrasShishkov, Varna 2010. Black is past the worst, having mobilized his forces without any material losses. White should have acted more sharply.

White to play Please think for a while about why that might be the case before reading further. Even if you are fairly new to chess, you will probably be aware that it is desirable for Black not to lock in the bishop on c8. It needs to have some prospects (on g4, f5, e6 or b7). After 10 Bd3 Nbd7 11 Bf4 Nxe5 12 Bxe5 (the exchange on e5 can be considered an achievement for Black but it doesn’t equalize, as the c8-bishop is still suffering) 12...Bd6 13 f4 Bd7 14 0-0 0-0-0 15 a5 c5 16 a6 b6 17 Nb5 I do not believe many people would be happy to play Black’s position. Of course, the light-squared bishop can be exchanged but the situation has considerably changed, and the c6and b7-squares are weak. 17...Bxb5 18 Bxb5 cxd4

We now return to 9 a5 (D):

32

Black to play

19...c5! 20 Qxb5 cxd4 is best, with the possible continuation 21 cxd4 Rab8 =+. I recall that I didn’t like 21 Qb7, though in that case 21...Qd7 (or 21...Rac8 =+) is quite good; after 22 cxd4 Rab8 23 Qxd7 Rxd7 White’s extra pawn will soon be lost while the flaws in his position remain. Furthermore, here the bishop, in combination with rooks, is preferable to the knight.

9...Nxc4 10 Bxc4 Bf5 11 Bg5 11 Qf3 is possible too, but I think this brings no special benefits. After 11...Bg4 12 Qf4 (12 Qg3 Qxg3 13 hxg3 Bf5 14 Bb3 e6 15 f3 Bd6 16 g4 Bg6 17 a6 b6 =) 12...Qxf4 13 Bxf4 e6 14 a6 b5 15 Bd3 0-0-0 = the sacrifice on b5 is not dangerous when the queens are off the board; e.g., 16 Bxb5? cxb5 17 Nxb5 Nd5 18 Nxa7+ Kd7 19 Bg3 (19 Bd2 Nc7 20 Bf4 Ra8 21 Bxc7 Kxc7 22 Nb5+ Kb6 –/+) 19...Bb4+ 20 Kf1 Ra8 21 Nb5 Bf5 –+.

20 Qe3 (D)

11...e6 12 a6 A weighty decision. The breakthrough doesn’t achieve anything clear-cut, whereas there is a risk that the far-advanced pawn can be surrounded. 12...b5 13 Bd3 Bxd3 14 Qxd3 Be7 15 Bxf6 Bxf6 16 Ne4 Be7 17 0-0 0-0 18 c3 Rfd8 (D) Black to play

Now White committed an inaccuracy but I failed to exploit it. After that things went from bad to worse.

20...Rd5 I was seeking to prepare a more favourable pawnbreak, thinking that my opponent would be unable to stop them in the long term, but this was in vain. The status quo hasn’t been broken yet, but the positional inaccuracy on the next move turned out to be hard to repair.

Before reading on, think about what White should play next.

Black’s game is wholly acceptable after 20...c5 21 dxc5 Bxc5 22 Nxc5 Qxc5 23 Qxc5 Rxc5 24 Rad1 Rcd5 =. 21 b4! h6?! This was Black’s last chance to free himself: 21...e5! 22 dxe5 (other continuations are weaker; e.g., 22 Qg3 is met by 22...f5 =+, and not 22...exd4?? 23 Nf6+ Bxf6 24 Qxc7 Rxc7 25 Re8#, while 22 Ng3 Bd6 gives Black an edge after 23 dxe5 Rxe5 =+ or 23 Nf5 exd4 =+) 22...Rxe5 23 f4 Re6 =.

White to play 19 Rfe1?! If White tries to prevent ...c5 by playing 19 b4 then after 19...e5 20 Qg3 f5 21 Nc5 Bxc5 22 bxc5 f4 Black enjoys a certain initiative. 19 Qe3 is probably best; e.g., 19...Rd5 20 b4 (20 f4 c5 gives Black the initiative) 20...e5 21 dxe5 Qxe5 22 f4 Qe6 =.

22 f4! g6 23 Rf1 Kg7 24 Rf3 (D)

19...Rac8?! This is somewhat sluggish.

33

Black can fly the coop by means of 41...a5 42 bxa5 b4 43 cxb4 Rxd4 44 Qf6 =. Armed with a computer engine, we can see that Black is alright here, but over the board and under time-pressure, everything seemed bleak. 33 fxe5 f5 34 Qh4 Qd8 35 Qxh6 Qf8 36 Qe3 Rh8 37 Qg5 Qg7 38 Rf1 Rdd8 More stubborn is 38...Rh5 39 Qxh5 gxh5 40 Rxg7+ Kxg7 41 Rf3 Rd8 42 Rg3+ Kh6 43 h4 +/–, though it is hardly possible to hold this ending. White has a clear plan: the king goes to d3, and then he plays c4; he can also transfer his rook to the c-file. 43...Rb8 is met by 44 Rg5 Rb6 45 Rg8 Rxa6 46 Re8 +–.

Black to play Harutjunyan has managed to clamp down on Black’s potential pawn-breaks and now his position is a little better. In addition, White has ways to strengthen his game (rook transfer to the f-, g- or h-files; advance of the h- and g-pawns...), while Black’s intentions are less distinct.

39 Rff3 Rdg8 40 Rh3 Ke8 41 Rfg3 Kd7 42 Kg1 Kc7 43 Rxh8 Rxh8 44 Qxg6 Qxg6 45 Rxg6 Re8 46 h4 Kb6 47 h5 Kxa6 48 h6 Kb6 49 Kf2 a5 50 bxa5+ Kxa5 51 h7 Rh8 52 Rxe6 c5 53 Rh6 b4 54 cxb4+ cxb4 55 e6 b3 56 e7 b2 57 e8Q b1Q 58 Qxh8 Qb2+ 59 Kg3 1-0

24...Re8 25 Nf2 Bd6 26 Nd3 Qc8 The only way to inconvenience White is by keeping pressure on the a6-pawn.

In the next game we shall see what nuances there are if Black first fianchettoes his dark-squared bishop, with the knight still on d7.

27 Kh1 Rg8 28 Ne5 f6 29 Nd3 Rh8 30 Qe4 Kf7 31 Rg3 Rg8 32 Ne5+?! (D) A key moment in the middlegame. I had managed to construct a rather solid defensive line, and my opponent, not seeing a ‘normal’ way to make progress, opted for an active but dubious idea.

Game 8 Radjabov – Nisipeanu Bazna 2009 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Nc4 Qc7 8 a4 g6 (D)

Black to play 32...Bxe5? 32...fxe5! is correct: 33 fxe5 Be7 34 Rf1+ Ke8 35 Rxg6 Rxg6 36 Qxg6+ Kd7 37 h3 (37 Rf7?! Qxa6 38 h3? {38 Qb1 –/+} doesn’t work due to 38...Qa1+ 39 Kh2 Qxc3 40 Qg7 Qxb4 –+) 37...Qxa6 38 Qh7 Qc8 39 Rf7 Qe8 40 Rg7 =. At the board I was misassessing positions like this since the d5-rook is ‘arrested’. After 40...Qf8 (40...Kd8?? 41 Rg8 Bf8 42 Qh8 +–) 41 Qxh6

White to play 9 g3 Or: a) 9 Bd3 is not a promising moment to develop the bishop to d3: 9...Bg7 10 0-0 0-0 11 Qf3 Nb6 12 Ne5 a5 13 Bf4 Qd8 14 Rad1 Be6 (a standard 34

b2) 9...Bg7

knight transfer with ...Nbd5-b4 is on the agenda; then the d3-bishop will be unable to retreat because of the c2-pawn) 15 b3 Nbd5 16 Nxd5 Nxd5 17 Bd2 (C.Balogh-Babula, European Team Ch, Porto Carras 2011) and now 17...Qb6! would even have given Black the better chances.

10 Bf4 Qd8 11 a5 (a well-founded decision; now the d7-knight is restricted, while the further advance to a6, destroying the base pawn on b7 is threatened; a question arises: why not carry out the leap on the 9th move by 9...Nb6 instead?) 11...Nh5 12 Be3 0-0 (D).

b) 9 Qf3 is another standard idea. Then: b1) 9...Nb6 10 Ne5 a5?! is an unfortunate attempt to combine ...g6 and ...a5. The intention is good – to bring White’s aggression to a halt. However, the cost in time is too great. Once in our local chess school a mom of a kid ‘informed’ me that the Scandinavian Defence is less sharp than the Sicilian, and tempi matter less. Well, I dare to venture that in almost any opening tempi are important, and this position is not very blocked. 11 Bf4 Qd8 (D).

White to play 13 g4 (I doubt that Sergei was satisfied with his position at this moment) 13...Nhf6 14 Bg2 (14 g5!?; after 14 a6 Nb6 15 axb7 Bxb7 16 Na5 Qd7 nothing fatal can be seen) 14...Nb8! (I like moves like these; note that Black has no pawn weaknesses and if he manages to exchange a few pieces, everything will be fine; meanwhile, it is still unclear where the white king should castle) 15 Ne5 Be6 16 0-0 Nd5 17 Ne4 Qc8 18 h3 Nd7 19 Nxd7 Bxd7 20 Nc5 f5! with counterplay, Naiditsch-Tiviakov, Montreal 2009. Sergei has overcome his mild discomfort and secured ample play.

White to play 12 d5! (exploiting Black’s unfinished development, White blasts open the game at what looks like the most heavily fortified point) 12...Nbxd5 13 Nxd5 Qxd5 (13...Nxd5 14 Nxf7 and now 14...Kxf7?? doesn’t work due to 15 Bc7+ +–, so Black must settle for 14...Qb6 +/–) 14 Bc4 (White has time to take aim at f7) 14...Qxf3 15 Bxf7+ Kd8 16 gxf3 (16 Nxf3 can be met by 16...e6) 16...Bf5? (16...e6? is also bad due to 17 00-0+ Ke7 18 Bxg6 +–; 16...Nh5 17 0-0-0+ Kc7 is necessary) 17 0-0-0+ Kc8 18 Rhe1 Nh5 19 Be3 Kc7 20 Nc4 Ra6 21 Bd4 Nf6 22 Kb1 (moves like these show how helpless the opponent is; despite having many pieces on the board, it is difficult for Black to find any useful move) 22...b5 23 Be5+ Kb7 24 Nd6+ +– led to a crushing defeat in Inarkiev-Ostojić, Ohrid 2009. Moral: in the Scandinavian Defence, just as in the Sicilian, inaccurate play can leave you facing a potent attack.

9...Bg7 (D)

White to play 10 Bg2 If White continues his pawn-push with 10 a5

35

, Black has a reasonable reply in 10...Nd5!, making use of the fact that the bishop has not yet taken control of d5 by moving to g2. After 11 Bg2 (or 11 Nxd5 cxd5 12 Bf4 Qd8 with rough equality) 11...N7f6 12 Nxd5 Nxd5 13 0-0 0-0 (as we already know, the exchange of even one knight considerably simplifies Black’s life) 14 Ne5 Bf5 15 c4 Bxe5! (D) we have:

10...Nb6 11 Nxb6 If White inserts 11 Bf4 Qd8 before playing 12 Nxb6 Qxb6, then the b2-pawn is under attack, and the game is about equal. 11 Ne5 was played in Darini-Tiviakov, Mashhad 2010: 11...a5 (that’s it!) 12 0-0 0-0 13 Bf4 Qd8 14 Qd2 Bf5 15 Be3 Nbd5 (everything occurs according to motifs we already know: the knights use the convenient bases on d5 and b4 for their manoeuvres) 16 Nxd5 Nxd5 17 c3 Nxe3 18 Qxe3 Qb6 =. 11...Qxb6! 11...axb6 looks suspicious because the b7- and b6-pawns are restrained by the single white pawn on a4. I personally witnessed Sergei’s discomfort during his game against Judit: 12 Bf4 Qd8 13 0-0 Be6 14 Re1 Qd7 (Tiviakov later chose 14...0-0 15 Be5 Re8 16 Qe2 Qd7 17 b3 Bg4 18 Qc4 Bf5 19 d5 cxd5 20 Nxd5 Nxd5 21 Bxg7 Kxg7 22 Qxd5 and after the general liquidation, the ending was approximately equal in Chytilek-Tiviakov, Leiden 2011) 15 Qd3 0-0 16 Rad1 Nd5 17 Nxd5 Bxd5 18 b3 Bxg2 19 Kxg2 += J.Polgar-Tiviakov, Hoogeveen 2009. Black is unwilling to play ...b5 because this would leave doubled weaklings on b7 and b5. But otherwise the b6-pawn is neither use nor ornament...

White to play a) After 16 dxe5 Nb4 17 Bh6 Rfd8 Black enjoys sufficient counterplay. You may be thinking that I have overlooked the bishop-pair in assessing this position. Two bishops are powerful – no one disputes this fact – but... in open positions. Here the e5-pawn hampers White’s own dark-squared bishop. And if White has to protect it with f4, this bishop will be impeded even more. Moreover, White’s pawn-structure on the queenside is inflexible, which creates strong points for the enemy knight, particularly b4.

12 a5 Qa6 (D)

b) 16 cxd5 Bf6 17 Bf4 Qd7 18 dxc6 bxc6 19 Rc1 Rac8 20 Be5 Bxe5 21 dxe5 Rfd8 22 Qxd7 Rxd7 23 f4 (23 Rxc6 doesn’t win a pawn because of 23...Rb8 and the soldier on b2 perishes) 23...Rb7 24 Rf2 Rb5 25 Rxc6 ½-½ Baklan-Epishin, Porto San Giorgio 2009. This is a high-quality game, and we can take Epishin’s play as a model. We now return to 10 Bg2 (D):

White to play Here the queen appears highly restricted and vulnerable, but matters are not so gloomy. 13 Qe2 13 Bf1?! suggests itself but is unpromising: 13...b5 (forced but sufficient) 14 Bg2 (14 axb6?? Black to play 36

Qxa1 –+) 14...Bg4 15 Qd3 Rd8 16 0-0 0-0 gives Black the initiative. White can first bring out the bishop, thus protecting the rook on a1: 13 Bf4 0-0 14 Bf1 b5 15 axb6 Qxb6 (Black’s queenside pawn-structure is damaged but this is counterbalanced by the fact that he is ahead of White in development) 16 Na4 (16 Bg2? Qxb2 =+) 16...Qa5+ 17 Qd2 Qd5 and again Black has the initiative. 13...Qxe2+ 14 Nxe2 (D) White to play The typical ‘unloading’ leads to complete equality. 21 Nxd5 exd5 = It is logical to capture with the e-pawn against the fianchettoed bishop because it lengthens the b7c6-d5 ‘breakwater’. 22 Bd2 Rxe1+ 23 Rxe1 Kd7 24 b4 Re8 25 Rxe8 Kxe8 26 Bf1 ½-½ Black to play

Conclusions

14...a6! =

When behind in development, it is risky for Black to combine two different ideas without backing it up with very accurate calculations (see InarkievOstojić).

It is never pleasant to make such moves, fixing your own pawns. Nevertheless, it was necessary because White was threatening a6. The fact that the queens and a pair of knights have been exchanged speaks in Black’s favour. Not 14...0-0? 15 a6 +=, nor 14...Bf5? 15 a6 0-0-0 16 axb7+ Kxb7 17 c3 +=.

For the umpteenth time, we saw Tiviakov defusing pressure from an elite opponent (Naiditsch-Tiviakov); here the key was the elegant manoeuvre ...Nb8!.

15 0-0 Bf5 16 c3 0-0-0

There was an instructive moment in the game Baklan-Epishin. Black made a well-timed exchange of his fianchettoed bishop for an e5knight (15...Bxe5!). We shall see further examples later in the book. This idea can also be used as a way to unbalance the game as it results in an asymmetrical set-up.

16...0-0 is a worthy alternative, putting the king on the other wing, but not the straightforward 16...Be4?! 17 Bxe4 Nxe4 18 Nf4 e6 19 Re1 Nf6 20 Nd3 +=. 17 Re1 Rhe8 18 Nf4 e6 19 h3 h5 20 Be3

In J.Polgar-Tiviakov, Black’s fixed queenside pawns came under pressure. Tiviakov managed to defend, though he had to struggle well into the endgame. That’s why the 11...Qxb6 played in the main game looks more logical. Obviously, Nisipeanu had worked out beforehand that, paradoxically, the queen is safe despite being stuck on the edge, while it performs a useful function by preventing White from castling.

20 f3 Nd5 21 g4 Nxf4 22 Bxf4 Bd3 =. 20...Nd5 (D)

37

So far we have examined games where White answered 6...Nbd7 by dropping his knight back to c4. We shall now move on to lines where White supports the e5-knight, by 7 Bf4 (Games 9-11) or 7 f4 (Game 12 ). Game 9 Oleksienko – Dražić Kavala 2007 White to play

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Bf4 (D)

9 Qd2 (9 Bd2!? leads to unclear consequences; then S.Ristoja-H.Pohjala, Jyväskylä 2013 featured accurate play even though the players’ Elo ratings were not very high: 9...Ne4 {9...Nd7? 10 e6! +/–} 10 Nd5 Qxd2+ 11 Qxd2 Nxd2 12 Nc7+ Kd8 13 Nxa8 Nxf1 14 0-0-0+ Bd7 15 e6 fxe6 16 Rhxf1 g6 with a position that is hard to assess) 9...Nd5 10 Nxd5 Qxd2+ and then:

This leads to a totally different type of game than 7 Nc4.

a1) 11 Kxd2 makes some sense as the king must be active in the ending. 11...cxd5 12 c4 dxc4 13 Bxc4 e6 14 Bb5+ Bd7 15 Bxd7+ Kxd7 16 Rhd1 Bc5 17 Ke2+ Ke7 18 Rd3 Rhd8 19 Rxd8 (19 Rg3 doesn’t yield any benefits after 19...Kf8) 19...Rxd8 20 Bg5+ f6 21 exf6+ gxf6 = MoldovanBurnoiu, Bucharest 2006. Black has absolutely solved his opening problems.

Black to play The bishop lies in ambush, creating threats against the black queen. For example, White may be able to win material by playing Ng6 or Nxf7 in some lines.

a2) 11 Bxd2 cxd5 12 c4! (the right decision; White should hurry, as otherwise Black will consolidate) 12...d4 13 Bb4 Bf5 14 0-0-0 0-0-0 15 Bc5 (White wins a pawn, for which Black gets too little in return) 15...d3 16 Bxa7 g5 (KarpatchevMann, 2nd Bundesliga 1999/00) 17 Kd2 +=.

7...Nxe5 This is the prosaic option, with which, generally speaking, Black is simply looking to equalize and draw. The main alternative is the more interesting 7...Nd5 – see Games 10 and 11.

b) Black usually trades queens in accordance with the general principle that exchanges ease Black’s defence. 8...Qxd1+ 9 Rxd1 Nd7 (D) and then:

8 Bxe5 White has a choice here. After 8 dxe5, whether Black exchanges queens or not, his prospects are joyless, only striving for a draw: a) 8...Qb4 (D).

White to play 38

secured equal play) 17 e6 (blasting open the game, denying Black the chance to set up a harmonious pawn-structure with ...e6; however, the long darksquare diagonal is also opened) 17...Bg7 18 Rd3 f5 19 Rhd1 (19 Qf2 is probably more accurate, vacating e2 for the knight) 19...d4 20 Nb1 Rd6 21 Nd2 Qa6 22 Re1 Kb8 (precise play) 23 Kb1 Rc8 24 Rh3 Qc6 25 Nb3?? d3! 26 cxd3 Rxe6 0-1 Inkiov-Therkildsen, Le Touquet 2001. One way or another, the white queen is distracted from c2, and then mate follows on b2.

b1) 10 Bc4 can be met with the tough-looking 10...g5, although with a good response from White it doesn’t guarantee equality: 11 e6! fxe6?! 12 Bxg5 Nb6 13 Be2 (with a mating threat of Bh5#) 13...h5 14 0-0 Bg7 (H.Hoffmann-Dražić, Dresden rapid 2010) and now the multi-purpose 15 Rd3 +/– looks powerful, preparing manoeuvres along the third rank and doubling rooks on the dfile. Black is in a really precarious state. b2) 10 Ne4 g6 (10...Nb6!?) 11 e6 fxe6 12 Ng5 Nf6 13 Bc7 (threatening to mate by Rd8# even without queens) 13...Bd7 14 Bc4 Nd5?! (better is 14...Bg7) 15 Be5 Rg8 16 Nxh7 Bh6 (preferring to give up the h7-pawn instead of e6; that’s a questionable decision) 17 h4 0-0-0 18 Ng5 Rdf8 19 Bd3 Bg7 20 Bxg7 Rxg7 21 g3 Kc7 22 0-0 +/– Larino-Khasangatin, Pardubice 2006. Black has no compensation for his weaknesses. I dare say that if such a tough player as Khasangatin failed to hold this position, then this line must be really unpleasant for Black.

b) 9 Bc4 (D) and here:

We now return to 8 Bxe5 (D): Black to play b1) 9...Nd7 looks somewhat fanciful – the only developed piece retreats! 10 Qf3 e6 11 0-0 (11 Qg3 +=) 11...Nxe5 12 dxe5 Qd4 13 Qe2 Be7 14 Rad1 Qc5 (Pap-Savić, Valjevo 2012) and after 15 Qe4 White is obviously better. The c8-bishop is passive, and the rook’s transfer to the kingside by Rd3-g3 is in prospect. 15...b5? doesn’t work due to 16 Bxb5 +/–. b2) After 9...Bf5 we shall follow the game S.Kasparov-Pytel, Metz 2007: 10 Bxf6 (now I am not sure this is the strongest move, but during the game I thought I should hurry to play actively because after ...e6 Black would smoothly complete his development; 10 d5!?) 10...gxf6 11 d5 Qc7 12 0-0 (better is 12 dxc6) 12...0-0-0 13 Qf3 Bg6?! (Black can secure counterplay by 13...e6! 14 dxe6 fxe6 15 Rfe1 Bd6) 14 dxc6 Qxc6 15 Bd5 += (White’s idea of a central break has paid off) 15...Qa6?! (better is 15...Qb6) 16 Rad1 Bg7 (Black must spend a tempo withdrawing the bishop from the f8-a3 diagonal; 16...e6?? doesn’t work due to 17 Bxb7+) 17 Ne4 (now Nc5 is threatened; that’s why 15...Qa6 was an inaccuracy) 17...Qb6 18 Rd3 (the rook is included in the frontal attack, bringing Black to the brink of disaster) 18...f5 (D) and now:

Black to play After the text-move (8 Bxe5) it is also not easy for Black to equalize despite some simplifications. Lack of development shows itself. In the main game, Dražić landed quickly in an indefensible position. 8...Qb4 8...Qd8 is a reasonably safe continuation. Then: a) 9 Bd3 Ng4 (this manoeuvre is similar to ...Nd7, but now the d4-pawn is attacked by the queen; nevertheless, the pawn may turn out to be poisoned) 10 Qe2 (10 Bg3!? Qxd4?? 11 Nb5 +–) 10...Nxe5 11 dxe5 Qa5 12 0-0-0 Be6 13 a3 (13 Kb1 +=) 13...0-0-0 14 f4 g6 15 Bc4 Bd5 16 Bxd5 cxd5 (Black has been playing confidently and 39

9 a3 Qa5 10 Bd3 10 Bc4!? develops the bishop more actively, but allows his black counterpart to deploy more actively too: 10...Bf5. However, there are some tactics: 11 Bxf6! exf6 (11...gxf6?? is impossible due to 12 Qh5, pinning down the f5-bishop and threatening a strike on f7, and 12...e6 allows 13 Bxe6 +–) 12 Qe2+ Be7 13 0-0 Rd8 14 d5 Qc7 15 dxc6 bxc6 16 Rfe1 (besides the fact that the pawnstructure is damaged, Black has a difficult time finishing his development) 16...Rd7 17 Rad1 Rxd1 18 Nxd1 h5 19 h3 Kf8 20 Ne3 Bc8 += G.Jones-Levi, Sydney 2010. Black’s defence has been quite reasonable, but the pressure continues.

White to play b21) I played 19 Ng5!?, which is a good move in itself, but clearly not the strongest. 19...Rd7 20 Bxf7?! (better is 20 Nxf7 or 20 Rb3) 20...Rxd3 21 Qxd3 Bf6 22 Rd1 and now with 22...Qxb2? the Polish grandmaster probably thought that if he was going to suffer anyway, he might as well grab some material. Such an approach is quite common, and went unpunished here, but it could have ended in disaster. After 23 Bxg6? hxg6 Black managed to hold his inferior position. I calculated the lines diligently, but unfortunately I failed to find the elegant 23 Bg8!! Qb6 (the bishop is untouchable: 23...Rxg8 24 Qc4+ Kb8 25 Qxg8+ +–) 24 Nxh7! (it is hard to believe that it is really worth a couple of tempi to grab a mere pawn in a position like this, but on the other hand the h-file is opened) 24...Qc7 (24...Rxg8? 25 Qc4+ +–) 25 Nxf6 exf6 (25...Qxh2+ 26 Kf1 +–) 26 Be6+ Kb8 27 h3 +–.

10...Bg4 11 Qd2 Nd7 (D)

White to play 12 Qf4 It is probably more precise to insert 12 b4 Qd8 before playing 13 Qf4 Nxe5 14 dxe5 Be6 15 Ne4 +=.

b22) Houdini claims victory after the impressive 19 Nf6!! Bxf6 20 Rb3 Qc7 21 Bxb7+ Kd7 22 Rd3+ (precisely this rook; after 22 Rd1+?? Ke8 23 Qc6+ Kf8 –+ White can’t capture the queen because of the weakness of his back rank: 24 Qxc7 Rxd1#) 22...Ke8 23 Qc6+ Rd7 24 Rfd1 +–.

12...Nxe5 13 dxe5 Be6 14 0-0-0 Now after 14 b4 Black can reply 14...Qc7. 14...g6 15 Bc4 Bxc4 16 Qxc4 Bh6+

We now return to 8...Qb4 (D):

Our silicon friend coolly recommends capturing the pawn: 16...Qxe5 17 Rhe1 Qa5, when White must prove his compensation. I wouldn’t dare to try such a provocative idea: White’s development advantage is huge, whereas Black’s material plus is slight. 17 Kb1 Qxe5 This way is no better. 18 Rhe1 Qa5? (D) White to play

Better is 18...Qf5. 40

Game 10 Anand – Tiviakov Wijk aan Zee 2006 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Bf4 Nd5! (D)

White to play 19 Qh4 By geometrical manoeuvres White emphasizes the instability of the bishop’s position. However, the computer points out that he could already have smashed through with 19 Rxe7+! Kxe7 20 Qd4!, threatening both 21 Re1+ and 21 Qd6+ Ke8 22 Re1+, and meeting 20...f6 by 21 Re1+ Kf7 22 Qd7+ Kf8 23 Qd6+ Kg8 24 Re7 Qf5 25 Qb4! Rd8 26 g4!, with decisive threats.

The idea is to establish the queen on the central square d5 after the exchange of knights (which is also desirable for Black in itself).

19...Bg5?

8 Nxd5

19...Qg5 could have prolonged the resistance: 20 Qh3 0-0 21 Ne4 Qh5 22 Qxh5 gxh5 23 Rd7 +/–. It is unpleasant to defend positions of this type, but the game is far from over.

Other continuations are weaker; e.g., 8 Nxf7??

20 Qd4 +–

Nxc3 9 bxc3 Nxe5 10 Bxe5 Qd5 11 c4 Qa5+ 12 Qd2 Qxd2+ 13 Kxd2 Bf5 14 Bd3 Bxd3 15 Kxd3 0-0-0 16 c3?! f6 17 Bf4 e5 18 Be3 Bc5 gave Black the initiative in Jerez Perez-S.Kasparov, Guben 2014; White had to struggle for a draw.

White to play

Qe6+ –+; 8 Nxd7 Nxf4 9 Nxf8 Rxf8 =; 8 Bg3

Now the bishop drops off. 20...0-0 20...f6 21 Qd7+ +–.

8...Qxd5 (D)

21 Re5 Qa6 22 Rxg5

Thus the queen escapes the bishop’s gaze and at the same time hinders the development of White’s kingside; it is unclear how the f1-bishop can join the struggle, as it is tied to the protection of the g2-pawn.

The remainder is clear. 22...Rad8 23 Qe3 Rxd1+ 24 Nxd1 Rd8 25 Nc3 f6 26 Qxe7 Qf1+ 27 Ka2 Qc4+ 28 b3 Qxc3 29 Qxd8+ Kf7 30 Qd7+ Kf8 31 Qd6+ Kf7 32 Rc5 1-0

Here White can choose between two options. If he wants to play aggressively, then I can recommend the line we examine in the current game, where he sacrifices the g2-pawn. However, Black’s chances are fully acceptable in the ensuing complications.

So, to my mind, Black’s optimal reaction in this variation is 7...Nd5!, which is the subject of our next game.

41

12 d5?! This looks menacing, especially when played by such a strong player as Anand, but Tiviakov shows that Black can keep matters well under control. The reserved 12 Qe2 is better. Then:

While this pawn sacrifice is an interesting idea, and forces both players to display accuracy, it is not surprising that it is not very popular.

a) The centralizing 12...Qe6? is bad in this case, and in such a sharp position one wrong move is enough to derail the game! 13 d5! cxd5 14 0-0-0 f6 15 Bxd5 Qxe5 16 Qc4! (White acts in ideal fashion; I don’t see a way for Black to hold) 16...e6 17 Rhe1 Qxd5?? (the most resilient path is 17...exd5 18 Rxe5+ fxe5 19 Rxd5 Be7 20 Rxe5 Rf8 +/–) 18 Qxd5 Be7 19 Rxe6 Bxe6 20 Qxe6 Rd8 21 Rxd8+ Kxd8 22 Qe4 1-0 DegraeveSmirnov, Cappelle la Grande 2009.

The next game features the more modest 9 Nf3.

b) 12...h5 (D) places g4 under control.

White to play 9 Be2!?

9...Nxe5 It makes sense to eliminate the active knight first. In general, simplifications tend to favour the defender. 10 Bxe5 Qxg2! The cowardly 10...Bf5?! is also possible. You might consider using it if you fear your opponent’s tactical skills, but objectively it is difficult to equalize after this move; e.g., 11 0-0 e6 12 c4 Qd7 13 Qb3 f6 14 Bg3 0-0-0 (Dolzhikova-S.Kasparov, Fagernes 2012) 15 c5! +/–.

White to play Then: b1) The prophylactic 13 h3 avoids the exchange of bishops but also provides Black with an extra tempo for consolidation. 13...f6 14 Bh2 and here:

11 Bf3 Qg6 (D)

b11) 14...Qf7? is insufficient; in such a position, each move needs to achieve as much as possible, in both a defensive and an attacking sense. 15 d5! Bd7 16 0-0-0 g6 17 Kb1 (it is not easy for Black to finish his development) 17...Bh6?! (17...Bg7 is better) 18 Rhg1 +/– (threatening Rxg6 and Bxh5) 18...Kf8 19 dxc6 Bxc6 20 Bxc6 bxc6 21 Rd7 Re8 (K.Szabo-Kernazhitsky, Zalakaros 2008) and here the precise 22 Qe4! would have placed Black in a critical position. His king is under attack and soon there will be nothing left on the queenside.

White to play It is possible to get lost analysing a labyrinth of variations from this position, so let’s focus on some key moments.

b12) 14...Qg5 (D) impedes queenside castling (and any castling, actually!). 42

victory, such as 23 Rhe1 Be7 24 Qf5+ Kd8 25 Qe6 Re8 26 Bd6 +–} 23 Rc1 +– with a murderous intrusion on the 7th rank) 22 Qc2 Bd6 23 d5?! (23 Qb3! is much harder to answer) 23...Bxh2 24 dxc6++ Kxc6 (it is rare for the king to be safe this far forward without any cover) 25 Qa4+ Kb6 and now 26 Qb4+ Kc6 27 Qa4+ Kb6 28 Qb4+ Kc6 29 Qa4+ ½-½ was the finish of AbrahamyanKosteniuk, Khanty-Mansiisk (women) 2012. White could have fought on, but the game will remain in equilibrium if both sides play accurately: 26 Rd7 Rab8 27 Qb4+ Kc6 28 Rhd1 Qc5 29 Qe1 Rhe8 (29...Be5? is too risky: 30 f4 Bxb2+? 31 Kxb2 Rhe8 32 R1d3 +–) 30 Qh1+ Kb6 31 Qxh2 a6 with a dynamic balance, as the king can hide on a7.

White to play 15 d5! cxd5 16 Rg1 (16 Rd1!?) 16...Qf5 17 0-0-0 Be6 18 Qb5+ Kf7 (I would call this whole variation ‘geometrical’, as the knights have left the board and there are only line-moving pieces left; White can’t carry out a pawn-storm on the kingside for the simple fact that he lacks a g- or epawn, so the black queen can use the light squares h3, f5, g6 and e6, while the king has a safe shelter on f7) 19 Bxd5 Rc8 20 Bxe6+ Qxe6 21 Qxb7 Qxh3 22 Rge1 g5 23 Bd6? (now White even stands worse; 23 Rd7!?) 23...Qf5 (after 23...Kg6! 24 Bxe7 Qf5 –/+ it is unclear how White might rescue himself) 24 Qb3+ Kg6 25 Bxe7 (Z.AlmasiZ.Varga, Hungarian Team Ch 2006/7) and here 25...Qf4+ –/+ would pose real difficulties for White.

b22) 19...Ke7 looks exotic, but is best. After 20 Rhe1, as in Caruana-Milanović, European Ch, Budva 2009, Black can play 20...Rh6!?. After the queen is chased away, the king goes to f7. White has compensation for a pawn, but no more than that. 12...Bg4! (D)

b2) 13 0-0-0 Bg4 14 h3 Bxf3 15 Qxf3 Qe6 16 Kb1 f6 17 Bh2 Qd5 18 Qd3 e6 19 Qg6+ (D) and then:

White to play 13 dxc6 bxc6 14 Qe2 14 Rg1? Bxf3 –+. 14...Bxf3 15 Qxf3 Rd8 15...Qxc2 16 0-0 Qg6+ 17 Kh1 Rc8 is unclear. Black to play

16 Rd1 Qe6

b21) 19...Kd7?!

16...Rxd1+!? is sufficient to reach a roughly level ending: 17 Kxd1 Qe6 18 Re1 Qd5+ 19 Qxd5 cxd5 20 Bd4 a6 21 Re3 e6 22 Rb3 Bd6 23 Rb6 Kd7 with counterplay.

is inaccurate. 20 Ka1?! (losing time, but White doesn’t wish to allow ...Qf5+ and a queen exchange; better is 20 Rhe1) 20...h4?! (vacating h5 for the major pieces, but 20...Re8 is more solid) 21 c4 Qh5 (21...Qxc4?? is inappropriate greed: 22 d5 cxd5 {after 22...exd5 many roads lead to

17 0-0 Rxd1 18 Rxd1 h5 (D) 43

33 Bd6+ Kxd6 34 Qxf8+ Kd7 35 Qxh6 Qb1+ 36 Kg2 Qxb2 =. 33...Qb1+ 34 Kh2 Qxb2 35 Qxa6 Kf6 Not 35...Qxc3? 36 Qd6+ Ke8 37 Qb8+ Ke7 38 Bd6+ +–. 36 Qxc4 Qxa3 37 Qd4+ Ke7 38 c4 Ke8! 39 Qb6 39 h5 Qc5 40 Qxc5 Bxc5 41 Kg2 Kd7 =. White to play

39...Qc5 40 Qxc5 Bxc5

This typical idea again! Besides ...Qg4+ (remember that Black has an extra pawn so an exchange of queens plays into his hands), Black will have the possibility of bringing the rook to h6.

The remainder is elementary. 41 h5 Bd4 42 Kg2 Kd7 43 Bf4 Bg7 44 Be3 Kd6 45 Kf1 e5 46 Ke2 f5 47 f3 Kc6 48 Kd3 Bf8 49 Bc1 Bg7 50 Ba3 Bf6 ½-½

19 Qd3 Qc8 20 h4! Rh6 21 Kh2 Rg6 22 Bg3 Rg4 Conclusions

22...e6 is also acceptable: 23 Qf3 Be7 24 Rd4 Kf8 25 Ra4 with compensation.

26 Re5 e6 27 Rxh5

In our main game, the great Vishy appeared to be worse prepared than the top Scandinavian expert. Quite soon Black got a slightly preferable position. Then gradually, Sergei lost his grip on the game, shedding a pawn, but not the game. The encounter finished in an honourable draw.

Black has passed up some good safe options, and now White can win the pawn back while keeping the queens on the board. It is not a disaster, but now Black is the one who is fighting for a draw.

The most critical lines can be found in the notes. First, I would suggest that Black accept the sacrifice; otherwise there is no path to equality (Dolzhikova-S.Kasparov).

27...Rh6 28 Rxh6 gxh6 29 Qf3 c4 30 Bf4 Qc5 31 Qa8+ Ke7 (D)

It is hard to explain the complications seen in the note to White’s 12th move with mere words, as they are so concrete. Therefore please examine the variations carefully and note the many subtle points highlighted in the notes. Opening preparation in this line must go rather deep, and pay particular attention to AbrahamyanKosteniuk.

23 a3 a6 24 Re1 c5 25 c3 Rg6 25...e6!?.

White can play it safe with the modest 9 Nf3, but this move is also less ambitious. It is the subject of our next game. Game 11

White to play

Ivanchuk – Tiviakov

32 Kg1

Hoogeveen 2009

32 Kg3 +=.

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 Bf4 Nd5 8 Nxd5 Qxd5

32...Qf5! 33 Bg3 44

White doesn’t have to make the pawn sacrifice that we saw in the previous game, but may confine himself to the modest...

a) 12 c4 Qd8 13 h3 Bxf3 14 Bxf3 Bd6 15 Bg5 0-0 16 Qb3 Rb8 (unlike the game we see in line ‘b’, Black hasn’t advanced the a-pawn) 17 Rad1 h6 18 Be3 Qa5 19 g3 Rfc8 20 Rfe1 b5 (Black attacks the light squares and secures a strongpoint on d5 for his knight; however, this isn’t sufficient to equalize) 21 c5 Bc7? (this subtle move is dubious for tactical reasons; however, the more standard 21...Be7 also allows White an obvious advantage since after 22 Bf4 the ‘bad’ bishop is well-placed, and after 22...Ra8 {not 22...Rb7? 23 Rxe6! fxe6 24 Qxe6+ +–} 23 Ra1 += White intends a4) 22 Bxh6 (22 Bxc6 is even stronger and more solid; then after 22...Bxg3 23 Bxh6! Black can’t destroy both bishops, and very likely one of them will be saved; e.g., 23...Qc7 24 Qxg3 Qxg3+ 25 fxg3 Rxc6 26 Bf4 +/– with an extra pawn) 22...gxh6 23 Rxe6! (an effective strike) 23...Bd8 24 Rxc6 Rxc6 25 Bxc6 Qc7 26 Bg2 +/– Külaots-Pytel, French Team Ch 2008/9.

9 Nf3 (D)

Black to play With this move he hopes for some advantage thanks to his preponderance in the centre (i.e. the d4-pawn). Let’s note that this knight first moved from f3 to e5 and has now reversed this manoeuvre. Even if White can claim that Black has had to play some less-than-ideal moves to bring about this situation, it is hardly an example of rapid classical development! Anyway, from a more concrete viewpoint, the g2-pawn is covered and so the bishop can leave f1.

b) 12 h3

9...Nb6

22 Rd3 h6 23 Bd2 Ra6 24 Re1 Bc7 25 Bb4 Qd7 26 Red1 (D).

Bxf3 (with this exchange, Black stabilizes the situation but can’t reach full equality) 13 Bxf3 Qd7 14 c4 Bd6 15 Bg5 Be7 16 Qb3 0-0 17 Rfd1 Rfd8 18 Bf4 a5 19 a3 a4 20 Qc2 Bd6 21 Bg5 (note that White constantly avoids the exchange of bishops) 21...Qe7

This destination makes sense as the knight takes control of c4 and opens the way for the lightsquared bishop. It is also possible to move the knight to the kingside, though this turns out less well: 9...Nf6?! 10 Be2 Bg4 11 0-0 e6 (D) and now:

Black to play We are following the game Ni Hua-Tiviakov, Reggio Emilia 2008. White’s main way to achieve active play is a well-timed d5 pawn-break; indeed, I don’t see any other way forward for White, since by playing ...a5-a4, Black has nipped in the bud the plan of advancing the b-pawn. 26...Bd6 27 Bc3 Qc7 28 d5 cxd5 29 cxd5 e5 (Tiviakov prefers

White to play 45

realize it because Black is a few tempi behind in development) 17 Rac1 Bg7 18 Rxc6 0-0 19 Rxb6 Rxa2 20 Nd4 Be4 (the knight is insecure on d4, and has no strongpoints to use) 21 Bc4? Ra4 22 b3 Rxc4 23 bxc4 Rd8 24 c5 Bxd4 –+ (two bishops are considerably stronger than a rook, and there are plenty of pawns left) 25 Kf1 e5 26 f4 Ra8 27 fxe5 Bxc5 28 Rb5 Be7 29 Rb2 Ra5 30 e6 Rf5+ 31 Ke1 fxe6 32 Rd7 Rf7 33 Rb8+ Kg7 0-1 ParligrasTiviakov, Bundesliga 2008/9.

to block the centre whenever possible, though the clumsy position of the rook on a6 may be a problem – this piece can’t come quickly to the rescue if there is some emergency in another part of the board; instead a policy of simplification would deprive White of the bishop-pair and feels like it ought to equalize, but matters are not so simple: after 29...Nxd5 30 Bxd5 exd5 31 Rxd5 Bh2+ 32 Kh1 Rxd5 33 Rxd5 += Black has a difficult time protecting the a4-pawn, while there is also some danger of a piece attack on his king – the c3-bishop already casts a predatory glance at the kingside) 30 Re3 Raa8 31 Qf5 Re8 32 Rde1 Nd7 33 Bh5 Nf8?! (33...Nf6 makes White’s task harder; e.g., 34 Bxe5 Bxe5 35 Rxe5 Rxe5 36 Qxe5 Qb6) 34 Rg3 Re7 35 Re4 Ra6?? (an oversight in a complicated position, and maybe in time-trouble) 36 Rxg7+! Kxg7 37 Rg4+ Ng6 38 Bxg6 Kf8 39 Bh7 Rd7 40 Rg8+ Ke7 41 Bg6 1-0. A heavy defeat. But the result is quite logical, as White had been exerting pressure the whole game.

b) 11 Bg3 (D) also solves the tactical problem, by removing the bishop from potential attack.

10 Be2 Bf5 (D)

Black to play Then: b1) 11...Qe4 12 Rc1 g6 (I don’t like the combination of moves ...Bf5 and ...g6; I would say ...Bf5 and ...e6 looks more reasonable) 13 0-0 Bh6 (the idea is to kick the rook away from c1 and capture on c2; however, it is easy for White to deal with this plan) 14 Bd3 Qd5 15 c4 Qd7 16 Bxf5 Qxf5 17 Rc3 0-0 18 Re1 Rae8 19 b3 Qd7 20 Rd3 += Zhao Jun-Wang Li, Hefei rapid 2010. The exchange of a couple of minor pieces can be considered an achievement by Black, but it is clear that his position is somewhat worse.

White to play Let’s note that Black threatens to win a pawn. Once you have seen the idea it is rather obvious, but for some reason it can prove hard to spot. Nowadays few people are caught in this trap, but still it happens from time to time, even with chessplayers of a high level.

b2) 11...e6 was the simpler choice of Milanović, a Scandinavian Defence expert. 12 0-0 Be7 13 a4 a5 14 Ne5 Qd8 15 c3 Nd5 16 Bf3 Ra6!? (this move really surprised me – I don’t recall seeing this manoeuvre before with this kind of pawn-structure; 16...0-0 17 Qb3 Ra7 is about equal) 17 Re1 0-0 18 Nc4 h6 19 Qb3 Bd3 20 Ne5 Rb6 21 Qd1 Bh7 22 Nc4 Ra6 23 Qb3 Bd3 24 Ne5 Rb6 25 Qd1 Bh7 26 Nc4 Ra6 and in Savanović-Milanović, Banja Vrucica 2012 a repetition of moves followed. White could have kept fighting by 27 h3!?, when I feel he stands a little better.

11 c3 White removes the target from c2. Otherwise: a) 11 0-0? blunders a pawn to a simple tactic, but it has been played several times, even by titled players. 11...Qe4 (attacking the c2-pawn and the f4-bishop) 12 Bc7 Qxc2 13 Bxb6 Qxd1 14 Rfxd1 axb6 15 d5 g6 16 dxc6 bxc6 (an extra pawn is a big material advantage, with the bishop-pair to boot, but good technical skills are required to

We now return to 11 c3 (D): 46

Black to play

White to play

11...e6

16 Qb3

11...Qe4 was tried in a later game between the same players as in our main game: 12 Qd2 (I equalized in an analogous manner after 12 Bg3 Qc2 13 Qxc2 Bxc2 14 Nd2 e6 15 Nb3 0-0-0 16 Na5 Bd6 = Sebbar-S.Kasparov, Chefchaouen 2014) 12...Qc2 (it is not so usual for the queen to penetrate into the enemy camp at such an early stage of the game; however, it is not with attacking intentions, but to force an exchange) 13 Qxc2 Bxc2 14 Ne5 h5 (a familiar motif; as normal, this advance insures Black against the aggressive g4 advance) 15 0-0 e6 16 Rfe1 Be7 17 h3 Rd8 with approximate equality, Ivanchuk-Tiviakov, Wijk aan Zee 2010. It was difficult even for Ivanchuk to put serious obstacles in Black’s way after the exchange of queens.

16 Ne5 Qe7 +=.

12 0-0 Be7

24 c4 Bd6

It is noteworthy that later Ivanchuk himself switched sides: 12...Qe4 13 Bg3 Qc2 14 Nh4 Qxd1 15 Raxd1 Be4 16 Rfe1 0-0-0 17 Nf3 Bxf3 18 Bxf3 Bd6 19 Kf1 Bxg3 20 hxg3 = Quesada Perez-Ivanchuk, Havana 2012.

The exchange of bishops doesn’t equalize: 24...Bh5 25 Bxh5 Nxh5 26 Qf3 Nf6 +=.

16...Bc7 17 Rfe1 0-0 18 Rad1 After 18 Ne5 Qe7 I don’t see anything terrible. 18...Nd5 Black has a wide choice of sound continuations: 18...Bxg3 19 hxg3 Nd5 +=, and 18...Rfe8 and 18...Rab8 can also be considered. 18...Rad8!? 19 Ne5 Qc8 += is reasonable as the queen’s rook has found employment in the centre. 19 Ne5 Qc8 20 Rd2 Rd8 21 Bh5 Bg6 22 Bg4 Re8 23 Bf3 Nf6 23...Bf5 and 23...Rd8 are also possible.

25 c5 Bxe5 26 Bxe5 Nd5 (D)

13 a4 a5 14 Bc7 Bd8 Not 14...Nc8? 15 Ne5 +–, when Black’s queen is in trouble. 14...Nd7 15 Nd2 +=. 15 Bg3 Qd7 (D)

White to play Can White count on anything real when Black has such a strong centralized knight? White can only 47

eliminate the knight at the cost of entering an ending with opposite-coloured bishops. 27 h4 h6 28 h5 Bf5 29 Rde2 29 Bxd5 exd5 30 Rde2 offers White little because Black can establish a defensive line on the light squares: 30...f6 31 Bd6 Rxe2 32 Rxe2 Be6 +=, planning ...Qd7 and ...Re8 (rather than 32...Bg4?! 33 Re7 Bxh5 34 f3 +/–). 29...Rd8 30 Bd6 Nf6 31 Qe3 Rd7 32 Qf4 Qd8 33 Qg3?!

Black to play

The aggressive 33 g4 is parried by the transfer of the bishop to the dominant height d5: 33...Bd3 34 Re3 Bc4 =.

38...Kf8?! Time-trouble. 38...Nf6 and 38...Kh8 are both equal.

33...Ne8 34 Be5 (D) 39 Bb3 Qg5 40 Qg3 Nf6 41 f3?! Missing the chance to exploit Black’s time-trouble inaccuracies by 41 Bd1 += and then f4. 41...Nd5? Tiviakov in his turn could have altered the unpleasant scenario with a positional exchange sacrifice: 41...Rxd4! 42 Bxd4 Rxd4 43 Qb8+ (43 Bxe6 fxe6 44 Qb8+ Kf7 45 Rxe6 Rxg4+ 46 fxg4 Qxg4+ =) 43...Ne8 with unclear play. The weaknesses on c5, b2 and a4 compensate for the minimal material advantage White enjoys.

Black to play 34...Qg5?!

42 Bxd5 Rxd5 43 f4 Qe7 44 g5 +/–

Missing a chance to take the initiative: 34...f6! 35 Bf4 Rxd4 36 Bxh6 (36 Rxe6!? Bxe6 37 Rxe6 offers White compensation, but Black is still slightly for preference) 36...Qc7 =+.

44 f5 is also good: 44...Kg8 45 fxe6 fxe6 +/–. 44...hxg5

35 Qh2

44...Bf5 45 gxh6 f6 46 Bd6 R5xd6 47 cxd6 Rxd6 +/–.

35 Qxg5 hxg5 36 g4 Bd3 37 Re3 Bc2 =.

45 fxg5 Bf5 (D)

35...Rad8 36 Re3 Qe7 Remarkably, the passive e8-knight considerably limits the capacities of the active white bishop on e5. 36...Bc2 37 b3 Nf6 38 Bf4 Qf5 39 Bc7 Ra8 is unclear. 37 g4 Bh7 37...Bc2 38 b3 Nf6 =. 38 Bd1 (D) White to play 48

can observe Black suffering in a struggle against the bishop-pair. The moral: if there is no special reason, you shouldn’t give up a bishop for a knight. These pieces are not of equal value.

46 h6 The assessment is clear, so I just add dry comments to the rest as it has little to do with the topic of this book. It is even stronger to put the rook on the f-file: 46 Rf3 +/– or 46 Rf1 +/–.

In the position after 9...Nb6 10 Be2 Bf5, assuming White does not blunder a pawn and prefers a ‘normal’ continuation, Black often penetrates into the enemy camp with the queen (...Qe4-c2), seeking an exchange of queens. Once this plan is carried out, it is difficult for White to blast his way through the enemy defences.

46...Rxd4 46...gxh6 47 Bf6 +–; 46...Kg8 47 Rf3 Rxe5 48 Rxe5 +/–. 47 Bxd4?

In our main game, Ivanchuk dropped his guard at one moment while searching for an advantage, and Tiviakov could have taken over the initiative.

47 Bxg7+ Kg8 48 Bf6 Qxc5 49 g6!! Bxg6 50 h7+ Kxh7 51 Bxd8 +–. 47...Rxd4 48 Re5?! Bg6

Now we come to the idea of White supporting the e5-knight by playing 7 f4. This natural and fundamental plan is the subject of our next game.

48...Qc7 =. 49 Qh3 49 Rxe6 fxe6 50 Qf2+ Bf5 51 Qxd4 Qxg5+ 52 Kf2 Qxh6 =.

Game 12

49...gxh6

Wan Yunguo – Tiviakov

49...Bh7 =.

Moscow 2011

50 Qxh6+ Kg8 51 Rxe6 fxe6 52 Qxg6+ Qg7 53 Qe8+ Qf8 54 Qxe6+ Qf7 55 Qc8+ Kg7 56 Qh3 Qg6

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5 Nbd7 7 f4 (D)

56...Kg8 =. 57 Kh1 57 Re2 Rd3 58 Qh4 Rd5 =. 57...Rd5 57...Qf7 =. 58 Re7+ Black to play

58 Qh4 Rxg5 59 Qd4+ Qf6 60 Re7+ Kg6 =; 58 Qc3+ Kg8 59 Qh3 Kg7.

This move looks very natural; why not support the advanced guard in the centre with a pawn? Now if Black wishes to carry out his plan of exchanging on e5, he will first need to avoid a pawn fork, and then have to find a way to counter White’s resulting spatial plus.

58...Kf8 59 Re1 Kg7 ½-½

Conclusions

By this move White:

On move 9 of our main game, it is best for Black to bring the knight to b6, clearing the way for the c8-bishop. If it goes to f6, the knight turns out not to have much to do on the kingside (see KülaotsPytel and Ni Hua-Tiviakov). In both examples you

1) consolidates the advanced post on e5; 2) grabs space; 49

Lvov 2013. White has a big advantage due to his much better pawn-structure.

3) plans a kingside advance with g4. However, White’s plan is not often applied, as it has some flaws that Black can exploit.

We now return to 7...Nb6 (D):

7...Nb6 This move is best; other moves lead to examples of how you should not play as Black: a) 7...g6?! 8 Bc4 (the bishop has time to leap to c4 and attack the f7-pawn) 8...e6 (the defect is obvious: the c8bishop is locked in, which results in problems with its development, and as is well-known, whenever one piece is bad, the whole position is bad too!) 9 0-0 Bg7 10 a4 0-0 11 b3 Qc7 12 Ba3 += Svetushkin-S.Kasparov, Vrachati 2011. During my preparation for this game late the night before, I had correctly guessed that this variation would be played. However, in the morning I was tired and mixed up the move-order at the board.

White to play 8 g4 White joyfully continues the offensive. While a direct mating attack is rather improbable, Black must now be highly alert, and take into account the g5 and f5 advances at every turn. However, despite the apparent danger, Black has several acceptable replies to this move, a factor that makes White’s opening preparation harder if he wishes to play this line.

b) 7...e6 (D) can hardly be recommended as the c8-bishop is hemmed in, and can only hope to find employment on the queenside.

It makes little sense for White to put his bishop on e3 as it would help Black arrange standard exchanges via the d5-square: 8 Be3 g6 9 Qf3 Be6 10 0-0-0 Nbd5 = 11 Bd2 Nxc3 12 Bxc3 Bd5 (more accurate is 12...Bh6! 13 Kb1 0-0; then after 14 g4? Qd5 15 Qxd5 Bxd5 –/+ the rook and the f4-pawn are exposed, while 14 Rg1 Qd5 attacks both a2 and {indirectly} f4; then 15 Qxd5 Nxd5 leaves Black with the initiative) 13 Qe3 Bh6 14 Kb1 Nh5 15 Qh3 (15 Bc4 is unclear) 15...Bxf4 16 Be2 Ng7 17 Rhf1 Bg5 18 a3 0-0 19 Bb4 (BrauerS.Kasparov, Apolda 2013) 19...Qc7, dropping the queen back to its standard safe-house, keeps an extra pawn in a safe position.

White to play 8 Nc4 (this looks strange, leaving the central advanced post; 8 Be2!?) 8...Qc7 9 Bd3 g6 (weakening the dark-square complex even more; I wonder what would follow in case of 9...a6, preparing an offensive on the queenside) 10 0-0 Nb6 11 Ne5 Bg7 12 Qe1 0-0 13 Qh4 Nbd5 14 Nxd5 cxd5 (14...exd5 seems more attractive as it opens a line for the bishop but then 15 f5 is unpleasant; all the white pieces, except the a1rook, are aiming at the kingside) 15 c3 Bd7 (15...Ne4!?) 16 Rf3 Qd8 17 Rh3 h5 18 Be3 Ne4?! 19 Nxd7 Qxd7 20 Bxe4 dxe4 21 g4! Qb5 22 gxh5 Qxh5 23 Qxh5 gxh5 24 f5! +/– Sivuk-Boricsev,

8...Nbd5 Other continuations are also possible: a) 8...Be6 is a surprising move, boldly bringing out the bishop, unafraid of the advancing enemy pawns. 9 Bg2 g6 10 0-0 Rd8 11 a4 Bg7 12 g5 Nfd7 13 a5 Nd5 14 Ne4 Qc7 15 Qe2 0-0 16 c4 Nb4 is unclear, Barrientos-Tiviakov, Bogota 2010. 50

Qc6 18 Be3 e6 19 a6 b6 20 0-0-0 0-0-0 (the position is kind of exotic, but we can state that White hasn’t achieved any particular dividends) 21 Qxc6+ (21 Rxd5!? Qxd5 22 Bg2 Qd7 23 Qc4 offers White compensation, but probably nothing more than that: 23...Be7 24 Qe4 Qd5 25 Qa4 Qd7 26 Qe4 Qd5 =) 21...Bxc6 22 Rxd8+ Kxd8 23 h4 Be7 (more precise is 23...h6 24 gxh6 Ne8! 25 h5 gxh5 26 Nxh5 Bxh6, when Black even stands better as the a6-pawn may be surrounded) 24 Be2 Kc7 25 h5 (Istratescu-Ardelean, Predeal 2007) and now 25...b5 gives Black the initiative.

b) 8...g6 (D) and then:

White to play

b2) 9 Bg2 (D) is more flexible than the straightforward advance of the g-pawn.

b1) The impatient 9 g5 brings up important questions of how Black should organize his knights: b11) With 9...Nfd5 10 Ne4 Qd8 11 c4 Nc7 12 Qe2 Bg7 13 c5 it seems that Black is under enormous pressure, with few viable moves. But in Shirov-Nisipeanu, Foros 2007, Black suddenly played 13...Ne6!?, with considerable complications. After 14 cxb6 Nxd4 15 Qf2 Bxe5 16 fxe5 Qd5! 17 Bg2 Qc4! 18 b3 Qd3 19 Qd2 Nc2+ 20 Kf2 Qxd2+ 21 Bxd2 Nxa1 22 Rxa1 axb6 the game was unclear.

Black to play 9...Bg7 10 0-0 0-0 11 g5 Nfd7 (the more active 11...Nfd5!? is also quite acceptable, because White won’t be able to carry out an effective c4 advance: 12 Ne4 Qc7 13 c4? Bxe5 14 fxe5 {after 14 cxd5 Bg7 –/+ White has too many weak spots} 14...Nxc4 =+) 12 Ne4 Qc7 and now:

b12) 9...Nfd7!? is somewhat more precise, intending to exchange the knight on e5. b13) 9...Nh5!? deserves attention, as the knight can manoeuvre to f5 via g7. After 10 Ne4 Qc7 11 Ng3 (D).

b21) 13 c3 was played in Sivuk-S.Kasparov, Donetsk 2011. Now 13...Nxe5 is quite acceptable; e.g., 14 fxe5 Bf5 =. b22) 13 Be3 looks less precise. I am not sure about the merits of 13...f6, as it weakens the lightsquare complex, though after 14 Nd3 Nd5 15 Qe2 Nxe3 16 Qxe3 Nb6 17 c3, as in Vitiugov-Lysy, Serpukhov 2008, 17...Nd5 would have been equal. Instead, 13...Nd5! is more accurate. It is then unclear where the bishop should move. If he plays by analogy with 14 Qe2 then after 14...Nxe3 15 Qxe3 Qb6 Black even stands better: he has two bishops and no weaknesses. A pair of pieces is exchanged and if necessary an exchange can be made on e5.

Black to play 11...Ng7 (the computer suggests 11...Nxg3 but I don’t like the fact that after 12 hxg3 the h-file is opened, creating pressure on h7; also the shift of the pawn to the g-file is quite significant, as it may perform a useful role on g4) 12 Qe2 Be6 13 a4 Nd7 14 a5 Nxe5 15 dxe5 Bd5 16 Rg1 c5 17 Qb5+

We now return to 8...Nbd5 (D): 51

White to play

Black to play However, even this wouldn’t have been so bad if Sergei had found 14...Bg7!. Few people would dare give up the c6-pawn, but 15 dxc6 Qxd1+ 16 Kxd1 b5 17 cxb6 axb6 18 Be3 Be6 offers Black counterplay because after 19 Bxb6? Nd5 only White has problems.

9 Bg2 9 g5 is ineffective due to White’s lack of development: 9...Nxc3 10 bxc3 Qd5! 11 Rg1 Nd7 12 Bc4 Qe4+ 13 Qe2 Qxe2+ 14 Kxe2 Nxe5 15 fxe5 and Black has no problems in the ending after 15...g6 or 15...Bf5.

Instead the game continued 14...cxd5? 15 c4 e6 16 Bb2 Bg7 17 Nc6 bxc6 18 Bxg7 Rg8 19 Be5 Bd7 20 0-0 +/– (material is equal, but the situation is extremely difficult; the dark squares are awfully weak, and the e5-bishop dominates the board) 20...Rb8 21 Qa4 Rb7 22 Rab1 Qc8 23 Rxb7 Qxb7 24 Rf2 d4 25 Qa5 Kd8 26 Be4 1-0.

9 Bc4 Be6 10 g5? is even worse: 10...Nxc3 11 bxc3 Bxc4 12 gxf6 gxf6 13 Nxc4 Qd5 14 Kf2 Qxc4 –/+. 9...g6 10 g5 10 Nxd5 Nxd5 11 c4 Nc7 12 c5 Qd8 13 Qb3 Nd5 is unclear.

We now return to 11...Nd7! (D):

10...Nxc3 11 bxc3 An important moment; it is easy for Black to go astray here. 11...Nd7! Continuing the main theme: exchanges are favourable for Black, though 11...Nh5!? also deserves attention. In Shirov-Tiviakov, Hoogeveen 2010 our hero came under unpleasant pressure after 11...Nd5?!, since his knight was chased by the enemy pawns: 12 c4 Nc7 13 c5 Qd8 14 d5! (D).

White to play 12 0-0 White played in a creative but unsuccessful manner in Didenko-S.Kasparov, Pardubice 2012: 12 Nc4?! Qc7 13 0-0 Nb6 14 Ne5 Bg7 15 c4? (this is too much; in such positions Black has the right to give up the dark-squared bishop, winning a pawn) 15...Bxe5! 16 fxe5 Nxc4 (the fact that the enemy pawns are situated on squares of the bishop’s colour – g5, e5 and d4 – allows Black to build an effective ‘Maginot Line’ on the lightsquare complex: f5, e6, d5 and c4) 17 Rxf7?! 52

(seeking complications) 17...Kxf7 18 Qf1+ Bf5 19 e6+ Kg7 (19...Kxe6!? –/+) 20 Qxc4 Rhd8 =+. 12...h6! (D) I also don’t see anything wrong with the straightforward 12...Nxe5!? 13 fxe5 Qc7 14 Qf3 (14 Bf4 Be6 =) 14...Bf5 15 Rb1 Rc8 = (15...Rd8!?).

Black to play e1) 13...Nxe5?! 14 fxe5 Qe6 15 a4 hxg5 16 Bxg5 Bh6 17 Bxh6 Rxh6 18 a5 (White wants to play a6, undermining the c6-pawn) 18...a6 19 Rab1 Rh5 20 Qf4 Kf8 21 Rf3 Kg8 22 Rbf1, SvidlerZolotukhin, Olginka 2011. A sore spot in Black’s fortifications is obvious: almost all his pawns are fixed on light squares – the same as his bishop. Also, the a8-rook still doesn’t participate in the struggle. Nevertheless, after 22...Rf5 23 Qh4 Rh5 24 Qe1 Rh7 although White stands a little better, I can’t find anything incisive for him.

White to play The far-advanced white g-pawn has certain negative sides. Before the f8-bishop has moved, Black attacks the opponent’s vanguard; White’s response may give the f8-bishop new possibilities. Also without having to make a move, the h8-rook may be granted an open (or half-open) file along which to threaten the white king.

e2) 13...hxg5 leads to messy complications: 14 Nxf7! Kxf7 15 fxg5+ Ke8 16 Qf7+ (16 Bf4 e5 17 Rae1 is less clear) 16...Kd8 17 Bf4 with compensation. e3) 13...Bg7 = maintains the status quo.

13 Qe2?!

We now return to 13 Qe2?! (D):

There are several alternatives but nothing that promises much of an advantage: a) 13 Ng4 Nb6 =. b) The opening of the h-file after 13 h4 hxg5 14 hxg5 Nxe5 15 fxe5 Qc7 =+ is unappealing for White. c) 13 Re1 hxg5 14 fxg5 Nxe5 15 Bf4 Bg7 16 Bxe5 Bxe5 17 Rxe5 Bf5 =+. d) I faced 13 gxh6 Bxh6 14 Qf3 in VedderS.Kasparov, Hoogeveen 2012. Now 14...Qe6 = looks reasonable, intending to eliminate the e5knight as 15 Bh3?? doesn’t work due to 15...Nxe5 –+.

Black to play 13...hxg5 14 fxg5 The 14 Nxf7? sacrifice is inappropriate: 14...Kxf7 15 fxg5+ Ke8 16 Bf4 e5 17 Rae1 (17 dxe5 Qe6 – /+) 17...Be7 18 dxe5 Qc5+ 19 Kh1 Nf8 –/+.

e) 13 Qf3!? (D) looks right:

14...Nxe5 15 Bf4 15 Qxe5?! Qxe5 16 dxe5 Bf5 –/+; 15 dxe5?! Qc5+ –/+ (15...Qe6 =+). 53

15...Bg7 16 Rae1 Bf5 (D) 16...Be6 is also not bad.

Black to play 21...0-0?! White to play

21...bxc6 would have kept an advantage: 22 Bxc6+ Kf8 23 Bxa8 Qxa2 =+.

Now that Black is fully mobilized, White stands somewhat worse because the inevitable exchanges on e5 will diminish his attacking potential. There are few threats to the black king, while the chronic defects in White’s pawn-structure (notably c2 and c3) remain.

22 Rc5 Or 22 Rc7 Rac8 23 Rxb7 Qxa2 =+. 22...Rac8 23 c4 23 c3 Qxf2+ 24 Rxf2 Rxc5 25 dxc5 Rc8 =+.

17 Qf2

23...Qxf2+!

17 Bxe5 Bxe5 18 Qxe5 Qxe5 19 Rxe5 e6 =+.

23...Qc3 24 Rxc8 Rxc8 25 c5 gives White counterplay.

17...Qa3 As usual, Tiviakov handles his queen – a central theme in this opening – with great skill. However, it is possible that 17...Rh5!? is even stronger: 18 Qg3?! 0-0-0 19 Bxe5 Bxe5 20 Rxe5 Rdh8 –/+.

24 Rxf2 Rxc5 24...b6 25 Rxc8 Rxc8 26 Bf1 =. 25 dxc5 Rc8 26 Rb2 Rxc5 27 Rxb7 a6! (D)

18 Bxe5 Bxe5 19 Rxe5 Qxc3 19...Qxa2?! 20 d5 gives White compensation. 20 Rc5 White also experiences difficulties after 20 Rb1?! b6 (20...0-0?! 21 Rxb7 is unclear) 21 Re3 Qc4 22 Qe2 Qxe2 23 Bxc6+ Kf8 24 Rxe2 Rc8 25 d5 Rh5 –/+. 20...Qb2 21 Rxc6 (D) White to play Nevertheless, Black retains at least a symbolic advantage due to the scattered white pawns. 28 Rxe7 28 Bf1 Be6 29 h4 (29 Rxe7 Rxg5+ –/+) 29...Bxc4 =+. 28...Be6 29 Bd5 Bxd5 30 Re5 54

sacrificed material to force it (as in Bologan’s gambit idea 10 h3!? from Game 3).

30 cxd5 Rxd5 31 h4 Rd4 –/+. 30...Rxc4 31 Rxd5 Rc1+ 32 Kg2 Rc2+ 33 Kg3 Rxa2

After 7...Nb6, the consistent 8 g4 gives Black several viable options. Even the provocative bishop placement in front of the oncoming enemy infantrymen (8...Be6) makes sense because on occasion this piece may move to d5, exploiting the fact that White’s pawn advances have weakened the h1-a8 diagonal (see Barrientos-Tiviakov). 8...g6 is more popular – just don’t repeat the author’s mistakes in the game Sivuk-S.Kasparov. If White plays the impatient 9 g5 (or this same move a little later), then situations arise where the black knights undertake very specific manoeuvres. For instance, quite often the retreat ...Nfd7, intending to exchange on e5, is stronger than the more obvious centralization ...Nfd5 because sooner or later the d5-knight may be pushed aside by the c4 advance. The logic here is that by playing g5, White changes the situation dramatically; there is then greater reason for Black to exchange on e5 by ...Nfd7xe5, as his bishop may occupy the strongpoint on f5.

I shall leave the remainder with just minimal notes, because it would be more relevant to an endgame book. 34 Rd8+! Kg7 35 Ra8 Ra3+ 36 Kg4 a5 37 h4 Ra1 38 Kg3 a4 39 Kg2 (D) 39 Ra7!?.

Black to play

A ‘coiled spring’ effect can be observed in ShirovNisipeanu, where a pawn-structure arose that is rather unusual for the Scandinavian Defence.

39...f5! 40 gxf6+ 40 Ra7+ Kf8 41 Ra8+ Ke7 42 Ra7+ Ke6 43 Ra6+ Ke5 44 h5 (44 Rxg6!?) 44...gxh5 45 g6 Rd1 =+.

51 Rxa2 Rxa2 52 Kf4 Kc5 53 h6 Kd6 54 Kf5 Ke7 55 Kg6 Kf8 –+; 51 Kg4 Rc4+ 52 Kf5 Rc5+ 53 Kg6 Ra5 –+.

The inappropriateness of ...Nfd5 was emphasized by the game Shirov-Tiviakov, where Black came under pressure. As usual, after the game Sergei found refinements in Black’s play, which he put to good use in our main game. Please pay close attention to the pawn-break 12...h6!, which shows that a rapid pawn advance can work against White, especially if his king is exposed. Our motto is simple: let’s follow Tiviakov along the twisting paths of the variations. His games are of a very high quality, and his defensive abilities are extraordinary.

51...Rc3+ 52 Kg2 Ra3 0-1

2: 3...Qd6: Other Lines

Conclusions

In this chapter we shall discuss several possibilities, for both Black and White, to avoid the lines we saw in the previous chapter (i.e. 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 Ne5). They fall into these categories: Fianchetto with 6 g3 (Games 13 and 14) Kovalenko’s 5...Bg4 (Games 15 and 16) 5...a6 (Game 17)

40...Kxf6 41 Ra5 a3 42 Kh2 Ke6 43 Kg2 Ra2+ 44 Kg3 Ra1 45 Kg2 Kd6 46 Ra6+ Kc5 47 Rxg6 Rc1 48 Ra6 Kb4 49 h5?? For example, 49 Kg3 =. 49...Rc2+! 50 Kg3 a2 51 h6

After 7 f4, Black’s best reaction is 7...Nb6; other continuations are weaker. My game against Svetushkin demonstrates that it can be very harmful to get the move-order wrong. I had to defend arduously and despite the fact that I went on to win this game, I don’t intend to repeat my error! 7...e6 is dubious as it blocks the c8-bishop. In general a set-up with bishop on c8 and pawns on e6 and c6 is unattractive, unless White has

Rare Lines for White on Moves 4, 5 and 6 (Games 18 and 19) 55

Fianchetto with 5...g6 (Games 20 -23) Fianchetto with 6 g3 Quite a popular line for White is, after 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 to fianchetto with 6 g3 (D).

White to play 8 0-0 8 h3 is similar to the main game. 8...Bh5 (8...Bxf3 should be compared with the game continuation) 9 0-0 Be7 Black to play

10 Ne2 (the same manoeuvre as in the main game; White intends to force the exchange of Black’s bishop without creating weaknesses by playing g4) 10...0-0 11 Nf4 Bxf3 12 Bxf3 Nbd7 13 Bg2 Rad8 14 c3 Rfe8 15 Nd3 Qc7 16 Qe2 Bd6 17 Be3 h6 18 Rad1 a6 19 Qc2 Nd5 20 Bd2 Qb8 21 Rfe1 ½-½ Chigaev-S.Kasparov, St Petersburg 2011. Both players have arranged their pieces harmoniously. Black is solid, but White’s game is appreciably more pleasant as he has the bishop-pair and enjoys a long-term initiative. White’s lower Elo rating may have played a role in his decision not to play on.

This is rather a non-committal plan, hoping in many lines to achieve a comfortable ‘two bishops advantage’. Let’s also note that White is preparing Bf4 as well as Bg2. In our first game, we shall take a look at a highlevel encounter where White was able to increase a very slight advantage and eventually turn it into a full point. This was despite stubborn resistance from Tiviakov, which should alert us to the dangers that Black faces in this modest-looking line.

8...Be7 9 h3

Game 13

This is the most testing. White used to apply less effective methods:

Caruana – Tiviakov Wijk aan Zee 2010

a) After 9 Bf4

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 g3

Qd8 10 h3 Bh5 11 g4 Bg6 12 Ne5 Nbd7 13 Qe2 Nxe5 14 dxe5 Nd7 15 Rad1 Qc7 16 Ne4 (D) Black equalized quite confidently in BaklanTiviakov, Wroclaw 2009:

White waits for the c8-bishop to choose a square, and may then chase it. Naturally, Black won’t play ...e6 before developing this bishop. 6...Bg4 This move looks logical, as it puts pressure on the d4-pawn. We shall also analyse the other main line, i.e. 6...Bf5 – see Game 14. 7 Bg2 e6 (D)

Black to play 56

b2) 14...Bd6

16...Bxe4 (Black shouldn’t let the knight reach d6) 17 Qxe4 0-0 18 Rd2 Rfd8 19 Rfd1 Nb6 20 c4 (not allowing the knight to go to d5, but now the darksquare complex is weakened, which allows Black to build a defensive line) 20...Rxd2 21 Rxd2 Rd8 (Tiviakov relieves the tension on the only open file) 22 b3 Rxd2 23 Bxd2 Nd7 24 f4 Bc5+ 25 Kf1 a5 26 Qd3 Bb4 27 Be3 Bc5 28 Bd2 Bb4 (the f8-a3 diagonal proves vital, as is often the case; White’s g-pawn being on a light square doesn’t help him, as it weakens the white kingside) 29 Be3 Bc5 30 Bd2 Bb4 ½-½.

15 Bxd6 Qxd6 16 Nh4 Bg6 17 Nxg6 hxg6 (D).

b) 9 Re1 0-0 10 Ne4?! (why does White want this exchange?) 10...Nxe4 11 Rxe4 Bf5 12 Re1 Nd7 13 c3 Qc7 14 Bf4 (D) and then:

White to play Rather than a full bishop-pair, White has only one bishop against a knight, which doesn’t promise special benefits. 18 Qb3 Rab8 19 Rad1 Rfe8 20 Qa4 Qc7 21 c4 (21 Qxa7?? Ra8 –+) 21...Nb6 22 Qa5 Rbd8 23 b3 (Godena-Tiviakov, Saint Vincent 2006) and now a simplifying manoeuvre is enough for equality: 23...Nd5 24 Qxc7 (24 Qd2 Nf6 = and at some moment ...e5 may follow, bisecting the dark squares) 24...Nxc7 =. We now return to 9 h3 (D):

Black to play b1) 14...Qb6 15 Qb3 Rfe8 16 Nd2!? (this looks strange, but it makes sense: the knight is heading for c4, bothering the black queen, and may then proceed to d6, a5, e5, etc.) 16...Qxb3 (Houdini suggests the sharp 16...g5!?, but it isn’t a very ‘human’ move) 17 axb3 Nb6 18 Nc4 Nd5 19 Bd6 a6 20 Bxe7 Rxe7 21 b4 Rd8 22 Na5 (White’s position is certainly better: given the bishops, his pawns are well-placed on dark squares, while the b2-c3-d4b4 structure solidly fixes the queenside; however, real winning chances are hard to come by) 22...f6 23 h4 Kf7 24 Bf3 h6 25 h5 e5! (Black has gradually rearranged his kingside pawns onto dark squares) 26 Rad1 e4! 27 Be2 Nb6 (the prophylactic 27...Rde8 is more accurate) and now rather than 28 f3?! (Fedorov-Caspi, Eforie Nord 2010) 28...Rde8 with good play for Black, White could have landed a blow by 28 Bxa6!? bxa6 29 Nxc6 Rde8 30 Nxe7 Rxe7 31 Ra1, when White has the initiative, but the position is not so easy to assess beyond that.

Black to play 9...Bxf3 9...Bh5 transposes to the note to White’s 8th move. 10 Bxf3 0-0 11 Ne2 B.Socko-Tiviakov, Bundesliga 2008/9 provided an instructive example of incisive action by Black on the dark squares: 11 Bf4 Qd8 12 Qd3 Nbd7 13 Ne2 Re8 14 Rfd1 Bf8 15 Bg2 Qc8 16 c4 e5! (D).

57

White to play

White to play

A skilful counterblow; don’t be put off by generalizations like open lines favouring the bishops – concrete considerations are more important. 17 Be3 exd4 18 Nxd4 Nc5 19 Qc2 a5 (note that since Black’s pawn-break was made with the e-pawn, the b7-c6 breakwater remains in place, limiting the activity of White’s lightsquared bishop; the c4-pawn is fixed on a light square, while in the fight for the dark squares – in particular b4 and c5 – Black effectively has an extra piece) 20 Re1 Qc7 21 Nf5 Ne6 22 Rad1 Bc5 23 a3 Bxe3 24 Rxe3 Red8 25 Rde1 Nd4 26 Nxd4 Rxd4 27 Qc3 Rdd8 28 Re7 Rd7 29 R7e5 Qd8 30 R5e3 Rd1 31 b4 (White finally mobilizes his queenside pawns, but it is too late – he has just one bishop left) 31...axb4 32 axb4 Rxe1+ 33 Rxe1 Qc7 34 Qe3 h6 35 Qc5 Re8 36 Rxe8+ Nxe8 37 Bf3 ½-½.

As a result of the opponent’s sluggishness, Tiviakov starts his own expansion. The game continued 20 dxc5 Qxc5 21 Qe4 Rfe8 22 Nd4 b4?! (I don’t think a reduction in the number of queenside pawns can be to Black’s benefit, because the opening of lines does benefit White’s bishops here; 22...Nc4 =) 23 cxb4 Qxb4 24 a3 Qc5 25 Qd3 Qc4 26 Qxc4 Nxc4 27 Nf5 Be5 28 Bxd5 exd5 29 b3 Nb6 30 Be3 +=. White has transformed the advantage from one type to another. Now his pawns are better arranged as the d-pawn is quite vulnerable. 12 Bg2 e5 13 c3! Rad8 13...Rfe8!?. 14 Qb3 Qc7 14...Nb6 15 a4 exd4 16 Nxd4 +=.

We now return to 11 Ne2 (D): 15 dxe5 Nxe5 16 Be3 (D)

Black to play Black to play 11...Nbd7 The centre is open, which is nice for White’s bishops. Black must keep the b7-c6 breakwater alive at any cost, and try to trade or at least to force back White’s dark-squared bishop.

11...Qc7 12 c3 Nbd7 13 Bg2 a6 14 Qb3 b5 (this move makes sense given that Black has no lightsquared bishop, whereas White does) 15 Bd2 Nb6 16 Qc2 Rac8 17 Rad1 Nfd5 18 Bc1 Bd6 19 Kh1 c5 (D) gave Black counterplay in Van KampenTiviakov, Montreal 2014.

16...Nfd7 17 Rad1 Nc5 17...Rfe8!?; 17...Bc5 18 Bg5 +=. 58

18 Qc2 Nc4 19 Bf4 A disruptive move: before stepping back to c1, White tries to extract a gain by drawing an enemy piece to a worse square. Such nuances are often vital to accumulate advantages, and can add up to the difference between winning and drawing. 19...Bd6 20 Bg5 Be7 21 Bf4 Bd6 22 Bc1 Rfe8 23 b3 Nb6 24 Nd4 Ne6 24...Be5!?.

White to play

25 Nxe6 Rxe6 26 Rfe1 Rxe1+ 27 Rxe1 Nd5 28 a3 Qc8 29 b4 (D)

A critical weakening, but there were no real alternatives; Black must restrain the h3-bishop. However, the black king is now exposed, which will inevitably become more important in due course. 38 Qa7 Qc8 39 Qd4 Bf8 40 Bf6 Ne6 41 Qe5 Kf7 42 c5 Qc7 43 Qc3 Bg7?! Perhaps 43...Be7 is more accurate. 44 Bxg7 Nxg7 45 Bf1! +/– Here are the repercussions of the ...f5 advance. Alas, it is not feasible to move the pawn back, but how will Black cover the a2-g8 and a1-h8 diagonals now?

Black to play Note how skilfully Caruana is playing, placing his pawns on dark squares (in harmony with the unopposed g2-bishop). Now a raid with c4-c5 will be a constant menace.

45...f4 46 Bc4+ Ke8 46...Kf8 47 Qf6+ Ke8 48 g4 +– is no different.

29...Re8 30 Rxe8+ Qxe8 31 c4 Nc7 32 Qd3! Qe1+

47 g4 Qe7 (D)

32...Be5 33 Be3 a6 +=.

There is no point in 47...f3+ as the pawn is doomed after 48 Kh3 +/–.

33 Kh2 Qe7 33...Qxc1? doesn’t work: 34 Qxd6 Ne6 35 Qb8+ +– and Black’s queenside falls apart. 34 Bg5 There are many other good options, such as 34 Be3 a6 35 Be4 g6 36 f4 +=. 34...Qe6 35 h4 += See the note to White’s 29th move.

White to play

35...g6 36 Qd4 a6 37 Bh3 f5 (D)

48 Kg1 +/– 48 g5 +/– is strong but Caruana probably didn’t want to let the knight escape from its cage by ...Nf5. 59

48...h5?

Conclusions

48...g5 is a better try: 49 Qd4 (49 h5 Kd8 +=) 49...h6 and now 50 h5?! is not dangerous in view of 50...Qe1+ 51 Kg2 Ne6, when Black is back in the game.

I think White’s best chance for advantage lies in simple play against Black’s light-squared bishop (see Fedorov-Caspi and Godena-Tiviakov). There are micro-nuances in the arrangement of White’s kingside pawns. In particular, the straightforward g4 advance also has a drawback – the weakening of the dark squares (Baklan-Tiviakov).

49 Qd3 Qf6 49...g5 50 gxh5 +–.

Please pay attention to the typical counterblow ...e5 in the fight for the dark-square complex. It is useful to know it not only in this opening, but in general. An often overlooked ‘mini-rule’: in the struggle for squares of a certain colour (in the given case, the dark squares) the player with the ‘two bishops advantage’ lacks one piece! The g2bishop can’t participate in this battle.

50 g5 +– Once again I call your attention to the placement of the white pawns. Every one is on a dark square – the ideal combination with the lightsquared bishop. In contrast, Black can’t defend against the numerous threats. 50...Qf5 51 Qd6 f3 52 Qb8+ Ke7 53 Qd6+

However, Black doesn’t have to bishop. Having found the lines of game rather uncomfortable, players the bishop on f5 instead of g4. Thus subject of the next game.

All roads lead to Rome; e.g., 53 Qxb7+ Kf8 54 Qb8+ Ne8 55 Bxa6 +–. 53...Ke8 54 Bd3 Qg4+ 55 Kf1 Qg2+ 56 Ke1 Qg1+ 57 Kd2 Qxf2+ 58 Kd1 Qg1+ 59 Kc2 Qf2+ 60 Kb3 Qe3 61 Ka4 (D)

give up the the previous began to put 6...Bf5 is the

Game 14 Van den Heever – S. Kasparov Cape Town 2014 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 g3 Bf5 (D) The previous game featured 6...Bg4, but having experienced discomfort in that line, Black started to prefer the text-move, which makes it harder for White to hunt down the bishop. Moreover, even if there weren’t an objective reason to prefer one move over the other, it is useful to include some variety in our repertoire so as to broaden our opponents’ workload in their preparation for the game.

Black to play It is curious to observe that the wandering white king has found a safe harbour on an exotic square – a4. 61...Qe7 62 Bxg6+ Kf8 63 Qb8+ Ne8 64 Qf4+ Kg8 65 Bxh5 Nc7 66 Bxf3 Nb5 67 g6 a5 68 h5 axb4 69 h6 1-0 Caruana’s play was almost faultless, exerting constant pressure from a strategically superior position. Even Tiviakov was unable to resist in the end.

White to play 60

The bishop-pair is nice, but the king’s residence is weakened, and development is not yet finished.

7 Bg2 Now Black can:

b) 8...Be7 (D) and then:

a) spend a tempo on a prophylactic measure, i.e., creating an escape-square (by ...h6) for the bishop and the king; or b) continue conventional development without being afraid of flank sallies like Nh4. But before moving on, let’s look at an unconventional alternative for White that was used by a strong player: 7 Bf4 Qd8 8 Bd3 (this seems a bit strange; was the move g3 played only to develop the bishop on f4?) 8...Bxd3 9 Qxd3 e6 10 0-0-0 Be7 11 Bxb8 (I think it’s dubious to exchange the already developed bishop for a passive knight without any special reasons) 11...Rxb8 12 Ne5 0-0 13 f4 Qa5 (hinting at ...Nd5 or ...Bb4) 14 Kb1 Rbd8 15 Qf3 Qb6 16 Ne2 c5! = (a typical Caro-Kann blow) A.Kovačević-Milanović, Kragujevac 2013.

White to play b1) 9 a3 0-0 (9...Nbd7?? 10 Bf4 +–) 10 Bf4 Qd8 11 h3 Nd5 12 Bd2 Nd7 13 Re1 h6 14 Ne4 N5f6 (otherwise c4 will follow, and the e4-knight will go to c3) 15 Nxf6+ Bxf6 16 Bf4 Re8 (better is 16...Qb6) 17 c3 a5 18 Ne5 Nxe5 19 Bxe5 Bxe5 20 Rxe5 a4 21 Qe2 Qb6 22 g4 Bh7 23 Re1 Ra5 = Meng-S.Kasparov, Delft 2012. If need be, the black rooks will be activated via a5; later ...Rb5 may follow with pressure on the b2-pawn.

7...h6 Making sure the bishop has a safe retreat-square. Practice has shown that this loss of time does not have bad consequences. Another, more ‘economical’ option is 7...e6 8 0-0 (D), and now:

b2) 9 Ne5 Nbd7 10 Bf4 (D) and then:

Black to play Black to play

b21) The simple 10...Nxe5 is enough for equality: 11 dxe5 Qxd1 12 Raxd1 Nd5 13 Nxd5 and now 13...cxd5?! 14 c4! += leaves Black facing unpleasant pressure on the h1-a8 diagonal. 13...exd5! = is an important nuance; with the bishop on g2, it makes sense to lengthen the breakwater (i.e. b7-c6-d5). Black then enjoys good play.

a) 8...Nbd7 9 Nh4 Bg4 (Polish IM Jacek Stopa is experienced in this variation) 10 f3 (this commits White to a subsequent f4 advance as one can’t leave the g2-bishop blocked in) 10...Bh5 11 g4 Bg6 12 f4 0-0-0 (although White’s pawn advance looks impressive, it is difficult to assess the situation unequivocally because Black’s position has no weaknesses and can expand like a spring) 13 f5 exf5 14 Nxf5 (Hernandez Guerrero-Stopa, Chicago 2008) and now 14...Bxf5 is right: 15 Rxf5 Qe6 16 Qf3 g6 with chances for both sides.

b22) 10...Qb4!? (Tiviakov manoeuvres his queen in exquisite fashion; he is also trying to keep winning chances alive, as a 2600+ GM must in open tournaments) 11 Nxd7 Nxd7 12 Ne4 Nb6 13 61

14 Bd2 a5 (I can conjecture why White seldom goes for this line: after the exchange of queens, White’s chances for victory are very slim; he can appear to be pressing for a long time, but real chances are hard to come by) 15 Nf3 Nbd7 16 Rac1 h6 17 h3 Ne4 18 Be3 Nef6 19 Rd3 e5!? 20 dxe5 Nxe5 21 Rdd1 Nxf3+ 22 Bxf3 Bc7 23 Rxd8+ Rxd8 24 Rc3 Rd7 25 Kf1 Be5 26 Rb3, Schut-Ushenina, European Women’s Team Ch, Porto Carras 2011. Now Black can mark time by 26...h5 followed by ...g6. It is unclear how White’s position can be strengthened. Meanwhile, if White overpresses, Black’s counterplay on the dark squares may become unpleasant, considering that the light-squared bishop cannot take part in that particular battle.

Rb1 0-0 14 c3 Qa5 = Cabezas-Tiviakov, Vila Nova de Gaia 2010. With Black taking no special measures to preserve his light-squared bishop, an attempt to hunt it down by Nh4 looks more reasonable. This can be done either at once, or after playing Bf4: b3) 9 Bf4 (D).

We now return to 7...h6 (D):

Black to play 9...Qb4 (the deeper I delve into the position’s essence, the less I like this thrust; 9...Qd8 is more prudent; e.g., 10 Qe2 0-0 11 Rad1 Qb6 12 Bc1 Qa6!? = Akash-Danielsen, New Delhi 2012) 10 Nh4 Bg4 11 Qd2 (it is not desirable to close the diagonal for the g2-bishop by 11 f3?!) 11...0-0 12 h3 Bh5 13 g4 Bg6 14 Nxg6 hxg6 15 a3 Qb6 16 Na4!? (a sound decision, seizing the opportunity to get the pawn to c4) 16...Qd8 17 c4 Nbd7 (17...Bd6!?) 18 Nc3 Re8 19 Rad1 Bf8 20 Rfe1 Nb6 21 Qe2 a5 22 h4 a4 23 Bf3 +/– TerentievOvechkin, Izhevsk 2010. It surprises me to state that Ovechkin handled the opening unsuccessfully. When playing blitz against him I can rarely get much advantage with White.

White to play 8 0-0 e6 9 a3 Setting a common ‘trap’. For some reason White hopes that the adversary will put the bishop on e7, the knight on d7 and after Bf4 will resign as the queen will have no place to go. Why else would he play a3? Sometimes it might support the b4 advance, but this isn’t always a good idea, as c4 is weakened, and it may give Black ideas of ...a5, breaking up White’s queenside structure.

b4) The immediate 9 Nh4 (D) is very seldom used.

9 Ne5 is more aggressive. Then: a) 9...Nbd7 might be a little inaccurate, but this is not 100% clear. 10 Nc4 Qb4 (the queen appears active, but it loses control over d5; 10...Qc7 is unpleasantly met by 11 Bf4 Qd8 12 Nd6+ Bxd6 13 Bxd6 – this is not fatal, but there is no good reason to give White the bishop-pair here) 11 Qe2 Nb6 12 Ne3 Bh7 13 d5! (otherwise White has nothing) 13...cxd5 14 Nexd5 Nbxd5 15 Nxd5 Nxd5 16 Bxd5 (Antonio-Dableo, Philippine Ch, Tagaytay City 2010) and now Black can get

Black to play 9...Bg4 10 Ne4!? (an interesting manoeuvre) 10...Bxd1 11 Nxd6+ Bxd6 12 Rxd1 0-0 13 c4 Rd8 62

We now return to 9 a3 (D):

acceptable play by 16...Rd8 17 Bg2 (17 c3!? is interesting, but amounts to an exchange sacrifice in view of the impending ...Bd3) 17...Be7. b) 9...Be7 seems the best move-order. 10 Bf4 Qd8 11 Re1 0-0 (D) and then:

Black to play 9...Be7 10 Qe2 0-0 11 h3 Qd8 White to play

This move is virtually obligatory, but equalizes. Not 11...Nbd7?? 12 Bf4 +–.

b1) 12 Nc4!?

12 g4 Bh7 13 Bf4 Nbd7 14 Bg3 = Nd5 15 Nd1 (D)

(rather an untypical idea) 12...Na6 (an oddlooking development, but the knight has the promising routes ...Nb4-d5 and ...Nc7-d5, which can hardly be impeded as White has two knights that are stopping him playing c4; 12...Nbd7?! 13 Nd6 gives White the initiative) 13 a3 Nc7 14 Na5 Qc8 15 Qe2 Ncd5 (my favourite fork guarantees the sought-for exchange) 16 Nxd5 Nxd5 (Houdini recommends 16...cxd5!, when 17 c4?! Qd7 =+ leaves the a5-knight out on a limb) 17 Bd2 Rd8 18 c4 Nf6 19 Bc3 Qc7 20 b4 Rac8 21 Rad1 (both players have finished their development; White appears to have the initiative, but Black has no weaknesses and has managed to exchange one piece) 21...Qd7 22 h3 b6 23 Nb3 = (White gains nothing with tactical blows on c6: 23 Bxc6? Qc7 – +; 23 Nxc6?! is preferable, but also to Black’s benefit: 23...Rxc6 24 Bxc6 Qxc6 25 g4 Ne4! 26 d5 Nxc3 27 dxc6 Nxe2+ 28 Kf1 Nd4 =+) 23...Qc7 24 Qb2 Rd7 25 Bd2 Qd8 = Provoost-S.Kasparov, Roosendaal 2012. There is a dynamic balance. If White overestimates his chances in positions of this type based on his space advantage, then he can get into trouble, which was indeed the case in this game.

15 Ne4 is met by 15...N7f6 =.

Black to play White obviously plans to play c4, and after that to bring his knight to c3 or e3. 15...N5f6 15...Bg5!? 16 Nxg5 hxg5. 16 Nc3 In case of 16 c4 Black’s knight can intrude into the centre: 16...Ne4 17 Bh2 Qb6 =.

b2) 12 h3 Nbd7 13 Nc4 Nb6 (precise timing: now the white knight can’t leap to d6) 14 Ne3 Bh7 = Mannion-Tiviakov, Douglas 2014. White lacks sufficient resources to make the d5 advance, and the standard ...N(either)d5 will follow. As you see, the bishop has no problems, as it has a retreatsquare on h7.

16...a5 There is no need to repeat moves. If Black wishes, he can continue the struggle, placing his hopes on slight nuances, such as the weakness of White’s kingside. 63

17 Rfe1 Re8 18 Ne5 Nxe5 Also acceptable is 18...Qb6 19 Nxd7 (19 Nxf7? Kxf7 20 Qxe6+ Kf8 –/+) 19...Nxd7 20 Na4 Qa6 =. 19 dxe5 Nd5 19...Nd7 20 Ne4 Bxe4 21 Qxe4 Qb6 =. 20 Ne4 (D) Black to play 29...Rb8 29...Rad8 is slightly more accurate, as it forces White to decide on his priorities on the d-file. 30 Rxd5 (30 Rd3 Ng6) 30...cxd5 31 Qa6 Qxa6 32 Bxa6 Rb8 leads to mutual chances in the ending. 30 Bf3 Rxd1+ 31 Bxd1 Nd5 32 Bg3 g6 33 Bf3 Black to play

Or 33 h4 Be3 34 Qxc6 Qxc6 35 Rxc6 Bf4 with compensation.

20...b5?!

33...Qb3

Black isn’t yet playing with fire, but this is also not really the best idea. 20...Bxe4!? 21 Bxe4 Bh4 maintains counterplay.

The queen or the rook can work both along the file and the rank from b3.

21 c4

34 Qe4

The principled 21 Nd6 is stronger. It wasn’t easy at the board to assess the consequences of an attempted encirclement of the advanced guard on d6: 21...Bxd6 (21...Rf8 22 c4 +=) 22 exd6 e5 23 c3 Qxd6 (23...f6? 24 Rad1 Qxd6 25 a4! +/–) 24 Bxe5 Qc5 25 Qd2 a4 with a rough balance, as White’s slight initiative comes at the cost of the bishop-pair.

34 Qxb3 Rxb3 35 Bg2 =. 34...Bf4! (D) Perhaps my opponent had missed this resource. It’s nothing catastrophic for him, but after the exchange, the initiative will be totally at Black’s disposal.

21...bxc4 22 Qxc4 Bxe4 23 Bxe4 Bh4 This is the right idea. I am fighting for control of the dark squares, taking into account the fact that the e4-bishop is light-squared and plays little role in this battle. If he exchanges it for the knight, Black’s pawn-structure will be improved by ...cxd5. 24 Bh2 Qb6 25 Re2 Bg5 26 Bd3 Red8 27 Rd1 Ne7 28 Rc2 Rd5 29 Be2 (D) White to play 35 Re2? The time-trouble inaccuracies begin. 64

Black is winning. The rest is more or less logical and doesn’t need special comments.

35...Bxg3 Black wishes to damage White’s pawn-structure and deprive him of the bishop-pair, but 35...g5 =+ or especially 35...Bc1! –/+ might have been wiser.

57 Ke2 Rc4 57...Rxg4 58 Rh7 Rxg3 –+.

36 fxg3 Rb5?!

58 Kd1 Rxg4 59 Rg6 Kf5 60 Rg8 Rxg3 61 Rf8+ Ke5 62 h6 Rh3 63 Rh8 Rh2 64 h7 Ke4 65 Rg8 Rxh7 66 Rxg5 Rh2 67 Rg4+ Kd3 68 Rg3+ Kc4 69 Rg5 Rd2+ 70 Kc1 Rd5 71 Rg4+ Rd4 72 Rg8 e5 73 Kc2 Rd2+ 74 Kc1 Rh2 75 Re8 Kd4

Stronger is 36...Qd1+ 37 Kg2 (37 Re1 Qd2 –+) 37...Kg7 38 h4 Rb3 –/+, when White is almost paralysed. 37 Kh2 Rc5 38 Bg2 Rc1 39 Qf3 Qxf3 40 Bxf3 Kf8 41 Be4 Ke7 42 Kg2

75...e4 76 Rxe4+ Kb3 –+.

The white king is approaching the centre, so Black must hurry, or else lose his advantage.

76 Rd8+ Ke3 77 Rc8 c2 78 Rc3+ Kf4 79 Rc8 e4 80 Rf8+ Ke3 81 Rf7 Rf2 82 Rh7 Kf3 83 Rf7+ Ke2 84 Re7 e3 85 Kxc2 Kf1+ 86 Kd3 e2 87 Rh7

42...Nb6 43 Bc2 Nd7 44 Kf3

After 87 Kd2, 87...e1Q++ comes with check.

A bit too late.

87...Rf3+ 88 Ke4 Rf8 89 Rh1+ Kg2 0-1

44...Nxe5+ I wasn’t sure that this leads to victory, but I saw nothing better.

Conclusions Moving the bishop to d3 after playing g3 doesn’t look very logical (A.Kovačević-Milanović).

45 Rxe5 Rxc2 46 Rxa5 Rxb2 47 Ra7+ Kf6 48 h4 Rb3+ 49 Kf2 c5 50 Rc7 Rb5 51 a4 Ra5 52 Kf3 g5 53 h5

After 7 Bg2 it is reasonable to provide the bishop with a safe haven at once by means of 7...h6 – a generally useful move which can also provide luft for the king.

This move fixes a ‘hook’ for counterplay – the h6pawn. 53...Ke5!? 54 Rxf7

Don’t forget that sooner or later White will play Bf4, so you must have a square for the queen, or a disaster will occur. Other things being equal, it seems to me that it is more reliable to move to d8, rather than to b4. However, the thrust ...Qb4 (and then to b6 or a5) may make the opponent nervous.

54 Kf2 f5 –/+. 54...Rxa4 55 Rh7 55 Rc7 –/+. 55...c4 56 Rxh6? c3 (D)

The knight’s transfer along the route Ne5-c4 will hardly bother Black, as his pieces are harmoniously placed (Provoost-S.Kasparov). The events in our main game demonstrate Black’s possibilities to play for a win, as the positions that arise are asymmetrical.

Kovalenko’s 5...Bg4 The Latvian grandmaster Kovalenko deals with the opening in an original and challenging way. Rather than spending time on ...c6, he immediately develops his bishop. An additional

White to play

65

and in this simplified position, White’s minuscule initiative is not enough to provide winning chances against such a high-level opponent, Movsesian-Tiviakov, Wijk aan Zee 2013.

bonus is that the black queen can still move freely along its third rank. This has its merits, but also its risks; we can say that Black creates additional problems both for his opponent and for himself. This may give Black more winning chances than he might expect in the standard lines, and it may be especially difficult for White to deal with in rapid or blitz games.

b) 6...a6 (D) is another possible approach.

Game 15 Laznička – Kovalenko European Ch, Erevan 2014 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 Bg4 (D)

White to play To me this move implies that Black is thinking in terms of active play on the queenside. The c-pawn is likely to advance two squares (...c5) at some point. But let’s not jump the gun. 7 h3 Bh5 8 Ne5 Bxe2 9 Nxe2 Nbd7 (the exchange of light-squared bishops has already eased Black’s task, and if the knights leave the board too, then the chances will become absolutely equal) 10 Nxd7 Qxd7 11 c3 e6 12 0-0 Bd6 (the ‘Caro-Kann-style’ 12...c5 is not bad either) 13 Bf4 0-0 14 Bxd6 cxd6 (as you see, I was mistaken in my comment about 6...a6... however, after 14...Qxd6!? Black could have carried out the standard ...c5 advance, with good counterchances; it’s a matter of taste) 15 Nf4 Rad8 16 Re1 Rfe8 17 Qb3 Kh8 18 Rad1 b5 = S.Chekhov-Kovalenko, Moscow 2012. It should be obvious to everyone that Black enjoys ample play here.

White to play Indeed, why must Black spend a tempo on ...c6? Can’t he save time and immediately put pressure on the d4-pawn? 6 h3 It makes sense to determine the bishop’s intentions immediately. 6 Be2 not only gives Black a freer hand, but also commits this bishop to a square where it may not be best placed (it makes more sense for this bishop to go to g2, after advancing the kingside pawns). Then:

c) 6...e6 7 0-0 Nc6 (D).

a) 6...c6 7 0-0 Nbd7 8 h3 Bh5 9 Bg5 e6 10 Bh4 Bxf3 (the prophylactic 10...Qc7 looks natural, to meet 11 Bg3 with 11...Bd6 =) 11 Bxf3 Qc7 12 Re1 Be7 13 Bg3 (I don’t understand White’s play; now he won’t have the bishop-pair) 13...Bd6 14 Bxd6 Qxd6 15 d5 cxd5 16 Nxd5 0-0-0 17 Nc3 Qxd1 18 Raxd1 a6 19 Rd4 Nb6 20 Rb4 Nbd5 21 Rb3 Rd7

White to play 66

Rdb1 Qxc2 15 Bd3! and the queen perishes due to 15...Qxc3 16 Bxh7+ +–) 14 b4 Qd8 15 Qg3 Rc8 16 Be5 a5 17 Ne4 axb4 18 axb4 (White has the initiative but Black still has his head above water) 18...Bxb4? (18...g6 19 Nxf6+ Bxf6 20 Bd6 Be7 21 Bf4 +=) 19 Nxf6+ Nxf6 20 Rab1?! (Black has major problems after 20 Rdb1! Be7 21 Rxb7 Ne8 22 Raa7 {the rooks’ intrusion on the seventh rank is an omen of a terrific attack} 22...Bf6 23 Rd7 Qb6 24 Bh5 g6 25 Bxf6 Nxf6 26 Bxg6 hxg6 27 Rxf7 +–) 20...Be7 21 Rxb7 Ne8 22 Ra1 +/– Pavasović-Karttunen, European Team Ch, Novi Sad 2009. It turns out that White has lost a tempo in comparison with the previous bracket. Nevertheless, Black’s position is unsatisfactory.

White’s modest treatment gives Black a wide range of plans to choose from; the knight’s development on c6 doesn’t look bad at all. 8 Bg5 Be7 9 Nb5 Qd8 10 c3 (in case of 10 c4 Black can, if he wishes, annihilate the d-pawn; e.g., 10...a6 11 Nc3 Bxf3 12 Bxf3 Nxd4 13 Bxb7 Ra7 14 Be4 c5 =) 10...0-0 11 h3 Bf5 12 a4 Nd5 13 Bxe7 Qxe7 14 Na3 Rad8 15 Nc4 Qf6 16 Re1 (DavidovichFinegold, Auburn Hills 2003) and after 16...Nf4 Black can’t complain; he can even generate some activity on the kingside. We now return to 6 h3 (D):

b) 9 0-0-0 e6 10 Kb1 Qc7 11 g4 Bb4 12 g5 += (White has gained a significant advantage by unsophisticated means) 12...Nd5 (D) and here:

Black to play 6...Bh5 6...Bxf3 is rarer and cannot be recommended. After 7 Qxf3, 7...Qxd4?? loses to 8 Qxb7 +–, so Black must spend time on the preventive 7...c6, but then it is not clear why he gave up the bishop. 8 Be3 Nbd7 (D) and then:

White to play b1) 13 Ne4 b5 14 h4 N7b6 15 Bc1 (neither player has any ‘hooks’ for their attacks as the enemy pawns are placed on their initial positions: h7-g7-f7 and c2b2-a2; however, White is a little ahead in development, and his long-range bishops can work on both flanks, while the black knights are less mobile) 15...0-0-0 16 c3 Bd6 (BaklanS.Löffler, Austrian Team Ch 2006/7) and now the simple 17 Nxd6+ Rxd6 18 b3 is sufficient. Bishops are considerably stronger than knights in an elastic position, and Black has no effective way to damage White’s pawn-structure: 18...b4? 19 cxb4 (19 c4? Nc3+) 19...Nxb4? 20 Bf4 +–.

White to play a) 9 Be2

b2) 13 Nxd5

e6 10 0-0 Be7 11 Rfd1 0-0 (the fact that both players have castled kingside suggests a positional struggle without swift attacks) 12 Bf4 Qb4 13 a3 Qa5 (the pawn is untouchable: 13...Qxb2?? 14

cxd5 14 h4 a6 15 h5 (D) is an even simpler and clearer approach: just advance the pawns.

67

dreams come true) 14 0-0 a6 15 Rd1 Qc6 (interesting is 15...Qe6!? 16 Qxe6 fxe6, which we can assess as +=, while noting that it is very unclear whether White will be able to make much of his positional advantage; it looks more like a draw) 16 c3 Bf6 17 Bf4 Rg8 18 f3 Rg6 19 Kh1 Rc8 20 Rd3 Kf8 21 Rad1 (Black’s uncoordinated rooks are definitely a problem) 21...Qc4 22 Be5 Qe6 23 Bxf6 Qxf6 24 Rd8+ Rxd8 25 Rxd8+ Kg7 26 Rd7, Kobalia-Ponkratov, Moscow 2013. Black lacks sufficient counterplay, and only White has winning chances.

Black to play In Zhigalko-Galego, Khanty-Mansiisk Olympiad 2010, Black found no real way to oppose this approach: 15...Nb6 16 g6 (16 Bf4!?) 16...hxg6 17 hxg6 Rxh1 18 Bf4 Qe7 (18...Bd6 is more stubborn, as after 19 Bxd6 Qxd6 20 gxf7+ {20 Qxf7+?? Kd8 –+} 20...Kd7 21 Qxh1 Rf8 += Black is hanging on somehow) 19 Qxh1 0-0-0 20 a3 Bd6 21 Bxd6 Qxd6 22 gxf7 Rf8 (the position is almost the same, except that the king has moved away from the centre; dark times are coming for the e6pawn) 23 Qh5 Qe7 24 Bh3 Rxf7 25 Re1 Rf6 26 Qe5 +/–. Black is on the verge of defeat.

7...Bg6 (D)

White to play

We now return to 6...Bh5 (D): 8 Bg2

For 8 Ne5, see the next game. 8...Nbd7 9 g5 Nd5 10 Nxd5 Qxd5 11 0-0 Qc4 12 c3 0-0-0 13 Nh4 e5 The quiet 13...e6 is inappropriate now: after 14 Nxg6 hxg6 15 Qf3 +/– both Qxf7 and Qxb7# are threatened. 14 Nxg6 hxg6 15 Qf3 c6

White to play

The queen protects f7.

7 g4

16 Rd1 Nb6 17 a4 (D)

And what if White returns to the idea of developing his bishop on e2? 7 Be2 Nc6 8 Nb5 Qd7 9 Ne5 Nxe5 (9...Bxe2 10 Qxe2 Nxe5 11 Qxe5 Rc8 drops a pawn: 12 Nxa7 Ra8 13 Nb5 +/–) 10 dxe5 Bxe2 (Black can’t exchange queens: 10...Qxd1+ 11 Bxd1 Bxd1 12 Nxc7+ +–) 11 Qxe2 c6 12 exf6 cxb5 13 fxg7 Bxg7 (the situation has significantly simplified, and I feel that Black’s position ought to be satisfactory: the rook can move to its third rank by ...Rc8-c6, and from there it can be brought into play on either flank; alas, in practice it is not easy to make these

Black to play 68

Look closer to the position. Laznička has been playing skilfully, launching his attack while Black is unable to make much headway on the kingside because of the ‘breakwater’ on g5. I’ll note that the g2-bishop performs two functions simultaneously: defence of his king and attacking the enemy residence. This is what distinguishes long-range pieces. The b6-knight can’t boast of the same skills.

Conclusions If White wishes to fight for the advantage, he must act aggressively after 5...Bg4. Quiet development, such as 6 Be2, is unlikely to embarrass Black (S.Chekhov-Kovalenko;Movsesian-Tiviakov). It makes more sense to ‘check the bishop’s documents’ at once with 6 h3. Black shouldn’t meekly give up the bishop (6...Bxf3) because he then comes under long-term pressure. Black’s torture was painful to behold in Pavasović-Karttunen,Zhigalko-Galego and Baklan-S.Löffler.

17...Qd5 17...a5!?. 18 Qe2 Qc4 19 Qg4+ Nd7 20 Be3 Be7 21 b4

It is more judicious to retreat with 6...Bh5, but then the white g-pawn is free to advance. Moreover, by running all the way to g5, it can actually paralyse Black’s kingside plans. In addition, there is the traditional advantage of the bishop-pair, which is quite appreciable with opposite-side castling, because long-range pieces are especially effective in a boardwide battle.

21 Qf3 is stronger; for instance, 21...f5 22 gxf6 gxf6 23 d5 +/–. 21...f5 22 gxf6 gxf6 23 d5 Qxg4 Now Black at least won’t be mated any time soon! 24 hxg4 cxd5 25 Rxd5 Rh4 26 g5 f5 27 Bxa7 Rc4 28 Rb5 e4 (D)

In the next game, we shall consider a more popular continuation for White: 8 Ne5. After 8...Nbd7, White has a very wide choice of viable lines, which raises an alert: does Black have acceptable play against all of them? We shall see.

Game 16 Ivanchuk – Kovalenko

White to play

Jurmala rapid 2012

The g2-bishop must be restrained, but this can’t solve all Black’s problems.

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 Bg4 6 h3 Bh5 7 g4 Bg6 8 Ne5 Nbd7 (D)

29 Bd4 Kc7?! Alas, the pawn can’t move backwards: 29...Bxg5?! 30 Bxe4 +–. 29...Nf8 is more resilient. 30 f3 exf3 31 Bxf3 b6 32 a5 Clear as day. The last reserve – the a1-rook – joins the attack. 32...bxa5 33 Raxa5 Bxg5 34 Rb7+ Kd6 35 Ra6+ Ke7 36 Rxg6 Bf4 37 Bf6+ Kf7 38 Bh5 1-0 A crushing defeat, even though Black didn’t make any clear-cut blunders. The variation is generally risky for Black.

White to play

69

9 Nxg6 This is the most popular continuation among strong players. But what of the other possibilities? a) 9 Bf4 Nxe5 10 dxe5 (10 Bxe5 is similarly met by 10...Qb4) 10...Qb4 11 Bd2 Ne4! (Black is operating ideally). After 12 Bg2? Nxd2 13 Qxd2 Rd8 14 Qe2 e6 =+ Khalifman-Kovalenko, Jurmala rapid 2012, the bishop-pair together with an elastic pawn-structure turned the evaluation in Black’s favour; Kovalenko gained a victory over an eminent adversary. With limited time, it is not that easy to find the right path in an unfamiliar area. Especially in blitz and rapid games, lines like 5...Bg4 are the most dangerous for your opponents! Instead, the position would be unclear after 12 Nd5 Qxd2+ 13 Qxd2 Nxd2 14 Nxc7+ Kd7 15 Nxa8 Nf3+ 16 Ke2 Nd4+ or 12 Nb5 Qxd2+ 13 Qxd2 Nxd2 14 Nxc7+.

White to play c1) 10 Bg2 0-0-0 11 Qf3 Qa6! (a strong manoeuvre; it’s a close-run thing, but Black is just about holding on) 12 Bf1 Qb6 13 Nc4 Qc6 14 Qxc6 bxc6 = A.Gallego-Perez Mitjans, Sabadell 2012. The queens have disappeared, and Black is even slightly ahead in development, while the doubled pawns can be exchanged off. c2) In reply to the terrifying 10 h4, 10...Be4 is more or less fitting; other options are complex but worse. 11 Rh3 Bc6 12 g5 (12 h5 +=) 12...Ne4 13 Nxc6 Qxc6 14 a3 (to avoid a pin by ...Bb4) 14...Nxc3 15 Rxc3 Qe4+ 16 Kf2 +=.

b) 9 Nc4 Qe6+ (D).

d) 9 Nb5!? (D) is a worrying-looking thrust.

White to play It is rare to find Tiviakov tangled up in such unclear lines, but sometimes it happens, and not always with a positive outcome. 10 Ne3 h5 11 g5 Ne4 12 Bg2 Nb6 (after 12...Nxc3 13 bxc3 Qa6 14 Rb1 c6 15 Bf1 White gradually starts pressing) 13 Qf3 Nd6?! (Black should capture on c3: 13...Nxc3 14 bxc3 c6 15 0-0 Qd7 16 a4 e6 with nearequality) 14 a4 a6 15 0-0 Qd7 16 a5 Nbc8 17 Re1 e6 18 d5 (this is already serious; Black hasn’t finished his development but is under intense pressure) 18...Be7 19 dxe6 fxe6 20 h4 +/– Iordachescu-Tiviakov, Nakhchivan 2013.

Black to play In a rapid game it is difficult for White to orientate himself in such unconventional positions without special preparation, but with a ‘classical’ timecontrol he can assess matters more calmly, and may choose this move. White seeks to exploit Black’s failure to play ...c6. After 9...Qb6 10 Nc4 Black must decide where to put his queen: d1) Of course, the knight doesn’t taste good: 10...Qxb5?? 11 Nd6+ +–.

c) 9 f4 looks critical, and has the simple idea of trapping the bishop. 9...e6 (D) and then: 70

d2) 10...Qe6+ 11 Qe2 (11 Be2 is unpromising since after 11...0-0-0, 12 Nxa7+? fails to 12...Kb8 13 Nb5 Qd5 –+) 11...Qxe2+ 12 Bxe2 Kd8?! (creative, but also a little artificial; perhaps 12...00-0!? is more logical; e.g., 13 Nxa7+ Kb8 14 Nb5 Bxc2 with approximate equality, Vishnu-Z.Varga, Budapest 2013) 13 Na5 c6 14 Nc3 Kc8 15 Bf4! (White must play very actively, with a willingness to sacrifice; otherwise he risks finding himself in a worse position) 15...e6 (Nayer-Kovalenko, Moscow 2010; 15...Bxc2 is risky due to 16 Rc1 with a possible attack on the queenside) 16 0-0-0 Bb4 17 Nc4 +=.

White to play 10 Qf3

d3) 10...Qc6!? (D) is possible, attacking the rook, but it invites a sharp response in the centre:

White easily obtained an advantage in the following example: 10 Bg2 c6 11 g5 Nh5 (I doubt this is a good idea; the centralizing 11...Nd5!? looks healthier, and at least there is no instant central break here; e.g., 12 Nxd5 {12 Ne4 Qc7} 12...cxd5 and then 13 c4? fails to 13...Qb4+ =+) 12 d5! (I am confused by the fact that White’s simplistic play proves effective; if it is so easy, doesn’t it mean that the whole line with 5...Bg4 is dubious?) 12...cxd5 13 Qxd5 Qxd5 14 Bxd5 Rb8 15 0-0 e6 16 Bg2 Bd6 17 Ne2 Ke7 18 Rd1 Rhc8 19 c3 += Y.Geller-Kovalenko, Sochi 2012. The two bishops provide a long-term initiative.

White to play

10...c6 11 Be3 e6 After 11...Nb6 12 0-0-0 Nbd5 13 Nxd5 Qxd5 14 Qxd5 cxd5 15 Bg2 e6 (D), if White plays passively, then as a rule Black can take over the initiative:

11 d5! (this leads to forcing play; 11 Rg1?! is weaker due to 11...e6 12 Na5 Bb4+ 13 Bd2 Bxa5 14 Bxa5 0-0 =+; also poor is 11 Ne3?!, as in Firman-Kovalenko, Yuzhny 2011, when Houdini recommends the uncompromising 11...Qxh1 12 Nxc7+ Kd8 13 Nxa8 e6 –/+; while this is complicated, it all ends well for Black) 11...Nxd5 12 Qxd5 (at first it seems that White wins a piece, but we must look further) 12...Qxd5 13 Nxc7+ Kd8 14 Nxd5 Be4 (the vital counter-blow) 15 Rh2 (15 Rg1 Bxd5 16 Bf4 +=) 15...Bxd5 16 Bf4 Rc8 17 Ne3 Bf3 (the smoke has cleared, though the situation remains unconventional; the black king is deprived of his castling rights, but the f3-bishop is excellent; overall, it looks about even) 18 Bg2 Bxg2 19 Rxg2 e5 20 Bh2 Bb4+ 21 Kf1 (21 c3? Rxc3 =+) 21...Re8 22 Rd1 Rc6 23 f3 Kc8 (the king has to move ‘on foot’, but since there are no queens on the board, there is little danger) 24 c3 Bc5 25 Nc4 f6 26 Rgd2 Nb6 27 Na5 Rc7 28 f4 exf4 29 Bxf4 Rce7 with full equality, EdouardMilanović, Deizisau 2012.

White to play a) 16 Rhe1?! Kd7 17 Kb1 Bd6 18 Bg5 b5 19 Rd3 Ne8 20 Rf3 f6 21 Bd2 g5 22 Rb3 a6 23 Bf1 Bf4, Grekh-Kovalenko, Mukachevo 2011. Black stands slightly better – this is the result of the fixation of dark squares, notably f4.

9...hxg6 (D) 71

b) 16 c4!? Rc8 17 c5 is a more aggressive approach, launching a pawn-storm on the queenside. 17...Be7 18 b4 Ne4 19 Bxe4 (19 Kb2!? seems more flexible, without revealing White’s intentions for a while) 19...dxe4 20 d5 Kd7 (20...exd5!?) 21 d6 (21 f3!?) 21...Bh4 22 Rhg1 Bf6 (despite the menacing pawn-chain, the situation is far from clear; the bishop intends to go to e5 and c3, and the pawn will play ...f5, and so on) 23 g5 Bc3 24 a3 a5 (I can’t see how Black can lose after 24...Rxh3!? 25 Kc2 Be5 26 Rg4 f5 27 gxf6 gxf6 28 Rxg6 f5, with counterplay) 25 Kc2 axb4 26 Kb3! Rxh3 27 axb4 Be5 28 Rg4 += Siugirov-Moranda, Kazan 2013. White went on to win.

Black to play 26...Rxe4! Creative and correct – especially with a rapid time-limit.

12 0-0-0 Nd5 13 Ne4 Qc7 14 c4 Nxe3 15 fxe3 (D)

27 Rxe4 Nxc5 28 Ree1 Na4 29 Bb3 Nb6 A pawn is quite sufficient compensation for the exchange, considering the opposite-coloured bishops. 30 Re4 a5 31 a3 a4 32 Bc2 Rd5 33 Re8+ Ka7 34 Be4 Rb5 35 Bd3 Rb3 36 Re4? A blunder, but let’s not be too carping; it’s not a game with a classical time-limit. I decided to feature this game because of the high level of the players. Besides, the opening part is more valuable for us, not the ending.

Black to play Exchanges are generally to Black’s benefit, but one should take into account the fact that the f-file has been opened. White’s pressure upon the f7pawn grants him an edge.

36...Nd5 37 Ka2 Qb6 37...Qd6! would have won more efficiently, thanks to the unavoidable threat of ...Nc3+ and ...Qxa3#. I’ll leave the rest without comment.

15...Be7 16 Kb1 f5 17 Nc3 0-0-0 18 c5 Rhf8 19 Rg1 f4!

38 Bc4 Ne3 39 Rc1 Bxd4 40 Rxd4 Qxd4 41 Bxb3 axb3+ 42 Kxb3 Qd5+ 43 Kc3 Qe5+ 44 Kb3 Qd5+ 45 Kc3 c5 46 Qe2 Qd4+ 47 Kb3 Qd5+ 48 Kc3 c4 49 Rb1 Qa5+ 50 Kd4 Qb6+ 51 Kc3 Qf6+ 52 Kd2 Qd4+ 53 Ke1 Qe4 54 Rc1 f3 55 Qd2 c3 56 Qxc3 Nd5+ 0-1

This is the right decision over the board. Black creates counterplay on the dark squares. 20 Bd3 The natural-looking 20 e4? is not good due to 20...e5 –/+, when the c5-pawn falls. 20...g5 21 Rge1 Kb8 22 Qe2

Conclusions

White could still have kept a better position by 22 exf4. For example, 22...e5 23 f5 += or 22...Rxf4 23 Qe3 Nf8 += (23...Bf6?! 24 Ne2 +/–).

In this line, White has a key decision on move 9, where he has a wide choice of options.

22...Bf6 23 Ne4 e5 24 Bc4 exd4 25 exd4 Rfe8 26 Qf2 (D)

9 Nb5 leads to an original dance of queen and knights – there is nothing similar in other lines. Here one needs to be specifically prepared. Theoretically White’s chances are better, but the 72

practical chances tend to lie with the player who is more familiar with the positions (which will tend to be Black). In a blitz game one can’t envy White at all, as every single step from the right path can result in disaster.

The fianchetto is considered the most promising reply. 6...Bg4 This is the most accurate option. Two other moves are rather dubious:

The struggle is also quite complicated in the case of 9 Nc4. At least Tiviakov couldn’t defuse the pressure exerted by Iordachescu, having allowed the ‘dyke breach’ by d5.

a) 6...Nc6?! 7 Bf4 Qb4?! 8 a3 Qb6 (8...Qxb2 9 Na4 +–) 9 d5 and now 9...Ng4 10 Qd2 e5 is losing, but what else should Black have done? If he had retreated the knight, he would have been at least four tempi behind in development! After 11 dxc6 exf4 12 0-0-0 Bd6 I played the imprecise 13 Re1+?! in S.Kasparov-Ramon Perez, Balaguer 2008 and the struggle dragged on. 13 Qe2+! would have ended the fight quickly: 13...Be6 (13...Kf8 14 Rxd6 cxd6 15 Nd5 Qd8 16 c7 Qd7 17 Bh3 h5 18 Re1 +–) 14 Nd4 f3 15 Qxf3 Ne5 16 cxb7 Rb8 17 Qe4 Qxb7 (17...Rxb7 18 Na4 Qa7 19 Nxe6 fxe6 20 f4 +–) 18 Qxb7 Rxb7 19 Bxa6 +–.

9 f4 also leads to hand-to-hand fighting with very unclear consequences. Please pay heed to an interesting queen manoeuvre to a6 (Gallego Alcaraz-Perez Mitjans). 9 Nxg6 was used in our main game. Ivanchuk had the initiative, but, probably in time-pressure, lost the thread. 9 Bf4 was successfully met by the variation’s expert in Khalifman-Kovalenko. The victim was a famous Russian grandmaster. I dare to assume that time-trouble played a significant role here too.

b) 6...b5?! (D), expanding on the queenside, looks too provocative, and with accurate play White not only stands better, but he is actually stifling his opponent:

5...a6 In the early days of the 3...Qd6 system, this was a main line, but nowadays it is considered less solid than 5...c6. Game 17 Svidler – Tiviakov Turin Olympiad 2006 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 a6 (D)

White to play b1) 7 Bg2 Bb7 8 0-0 e6 and now: b11) 9 Bf4 can be played before putting the knight on e5. After 9...Qd8 10 Ne5 Black can contest the h1-a8 diagonal by 10...Qc8, but it doesn’t bring much relief after 11 d5!. Usually this move means trouble for Black if it opens the central files while the black king is still on e8 and the white king is safe. After 11...Bb4 12 dxe6 fxe6, KhalifmanZsu.Polgar, Internet 2005 featured 13 Bxb7?! Qxb7 14 Bg5 0-0 15 Bxf6 Rxf6 (not, of course, 15...gxf6?? 16 Qg4+ Kh8 17 Qxb4 fxe5 18 Qxf8#) 16 Qd4 Be7 17 Qe4 Qxe4 18 Nxe4 +=, when Black was still on the board, but White’s advantage was indisputable. 13 Ne4! is even

White to play

6 g3 73

stronger in view of 13...Nxe4? 14 Qh5+ +– or 13...Bxe4 14 Bxe4 Nxe4 15 Qh5+ +–.

(S.Kasparov-Akbarinia, Urmia 2008) and now 16 Ne5! looks winning, as Black’s pieces can’t disentangle. The a5-knight is endangered, c4 is threatened all the time, etc. 16...Be7 17 Nc5 Bxc5 18 Qa4+ c6 19 dxc5 +–.

b12) 9 Ne5! Bxg2 10 Kxg2 (D).

b212) White should capture the pawn by 12 Bxc7!, because 12...Bxf3?? is losing: 13 Qxf3 Qxc7 14 Nb5! (alas, I didn’t find this resource) 14...axb5 (14...Qd7 15 Qxa8 Qxb5 16 c4 bxc3 17 bxc3 +–) 15 Qxa8 +– and the passed a-pawn provides the victory. Thus, Black has to play 12...Nbd7 13 Na4 Nd5 14 Bb6 N7xb6 15 axb6 Nxb6 16 Nxb6 Qxb6 17 c4 +/– (Ne5 and Qa4+ are threatened, while it is not so easy for Black to castle kingside, as there is a fork by Nd7) 17...Be7? 18 Ne5 Qc7 19 Qa4+ Kf8 20 Qd7 +– Korneev-L.Galego, Portuguese Team Ch 2010.

Black to play Now the bishop will be replaced by the queen on the long light-square diagonal) 10...b4?! (overoptimistic) 11 Qf3 c6 12 Bf4! (White plays vigorously, spending no time retreating the knight) 12...bxc3 13 Nxf7 Qd5 14 Nxh8 Bb4? (after the more resilient 14...cxb2 15 Rab1 g5 16 Qxd5 cxd5 17 Be5 Bg7 +/– the fight continues, although White obviously has the advantage, as his rooks can burst into the enemy camp via the b-file, which may impact the health of the a-pawn too) 15 Bxb8 Rxb8 16 bxc3 Bxc3 17 Rab1 Rd8 18 Rb7 Nd7 19 Rd1 Qxf3+ 20 Kxf3 +– Solodovnichenko-Panbukchian, Rochefort 2014. On the agenda is Rd3, etc.

b22) 10...Bd6 (D) seems more accurate, but life is not easy for Black here either.

b2) 7 Bf4 Qb6 8 Bg2 Bb7 9 0-0 e6 10 a4 (D) (a common way to create play on the queenside when the black queen sits on b6) and then:

White to play For instance, 11 Be3 c5 (a high-level encounter is instructive: 11...b4 12 a5 Qa7 13 d5 Bc5 14 dxe6! Bxe3 15 exf7+ Kxf7 16 Ne5+ Ke7 17 Bxb7 Qxb7 18 Qd3 with a strong attack, Sutovsky-Sermek, Terme Zrece 2003; it is not surprising that Black failed to repel the attack, as his king is absolutely exposed in the centre) 12 axb5 axb5 (Z.AlmasiKurajica, Bled Olympiad 2002) 13 dxc5 Bxc5 14 Bxc5 Qxc5 15 Qd4! Qxd4 16 Nxd4 Bxg2 17 Kxg2 Rxa1 18 Rxa1 b4 19 Na2 Nd5 20 Nc1 +=. The b4-pawn is half-dead, while the rook is about to invade on the seventh rank.

Black to play

We now return to 6...Bg4 (D):

b21) 10...b4?! 11 a5 Qa7 and here in one game I had a long and diligent mediation but still didn’t find the best variation: b211) However, my 12 Na4 is not bad. 12...Nd5 13 Bd2 Nc6 14 c4 bxc3 15 bxc3 Nxa5 74

White to play

Black to play a) 9...Nb4 10 Bf4 Qc5 11 Be3 Qd6 12 h3 Bh5 13 Qe2 and now:

7 Bg2 7 h3 Bh5 8 Bg2 Nc6 9 0-0 0-0-0 10 Bf4 Qb4 11 g4 Bg6 12 a3 occurred in a later game between the same players as in our main game. White is attacking everything in his way. That can be an effective way to fight for the advantage, with the motto: ‘everyone – forward into action!’. Now:

a1) The greedy 13...Nbxd5? led to a crushing defeat in Rašik-Antoniewski, Czech League 2006/7: 14 Nxd5 Nxd5 15 Rad1 Qf6 16 c4 (Black’s position is already hopeless; his army on the kingside, as if paralysed, can merely observe the disaster) 16...Nxe3 17 Rxd8+ Kxd8 18 Qxe3 c6 19 g4 Bg6 20 Qb6+ Kc8 21 Rd1 e5 22 Nxe5 +–.

a) 12...Qxb2 13 Qe1 (White has significant threats; the black king may come under attack on the band c-files and from the f4- and g2-bishops) 13...e6 14 Rb1 Qxc2 15 Ne5 Nd5 16 Rc1 Nxf4 17 Rxc2 Nxg2 18 Kxg2 Nxe5 (Hraček-Muse, Hamburg 2013) 19 Ra2 +=. The correlation of forces is quite unconventional. I like to play ‘on points’ with Black, but objectively White’s chances are better.

a2) 13...e5 would be a nice idea, if there were no such thing as the en passant rule. But after 14 dxe6 Qxe6 15 g4 Bg6 16 Nd4 Qe8 17 Rfe1 h5 18 Qf3 c6 (Fressinet-M.Wong, Calvia Olympiad 2004) White can give Black major problems by 19 g5! Nfd5 20 h4 (now Bh3+ is an unpleasant threat as it is impossible to move to b8 because the queen drops after Bf4+) 20...Be7 21 Nxd5 Nxd5 22 Rad1 +– with deadly pressure on the central files.

b) 12...Qc4 13 g5 Nd5 14 Nxd5 Rxd5 15 c3 Rd8 (SvidlerTiviakov, Wijk an Zee 2007) 16 Nd2 +=. The initiative is on White’s side, even though I don’t see anything crucial.

b) 9...Ne5 (D).

7...c6 This raises the question of why Black played ...a6 at all. It looks like a tempo-loss... The consistent 7...Nc6 is more critical for the assessment of this variation. White must act resolutely, as otherwise he will have problems with his d-pawn. 8 0-0 0-0-0 9 d5 (D) and then:

White to play This looks more solid, starting an exchanging operation. 10 Bf4 Nxf3+ 11 Bxf3 Bxf3 12 Qxf3 e5 (best; otherwise Black would be too passive) 75

13 dxe6 Qxe6 14 Bg5 (after this simple move White gets at least some positional advantage by destroying Black’s pawn-structure) 14...Bd6 (14...Be7? doesn’t work due to 15 Rfe1 Qd6 16 Bf4 Qd7 {16...Qc5?? 17 b4 Qxb4 18 Rab1 +–} 17 Rad1 +/–) 15 Rae1 Be5 16 Na4 Rhe8 17 Nc5 (threatening 18 Qxb7#) 17...Qd5 18 Qxd5 Rxd5 19 Bxf6 gxf6 (19...Rxc5? 20 Bxg7 +–) 20 Nd3 += Hraček-Tomczak, Wroclaw 2011.

14...cxd5 It looks like Tiviakov is seeking to make use of the c-file for his own active plans. It is tempting to create a ‘breakwater’ on b7-c6-d5 against the g2-bishop by 14...exd5. But where is Black’s counterplay then? Meanwhile, White could press forward with his kingside advance. 15 c4! Nb6 16 c5 Nc4 17 Bf1 Na5 18 Qg3 Bh4 19 Qe3 h6 20 Be5 Bf6 21 Bxf6 Qxf6 22 f4 Nc6 23 g5 Qe7 24 h4 g6 (D)

We now return to 7...c6 (D):

White to play White to play

8 Bf4 Qd8 9 h3 Bxf3 10 Qxf3 e6 25 Rh3

As usual, 10...Qxd4 is risky given that Black is three moves behind in development.

Or 25 gxh6 0-0-0, with counterplay.

11 0-0-0 Be7 12 g4 (D)

25...0-0-0 It turns out that, on the sly, Tiviakov has been preparing to hide his king on the queenside. 25...h5!? is another idea. 26 Be2 hxg5 27 fxg5 27 hxg5 Qc7. 27...e5 ½-½

Black to play

Conclusions

You don’t need to fire up your finest analysis engine to see that Black has failed to equalize. White’s bishop-pair and lead in development make his advantage clear. On the other hand, Black is solid, and it will be difficult for White to win against such a ‘fireproof’ player as Tiviakov.

In our main game, Tiviakov managed to neutralize Svidler’s pressure, albeit with certain difficulties. However, a year later, Svidler succeeded in breaking Tiviakov’s defences. On that second occasion, Sergei decided not to capture on f3, but to retreat the bishop. But instead of 7...c6, what happens after the principled 7...Nc6? If White replies well, Black can at best reckon on a slightly worse ending (Hraček-Tomczak). Terrible problems can be observed in the games RašikAntoniewski and Fressinet-M.Wong. All this

12...Nbd7 13 Kb1 Nd5 14 Nxd5?! 14 Bc1!? += denies Black the chance to put a pawn on d5. 76

suggests that the 5...a6 line is weaker than the other main continuations

4 d4 Sometimes, not wishing to be engaged in a theoretical argument, White refrains from playing d4:

Rare Lines for White on Moves 4, 5 and 6 Let’s take a look at what happens if White plays a relatively rare continuation, that is, after 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6, any deviation by White from the sequence 4 d4 Nf6 5 Nf3 c6 6 Ne5 (or 6 g3).

a) 4 Nf3 Nf6 5 g3 (combining a fianchetto with avoiding d4 provides Black with additional possibilities) 5...Bg4 6 Bg2 Nc6 (if White doesn’t exert any pressure, Black might as well develop more actively; 6...Qe6+!?) 7 0-0 Nd4 8 d3 0-0-0 9 Be3 e5 10 Re1 h5 11 Bxd4 exd4 12 Ne4 Nxe4 13 dxe4 (R.Barski-Neiksans, Warsaw rapid 2010) 13...h4 =+.

Straight away, we should note that ‘rare’ doesn’t necessarily mean ‘bad’! Some are rightly unpopular, while others are fully playable and simply haven’t gained a following. In my own games in these lines, White’s average Elo rating was around 2450.

b) 4 Bc4 Nf6 (D) and now we shall look at structures where White plays d3:

First of all, there are plans where White puts his pawn on d3 rather than d4; they are covered in the notes to Game 18. Once the moves 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 have been played, there is only one significant deviation on move 5, that is 5 Nge2 intending to bring the bishop to f4 with tempo. This is also covered in the notes to Game18. Finally we come to move 6, after 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6. We have already discussed the main moves, 6 Ne5 (Chapter 1) and 6 g3 (the first section of this chapter). Game 18 features some rather unsophisticated bishop moves: 6 Bd3, 6 Be3 and 6 Be2. The more subtle 6 h3 is the subject of Game 19.

White to play b1) 5 Nge2 a6 6 d3 b5 7 Bb3 Bb7 8 0-0 e6 9 f3 (building a breakwater against the b7-bishop; I am sure that White should get no advantage with such methods) 9...Qb6+ 10 Kh1 Nbd7 (in this specific case, bringing the knight to d7 seemed most flexible to me) 11 a4 Be7 12 axb5 axb5 13 Rxa8+ Bxa8 14 Qe1 0-0 15 Ne4 c5 (15...Bd5!? =+) 16 c3 Rd8 (this pawn-structure is discussed in more detail in line ‘b2’; White’s problem is that the d3-pawn is subject to pressure; 16...Bd5 17 Bc2 Ra8 =+) 17 Qg3 Bd5 18 Bc2 Qb8 (it makes good sense here to move into an endgame) 19 Bf4 e5 (White has enticed this pawn to e5, but does it feel bad there?) 20 Nxf6+ Bxf6 21 Be3 Qb7 22 Ra1 g6 23 Bg5 Bxg5 24 Qxg5 f6 (keeping the pawn-structure flexible; the bishop controls the light-square complex so the pawns are best placed on dark squares) 25 Qe3 Ra8 26 Rb1 Qc7 =+ S.SahlS.Kasparov, Leros 2010.

Game 18 Vorobiov – S. Kasparov Bhubaneswar 2010 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 (D)

White to play 77

despite also being on a half-open file, is not so vulnerable because it is safely protected by its colleague on f7) 15 Qe3 Bf6 16 Bd1 0-0 17 Bf3 Rfd8 18 0-0 Rac8 19 Rad1 (D).

b2) 5 d3 a6 (D).

White to play With the bishop on c4, this is the most appropriate way to start a queenside pawn advance, which can be continued in standard fashion with ...b5 and ...c5. The more conservative plan of setting up a ‘breakwater’ with pawns on b7 and c6 is more effective when White fianchettoes his bishop on g2. 6 Bb3 e6 7 Qe2 c5 8 Bd2 Nc6 (with this pawn-structure, there is no reason not to develop the knight to an active post on c6; in some cases it will be happy to leap to d4) 9 Nf3 Be7 10 Ne4 (D).

Black to play 19...Ba4 (as an aside: if you wish to put the bishop on c6, but can first make a disruptive move of this type, you should normally seize this opportunity; little nuances like this can serve to weaken the enemy pawn-structure or misplace his pieces, and can play a vital role in the fight for an advantage) 20 Rc1 (20 b3 weakens the c3-pawn) 20...Bc6 21 Rfd1 Bxf3 22 Qxf3 Rd5 23 Be3 Rcd8 24 Qe2 h6 25 Rd2 Bg5 26 Bxg5 Rxg5 27 Rcd1 Rgd5 28 g3 R8d7 29 Qe3 Qd8 (frankly speaking, this is nothing terrible for White, who could passively wait for his opponent to reveal a plan; however, purely passive defence is quite tiring and unpleasant) 30 c4?! Rd4 31 b3 b6 32 f3 Qb8 33 Kg2 Qd6 34 Kf2 e5 35 Re2 f6 36 Red2 Kh7 (36...f5!? is a noteworthy idea, with 37 Re2 e4 38 fxe4 fxe4 39 Rf1 Rf7+ 40 Kg2 Rxf1 41 Kxf1 exd3 42 Qe8+ Qf8+ 43 Qxf8+ Kxf8 44 Rf2+ Ke7 45 Ke1 b5 –/+ a possible continuation) 37 g4 e4!? 38 fxe4 Qxh2+ 39 Kf1 Qh1+ 40 Kf2 Qh2+ 41 Kf1 Qe5 –/+ E.Heinemann-S.Kasparov, Bad Zwischenahn 2011. It might be that Black’s advantage won’t be enough for victory, but the assessment of the opening is definite: White has not equalized!

Black to play This is White’s extra resource granted by putting his pawn on d3: he can offer an exchange of knights on e4, when recapturing with the pawn may prove a favourable alteration in the structure for White; however, Black is not obliged to capture on e4, as the knight doesn’t cause Black as many problems from e4 as a black knight on d4 would cause White. 10...Qc7 11 Nxf6+ (11 Bc3 b6 12 Nxf6+ Bxf6 13 Bxf6 gxf6 is approximately equal) 11...Bxf6 12 c3 Bd7 13 Bc2 Ne5 14 Nxe5 Bxe5 (with the knights gone from the board, we should recalibrate our assessment of the position; Black has a long-term initiative thanks to his somewhat better pawn-configuration: the d3-pawn is potentially weak, as it can be subject to frontal pressure on the d-file, whereas the black e-pawn,

4...Nf6 (D)

78

looks reasonable. Let’s not forget that the knight is more passive on e2 than on its standard f3square. c) From time to time 5 Nge2 is played, with the obvious idea of bringing the bishop to an attacking post on f4 with gain of time. But note that the knight would be more active on f3, where it controls e5 and is ready to jump to this square. Overall the positive and negative features of the knight’s development on e2 roughly balance each other out. The most principled reply is 5...Bf5; e.g., 6 Bf4 Qb6 (D) and then:

White to play 5 Nf3 Or: a) The early 5 g3 gives Black an extra option besides the standard set-up with ...c6; he can also play the disruptive 5...Bg4 6 f3 (now the bishop will be ineffective on g2 unless White plays f4, which has its own drawbacks) 6...Bd7!? 7 Bf4 Qb6 8 Qd2 e6 (the bishop is not too bad on d7 here; it performs a useful function on the e8-a4 diagonal and can always be activated via c6) 9 0-0-0 Bb4 10 a3 Bxc3 11 Qxc3 Nd5 12 Qd2 (Jamrich-S.Kasparov, Paks 2010) and now Black can equalize by 12...Nc6 13 Ne2 (13 c4 Na5 14 Kb1 Nxf4 15 Qxf4 0-0 =+) 13...0-0 =.

White to play c1) 7 Qd2 Nc6 8 Ng3 Bg6 and now SolodovnichenkoS.Kasparov, Fagernes blitz 2012 continued 9 d5? (White was making it up as he was going along, whereas the author was aware of these lines; 9 Nb5 Nd5 {9...e5!?} 10 c4 Nxf4 11 c5 Qa5 =+) 9...0-0-0 10 0-0-0 e6 11 Na4? Qb4 –/+. Even in a quiet-looking opening like the Scandinavian, it is wholly possible to get into trouble as White if you ‘ad lib’ carelessly. It’s just move 11, and White might as well have resigned...

b) With 5 Bd3 (D), White wishes to place the knight on e2 after developing the bishop to d3.

c2) 7 Na4 Qa5+ 8 c3 c6 9 Nc5 Qb6 10 Qb3 and now 10...Bc8?! 11 Nc1 Nbd7 12 N1d3 allowed White the initiative in Chrigui-S.Kasparov, Chefchaouen 2014. The best path was 10...Nbd7! 11 Nxb7 (11 Ng3 Nxc5 12 Qxb6 axb6 13 Nxf5 {13 dxc5 Bg6 14 cxb6 Nd5 =} 13...Na4 =) 11...Be6! 12 c4 Bxc4 13 Qxc4 Qxb2 14 Rc1 Qxb7 =.

Black to play Black has a useful disruptive response: 5...Bg4 6 f3 Bh5 7 Nge2 Nc6 8 Bf4 Qd7 9 Bg3 e6 10 Nb5 Rc8 11 c3 a6 12 Na3 Bxa3 13 bxa3 (NarmontasVysochin, Warsaw rapid 2007) 13...Bg6 gives Black the initiative. Without doubt White can improve his play, but in general Black’s plan

d) The game takes a peculiar course after 5 Bc4 a6 (this move has a prophylactic function, placing b5 under control; this fact can be used in a number 79

of different ways by Black) 6 Nge2 (D) (with the idea of Bf4) and now:

Black to play This is not an attacking thrust, but seeks a positional goal, namely to place the pawn on c4, so that it works in harmony with its colleague on d4; the knight can then drop back to c3 when Black kicks it with a pawn. However, this idea has found few followers, as it is rather timeconsuming. After 5...Qb6 6 c4 c6 7 Nc3 Black has a choice:

Black to play d1) 6...b5 7 Bb3 e6 represents a calm and safe approach by Black, advancing his queenside pawns with gain of time, and seeking normal development and counterplay. 8 Bf4 Qb6 9 d5 c5! (intending to ‘entomb’ the bishop by ...c4) 10 dxc6 (E.Berg-Pohjala, Västerås 2013) 10...Nxc6 =+.

e1) 7...g6 (as in Lahno-Motylev, Wijk aan Zee 2012) is a calm continuation.

d2) 6...Qc6!?

e2) 7...Bg4 leads to acceptable play: 8 Be2 Bxe2 9 Ngxe2 e6 10 0-0 Be7, Amonatov-Zhigalko, Moscow 2009. Here we have a normal ‘CaroKann⠳tructure. White’s assets are the pawn-pair on d4 and c4 and the knight on c3. However, the knight is more passively placed on e2 than it would be on f3. Furthermore, one set of bishops has been exchanged, which gives Black more freedom.

is the more adventurous approach: not needing to fear a pin by Bb5, Black attacks both g2 and c4. In the ensuing complications, Black will have to defend very carefully, as any mistake will lead to disaster. Therefore I wouldn’t recommend playing this way in a rapid game, but rather in a game with a classical time-limit. 7 Qd3 Qxg2 8 Rg1 Qh3 9 Qxh3 Bxh3 10 Nf4 Bf5 11 Nfd5 Kd8 (11...Nxd5!? and 11...Kd7!? are also quite possible) 12 Bf4 (12 Nxf6 exf6 =+) 12...Ne8 13 Ne3 Bg6 14 0-0-0 (Ranft-S.Kasparov, Guben 2013) 14...Nd7 leads to an unclear and very curious position, ideal for those who like exotic chess. Almost all the black pieces are huddled on their back two ranks, but he has absolutely no weaknesses and still an extra pawn. The fact that the queens have left the board eases Black’s defence.

e3) Black can react in gambit style with the more ambitious 7...e5!? (D).

e) White can bother the queen with 5 Nb5 (D).

White to play I suspect this is the strongest move, and may prove quite unpleasant for White. 8 dxe5 Ng4 9 Qe2 Bc5 10 Ne4 (the only move) 10...0-0 11 Nxc5 Qxc5 12 f4 Qb4+ 13 Qd2 Na6 14 a3 (Khruschov80

Sulava, European Ch, Ohrid 2001) 14...Qb3 is unclear.

with ample play for Black. 5 days later this idea was implemented...

5...c6 (D)

a2) 8 Qd2 g6 (this is the right line) 9 0-0-0 Bg7 10 Bf4 Qd8 11 Kb1 (Cawdery-S.Kasparov, Cape Town 2013) and now Black could have played 11...b5!?. Take a look at the h4-knight; I am not sure that it is well-placed. b) 6 Be2 looks quite modest. If it is followed up with simple development, then Black shouldn’t expect any serious trouble, but he must be alert to the possibility of a swift advance of the kingside pawns (g4, etc.). 6...Bf5 7 Ne5 Nbd7 (D) and then: White to play

6 Bd3 An unpretentious move, directly preventing ...Bf5. The next game features the somewhat popular idea 6 h3, planning to counter this move with a g4 advance. Here we should also examine two other bishop moves: a) 6 Be3

White to play

was used twice against the author in a tournament in Cape Town. Both games ended in victory for Black but during the opening White had been striving for an advantage. 6...Bf5 7 Nh4 (too straightforward; if White could gain the bishoppair so easily, the Scandinavian would be a dubious opening) 7...Be6 (D) and then:

b1) 8 Nc4 Qc7 9 g4 Be6 10 Ne3 Nb6 11 g5 Nfd5 (our standard manoeuvre is almost a universal panacea) 12 Nexd5 Nxd5 13 Nxd5 Bxd5 14 Bf3 e6 =+ Aderito-S.Kasparov, Luanda 2012. I then gradually exploited the dubious position of the g5pawn, without needing to resort to any drastic measures. b2) 8 Bf4 Nxe5 9 Bxe5 (9 dxe5 Qxd1+ 10 Rxd1 Nd7 11 Bd3 Bxd3 12 Rxd3 g6 =) 9...Qd7 10 0-0 e6 = Lindemann-S.Kasparov, Bad Wörishofen 2011. b3) In reply to 8 f4 I don’t see anything wrong with the prophylactic 8...h5 9 0-0 e6 =. We now return to 6 Bd3 (D):

White to play a1) 8 Bd3 Nbd7 9 Qd2 Nb6 10 Bf4 Qd7 11 Nf3 and now I chose 11...Bf5?! in Gwaze-S.Kasparov, Cape Town 2013. Usually I prefer a ‘central’ strategy (...e6), but in this case the fianchetto by 11...g6!? deserves attention; e.g., 12 0-0-0 Bg7 81

Black to play

Black to play

A natural reply; while f5 is denied to this piece, g4 is also a good square.

Another important moment. Despite missing some better options over the last few moves, I could now have retained an acceptable position by securing the knight’s central post.

7 Be3 Nbd7 8 h3 Bh5 9 Qe2 e6 10 0-0-0 Be7

16...0-0-0?!

10...Qc7!? is worth considering.

16...b5!? is preferable. I was concerned about 17 Nxe6 (17 c4 bxc4 18 Qxc4 Rc8 = yields White no benefits) 17...fxe6 18 Qxe6+ Ne7, but actually this is no more than unclear for White.

6...Bg4

After 10...Nd5, Vorobiov couldn’t get anything special: 11 Nxd5 cxd5 12 g4 Bg6 13 Ne5 Nxe5 14 dxe5 Bxd3 15 Rxd3 (Vorobiov-Bibik, Pardubice 2011) 15...Qxe5!?, with unclear play.

17 c4 Nc7

11 g4 Bg6 12 Bxg6 hxg6 13 Bg5 (D)

Usually the knight is passive here, as it lacks prospects. 18 Qe3 Qxe3 19 fxe3 Rdf8 20 Ne4 f6 Black is very solid, with no weaknesses and the queens exchanged. 20...Ne8 21 d5 exd5 22 cxd5 +=. 21 Nd6+ Kb8 22 e4 Ne8 The game is very close to equality after 22...e5. 23 c5?!

Black to play

23 e5 is better but very far from clear after 23...Nxd6 24 exd6 c5 25 dxc5 Nxc5 26 b4 Ne4.

13...Nd5 Again the prophylactic 13...Qc7!? is good. This move makes sense if Black plans to capture the white knight if it moves to e5, as he then avoids a pawn fork after dxe5 (attacking f6 and d6). There is also the idea of meeting a later Bf4 with ...Bd6. 14 Ne5 (or 14 Qe3 0-0-0 15 Bf4 Bd6, when 16 Ne5?! Nd5 17 Nxd5 exd5 18 Bg3 Nf6 gives Black the initiative) 14...Nxe5 15 dxe5 Nd5 16 Bxe7 Nxc3 17 bxc3 Qxe7 =.

23...Nxd6 24 cxd6 e5 ½-½ Black’s position is already slightly better as the passed d6-pawn is firmly blockaded, and the black pieces have healthy prospects. Conclusions Lines with d3 Combining a kingside fianchetto with d3 is unpromising for White. Besides the typical set-ups (...c6, etc.) Black can develop the knight on c6,

14 Ne4 Bxg5+ 15 Nexg5 Qf4+ 16 Kb1 (D)

82

intensifying his control over d4 (R.BarskiNeiksans). Once the knight is established there, White’s game can become unpleasant.

Town was obviously not ideal but in both cases I got more or less adequate play. The modest 6 Be2 is even more harmless.

In lines with d3, White sometimes puts the knight on e4, provoking an exchange on this square. This would lead to the d3-pawn or even the f3-pawn (see S.Sahl-S.Kasparov) shifting to the centre with all the benefits that entails. Black does best to ignore the knight and continue developing.

6 Bd3 was played in our main game. Since f5 is inaccessible now, Black puts his bishop on g4. After the predictable pawn advance with h3 and g4, this piece ends up on g6, where it joyfully meets eyes with its white counterpart. An exchange on g6 only serves to consolidate Black’s fortifications on the kingside, and to my mind there are no negatives for Black.

On d3, the pawn can prove a little weak and can come under pressure from the black pieces (E.Heinemann-S.Kasparov).

The next game features the plain-looking 6 h3. However, this move is quite popular, since there is an idea behind it: if Black replies 6...Bf5, White plans to chase this bishop by playing 7 g4, with a view to burying this piece on g6.

Lines with d4 An early g3 provides Black with additional resources, such as the disruptive ...Bg4 (JamrichS.Kasparov). Having caused disharmony in the white pawn-structure, the bishop can return to its natural role on the queenside, satisfied with a job well done.

Game 19 Neiksans – S. Kasparov Bergamo 2012

In reply to an early Bc4 I would generally recommend a set-up with ...a6 and play on the queenside (...e6, ...b5, ...c5, etc.). It looks logical because the bishop has to retreat, wasting time. With the bishop on the a2-g8 diagonal, the move ...e6 is more important for the defence than ...c6. Black can do without the latter move in this structure. Black also has some tactical options vs an early Bc4, based on the move ...Qc6 (seeRanft-S.Kasparov).

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 c6 6 h3 (D)

Several times I have faced 5 Nge2, putting the knight on a less-than-ideal square in order to bring the bishop to f4 with tempo. A typical set-up with ...c6 is possible, but Black can also play the ‘stubborn’ 5...Bf5, unafraid of Ng3, etc. The line demonstrated in Chrigui-S.Kasparov looks safe for White but even there, with the proper reaction, Black is alright. Even without taking into account concrete variations, from logical considerations a deployment with a bishop on f1 and a knight on e2 can hardly trouble Black. The misfortunes which may befall White (especially in a rapid or blitz game) if he doesn’t have specific knowledge can be observed in Solodovnichenko-S.Kasparov.

Black to play An unpretentious but rather venomous move. White as if invites the enemy bishop to move to f5, where it will be subjected to severe persecution. We shall examine several responses for Black. 6...Bf5 This is natural and probably best, though there are some rather critical lines that Black needs to know, in particular where White launches a kingside pawn advance with the intent of imprisoning the bishop on g6. However, White is taking quite a risk here because his own king is exposed and it is absolutely unclear how the hand-to-hand fight in the centre and on the kingside will finish. Meanwhile, if White settles for a reduction in the

Lines with d4 and Nf3 In this game we examined a number of unsophisticated bishop developments for White on move 6. I don’t think that 6 Be3 should give Black problems. My reaction in the two games in Cape 83

b1) 7 Bd3 was played in Narayanan SrinathS.Kasparov, Anogia 2014. My opponent was one of the winners of this tournament and he had prepared some interesting ideas for this encounter. Now the bishop sally ...Bf5 will entail damage to the pawn-structure. 7...Bg7 8 0-0 0-0 9 Re1 Nbd7 (9...Na6!? is interesting but it allows the creation of ‘splinters’, this time on the queenside: 10 Bxa6 bxa6 11 Na4 +=) 10 Bg5 Re8 11 Qd2 (now how should Black continue development?) 11...e6 (other moves don’t equalize: 11...e5?! 12 Bxf6 Qxf6 13 Bc4 +/– or 11...Nf8 12 Bxf6 Bxf6 {12...Qxf6 13 Ne4 +=} 13 Ne4 Qd8 14 Nxf6+ exf6 +=) 12 Rad1 Qf8 (taking control of h6 and escaping the gaze of the d1-rook) 13 a4 a5 14 Ne5 Nb6 (14...Qb4!?) 15 Bxf6 Bxf6 16 Ne4 Bg7 17 c4 f6 18 Nf3 f5! (D).

tension with the banal exchange Ne5xg6, then the situation will generally be roughly equal as it is quite difficult for White to maintain his bishoppair in any useful way. Conceptually, Black’s other option is to fianchetto his king’s bishop, rendering the move h3 almost a loss of tempo: a) 6...Nbd7 (D) places e5 under control.

White to play Then: a1) 7 Bg5 Nb6 8 Bd3 g6 9 Qd2 Bg7 10 0-0 0-0 11 Bf4 Qd8 12 Bh6 Be6 13 Rfe1 Bd5 14 Ne5 gave White a very slight initiative in Jaracz-Tiviakov, Bratto 2007. Nevertheless, Black’s fortifications are sturdy and White’s chances of overcoming the ‘expert’ from such a position are very slim.

White to play Houdini approves of Black’s strategy; indeed, the weakening of the dark squares – e5 and g5 – is not much of a problem as Black has an unopposed dark-squared bishop. 19 Nc3 e5 (maybe not best; there were several worthy alternatives: 19...Bd7!?, 19...c5!? or 19...Qb4!?) 20 Nxe5 Bxe5?! (20...Qb4! 21 b3 Be6 is more precise; the two bishops – the dark-squared one is especially strong – partly compensate for the missing pawn) 21 Rxe5 += (better is 21 dxe5 +/–) 21...Rxe5 22 dxe5 Be6 23 Qe3 Qb4 and at the moment of truth, it was now the turn of the Indian player to make a slight mistake: 24 Bf1?! Nxc4 25 Bxc4 Bxc4 += and the whole struggle lay ahead. The correct 24 Nb5! would have driven Black into a precarious situation: 24...Qxa4 25 Rc1 cxb5 26 Qxb6 Re8 27 cxb5 Qb3 28 Bf1 +–.

a2) 7 Bd3 g6 8 0-0 Bg7 9 b3 0-0 10 Re1 Re8 11 Bb2 a6 (this prophylactic move introduces an interesting plan; it turns out that activity on the queenside is quite appropriate) 12 Ne5 c5! 13 Bc4 e6 (now the bishop on c8 won’t suffer, as it intends to move to b7) 14 dxc5 Qxc5 = BulskiLanda, Copenhagen 2010. b) The immediate 6...g6 (D) is also viable, but in this case White also has the idea of planting his knight on e5:

b2) 7 Ne5 Nbd7 8 Bf4 Nd5 (D).

White to play 84

keeping control over d5) 11 Bd3 g6 12 0-0 Bg7 13 Rfe1 0-0 14 Rad1 Nb6 15 Ne3 Re8 16 Be5 Nbd7 17 Bh2 Nb6 gave Black ample play in StrikovićS.Kasparov, Casablanca 2015. The key point d5 is guarded, and White can’t easily play c4. 7...Bg6 8 Ne5 Nbd7 9 Nxg6 Baffling and unclear complications arise after 9 f4 (D). White to play 9 Nxd5 Qxd5 10 Qf3 Qxf3 11 Nxf3 Nb6 12 Bd3 Be6 13 0-0 Bd5 14 Nd2 Bg7 15 c3 0-0-0 16 Rfe1 Rhe8 brought Black very close to equality in Hrisanthopoulos-Tiviakov, Vrachati 2011. It’s true that it is hard for Black to create winning chances against a weaker opponent in a line like this. Then again, in which opening is that an easy matter? Even in the Sicilian Defence, White has some very safe options (for instance, some lines with c3, and the Bb5(+) lines) where, if he doesn’t take many risks and avoids substantial mistakes, he won’t lose. Having said all that, Tiviakov did in fact go on to win this game.

Black to play The true value of this whole line may lie in their accurate evaluation. However, in practice White normally doesn’t carry out this ‘rash’ pawn advance, and I have found only one game in which Tiviakov faced this line. 9...e6 10 Bg2 (if White goes berserk with 10 h4 Nd5 11 f5? {11 Bg2 f6 12 h5 Bxc2 is unclear} 11...exf5 12 h5 00-0, then Black’s chances are better as the white king may find itself in danger) 10...Qc7 and now:

We now return to 6...Bf5 (D):

a) 11 Qe2 Nxe5 12 dxe5 (Sulskis-Tiviakov, Neustadt an der Weinstrasse 2009) and now the natural 12...Nd5 13 Bxd5 cxd5 leads to approximate equality. b) 11 h4 is met by 11...Bb4, when e4 becomes available to the black pieces.

White to play

c) 11 0-0 0-0-0 12 Qe2 (D).

7 g4 This is the reason why White played h3. A bishop exchange can hardly give Black problems: 7 Bd3 Bxd3 8 Qxd3 Nbd7 9 Bg5 e6 and it is unclear how White might drum up much play, Inarkiev-Tiviakov, Moscow 2009. Recently 7 Ne5!? has also been used; e.g., 7...Nbd7 8 Nc4 Qc7 9 Qf3 (the queen deploys with tempo, as it attacks the bishop) 9...Be6 (9...e6 10 Bf4 Qd8 11 0-0-0 gives White the initiative) 10 Bf4 Qc8 (perhaps 10...Qd8!? is more accurate,

Black to play 85

12...Qb6 (in Linchevsky-S.Kasparov, Pardubice blitz 2012, I didn’t get to the heart of the position and lost the game after 12...Nxe5? 13 dxe5 Bc5+ 14 Kh1 Nd7 15 f5 +/–) 13 Nc4 (13 Be3?! Nxe5 gives Black an edge after 14 fxe5 Qxb2 =+ or 14 Na4 Nf3+ 15 Bxf3 Qb4 =+) 13...Qa6 is unclear; nothing terrible is in sight for Black.

27 Red1 Na8 28 Be4 Kf7 29 d5 cxd5 30 cxd5 e5 (D)

While 9 f4 is critical, I doubt that you are likely to face a really well-prepared opponent in this line since it would require them to learn some long and unclear variations by heart, in a line that they are not going to reach very often or with much certainty against any particular opponent.

White to play

9...hxg6 10 Be3 e6 (D)

A logical move; so far I have played reasonably well. 31 h5 Ke7 31...gxh5 is interesting but risky: 32 d6 hxg4 33 Bf5 g6 34 Bxd7 Rxd7 +=. White is an exchange up, has a passed pawn, and Black’s queenside weaknesses remain. 32 d6+ Rxd6 33 Rxd6 Rxd6 34 Rxd6 Kxd6 35 Bxb7 Nc7

White to play

For those of you who are solely interested in the opening, I recommend moving on to the next game, or whichever section of the book you wish to study next. If first you would like to improve your general chess knowledge, please stay with us here. This ending is very instructive.

White has obtained the bishop-pair, but not for free! His kingside pawn-structure is somewhat compromised. 11 Qf3 Nd5 12 Nxd5

It is well-known that a bishop is stronger than a knight in a fight on two flanks. But here the realization of this slight positional advantage is complicated by the fact that the promotion square of a passed white h-pawn on is a dark square, while his bishop is light-squared.

12 Ne4? Qb4+ –/+. 12...Qxd5 13 Bg2 Qxf3 14 Bxf3 Nb6 15 0-0-0 Nc4 16 Bf4 Bd6 With a number of simple and accurate moves, Black has exchanged almost all the pieces and is approaching full equality. However, the position is far from dead, and the struggle was destined to continue for many hours into a lengthy endgame.

35...Nb6 is also interesting, but after 36 b4 Nxa4+ 37 Kb3 axb4 38 Kxa4 gxh5 39 gxh5 Kc5 it’s hard to determine the outcome; it could well be a draw.

17 Bg3 Rd8 18 h4 Ke7 19 b3 Bxg3 20 fxg3 Ne3 21 Rd2 Nd5 22 c4 Nb6

36 b4!

22...Nc7!?.

36 hxg6?! is clearly weaker, as 36...Kc5 builds an impenetrable fortress on the dark-square complex.

23 Kc2 Rd7 24 Kc3 Rhd8 25 a4 a5 26 Re1 f6

36...axb4+ 37 Kxb4 gxh5 38 gxh5 (D)

Otherwise the rook will go to e5, attacking the queenside pawns. 86

Since we have identified the point where White missed a win, I shall leave the rest with just minimal comments. 40...Ke5 41 Kd2 Kf6 42 g4 fxg4 43 Ke3 Ke5 44 Bxe4 Na6 45 Bb7 Nb4 46 a6 Nxa6 47 Bxa6 Kf6 48 Bd3 Kf7 49 Bh7 Kf6?! A simpler way to draw is 49...g6 50 h6 (after 50 hxg6+ Kg7 it’s a draw, as we have already seen above) 50...Kf6 (with the idea ...Kg5xh6) 51 Bg8 (51 Kf4 g3 52 Kxg3 Kg5 =) 51...g5 52 Bh7 g3 (now nothing can prevent ...g4, after which the king will go to g5 and h6) 53 Kf3 g4+ 54 Kxg3 Kg5 =.

Black to play 38...f5 The pawn ending after 38...Nd5+? 39 Bxd5 Kxd5 40 a5 Kc6 is losing:

50 Kf2?!

a) The hasty 41 Kc4? is a mistake in view of 41...f5! 42 a6 Kb6 43 Kd5 e4 44 Kd4 Kxa6 45 g4 fxg4 46 Kxe4 Kb6 47 Kf4 Kc6 48 Kxg4 Kd7 49 Kg5 Ke8 50 Kg6 Kf8.

White can try 50 Bg6!?, when 50...Ke7?? loses to 51 Kf2 Kf8 52 Bh7 g6 53 h6 +–, since the black king is too far from g5. However, Black draws by 50...Kg5; e.g., 51 Ke4 Kh4 52 Kf4 g3 53 Ke3 Kh3 54 Be4 Kg4.

b) 41 g4! +– is an important nuance. Now the white king has time to annihilate the enemy pawns on the kingside: 41...Kb7 42 Kc4 Ka6 43 Kd5 Kxa5 44 g5 fxg5 45 Kxe5 Kb5 46 Kf5 Kc5 47 Kxg5 Kd6 48 Kg6 Ke7 49 Kxg7 +–.

50...g6 51 hxg6 g3+ 52 Kxg3 Kg7 53 Kg4 Kh8 ½-½ Don’t rush to criticize the players for the inaccuracies in this endgame. We really put in a lot of effort, but this event was two gruelling rounds per day, often against very strong opponents... Physical strength isn’t endless.

39 a5 e4 (D)

Fianchetto with 5...g6 To conclude the chapter, let’s discuss a ramification which is quite popular nowadays – the immediate fianchetto with 5...g6. In our first two games, both played at a very high level, White used the knight raid 6 Nb5 – not for attacking purposes, but so as to advance the cpawn. Then in Game 22 we examine 6 g3, while Game 23 deals with 6 Bc4 and 6 Ne5.

White to play 40 Kc3? It looks like the prosaic 40 a6 would have led to victory: 40...Nxa6+ 41 Bxa6 Ke5 42 Kc3 f4 43 gxf4+ Kxf4 44 Bc4 Kg5 45 Bf7 Kf6 (45...Kh6 46 Bg6! +– denies the king access to h8) 46 Bg6 Ke7 47 Kd4 Kf8 48 Bh7 g5 49 h6! (cutting the black king off from the h-file; after 49 hxg6? Kg7 50 Kxe4 Kh8 51 Kf5 Kg7 52 Kg5 Kh8 White is unable to force the black king out of the safe corner).

Game 20 Bacrot – Tiviakov Bundesliga 2012/13 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 g6 (D)

87

Or immediately 7 c4 (D).

White to play Personally I think that in this line Black feels more uncomfortable than in case of the more ‘central’ play following 5...c6. However, this is a subjective matter, so I invite you to form your own opinion.

Black to play Then: a) 7...c6 8 Nc3 Bg7 9 Be2 0-0 (a modest but noteworthy approach) 10 0-0 Bg4 11 Be3 Nbd7 12 Qb3 Qc7 13 Rfd1 Rfe8 14 d5?! Bxf3 (Black must damage White’s structure) 15 gxf3 (Wei YiBenidze, Turkish League, Kocaeli 2014; 15 Bxf3 Ne5 16 Be2 Neg4 =+ turns out to be rather unpleasant for White) and now 15...Nh5 takes control of f4.

The following cavalry raid is quite popular, the point being to make way for the c-pawn. 6 Nb5 With 6 h3 Bg7 7 Bc4 a6 8 0-0, White opts for a modest setup, not trying to grab the centre with his pawns. Then:

b) 7...Bg7 8 h3 0-0 9 Be2 c5! (not wasting time on ...c6, Black immediately attacks White’s fortifications in the centre) 10 d5 (10 dxc5 Qxd1+ 11 Bxd1 Na6 12 Be3 is unpromising as long as Black plays 12...Nd7! = rather than 12...Ne4?! 13 Bc2 Nexc5 14 0-0-0 +=) and now 10...a6 was played in Motylev-Tiviakov, Wijk aan Zee 2012. I would prefer first playing 10...Ne4!, intending only then to determine the knight’s intentions by playing ...a6. After 11 Nd2 Nd6 12 Nc3 e6 the game is approximately equal; White may get a vanguard on d5, but Black may also have one of his own on d4 (after the exchange on e6).

a) 8...Nc6 9 a3 0-0 10 Re1 Rd8 11 Be3 Be6 12 Bxe6 Qxe6 13 Qe2 Qf5 14 Rad1 (L.DominguezJ.Polgar, FIDE World Cup rapid, KhantyMansiisk 2011) 14...e6!? and after the exchange of the light-squared bishops, Black can sigh with relief. b) 8...b5 9 Bb3 Bb7 10 Ne5 0-0 11 Bf4 Qd8 12 Re1 Nbd7 13 Qe2 c5 14 dxc5 (14 Bxf7+!? Rxf7 15 Nxf7 Kxf7 16 Qe6+ gives White the initiative) 14...Nxc5 15 Rad1 (L.Dominguez-J.Polgar, FIDE World Cup rapid, Khanty-Mansiisk 2011) and after the careful 15...Qe8 Black has nothing to worry about.

We now return to 7 Bf4 (D):

After the text-move (6 Nb5), the queen has to move somewhere. In this game we analyse the ‘universal’ option, which avoids immediate complications but at the same time is a little passive. 6...Qd8 In the next game, we shall examine the more active 6...Qb6.

Black to play 7...Na6 8 c4

7 Bf4 88

25 h5 Rd8 26 Rxd8+ Rxd8 27 Bf4 Nd3 28 Rd1 Kg7 29 Be3 c5 30 Bd2 (Vocaturo-Tiviakov, Wijk aan Zee 2012) and the cautious 30...Rd7 brings the rook under the guard of the bishop, with about equality; e.g., 31 Bc3 f6 32 exf6+ exf6.

A modest set-up with a pawn on c3 can hardly claim an advantage: 8 Na3 c6 9 c3 Nc7 10 Nc4 Bg7 11 Be5 0-0 12 a4 Ncd5 13 a5 Bh6! (escaping the pin; it is a reasonable decision not to exchange the fianchettoed bishop) 14 Be2 Nh5 = HessTiviakov, Groningen 2011. Black plans ...Nf4 or even ...f6 with ample play.

11...Re8 12 h3 c6 13 Nc3 (D)

8...Bg7 9 Be2 Or 9 h3 0-0 10 Be2 c5 11 0-0 (11 d5 could be met by 11...Ne4; e.g., 12 Qc1 e6 13 dxe6 Qa5+ 14 Nd2 Bxe6 with counterplay) 11...cxd4, and now White must decide with which knight to capture: a) 12 Nfxd4 Ne4 13 Bf3 (Grandelius-Tiviakov, Wijk aan Zee 2013) 13...e5 14 Bxe4 exf4 and Black’s chances are no worse. The dark-squared bishop is very powerful and White has no resources to restrain it.

Black to play White has arranged his pieces very harmoniously, having established control over the d5-square. If he does nothing radical, Black risks being suffocated. In particular, I don’t see where the queen should be put so as to employ the rook by ...Rad8. The a6-knight also lacks obvious prospects at present.

b) 12 Nbxd4 Nh5 13 Bh2 Nc5 14 Qd2 b6?! 15 Rad1 Bb7 16 b4! (the Ugandan master plays with inspiration, or maybe he was simply well-prepared) 16...Ne4 17 Qe3 Qc8 (Emojong-Tiviakov, Cape Town 2013) and after 18 Ne5 +/– Black would have to work hard, as he stands considerably worse.

13...Ne4

We now return to 9 Be2 (D):

Exchanging at least one piece is necessary. 14 Qb3 Bxd4 It’s a pity to give up this excellent bishop, but it is the right decision. 15 Nxd4 Qxd4 16 Be3 Qd7 16...Qe5!?. 17 g4 Nxc3 18 Qxc3 Be6 19 Rad1 Qc7 20 Bh6 f6 21 Bf1 Bf7 22 g5 (D)

Black to play 9...0-0 10 0-0 Bf5 11 Re1 It is possible to activate the queen at this point: 11 Qb3 c6 12 Nc3 Qb6 13 Qxb6 axb6 14 a3 Rfd8 15 h3 Be4 (putting pressure on the d4-pawn) 16 Rfd1 h6 17 Rac1 g5 18 Bg3 Bf5 19 Na4 Nd7 20 Kf1 (better is 20 Rd2) 20...Rdc8 (by 20...Nac5! Black could have even grabbed the initiative: 21 dxc5 Rxa4 22 Rd2 Re8 23 cxb6 Nc5 24 Bd1 Bd3+; it’s strange that this was missed by both players) 21 h4 g4 22 Ne5 Bxe5 23 dxe5 Nac5 24 Nxc5 Nxc5

Black to play 89

White exerts pressure on the dark squares. The f6pawn is the last defensive bastion on the approach to the black king.

White threatens Bxh5. Black will have great difficulties in the bishop ending due to zugzwang possibilities.

22...Nc5 23 gxf6 e5

51...Ke6 52 Bc4+ Ke7 53 a5 Bc2 54 Kf4 Kf6 55 b5! (D)

Lacking resources to cover the weakened g7square, Black tries to use the f6-pawn (now a white one!) as a shield. 24 Qe3 Ne6 24...b6!?. 25 Qxe5 Qxe5 26 Rxe5 Rad8 27 Rxd8 Nxd8 28 Re7 Rxe7 29 fxe7 Ne6 30 Be3 a6 31 Bb6 Be8 32 Bd8 Kf7 (D)

Black to play Impressive! But is it enough to win? 55...axb5 56 Bxb5 Bf5 56...cxb5? loses because the e4-square is controlled by the white king and the bishop won’t be able to stop the passed pawn after 57 c6 bxc6 58 a6 +–. White to play

57 Bf1 Bd7 58 a6 bxa6 59 Bxa6 Be6 60 Bf1 At a glance this might appear tenable for Black, but in the long run he will have to make concessions due to zugzwang.

Black has managed to exchange pieces and gradually neutralize the pressure. The only thing left is to surround and destroy the e7-pawn. It’s still unclear exactly why Tiviakov went on to lose...

60...Bc8?! After 60...Bd5 61 Bd3, Black will be zugzwanged in a line such as 61...Bb3 62 Be4 Ba4 63 Bg2 Bb5 64 Bh1 Ba4 65 Ke4 Bb5 66 f4 Ke6 67 Ke3 followed by Be4 and bringing the king to the queenside.

33 Kg2 Bd7 34 Kg3 Ke8 35 h4 Nxd8 36 exd8Q+ Kxd8 37 Kf4 Ke7 38 Kg5 Bf5 39 c5! Fixing the queenside structure. In endings with same-coloured bishops it is advisable to put your pawns on squares of the opposite colour.

61 Bg2 Bd7?! 62 Bf3 Be8 63 Be4 Bd7 64 Bg2 Be8 65 Bh3 Bf7 66 Bd7 Bd5 67 Be8 +–

39 Kh6 is not deadly because after 39...Ke6 capturing by 40 Kxh7? is bad due to 40...g5+ =+.

The dance of the bishops is over, and Black is in zugzwang.

39...h5 I shall not analyse the ending in detail, and just point out some important moments.

67...Kg7 Alas, the bishop can’t move; for instance, 67...Bg2 68 f3 and the c6-pawn perishes.

40 b4 Kf7 41 Bc4+ Ke7 42 a3 Bb1 43 Be2 Ke6 44 Bd1 Bd3 45 Bb3+ Ke7 46 a4 Bb1 47 Bc4 Bc2 48 Be2 Ke6 49 Bc4+ Ke7 50 Bf1 Bf5 51 Be2

68 Kg5 Be4 69 f4 Kh7 70 Kf6 Kh6 71 Ke5 Bg2 72 Kd6 1-0 90

Black’s defence looked quite reasonable, but for the whole game, White was the only side with any realistic winning chances.

c6 8 a5 (this move is very natural: grabbing space with tempo and kicking back Black’s most powerful piece) 8...Qd8 9 Nc3 Bg7 10 Bc4 0-0 11 0-0 Bg4 12 Be3 (D) and then:

Conclusions In this game, we examined the cautious 6...Qd8. If White pushes the c-pawn, it makes sense to advance it two squares. Set-ups with c3 look solid but should hardly trouble Black. However, after the vigorous c4 Black often has problems equalizing (see Grandelius-Tiviakov and Emojong-Tiviakov). In our main game, Tiviakov fought hard but the long tiring struggle ended unsuccessfully for him.

Black to play

In the next game, we shall examine, instead of retreating by 6...Qd8, the more combative idea of moving the queen along its third rank, viz. 6...Qb6.

c1) 12...Na6!? is worth noting as the knight is quite well placed in front of the pawn and has prospects of moving onward to good pastures. The main negative is that White has the opportunity to give Black doubled isolated a-pawns by 13 Bxa6 bxa6. These structures can be hard to assess, as there are several nuances. The b-file is opened for the black rook, while the white a5-pawn is fixed on a dark square (the same colour of the e3bishop). Black also has the bishop-pair. It certainly adds up to at least some compensation (and maybe more) for the damaged structure.

Game 21 Leko – Caruana Tashkent 2012 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 g6 6 Nb5 Qb6 (D)

c2) 12...Nd5 13 Bg5 h6 14 Bh4 g5 15 Bg3 e6 16 h3 Bxf3 17 Qxf3 Bxd4 18 Rad1 (KülaotsKovalenko, Liepaja 2014) and after 18...Qf6 White has compensation for the pawn, but no more than that. d) 7 c4 looks natural, with the familiar idea of uniting the c- and d-pawns before dropping the knight back to c3. 7...c6 (D) and then: White to play 7 Na3 Some other options: a) 7 Bf4 Nd5 8 Be5 f6 9 Bg3 a6 10 c4 axb5 11 cxd5 with unclear play. b) 7 Be2 c6 8 Na3 Bg7 9 0-0 0-0 10 c3 (as usual, putting the pawn on c3 is safe but too modest to give Black problems) 10...Qc7 11 Bg5 Bf5 12 Qc1 Nbd7 = Andreikin-Caruana, Baku 2014.

White to play d1) The knight doesn’t have to retreat immediately, as White can grab space first by 8 c5

c) A speedy march of the a-pawn is possible: 7 a4 91

d22) 9 Be2

Qd8 9 Nc3. Now Black, if he wishes, can postpone ...Bg7 for a while:

0-0 10 0-0 Bg4 11 a3 Nbd7 12 Be3 Qc7 13 h3 Bxf3 14 Bxf3 Rfe8 15 Qb3 e5! (D).

d11) 9...Be6 (it makes sense to take control of the key d5-square, while not allowing the white bishop to c4) 10 Be2 Bg7 11 0-0 0-0 12 Re1 b6 (this undermining move is the scourge of White’s c4-c5 advance in many openings; if White lacks resources such as b4 and a3 to support the advanced post on c5, he may have to release the tension in an uncomfortable manner) 13 cxb6 axb6 14 Ne5 Qc8 15 Bf3 Rd8 16 Be3 Nd5 = Rohl-Rakhmanov, Caracas 2014. Black has no weaknesses which also means no problems. d12) 9...Bg7

White to play

10 Bc4 0-0 11 0-0 Bg4 12 Be3 b5! 13 cxb6 axb6 with an acceptable position for Black, BartelMotylev, European Rapid Ch, Warsaw 2012. After the destruction of the c5-pawn that restrained Black, his life is easier and happier.

This strike is easily predictable because Black doesn’t have any other counterplay; for instance, ...c5 would significantly weaken the light squares to the benefit of White’s f3-bishop. 16 d5 (at least this is consistent; 16 dxe5 doesn’t look good since after 16...Nxe5 Black again has strong play on the dark-square complex, while the f3bishop is restrained by the c6-pawn) 16...e4 17 Be2 Bf8 (if Black manages to exchange bishops by ...Bc5, he will start to take over the initiative as then the e2-bishop will remain alone; therefore White’s next move prevents this) 18 Na4 was played in Tarlev-Tiviakov, Erbil 2013. Now Black had a chance to get perfect play after the pawn exchange 18...cxd5 19 cxd5 Qe5. As they say, “centralization is a safeguard against all troubles” because 20 Qxb7? fails to 20...Nxd5! 21 Qxd7 Bd6 –+.

d2) With 8 Nc3 Bg7 (D) White keeps his structure flexible and doesn’t shoulder any extra responsibility by pushing his pawn to c5.

We now return to 7 Na3 (D):

White to play Now: d21) 9 Qb3 offers an exchange of queens, which may be a reasonable policy if White doesn’t mind a draw. Nevertheless, the goal is not guaranteed: 9...0-0 10 Be3 Na6 11 Qxb6 axb6 12 Rc1 Bf5 13 a3 Ng4 (the ‘pestering’ begins before White has even finished his development) 14 Be2 Nxe3 15 fxe3 e5!? (slicing through the dark squares, which are already a sore spot for White) 16 d5 (White stands stably worse after 16 Nxe5 Bxe5 17 dxe5 Nc5) 16...Nc5 17 Rd1 Nb3 18 Nd2?! (18 e4) 18...Bc2 19 Nxb3 Bxb3 20 dxc6 bxc6 21 Rd6 Rac8 22 Ne4 Rfd8 23 Rxd8+ Rxd8, BagiKovalenko, Zalakaros 2014. Black has the bishoppair and, accordingly, better play.

Black to play 7...c6 8 Nc4 Qd8 9 g3 Be6 10 Qe2 Bg7 11 Bg2 00 12 0-0 Qc8

92

So far, White’s main achievements are his control over e5 and the fact that he has avoided exchanges that would free Black’s game.

Black has managed to secure reasonably solid strongpoints for his knights. His bishop is active, and his pawns well-placed.

...Bxc4 is inappropriate because it provides White with the bishop-pair ‘for free’. Black would like to play ...Bh3, but won’t get the chance.

31 Qd2

13 Ng5 Bd5 14 Bh3 Bxc4 15 Qxc4 (D)

31...Qxd2 32 Rxd2 Red8 33 Red1 Rxd2 34 Rxd2 Ne4

Better is 31 Rdd1.

34...Nf5 is good and natural: 35 g4 Bh6! =. 35 Rc2 Nc5 36 Rd2 Ne4 37 Rd7 Nc5 38 Rd1 f6 This slightly weakens the light squares g6 and e6. 38...Nf5 looks safer, with approximate equality after 39 g4 hxg4 40 hxg4 Ne7 41 Bd6 Bf8. 39 Be3 Nf5 40 Bxc5 Rxc5 41 Rd8+ Kh7 42 Bf1 Rc7 43 Ra8 e5 44 Rxa4 (D) Black to play Now the two bishops ensure long-term pressure; White is somewhat better. 15...Nbd7 16 Qe2 Qd8 17 c4 h6 18 Nf3 Re8 19 Rd1 e6 20 Bd2 a5 21 Bc3 Qc7 22 Rac1 a4 23 Bg2 Rad8 24 Rd3 Ng4 24...c5!?. 25 Re1

Black to play

The rash 25 h3 would be met by the surprising 25...e5!; e.g., 26 dxe5 (26 hxg4? exd4 –/+) 26...Ngxe5 27 Nxe5 Nxe5 28 Re3 Re6 29 Re1 (29 f4 Nd3) 29...Rde8 with good play for Black.

White has won a pawn, but Black’s counterplay (and opposite-coloured bishops) makes it hard to exploit this material advantage.

25...h5 26 h3 Nh6 27 Bd2 c5

44...e4 45 Nd2 Nd6

27...Nf5 28 Bg5 Bf6 29 Bxf6 Nxf6 =.

45...e3!? 46 Ne4 h4 also gives Black counterplay.

28 Bg5 Rc8 29 Bf4 Qa5 30 dxc5 Nxc5 (D)

46 Ra5 f5 47 Rd5 Nc8 48 b4 Bc3 49 a3 Ne7 50 Rd6 Bb2 This is typical ‘harassment’ of the pawns from behind. With opposite-coloured bishops it’s rather effective, since if the pawns advance to squares where they can’t be attacked by the enemy bishop, they will obstruct their own bishop, and in some cases allow a firm blockade to be set up. 51 Nb1 Not 51 a4?! Ba3 52 Rb6 Nc8 53 Rb5 Nd6 54 Rb6 Nc8 =.

White to play 93

51...Nc6 52 c5 Re7 53 Kg2 h4 Also a typical motif, slicing through the dark squares. 54 b5 54 gxh4!?. 54...Nd4 55 Nd2 hxg3 56 Kxg3 Bc3 56...Ne6 57 Nc4 Bc1 +=. Black to play 57 Nb1 Bb2 58 Nd2 Bc3 59 Nb1 Bb2 60 Nd2 ½½

As in other lines, the fianchetto is a reasonable option for White.

Conclusions

6...Bg7 7 Bg2 c6

In our main game, White spent several tempi on the intricate manoeuvre Nc3-b5-a3-c4, but still he managed to create pressure after Black exchanged his bishop for the wandering knight. Caruana had to work hard, but he did hold the draw.

Let’s have a close look at the peculiar manoeuvre 7...Qa6!? (D), using the fact that Black’s third rank isn’t blocked by a pawn on c6. This ‘geometrical’ idea looks attractive, as White can’t castle for the time being.

The pawn march a4-a5 didn’t yield tangible benefits in Külaots-Kovalenko. After playing c4 White can either continue the advance to c5, restricting Black further but granting him use of the d5-square (BartelMotylev), or else keep his pawns on d4 and c4 (Tarlev-Tiviakov), retaining the possibility of pushing on to either d5 or c5. As in the lines after 5...c6, White can also opt for a fianchetto versus 5...g6. But while 6 g3 is acceptable here, I don’t feel that it gives Black any special problems. Note that Black has an extra resource – ...Qa6 – that doesn’t exist when the pawn is already on c6. Anyway, the counterfianchetto 5...g6 6 g3 is the subject of our next game.

White to play Now: a) 8 Qe2 is of course possible, but the exchange of queens spoils White’s chances of achieving any appreciable initiative and leads to approximate equality. The simplest and best decision is to trade queens; e.g., 8...Qxe2+ 9 Nxe2 Bf5 10 c3 Be4 = Cheparinov-Enchev, Bulgarian Ch, Kozloduy 2014.

Game 22 Van Kampen – Tiviakov Hoogeveen 2011

b) If White kicks the queen away by 8 Bf1 Qa5 and then comes back with 9 Bg2

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 g6 6 g3 (D)

(interestingly, this position arose via the moveorder 3...Qa5 4 g3 g6 5 Bg2 Bg7 6 d4 Nf6 7 Nf3 in Schmittdiel-Wahls, German Ch, Gladenbach 1997), Black can of course repeat moves. However, he isn’t obliged to do so, but can continue developing: 9...0-0 10 0-0 Rd8 11 Bd2 94

c) 8 Bf4 c6 and then:

to prove dangerous with his own king in the centre, and without open h- or g-files. After 14...Bd5 Black’s chances are at least no worse. Queenside castling is unlikely as it may lead to the loss of the a2-pawn.

c1) 9 a4

c22) 10 0-0-0

0-0 10 a5 is an interesting plan, seeking to trap the black queen, or at least to force some positional concessions from the threat to do so. 10...Be6 11 Bf1 Bc4 12 Bxc4 Qxc4 13 Qe2 Qxe2+ 14 Kxe2 Na6 (Black is OK here; the a6-knight feels quite comfortable, blocking the enemy vanguard) 15 Ra4 Rfd8 16 Rd1 Rac8 17 Be5 Nd5 18 Bxg7 Kxg7 = Oleksienko-Rakhmanov, Abu Dhabi 2013.

(D) is of course possible, but the white king runs the risk of getting attacked.

Nc6 12 Ne2 Qa6 13 Re1 Bg4 with pressure on d4 and the counterstrike ...e5 in prospect, CaspiMotylev, European Clubs Cup, Eilat 2012.

c2) 9 Qd2 0-0 (D) and then:

Black to play Meanwhile the fianchettoed bishop on g2 doesn’t fit in well with an all-out attack on the black king; for that purpose it would be better placed on the b1-h7 or d1-h5 diagonal. 10...b5 11 a3 b4! (Black operates vigorously) 12 axb4?! Qa1+ 13 Nb1 a5! 14 b5 cxb5 (14...Be6!?) 15 Ne1 Ra6 16 Bxb8 Be6 –/+ Wei Yi-Miroshnichenko, Turkish Team Ch, Kocaeli 2014. A picturesque position! Black carries out a powerful attack ‘for free’ since he can, if he wishes, regain material by ...Rxb8 or ...Ba2.

White to play c21) 10 Bh6 seems like a logical attempt to exchange the dark-squared bishops. However, if Black reacts appropriately, then he is perfectly safe: c211) 10...Be6?! is a sluggish response and led to trouble in Azarov-Bodiroga, Rijeka 2010: 11 h4 Qa5 12 h5 Nxh5 13 Bxg7 Kxg7 14 Ne5 f6 15 g4 fxe5 16 gxh5 Rf4 17 hxg6 hxg6 18 0-0-0 Nd7 19 Qe3 +/–. Note how swift White’s attack was. I don’t think it’s because Azarov is a fan of Mikhail Tal, but rather that he is a ‘universal’ player, and considered this to be the most appropriate way to handle the position.

We now return to 7...c6 (D):

c212) The computer suggests the ‘cold-blooded’ 10...Bxh6!? 11 Qxh6 Bf5. As mate does not seem in the offing, Black will win some time if the queen finds nothing better than returning to d2 to protect the c2-pawn.

White to play 8 0-0 0-0 9 Re1

c213) 10...Bf5 11 Bxg7 Kxg7 12 Nh4 Be6 13 f4 Nbd7 14 f5 was played in Andriasian-Bodiroga, Rijeka 2010. However, White’s attack is unlikely

After 9 h3 Black can reach an acceptable position by putting the bishop on d5: 9...Be6 10 Re1 Bd5. 95

We shall come to see how a bishop exchange on d5 differs from, e.g., an exchange on f3. Now:

Black has failed to equalize. No pieces have been exchanged, while White has two pawns in the centre with all the comfort that it entails. In particular, the black pieces lack access to the d5square.

a) After 11 Nxd5 cxd5 the pawn shifts from c6 to d5 and finds itself in the centre of the board. I remind you that hitherto only a white pawn was on one of the four centre squares. This white pawn on d4 (or in the best case, a pawn-pair on d4 and c4) typically provides White with a slight spatial plus in openings like the Caro-Kann and the Scandinavian. Usually Black must fight for equality with an eventual ...c5 or ...e5. Here matters are different. Play may continue 12 Bf4 Qd8 13 Ne5 Nc6. In this structure, the pawn advance c4 seeks to increase the role of the g2bishop, but also has some negative factors, such as the appearance of an isolated pawn on d4, which is already under fire. More reserved play with c3 may provoke a minority attack with ...b5-b4 at some point.

18...bxc4 19 bxc4 Qb7 20 Nh4 Be6 21 Rb1 21 Qc2 +/–. 21...Qc8 22 Qc1 h6 23 Nd3 g5 24 Nf3 Bf5 25 Rb3 Rfe8 26 Rd1 Be4 (D)

b) 11 Bf4 Qd8 12 Qd2 e6 13 Rad1 Nbd7 14 b3 (planning the c4 advance) 14...b5 15 a4 a6 16 axb5 axb5 17 Nxd5 cxd5 (this version of the exchange is even more favourable for Black) 18 Ne5 (I.Popov-K.Grigorian, Erevan 2014) and here the aggressive 18...Ra2 19 Ra1 Rxa1 20 Rxa1 Nxe5 21 Bxe5 Qb6 22 Qb4 Ne4 provides counterplay, and the situation is close to equality.

White to play Black has weakened his king’s shelter by ...h6 and ...g5, while the knight on a6 has few prospects, and the b-file is in enemy hands. 27 Nde1

We now return to 9 Re1 (D):

27 Qb2 +/–. 27...Qf5 28 Qc3?? A strange mistake as Black’s reply is hardly a deeply concealed idea. 28...Nc5 –/+ Of course, because the d1-rook is unprotected. About-turn! 29 Rb4 a5 30 Rb2 Na4 31 Qc1 Nxb2 32 Qxb2 Bf8

Black to play

It would be more reasonable to exchange the darksquared bishops by 32...Ng4 33 Bxg7 Kxg7 –/+ as the difference in their activity is evident. The e5bishop keeps the centre under control, restraining the black rooks.

9...Na6 10 Bf4 Qd8 11 Qd2 Bf5 12 a3 Nc7 13 Be5 Qd7 14 Na4 Na6 Not letting the knight reach c5. 15 Rad1

33 h3 Qe6 34 Qc3 a4 35 Nd3 Nd7?

It’s strange that Van Kampen didn’t play the natural 15 c4 +=, reinforcing the centre.

Now Black goes wrong. It is unclear what Tiviakov had in mind, as White’s reply is also not too intricate.

15...Rad8 16 b3 b5 17 Nb2 Qc8 18 c4 96

The idea of combining Bf1 with a quick march of the a-pawn (a4-a5) is not very dangerous either as Black has time to bring his bishop into play with ...Be6-c4 (Oleksienko-Rakhmanov).

36 Nxg5! hxg5 37 Bxe4 Nxe5 38 dxe5 Bg7 38...Qxh3? 39 Bxc6 +/–. 39 Re1 Qd7?! 40 Qc2 Qxh3 41 Bxc6 Rf8 (D)

With opposite-side castling it’s unclear what the bishop is doing on g2, since it can’t participate in a kingside attack, whereas both black bishops are pointing in the right direction (Wei YiMiroshnichenko). In the main game, where ...c6 was played, White achieved some advantage after establishing pawns on d4 and c4. In our final game we shall take a look at the ideas 6 Bc4 and 6 Ne5.

White to play

Game 23

The situation has become rather unclear, though White’s position is already preferable. The g7bishop is fettered by the e5-pawn, which is difficult to remove.

A. Vovk – Tiviakov Altenkirchen 2012 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6 4 d4 Nf6 5 Nf3 g6 6 Bc4

42 c5 Rd4 43 Bxa4 Qf5 44 Bb5 Rb8 45 Bc4?! 45 Qc3 is strong, with the following possible line: 45...Rbd8 46 c6 Rxd3 47 c7 Rxc3 48 cxd8Q+ +=.

As always, placing the knight in the centre by 6 Ne5

45...Re4 46 Rc1 Bxe5 47 Bd5 Rd4 48 Nxe5 Qxe5 49 Qg6+ Qg7 50 Bxf7+ Kf8 51 Qh5 Rbd8?!

looks appealing. White can seek to castle queenside, exchange dark-squared bishops with Bh6 and attack by h4-h5. 6...Bg7 (D) and then:

After some complications the ending is now promising for White. The rest has nothing to do with the subject of this book, so I won’t analyse it in detail. 52 Bc4 Qf6 53 c6?! Rd2?! 54 Rf1? Kg7 55 Qg4 Rf8 White has not acted in the best way and the initiative is in Black’s hands again. 56 Be2 Rc2 57 Qe4 Rxc6 58 Bd3 Rfc8 59 a4 Rc5 60 Re1 Rc1 61 a5 Rxe1+ 62 Qxe1 Rc3 63 Bf1 Ra3 64 a6 Kf8 65 Qe4 Ra2 66 Be2 Kg7 67 Kg2 Ra5 68 Bd3 Ra2 69 Be2 Ra5 70 Bd3 Ra2 ½-½

White to play a) 7 Bc4 0-0 8 0-0 (castling kingside doesn’t mean that White has renounced the fight for the initiative) 8...Nc6 9 Nb5 Qd8 10 h3 (10 Nxc6 bxc6 offers White little; the doubled pawns aren’t pretty, of course, but the b-file is opened and White no longer has his centralized knight; after 11 Nc3 Rb8 Black has counterplay, with ...c5 an idea for the future; 10 Re1 could be considered) 10...Ne8

Conclusions White clearly can’t hope for any advantage by chasing the queen with Bf1 and then returning to g2; moreover, Black is not even obliged to repeat the position (Caspi-Motylev). 97

11 Bg5 h6 (I don’t like moves like this, loosening the light-square complex, i.e. g6 and f7; 11...Na5!?) 12 Be3 a6 13 Nxc6 bxc6 14 Na3 Nd6 15 Bb3 Nf5 16 c3 a5 17 Bf4 Ba6 18 Re1 Qb8 19 Rb1 Qb6 20 g4! Nd6 (Varavin-Landa, Sochi 2012) and now the simple 21 Rxe7 +/– gives White a huge advantage. b) 7 Bf4 (D) and here: White to play 8 Nxd5 Qxd5 9 f3 0-0 10 Bc4 Qa5+ 11 Qd2 Qxd2+ 12 Kxd2 Nd7 13 Rae1 (I doubt that White has anything real here) 13...e6 (13...Nb6!?) 14 c3 Nb6 (the prosaic 14...Nxe5 is also not bad since after 15 Bxe5 Bd7 the pawn is poisoned: 16 Bxc7?? Rfc8 –+) 15 Bb3 a5 16 a3 a4 17 Ba2 Nd5 18 Bxd5 exd5 19 Nd3 c6 20 g4 (here White at least has an idea – play against the bad bishop: all the black pawns are on light squares, so White has more forces fighting for the dark squares) 20...Be6 21 h4 Rfe8 22 h5 b6 23 Bc7 (see the previous comment; White urges the enemy pawn to a light square) 23...b5 24 Nc5 Bh6+ 25 Kd1 g5 26 Be5 Bf8 27 h6 Bxc5 28 dxc5 d4! (skilfully played: Black opens a diagonal for his bishop) 29 cxd4 (Kurnosov-Rakhmanov, Abu Dhabi 2013) and after 29...Bd5 30 Rh3 (30 Re3?? f6 –+; 30 Kd2!?) 30...Re6 a draw is likely. However, Rakhmanov was short of time and lost after several mistakes.

Black to play b1) 7...Qd8 8 Qd2 0-0 9 0-0-0 Nc6 10 Bc4 Nb4?! 11 Bh6 (Black should hurry with his counter-action, as otherwise it will be too late) 11...Bf5 12 Bb3 a5 13 a3 Nbd5 14 Nxd5 Nxd5 15 h4 a4 16 Ba2 (the a2-bishop perfectly combines two functions: it protects the king, and it participates in the kingside attack) 16...Nf6?? (much too slow, which is fatal here; better is 16...f6) 17 h5?! (it is even stronger to play first 17 Bxg7 Kxg7 before ploughing ahead with 18 h5 Ne4 19 Qe3 +–) 17...Ne4 18 Qe3 Qd6? (both players underestimated {or missed} the defensive resource 18...Bxh6! 19 Qxh6 g5 20 g4 {20 Ng6!?} 20...Qd6 21 Qxd6 cxd6 +/–, when Black’s game is unenviable but the struggle continues) 19 hxg6 Bxg6 20 Bxg7 Kxg7 21 Rde1 Nf6 22 Qh6+ Kh8 23 Rh4 c5 24 Reh1 1-0 Sarić-Jovanović, Stari Mikanovci 2010. Absolutely all the white pieces are aimed at the black king.

We now return to 6 Bc4 (D):

b2) 7...Nd5 (D) can lead to exchanges that appear to bring Black very close to equality.

Black to play 6...Bg7 7 0-0 0-0 8 h3 With the bishop on c4, the knight leap 8 Nb5 doesn’t promise anything special. After 8...Qd8 White can’t follow up with the c4 advance, so he must justify his knight manoeuvre using pieceplay: 9 Bf4 (this makes sense, but Black has time to finish his development without creating 98

Gonzalez de la Torre-Kurajica, European Clubs Cup, Plovdiv 2010. My impression is that in the 5...g6 line, Black often suffers in a worse position. That’s why I prefer the ‘central strategy’ where Black’s king’s bishop develops to the centre after ...e6.

weaknesses) 9...Na6 10 a4 c6 11 Nc3 Nc7 (11...Nb4 = is a bit more active) 12 Be5 Ncd5 13 Nxd5 cxd5 14 Bd3 Bh6!? (D).

11 Ba3 Qd8 (D)

White to play This idea is seen in several openings; the point is to move the knight without allowing an exchange of bishops. 15 Re1 (White can double Black’s pawns by 15 Bxf6 exf6 but it is double-edged, as Black’s bishops compensate for the weakened structure; if White overestimates his chances, Black may quickly seize the initiative) 15...Nh5 16 h3 Nf4 17 Bf1 Be6?! (after 17...f6 18 Bxf4 Bxf4 19 c4 White has a very slight initiative thanks to the weakened light squares e6 and f7) 18 Nh2 (the interesting resource 18 g4! goes unnoticed; then 18...g5 is forced, and White keeps an advantage by 19 Bxf4 gxf4 20 Bd3) 18...Rc8 19 a5 Qd7 20 c3 a6 21 Ng4 Bxg4 22 Qxg4 Qxg4 23 hxg4 Ne6 24 g3 += Zelčić-Plenković, Croatian Team Ch, Šibenik 2009.

White to play Tiviakov must have known what he was doing. However, from general considerations alone, manoeuvres like ...Ne8 and ...Qd8 can hardly provide equal chances in the fight against White’s mobilized army. 12 Re1 White counterattacks the e7-pawn rather than protecting d4. That’s a reasonable idea. 12...Nd6 13 Bd5 The modest 13 Bf1?! is ineffective: 13...Nf5 = 14 Ne2 Re8 15 Bb2 b6 16 Qd2 Bb7 17 Rad1 a5 18 c3 (the set-up with pawns on d4, c3, b3 and a4 is not very good, whereas Black looks ready to strike here, as his bishops are not blocked by their own pawns, and ...e5 may follow at some point). Now 18...e6 = was played in Zelčić-Stević, Stari Mikanovci blitz 2011, but already 18...e5!? could have been considered.

We now return to 8 h3 (D):

13...Bd7 14 Qd3 e6 14...a5!? is probably stronger since after the exchange on c6 and the retreat of the black bishop from c6, the pawn is on e7, rather than e6.

Black to play 8...a6 9 a4 Nc6 10 b3 Ne8

15 Bxc6 Bxc6 16 Ne5 Be8

10...Nb4

With the pawn now on e6, White has possibilities of opening the centre by playing d5.

looks like a better square, but note the following disaster: 11 a5 c5?! (11...Bf5 is better) 12 Na4 b5 13 axb6 cxd4 14 Ba3 a5 15 Nxd4 Bb7 16 c3 +–

17 Rad1 99

17 Ne4 b5 18 axb5 axb5 19 Qf3 +=. 17...b5 18 axb5 axb5 19 Bb4 (D)

White to play With considerable effort, Tiviakov has for now managed to fill in the holes emerging in various parts of the sinking ship.

Black to play White has a slight but comfortable advantage. Almost all the black pawns are on light squares, thus ‘blunting’ the e8-bishop.

28 Ra8 Rf8 29 Ra7 Rf7 30 Nh4 Ne4 31 Ra8 Nxc5 32 Rxe8+ Bf8 33 Nxf5

19...Ra6

However, it seems that the patient has got worse. White has an extra pawn and a convenient strongpoint on f5.

19...f6!? 20 Nf3 Bd7 21 Ne4 (I would like to point out that grabbing the b5-pawn leads to a position with two bishops against two knights where Black has compensation: 21 Bxd6 cxd6 22 Nxb5 Qb6) 21...Nxe4 22 Bxf8 (22 Qxe4 Re8 =) 22...Nxf2 23 Kxf2 Bxf8 24 Ra1 Rc8 += gives Black almost enough for the exchange.

33...Rd7 34 Ne3 b4 35 Ng3 Kf7 36 Rc8 Stronger is 36 Rb8 +–. 36...Nb7?! 36...Bh6 37 Nef5 (37 Kf1 +–) 37...Bf8 38 Kf1 Rxd5 39 Rxc7+ Rd7 +/–.

20 d5 Qh4 21 Bc5 Qf4 22 Nf3 e5

37 Rb8 Nd8 38 Ngf5

It may seem that Black has somehow employed his pieces but after...

Better is 38 Ne4.

23 Ne2! ...it turns out that the queen is short of squares.

38...c6 39 d6 Bxd6 40 Nxd6+ Rxd6 41 Rxb4 Ne6 42 Rb7+

23...Qf5

42 Rb6!?.

23...Qf6 24 Ng3 +/–.

42...Kg6

24 Qxf5 gxf5 25 Ra1!

The situation is difficult for Black but the exchange of his bad bishop increases his chances of salvation.

By exploiting the passive position of the bishop on e8, White seizes control over the a-file. When White’s rook enters the black position, Black will have too many weaknesses to hold.

43 b4 Nd4 44 Kf1 f5 45 Rb6 Kf6! 46 c3 Nb5 (D)

25...Rxa1 26 Rxa1 f6 Forced. Not 26...Bd7? 27 Ra7 +–. 27 Ra7 Rf7 (D)

100

‘losing’ for Black, but with a superhuman effort, and some inaccuracies from his opponent, Tiviakov managed to save a half-point. In one example (Gonzalez de la Torre-Kurajica), Black quickly got into a bad position after he put his knight on b4, where it was much stressed. So, does Black get a worse position in all the main lines after 5...g6? Well, it’s certainly a difficult line for Black, and as things stand at present I can’t recommend it.

White to play 47 c4?! A logical move, but there was a better option. White should have broken up the pair f5-e5 as it is well-known that connected passed pawns are stronger than isolated ones: 47 g4! fxg4 (47...f4 48 Nf5 Re6 49 c4 +–) 48 c4! Nd4 49 hxg4 +–.

Conclusions on 3...Qd6

47...Nd4 48 b5?! Ke6 49 bxc6 Rxc6 50 Rxc6+ Nxc6

When playing 3...Qd6, Black may come under some pressure, but it’s still less than in case of other queen moves (i.e. to a5 or d8). After the standard 4 d4 Nf6 5 Nf3, I recommend the ‘central strategy’ and the move 5...c6. This line is the safest option, whereas the fianchetto set-up with 5...g6 looks more vulnerable. Kovalenko’s 5...Bg4 may confuse an opponent who is not specially prepared for it. It’s an interesting line but not for everyone. Finally 5...a6 is probably playable, but it doesn’t appear to equalize.

Having played this line many times and studied the theory and games of other players, it is time to present my overall assessment of the line.

In my childhood I heard the assertion: “If the pawn ending is won, then the knight ending is also won”. Well, in practice it doesn’t always work... 51 Ke2 e4 52 f4 exf3+ 53 Kxf3 Ke5 54 Nc2 Na5 55 c5 Nc6 56 g3 Nd8 57 Nb4 Ne6 58 c6 Kd6 = 59 h4 Kc5 60 c7 Nxc7 61 Kf4 Kxb4 62 Kxf5 Kc5 63 g4 Kd6 64 Kf6 Nd5+ 65 Kg7 Ne3 66 g5 Nf5+ 67 Kxh7 Nxh4 ½-½

3: 3...Qa5 Conclusions 6 Ne5 may be connected with opposite-side castling and a kingside attack with h4, etc. The black pawn on g6 provides a ‘hook’ for White, as he can seek to open the h-file by playing h5. There is also the idea of playing Bf4, Qd2 and Bh6 (Sarić-Jovanović), exchanging the main defender of the black king. This also helps relieve the pressure against White’s queenside on the long dark-square diagonal.

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 (D)

Obviously, it’s better for Black to play by analogy with Kurnosov-Rakhmanov, where exchanges led to a more or less safe position for Black. White to play

To be honest, White can also pose Black problems by castling kingside (Varavin-Landa).

Before the emergence of 3...Qd6 as the modern main line of the Scandinavian, putting the queen on a5 was almost automatic. Indeed, there are good reasons for playing 3...Qa5. In contrast to the placement on d6, the queen doesn’t hinder the

In the main game after 6 Bc4 Vovk developed the other bishop on a3 and exerted strong pressure on Tiviakov’s fortifications in the centre. The ending that emerged was initially ‘worse’, and then 101

development of the king’s bishop, which, as we shall see, quite often appears on b4. The queen can also do useful work on its fourth rank. White in his turn tries to cause Black discomfort, often putting his bishop on d2 to create threats against the enemy queen. The most fundamental plan for Black, to my mind, is the ...c6 advance, which secures squares for the queen’s possible retreat to c7 or even d8. Please don’t get the mistaken impression that Black is just wasting time with these queen moves. Although the knight moves to c3 with tempo (by 3 Nc3), it also blocks the c-pawn’s strategically-desirable advance. Future moves by this knight (such as Nd5) usually lead to an exchange of knights, which is rather welcome to Black, who is a little short of space. In short, the strategy for both sides is based on positional nuances and it is rather difficult to give an unequivocal verdict like ‘White stands better’ or ‘equality’.

White to play 7 Bd2 This is one of the negatives of the queen’s placement on a5. The bishop lies in ambush and, from now on, Black has to take into account discovered attacks by the c3-knight. Generally we are talking about Nd5 here; there is little point in the knight going to b5 because after ...Qd8 it is unclear what to do next. Moving to e4 is possible but seldom used, and in any case generally has no advantages over moving to d5. I wouldn’t say the Nd5 idea is highly dangerous but it can be used to create positional pressure.

We shall not be examining the 3...Qa5 lines in as much detail or as systematically as we did the 3...Qd6 main line. Instead we shall look into a number of specific strategic themes, while naturally taking in most of the key variations along the way (please refer to the

In the following example White did not handle the position in the best way: 7 Ne5

Index of Variations

Black Plays ...Bb4 6 Ne5 Rare Plans for White

e6 8 Bd2 Nbd7 9 Nd5 Qd8 10 Ne3 (a creative idea, but it doesn’t disturb Black) 10...Be4 (10...Nxe5 11 dxe5 Ne4 =) 11 0-0 Bd6 12 N3g4 h5 (forcing the knight to reveal its intentions) 13 Nxf6+ gxf6!? (opening the g-file, which can have an impact on the health of the white king) 14 Nxd7 Kxd7 15 Qe2 Bf5 (15...Bxc2!?) 16 Bd3 Bxd3 17 Qxd3 Qc7 18 h3 Rag8 and later on Black created pressure on the kingside in Toma-S.Kasparov, Wroclaw 2009.

Doubled Black f- or g-Pawns

7...e6 8 Nd5 Qd8 (D)

if you have trouble locating a specific sequence that you are interested in). These themes are: Doubled Black f- or g-Pawns Black Plays ...Bg4 Before Nf3

Firstly, let’s acquaint ourselves with both sides’ plans when Black has the f7-f6-e6 pawn-structure. This typically emerges after the knight’s jump to d5 or (more seldom) to e4, when White has played an early Bd2 and the knight move discovers an attack on the black queen. Game 24 Alavi – S. Kasparov Ahvaz 2007 White to play 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 c6 5 Nf3 Nf6 6 Bc4 Bf5 (D) 102

This retreat is one of the reasons why ...c6 is played.

Ch, Porto Carras 2011) 12 Bg5 Qg6 13 Bxe7 Kxe7 14 h4 +=.

9 Nxf6+ gxf6

We now return to 9...gxf6 (D):

The most solid reply: Black keeps full control over his fourth rank. He plans to castle queenside and the half-open g-file can prove useful for his rooks. Capturing with the queen is less popular. Although this avoids damage to the pawnstructure, the queen is exposed on f6. 9...Qxf6 10 Qe2 (D) and now: White to play 10 c3 Or: a) 10 0-0 Nd7 11 Nh4 Bg6 12 c3 Qc7 13 Qf3 0-0-0 14 Bf4 Bd6 15 Bxd6 Qxd6 16 g3 (White is better, and plans the unhurried advance a4-a5; meanwhile Black’s prospects are unclear – he tries to create counterplay but this just aggravates the situation) 16...c5 17 Rfd1 Bc2 18 Rd2 Ne5? (more prudent is 18...Ba4, trying to redeploy the bishop to c6) 19 Qxf6 Nxc4 20 Rxc2 +/– Caruana-Van Wely, Wijk aan Zee 2012.

Black to play a) 10...Bxc2?! 11 d5 (11 Bc3!?; White could also prepare d5 by playing 11 Rc1 or 11 0-0) 11...Qxb2?! 12 0-0 +/– gives White a huge superiority in development. b) 10...Bg4 looks principled as it inevitably ‘spoils’ White’s pawn-structure, but it fails to equalize: 11 d5 Bxf3 12 gxf3 cxd5 13 Bxd5 Nd7 (pawngrabbing is risky when you are behind in development: 13...Qxb2?! 14 0-0 Nc6 15 Rab1 Qxc2?! 16 Be4 +/– is unlikely to find any takers as Black) 14 0-0-0 (obviously bad is 14 Bxb7? Qxb2 15 Bxa8 Qxa1+ 16 Qd1 Qxd1+ 17 Kxd1 Bd6 =+) 14...Ba3 (prosaic play doesn’t bring equality either: 14...0-0-0 15 Be4 Qe5 16 Bc3 Qc7 17 h4 f6 18 Rhg1 g6 19 Kb1 Bd6 20 h5 {the doubled pawns provide a ‘hook’ for White’s advance} 20...g5, Landa-Bosch, Rosmalen 2014, 21 Rd3 +=) 15 c3 0-0 16 Be4! += Shirov-Salov, Madrid 1997.

b) The immediate 10 Qe2 offers a pawn sacrifice. Then: b1) 10...Qd6!? looks odd, putting the queen in front of the bishop, but isn’t obviously bad. White can use his advantage in development to launch an attack: 11 0-0-0 Nd7 12 Nh4! Bg6 13 f4! (clearly intending f5) 13...f5 (necessary to prevent the further advance of the white f-pawn; it looks like the g6bishop is entombed, but with ...Nf6 and maybe ...Ne4 on the agenda, matters are not so simple) 14 g4! Be7 15 gxf5! Bxh4 16 fxe6 0-0-0 17 Qg4 fxe6 (stronger is 17...Nb6 18 exf7+ Rd7 {18...Kb8? 19 f5 +–} 19 Be6 Bf6, with unclear play) 18 f5! Bxf5 19 Qxh4 += Vachier-Lagrave – Hamdouchi, French Team Ch 2010.

c) 10...Nd7 11 0-0-0 (11 d5 is sharper: 11...cxd5 12 Bxd5 Be7!? {12...Qxb2?! 13 0-0 gives White the initiative} 13 Bc3 Bb4! 14 Bxb4 Qxb2 15 0-0 Qxb4 16 Rab1 Qa4 with very unclear consequences, David-Tkachev, Cannes 1999) 11...Be7 (D.Howell-Papaioannou, European Team

b2) 10...Bxc2 (D).

103

22 c4! (after the annihilation of the g6-pawn the soldier on f7 is somewhat weakened, but right now it can only be attacked by the rook – and the immediate 22 Rh7 has little impact due to 22...Nf8 – so White intends to play d5, ripping through Black’s central structure; then the bishoppair will be more effective) 22...f4 23 c5 Be7 24 gxf4 Bxh4 25 Rxh4 += (White has won a pawn, albeit doubled, and has total control over e5; soon it will be possible to destroy the enemy structure by playing f5) 25...Qd8 26 Rh6 Ka7 27 f5! exf5 28 Qf3 +/– Kasimdzhanov-Papaioannou, Bled Olympiad 2002. You don’t need to be a grandmaster to understand that Black has serious problems here.

White to play 11 0-0 (11 Rc1 Bg6 12 0-0 Be7 13 Rfe1 0-0 14 Nh4 Re8 15 Nxg6 hxg6, Zhang Zhong-Nisipeanu, Cap d’Agde rapid 2000, 16 h4!? gives White the initiative) 11...Bg6 12 Rfe1 Qd7?! (12...Qd6!? intending ...Nd7 is far less clear) 13 Rad1 b5?! 14 Bb3 Na6 15 d5! +– (White must strike while he has a lead in development and the black king is still in the centre; this is not a question of style, but rather of chess logic – even for such a cautious player as the author, it is obvious that White must attack in a position like this, regardless of the material cost) 15...cxd5 16 Bc3 (Black has two extra pawns but they are unprotected) 16...Be4 17 Bxf6 Rg8 18 Ng5 Nc5 19 f3 Nxb3 20 fxe4 Be7 21 exd5 Nd4 22 Qh5 +– Negi-Van Wely, Spanish Team Ch, Leon 2012.

10...Nd7 11 Qe2 Qc7 12 Nh4 Bg6 13 f4 (D)

Black to play

c) 10 Bb3

13...0-0-0

Nd7 11 Qe2 Qc7 12 Nh4 Bg6 13 0-0-0 0-0-0 14 g3 (D) sees White using a plan that is typical for this kind of structure.

13...f5 seems to be the most precise move-order. It prevents the f5 advance and led to a valuable example from a theoretical point of view in Z.Varga-Friedrichs, Dortmund 2001: 14 Nxg6 (the sacrifice 14 Bxe6 looks scary, but White lacks the resources to back it up; after 14...fxe6 15 Qxe6+ Kd8 16 Nxg6 hxg6 17 Qxg6 Qa5 –/+ the white queen on its own doesn’t pose much danger to Black) 14...hxg6 15 g3 0-0-0 16 0-0-0 Kb8 17 Kb1 Nf6 and Black’s position is at least no worse. 14 0-0-0

Black to play

Black’s move-order means that 14 f5!?

After 14...Bd6, the exchange 15 Nxg6 looks strange as it ‘glues’ Black’s kingside pawns back together again. However, matters are not that simple. 15...hxg6 16 h4 f5 17 Bg5 Rde8 18 h5 (that’s the idea: the black knight didn’t reach f6 in time, so this flank breakthrough is possible) 18...gxh5 19 Rxh5 Rhg8 20 Bh4 Kb8 21 Kb1 a6

is possible here. It is not deadly but it changes the pattern of the struggle: 14...exf5 15 0-0 Bd6 16 g3 Nb6 17 Bb3 += de la Riva-Am.Rodriguez, Catalonian Team Ch 1999. 14...f5 15 g3 (D) 104

21...Nd7 22 Qf3 Be7 23 h4 Nf6 Don’t consider the manoeuvre ...Nf6-d7-f6 as a waste of time; there is a point to it as the position has changed. The white pawn has shifted to h4 and intends to advance with destructive effect; if White manages to open up the h-file and bring his rook to the seventh rank, he will gain the initiative. This leaves the black knight with more to do on the kingside, both due to the need to parry that idea, and because there are more potential weaknesses in the white kingside now.

Black to play The key point here is that the g6-bishop is not as bad as it may seem at first sight. In case of Nf3 it can immediately jump to h5. It wants to make that move anyway after the intended ...Nf6. The black knight also has a potential outpost on e4.

24 Ka1 Rd7 25 Rhe1 Red8 26 a4?! Personally, I doubt the appropriateness of moves like this. I can understand the desire to create a broad pawn-front (d4-c4-b4-a4), but it is a bad idea to put the pawn on a light square at this point.

15...Bd6 26...a5!? (D) 15...Bg7 looks interesting, with latent pressure on the white queenside. 16 Bb3 Kb8 17 c4 This advance is a significant decision. While it is a reasonable and usually desirable move as it establishes control over d5, the weakening of d4 may become significant. 17...Nf6 18 Nxg6 hxg6 19 Bc2 (D) White to play The ‘dark-square symphony’ begins. In this book I aim not only to teach the Scandinavian Defence but also, when possible, to widen your general chess arsenal. Here in the fight for the dark-square complex Black actually has an ‘extra’ piece because the c2-bishop can play little role in this particular struggle. 27 Qe2 Bb4 28 Qe3 Ng4 29 Qf3

Black to play

Now Black faces a pleasant dilemma. I probably did not choose the strongest continuation.

19...Rhe8 The straightforward 19...Rh3!? is worth noting because it both physically prevents the advance of the h-pawn and prepares frontal pressure on it (...Rdh8, ...Ng4, etc.).

29...Rxd4 Better is 29...Bxc3! 30 Qxc3 (30 bxc3 c5 31 d5 exd5 32 cxd5 c4 –/+ is no improvement for White) 30...Qb6 31 c5 Qb4 –/+.

20 Bc3 Ka8 21 Kb1

30 Rxd4 Rxd4 31 Rxe6

Of course, 21 d5? is bad because 21...exd5 uncovers an attack on the queen. 105

Of course, the rook doesn’t want to die ‘for nothing’ and takes with him a black pawn. 31 Bxd4 Bxe1 –/+.

greedy when behind in development. In Negi-Van Wely we saw a crushing attack in the centre. In other notes we saw typical pressure on the kingside. The placement of the black queen on d6 (after the lengthy manoeuvre ...Qxd5-a5-d8-d6) is seen in Vachier-Lagrave – Hamdouchi.

31...fxe6 32 Bxd4 Qd6 33 Bc3 Qc5 34 Bxb4 axb4 (D)

In Caruana-Van Wely, White preferred a slight but stable plus. Van Wely couldn’t bear the pressure and rashly sought central counterplay. White’s play in Kasimdzhanov-Papaioannou is instructive: he used a typical plan with an aggressive advance of his h-pawn. He didn’t win the game but got an extra pawn and gave his opponent significant problems. Black is well-advised to use a move-order that prevents White from carrying out a powerful f5 advance (de la Riva-Am.Rodriguez).

White to play The right decision. Now Black stands very solidly on the queenside and has a slight but comfortable advantage.

If Black gets a ‘pawn-diamond’ on f7, g6, f5 and e6, then if he places a knight on f6, he can generally stabilize the kingside in a satisfactory manner (see both Z.Varga-Friedrichs and the main game).

35 Qd3 Ka7 36 Bb3 e5 37 fxe5 Nxe5 38 Qd8 Nf3 39 Ka2 Nd4 40 Qc7?? A mistake in a cheerless situation. It is wellknown that it is psychologically difficult to defend prospectless, passive positions.

Clearly, there are certain pros and cons in positions with doubled black f-pawns for both players.

40...Ne2

In the next game we shall move on to a structure where Black has doubled g-pawns, generally following an exchange of white knight for black bishop on g6.

...Nc1+ is threatened, and the way to g1 is opened for the queen. 41 Bc2 b3+ 0-1 Material losses are unavoidable; e.g., 42 Bxb3 Nc1+. 42 Kxb3 is no better due to 42...Nc1+ 43 Kc3 Qe3+ 44 Kb4 Na2+ 45 Ka5 Qc5#.

Game 25 Lupulescu – S. Kasparov Bucharest 2008

Conclusions

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 Nf6 5 Bc4

The combination of different plans was not very successful in Toma-S.Kasparov. White played in a creative way but that is all that can be said. In the end the black pawns were doubled on the f-file but with the king on g1 it gave rise to an attack along the g-file.

White is not obliged to bring the knight to f3 immediately. It can be delayed while he seeks to exploit his activity on the queenside and in the centre. Similar positions emerge after 5 Bd2, but I want to acquaint you with the dangers of combining the ideas ...g6 and ...e6. 5...c6 (for 5...Bg4, see the next game) 6 Bc4 (D) (6 Bd3 is discussed in the next game) and then:

White usually castles queenside, and then seeks kingside activity. Of course, this scenario is radically altered if Black accepts the sacrifice of the c2-pawn. On the other hand, it is risky to be 106

complete development and castle. 10 h4 b5 11 Bd3 Nbd7 12 Nf3 Bb7 (if only Black also had time for ...a6 and ...c5, everything would be fine...) 13 Ng5 Nb6?! (13...0-0 14 h5 gives White an attack) 14 h5 gxh5 (by the 14th move Black already has a lost position; there is no chance to hide on the queenside: after 14...0-0-0? 15 h6 Bf8 16 Qe5 +–, the points f7 and f6 crumble) and in the game 15 Rh4?! (a good move, though not the strongest) 15...Qe7 +/– enabled Black to put up some resistance and prolong the fight. There was a cleaner win by 15 d5!, a breakthrough in what appears to be the best-fortified spot: 15...h6 (15...cxd5 16 Nxe6 fxe6 17 Qxe6+ Kd8 18 Bg5 Rf8 19 Bxb5 +–) 16 Nxe6 fxe6 17 Qxe6+ Kd8 18 d6 Qd7 19 Qe7+ Qxe7 (19...Kc8 20 Bf5 Qxf5 21 Qc7#) 20 Rxe7 +–.

Black to play a) 6...Qb6 7 Nf3 Qxb2 8 Ne5 e6 9 Rb1 gives White the initiative, Degraeve-Tomczak, Cappelle la Grande 2006. b) 6...Bf5

d12) 9...b5! is correct, immediately starting queenside counterplay. After 10 Bd3 (or 10 Bb3 a5 with counterplay) 10...b4 11 Na4 Nd5 the knight protects the e7-pawn and Black will castle next move.

allows the bishop to be hunted down with 7 Nge2 e6 8 Ng3 Bg6 9 h4 h6 10 Nd5 Qd8 11 Nf4 Bh7 12 Qe2 Bd6 (12...Be7!?) 13 Ngh5 Nxh5 14 Nxh5 Rg8 15 0-0-0 += Nedev-Slovineanu, Turin Olympiad 2006. White is definitely ahead in development. The rook may go to the third rank, where it will have nuisance value to Black on both flanks.

d2) 7...Bf5!? (D) is stronger.

c) 6...Nbd7 7 Nf3 e6 8 Qe2 Bb4 += HendriksTiviakov, Dutch Ch, Hilversum 2007. d) 6...Qc7 (this seems best) 7 Qe2 and then: d1) 7...g6?! 8 0-0-0 Bg7 9 Re1 (D) latches onto the e7-pawn, not allowing Black to castle out of trouble. White to play However, Black must be ready for some sharp lines; if he follows up carelessly, he can come under attack or end up in a stably worse position. 8 Nf3 e6 and now: d21) 9 Ne5 is imprecise: 9...Nbd7 10 Nxd7 (expansion with 10 g4? doesn’t work: 10...Nxe5 11 dxe5 {11 gxf5 Nxc4 =+} 11...Nxg4 –/+) 10...Qxd7 11 d5 0-0-0 12 dxe6 (12 dxc6 Qxc6 =) 12...Bxe6 13 Bxe6 Qxe6 14 Qxe6+ fxe6 =.

Black to play Now:

d22) 9 Nh4 Bg6 (when behind in development, it would be hard to venture 9...Bxc2 10 Rc1 Bg6 11 d5; even if objectively {from the computer’s point of view} there is nothing terrible, it is unpleasant for a human to defend as Black here) 10 f4 Bd6 11

d11) In Abergel-S.Kasparov, Toulouse 2008 I grudgingly played 9...e6?!. Of course, things are not too bad after this, but it’s better not to shut in the c8-bishop. Alas, I didn’t see how else to 107

f5 += A.Muzychuk-Chasovnikova, Moscow 2009. The e6-pawn is under threat and the g6-bishop also feels uneasy.

However, Black must demonstrate good knowledge of the variations as each move is rather committal. Now:

We now return to 5 Bc4 (D):

b1) 6 f3 Bf5 7 Ne2 failed to give White an ounce of advantage in the following example: 7...Nbd7 8 00 e6 9 Ng3 Bg6 10 Qe2 0-0-0 11 Bd2 Qb6 12 Be3 e5! (very strong – after the disappearance of the d4-pawn, the black pieces become active) 13 dxe5 Bc5 (the point: White doesn’t have time to capture on f6) 14 Bf2 Bxf2+ (Svetushkin-Cornette, Belfort 2012) 15 Qxf2 Qxf2+ 16 Rxf2 Nxe5 with absolute equality. b2) 6 Ne2 Nc6 7 f3 Bh5 8 Bd2 (D).

Black to play Now Black must make a major decision. 5...c6 Or: a) 5...Nc6 is a highly combative move, putting pressure on the d4-pawn and with ideas of ...0-0-0. The drawbacks are that the queen won’t have any way home (as it does with ...c6 and ...Qc7/d8), so the white knight will have opportunities to jump to b5 or d5. 6 Ne2 Be6 7 Bxe6 fxe6 (clearly this is a dubious pawn-structure, and it doesn’t seem that Black gets enough dynamic play in return) 8 0-0 0-0-0 9 Qd3 e5 10 Qh3+ Rd7 11 Rd1 += DudaReinderman, Wijk aan Zee 2014. One way or another, Black is saddled with an isolated e-pawn.

Black to play It may be quite disturbing for many of us to have a queen opposite the d2-bishop without a safe shelter. 8...0-0-0 9 a3 Nxd4! (principled and strong; obviously Tkachev was familiar with this variation) 10 Nb5 (10 Nxd4 Rxd4 11 Nb5 may be preferable) 10...Qb6 11 Nbxd4 e5! (White has a knight for a pawn, but Black has excellent compensation) 12 g4?! (better is 12 Nf5) 12...Bg6 13 g5 (Morozevich-Tkachev, FIDE Knockout, New Delhi 2000) and now 13...Nh5 14 Nb3 Be7 gives White real difficulties. Black threatens ...Bxg5 and ...Qc6, attacking f3 and c4, and it is not clear how White should finish his development.

b) 5...Bg4 (D) gives White, in my view, the greatest trouble securing an advantage.

b3) 6 Nf3 looks more natural, but then the pressure on the d4-pawn is even more pointed, as it is not possible to push the bishop aside with f3 (as it is when the knight is on e2). 6...Nc6 (D) and now:

White to play 108

9...Bb4 also doesn’t equalize: 10 Ne5 Nbd7 11 Nxd7 Nxd7 12 g4 (one should always consider this follow-up to Ne5) 12...Bg6 13 h4 h5 14 Bxe6 0-00 15 Bf5 Bxf5 16 gxf5 += Sadvakasov-Megaranto, Dubai 2007. 9...b5!? could be considered. 10 Nh4 Bg6 11 Nxg6 hxg6 (D) White to play b31) 7 d5?! looks critical but proves ineffective: 7...0-0-0 (as often happens in the Scandinavian, the faradvanced white d-pawn is subject to strong pressure) 8 Bd2 Ne5 (8...Nd4!? is even tougher) 9 Bb3 Qa6 10 Qe2 Qxe2+ 11 Kxe2 a6 12 Bf4 Nxf3 13 gxf3 (Edouard-Hamdouchi, Sestao 2010) 13...Bf5 and if anyone stands worse, it is not Black.

White to play

b32) 7 Bb5

At a glance it seems that any disadvantage for Black should be small and manageable. White has the bishop-pair but Black has total control over d5, a solid structure, an active h8-rook and so on. However...

Nd5 8 Qd3 (8 Bd2 seems sturdier, but even then 8...Nxc3 9 Bxc6+ bxc6 10 Bxc3 Qd5 is slightly better for Black) 8...Nb4 9 Qe4 Bxf3 (opting for simplification; there was a sharp option in 9...f5!, fighting for better chances by distracting the queen from the protection of the c2-pawn) 10 Bxc6+ bxc6 11 gxf3 e6 12 Bd2 Qa6 13 Ne2 Nd5 = Nakamura-C.Bauer, Le Port Marly 2009.

12 d5! Logically, having the two bishops White opens the centre. This provides a slight but comfortable advantage.

6 Qe2 (D)

12 h4 0-0-0 13 Bg5 Nb6 14 Bb3 Nbd5 15 Ne4 Nxe4 16 Qxe4 Be7 = Ganguly-Mohota, Gibraltar 2006. 12...cxd5 13 Bxd5 13 Nxd5 leads to similar play, but I don’t want to overburden you with a huge number of lines. 13...Rc8 Of course, it is tempting to eliminate the bishop, but it is no picnic either: 13...Nxd5!? 14 Nxd5 Qc6 15 Bc3 0-0-0 16 Ne3 leaves White with the initiative; for instance, 16...Nf6 17 Rxd8+ Kxd8 18 Ng4 +=.

Black to play 6...Bf5 Shutting in the bishop with 6...e6 is not very attractive.

14 Bf3 (D)

7 Bd2 Qc7 8 Nf3 e6 9 0-0-0 Nbd7 109

27 cxb4 Bxb4 28 Rc1 (D)

Black to play Black to play

14...Be7 14...a6 is probably more accurate: 15 g4 Nb6 with counterplay.

White releases the tension by exchanging rooks; now the white king will be active.

15 Qb5!

28...Bd6 29 Rxc7 Rxc7 30 Rc1 Rxc1+ 31 Kxc1 Nc5 32 Kc2 f5 33 g3 Kf7 34 b4 Nd7 35 Kb3 Kf6 36 f4! +/–

With an elegant manoeuvre, White causes a weakening of the light-square complex on the queenside. 15...b6 16 Qa4 0-0 17 Nb5 Nc5

The end is already close. The white a- and bpawns are much more dangerous than the central passed d-pawn.

Forced, but sufficient for the time being.

36...g5?

18 Nxc7

While this is a mistake, frankly speaking, I considered the position lost and could only seek practical chances.

18 Qa3? Qd7 =+. 18...Nxa4 19 Nb5 a6 20 Na7 Rc7 21 Nc6

37 fxg5+ Ke5 38 Bf4+ Kxd4 39 Bxd6 Ke4 40 h4 g6 41 a4 d4 42 a5

And now, a knight pirouette. All this is not deadly for Black, who is defending skilfully, but these nuances will add up to an appreciable advantage for White.

In this case I can’t say much about placing pawns on squares of the bishop’s colour. There are exceptions to every rule, where concrete considerations trump general ones. Here it is more important that h5 and b5 will be threatened.

21...Bd6 22 Be3 Nd5 Black’s situation is not very pleasant. Now White starts to transform his advantage. In reply to 22...Bc5 White would, of course, avoid the exchange with 23 Bf4 +=.

42...d3 43 Bf4 Kf3 44 Kc4 d2 45 Bxd2 Kxg3 46 b5 axb5+ 47 Kxb5 Nb8 48 h5 1-0 That was a very pure and instructive performance by Lupulescu.

23 Bxd5 exd5 24 Nd4 += One of the terrible bishops has left the board but there is another problem – the isolated d5-pawn.

Conclusions

24...Rfc8 25 Kb1 b5

In our main game, White, a strong Romanian player, operated in sophisticated fashion. Thanks to an accurate move-order he managed to blast open the centre with d5. While this wasn’t a disaster for Black, he came under pressure that lasted until the very end of the game.

Or 25...f6 26 c3 Be5 27 Rd2 +=. 26 c3 b4 This seems logical: Black searches for counterplay on the queenside. 110

Game 26

I also failed to equalize against Abergel. My opponent managed to exert pressure along the efile. In order to castle I had to put pawns on g6, e6 and c6, despite having a ‘bad’ c8-bishop. Black was behind in development and the game should, logically, have been decided by a d5 breakthrough. But I was granted amnesty and the outcome was different.

Anand – Van Wely Wijk aan Zee 2013 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 Nf6 5 Bd2 (D)

However, an early development of the bishop to f5 doesn’t solve all Black’s problems as this piece can be attacked and hunted down (A.MuzychukChasovnikova).Caro-Kann motifs can be seen in Nedev-Slovineanu: White advanced his h-pawn to attack the bishop, gaining a space advantage and a lead in development. Putting the bishop on b4 is no panacea (Sadvakasov-Megaranto), while taking on doubled e-pawns can hardly be recommended (DudaReinderman).

Black to play 5...Bg4

So what should Black do? Probably, the best plan is an early ...Bg4. After that the knight goes to c6, attacking the white d-pawn. Then White must play very precisely too; otherwise he could find himself in a worse position even if he modestly places the knight on e2 and follows up by playing f3 (Morozevich-Tkachev). In Svetushkin-Cornette we saw a similar situation where White was more careful, but still got no more than equality.

Surprisingly, White’s delay in playing Nf3 actually invites this apparent thrust into mid-air. White has no completely ideal way to parry the bishop’s random-looking stab. Let’s have a look at a similar theme with the inclusion of ...c6: 5...c6 6 Bd3

Responding with Nf3 is more natural, but doesn’t guarantee an advantage either, as the pin is annoying and a damaged pawn-structure can cause problems in many lines (Nakamura-C.Bauer).

Bg4 7 f3 Bh5 (when there is already a pawn on c6, the bishop has nothing to do on d7 – compare the main game) 8 Nge2 Nbd7 9 Nf4 Bg6 10 Nxg6 hxg6 11 Qe2 e6 (Black’s position is quite sturdy, and the advance f3 has weakened the dark squares f4, g3 and h2 in a way that, considering the fact that the h-file is half-open, may cause some discomfort for White) 12 Ne4 Bb4 13 c3 Be7 14 g3 Nxe4 15 fxe4 Bg5! (Ivanchuk is right, as always; one should deprive the opponent of the bishop-pair) 16 0-0 Bxd2 17 Qxd2 c5! 18 Qf2 (18 d5?! Ne5 19 dxe6? doesn’t work due to 19...0-00 –/+) 18...0-0 19 e5! Rad8 20 Be4 cxd4 21 cxd4 (L.Dominguez-Ivanchuk, Havana 2012) 21...Qb6 with satisfactory play for Black.

If White pushes his pawn into the crossfire on d5, he may even fail to equalize (EdouardHamdouchi).

Black Plays ...Bg4 Before Nf3 In our previous game we witnessed the effectiveness of Black playing an early ...Bg4 before the white knight appears on f3. This idea deserves a closer look, and in other scenarios. Our next game features the move-order 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 Nf6 5 Bd2 Bg4.

6 f3 After 6 Nf3 (D), Black can manage without playing ...c6:

111

diagonal, both obstructing White’s ideas and possibly becoming active there later on (e.g., on c6). In positions like this, 6...Bf5 always invites a further advance by the white pawns: 7 g4 Bg6 8 f4! e6 9 f5 exf5 10 g5 (a typical idea: the bishop is still restricted while the knight is kicked away into the unknown; 10 Qe2+ Kd8 11 g5 Nfd7 12 0-0-0 gave White the initiative in Houard-Prié, Saint Affrique 2007) 10...Nd5 (10...Ne4?? 11 Nxe4 +–) 11 Qe2+ Be7 (NatafJ.Sanchez, French League 2012) 12 Bg2 +/–.

Black to play a) In Edouard-Prié, Bergara 2012 a long-standing Scandinavian Defence specialist handled his queen skilfully: 6...Qf5!?

7 Bc4 Qb6 8 Nge2 e6 (D)

7 Bd3 Bxf3 (7...Qd7 8 h3 Bh5 proved less successful in Terrieux-Prié, Belfort 2010; after 9 Be3, intending g4, White is obviously better: 9...Bxf3 10 Qxf3 Nc6 11 0-0-0 +=) 8 gxf3 Qd7 9 Be3 Nc6 10 Ne4 0-0-0 (remarkably, all the black pawns are placed on their initial positions, so there are no weaknesses; this strategy appeals to me) 11 c3 e5 12 Nxf6 gxf6 13 Qc2 Kb8 14 Bf5 Qd5 15 Be4 Qc4 16 d5 Ne7 17 Qb3 Qxb3 18 axb3 b6 19 c4 f5 20 Bc2 f4 21 Bd2 c6. The reduced material and White’s less-than-ideal pawn-structure allow one to assess the position as equal.

White to play Don’t be put off by the bishop on d7. It is not especially bad as it does useful work on the e8-a4 diagonal (restraining the c3-knight), and one day can leap to c6 or b5.

b) 6...e6 7 h3 Bh5 8 g4 Bg6 9 Ne5 Qb6 10 Qe2?! Nc6! 11 Nc4 Qa6 12 Nb5 0-0-0 (12...Qxb5?? 13 Nd6+ +–) 13 c3 (N.Kosintseva-C.Bauer, Biel 2010) and now 13...h5!? 14 g5 Ne4 gives Black the initiative.

9 Bb3 Or:

We now return to 6 f3 (D):

a) 9 0-0 Be7 and now: a1) 10 a4 Nc6? (Black mustn’t let the white pawn advance too far; the right thing was to halt it immediately by 10...a5) 11 a5 Nxa5 12 Rxa5 Qxa5 13 Nd5 (Black has to give up his queen) 13...Qxd5 14 Bxd5 Nxd5 15 c4 Nf6 16 Bf4 +/– Nijboer-Van Wely, 2nd Bundesliga 2012/13. Surprisingly, Black won this game. However, the assessment is definite; with roughly equal material, White enjoys a sound initiative. Black to play

a2) 10 Kh1

6...Bd7!?

0-0 11 a4 a5 12 Ne4 Nc6 (12...Bc6, followed by developing the knight on d7, may be more precise) 13 c3 Rfd8?! (Black seems to be playing logically, but he runs into increasing problems) 14 Qc2 Be8

Having induced f3, the bishop steps back to d7, where it performs a useful role on the a4-e8 112

15 N2g3 Nd5 16 Rae1 g6?! 17 Nf2 Kg7 18 Qc1 Na7 19 b3 Nc8 20 Re2 c6 (20...Nd6?? 21 Bxd5 exd5 22 Rxe7 +–) 21 Bg5 Qc7 22 Ng4 +– Gharamian-Prié, Belgian Team Ch 2010/11. Without any special effects or sacrifices, White has created irresistible threats on the kingside dark squares.

10 Be3 Na5 See the previous comment! 11 0-0 Nxb3 12 axb3 Be7 13 Nf4 0-0 14 Re1 Rfd8 15 Nd3 (D)

b) 9 Be3 Nc6 10 a3 Ne7!? (D).

Black to play 15...Qd6 White to play

How else can the bishop be brought to c6? 15...Nd5 16 Bf2 Nxc3 17 bxc3 gives White the initiative, while 15...Bc6?? fails to 16 d5 +–.

Since d5 is not occupied by the bishop yet, Black transfers his knight there; its normal route was denied to it by the move a3. 11 0-0 Ned5 12 Bf2 Be7 13 Ne4 (of course, there is no point in capturing on d5; White should keep pieces on the board) 13...Bb5 14 b3 Bxc4 15 bxc4 Nxe4 16 fxe4 Nf6 17 Nc3 0-0 18 Qe2 Qa5, Sarić-Papaioannou, Istanbul Olympiad 2012. White stands better but Black’s game is tenable. Two pairs of minor pieces have left the board, which eases Black’s manoeuvres; also, he has no weaknesses.

16 Bf2 Bc6 Black can also play more modestly: 16...Be8 17 Ne4 Nxe4 18 fxe4 a5 19 c4 b6 20 Qg4, though White has the initiative here. 17 Ne4! Nxe4 18 fxe4 f5?! A committal continuation. As a rule, unless it makes concrete gains, Black should avoid this advance as its positional drawbacks are clear: it weakens e5 and places the pawn on a light square – the same as Black’s unopposed bishop.

We now return to 9 Bb3 (D):

18...b6 19 c4 Bb7 is more accurate. 19 exf5 exf5 20 c4 Be4 The bishop must hurry, before White plays d5, but it also has bad consequences. 21 Nc5 Qg6 22 Nxe4 fxe4 23 Qb1 +/– As might have been expected, the pawn that lost connection with its own forces perishes. Not 23 Qc2? e3! =.

Black to play 9...Nc6

23...Bf6 24 Qxe4 Qxe4 25 Rxe4 (D)

Given that the bishop is on d7, the knight is left with the c6-square, where it exerts pressure on the d4-pawn and can also jump to a5 or b4. 113

and from what looked like rather an equal position, the situation became hopeless for Black. We saw a successful bishop retreat along the route ...Bg4-h5-g6 in L.Dominguez-Ivanchuk. With White’s bishop on d3, this makes a lot of sense. Black has no objection to an exchange of these bishops. The ‘principled’ 6 Nf3 may stimulate Black’s appetite for the ambitious set-up with ...Nc6. I doubt that it is pleasant for the d4-pawn (N.Kosintseva-C.Bauer). Moreover, there are several ways for Black to reach equality. Black achieved a pleasant position quite soon after 10...0-0-0 in Edouard-Prié.

Black to play The rest is just technical work, which Anand handles comfortably, as usual. 25...c6 26 Kf1 Rd7 27 Ke2 a6 28 Kd3 Rad8 29 Rae1 Kf7 30 Be3 h5

In some cases, the presence of the bishop on f5 or g6 provokes a rapid advance by White’s kingside pawns; this can cause Black some discomfort (see Nataf-J.Sanchez). But after studying both sides’ plans, one can form the impression that putting the bishop back on d7 (i.e. ...Bg4-d7) also does not bring easy equality. However, let’s not overdramatize; Black has reasonable prospects of gradually equalizing – the fact that he has no weaknesses is very helpful in this respect (SarićPapaioannou).

Exchanging brings no relief: 30...Bxd4 31 Rxd4 Rxd4+ 32 Bxd4 c5 33 Re4 Rd7 34 b4 cxd4 35 Rxd4 +/–. 31 Rf1 Kg6 32 b4 Rd6 33 h3! R6d7 34 g4 hxg4 35 hxg4 Rf8? A mistake, albeit in a miserable position. The resistance could have been continued by 35...Be7 36 Re6+ Kh7 37 Rxe7 Rxe7 38 Bg5 Rde8 39 Bxe7 Rxe7 40 d5 +/– or 35...Kh7 36 g5 Be7 37 Kc3 (37 c5) 37...Rc8 38 Re6 Kg8 39 Rf5 Rf8 40 Rfe5 Bd6 41 Re4 +/– and everywhere there are problems which can hardly be solved.

Black Plays ...Bb4 We focus here on Black playing 8...Bb4 in the main-line position after 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 Nf6 5 Nf3 c6 6 Bc4 Bf5 7 Bd2 e6 8 Qe2. This is the most logical way for Black to finish his development.

36 g5! Bxd4 37 Re6+! Not 37 Rxf8? Bc5+. 1-0

We should bear in mind that sooner or later the bishop will have to be exchanged for the c3-knight. This strategy is based on securing a firm grip on the d5-square. No doubt you are sceptical about the idea of giving White the bishop-pair. However, under the right conditions the notion of the ‘two bishops advantage’ is ambiguous. Situations exist where knight and bishop are preferable to two bishops. Don’t believe me? Then look up in your database the following three games from my own practice: Moskalenko-S.Kasparov, Minsk 1996; Nosenko-S.Kasparov, Kiev 2001; and S.Kasparov-S.Ernst, Hoogeveen 2014. In each of these games, the two bishops proved less effective than the enemy knight and bishop.

37...Bf6+ 38 Kc2 Kf7 39 Re4 +–.

Conclusions In our main game, Black secured a solid position from the opening, with his bishop-pair largely compensating for his lack of space. The logical current of the struggle was broken by the premature 18...f5?!, which led to the loss of a pawn and gradual defeat. The good strategic model for White was demonstrated by White’s impeccable play in Gharamian-Prié. Without any blunders by Black, 114

We now return to 8...Bb4 (D):

Both our main game and many of the games embedded within it feature Rustem Dautov. It is no coincidence that he is also an expert in the Caro-Kann Defence.

Game 27 Vogt – Dautov Altensteig 1994 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 Nf6 5 Nf3 c6 6 Bc4 Bf5 7 Bd2 e6

White to play 9 Ne5

Here White is not obliged to play Nd5 or to prepare the d5 advance. He can just continue development by...

Let’s have a look at other continuations: a) Here is an example of how events may unfold if White doesn’t hurry to put the knight on e5: 9 0-00 Nbd7 10 Kb1

8 Qe2 (D)

(Black easily equalizes after 10 Ne5 Nxe5 11 dxe5 Nd5 12 Ne4 Bxe4 13 Qxe4 Bxd2+ 14 Rxd2 0-0-0 15 Rhd1 Nb6 {by exchanging the rooks on the only open file, Black will relieve the pressure} 16 Bb3, Rogers-Wahls, Bundesliga 1996/7, 16...Rxd2 17 Rxd2 Nd7 18 f4 Nc5 19 Qd4 Rd8 =) 10...Nb6 11 Bb3 (11 Ne5!?) 11...Bxc3 12 Bxc3 Qb5 (a typical manoeuvre in structures of this kind: Black intends ...a5-a4) 13 Qxb5 cxb5

Black to play

(I view this odd-looking structure as quite pleasant for Black; note Black’s total control over the light squares d5, c4 and a4, while White’s dark-squared bishop can do little to change that) 14 Ne5 a5 15 a3 Be4 16 Rhe1 0-0 17 f3 Bd5 18 Bxd5 Nfxd5 gave Black the initiative in Hjartarson-Cu.Hansen, Reykjavik 1995.

8...Bb4 White will sooner or later play a3, forcing the bishop to reveal its intentions. We shall see later why Black is not too perturbed by the idea of giving up his bishop for the c3-knight. Other moves are less appealing. For example, after 8...Be7

b) 9 a3 Nbd7 (D).

the d5 advance will present White with the bishop-pair in an open position – a very different proposition. A possible variation is 9 d5 cxd5 10 Nxd5 Qd8 11 Nxe7 Qxe7 12 Nd4 +=. Greed is bad here: it is instant disaster after 8...Bxc2?? 9 d5! cxd5 10 Nxd5 Qd8 11 Nxf6+ gxf6 12 Bb5+ Nd7 13 Bc3 Bf5 14 Nh4 +– Liou-Steinfl, Los Angeles 2011. When White put his queen on e2, it was not merely routine development; it also pinned the pawn on the e-file.

White to play 115

b1) The simple 12...Nxe5

what happens) 15...Nd5 (15...Nxe5?! is weaker due to 16 dxe5 Nd5 17 Qxf5 Nxc3 18 bxc3 Rad8?! {18...c5!?} 19 Kb2 +=) 16 Qxf5 Nxc3 17 Nxd7 (17 bxc3 allows Black to keep his knight by 17...Nb6, intending ...f6 and ...Nc4 with good counterplay) 17...Nxd1 18 Nxf8 Rxf8 (18...g6? 19 Ne6 Qe7 20 Qe5 f6 21 Qe2 Nxf2 22 Re1 +–) 19 Rxd1 (19 Kxd1 Rd8 20 Qe4 c5 =) 19...Qxh2 = Z.Almasi-Dautov, Altensteig 1994. The smoke has cleared, and Black has coped with his opening difficulties.

13 dxe5 Nd5 is admissible: 14 Bxd5

We now return to 9 Ne5 (D):

Black uses the pin against the a1-rook to delay having to decide what to do with this bishop; while Black is generally happy enough to exchange on c3, there may be something to gain by keeping options open while it is feasible to do so. 10 0-0-0 (queenside castling looks doubleedged, but is quite an appropriate way to fight for the advantage) 10...Bxc3 11 Bxc3 Qc7 12 Ne5 and now:

(White can also seek a slight superiority by playing 14 Bd2, but you’d need the technical skills of Karpov to gather much advantage, grainby-grain, in that case; anyway, it is the subject of Game 28) 14...cxd5 15 Rd4 Rc8 16 Rhd1 Bg6 17 R1d2 0-0 18 h4 h6 19 h5 Bh7 20 Rg4 (20 g4 a5 21 g5 hxg5 22 Rg4 Bf5 23 Rxg5 Qe7 24 Qe3 Rxc3 25 bxc3 Qxa3+ gives Black compensation) 20...Bf5 21 Rf4 Rfd8 = Sai Krishna-S.Kasparov, Sort 2008. The player with White was a small Indian boy who was satisfied with a draw. Well, Black also shouldn’t complain. It is possible for him to seek winning chances by acting correctly on both flanks, combining attack (...a5, ...b5) with defence.

Black to play 9...Nbd7 10 Nxd7 Nxd7 10...Kxd7!? was an interesting attempt to create winning chances by the higher-rated player in Trabert-Nisipeanu, Naujac 2000. Personally I like such moves although objectively it may not be the strongest. 11 0-0-0 Rad8 12 a3 Bxc3 13 Bxc3 Qc7 14 f3 Kc8 (castling by hand is completed) 15 Kb1 h5 16 Be1!? (a good idea: if the bishop appears on the h2-b8 diagonal, Black will have a difficult time) 16...h4 17 g3 hxg3 18 Bxg3 Qb6 and Black had saddled his opponent with isolated h- and fpawns (and was probably content thereby to have unbalanced the game), but in general White’s chances are better.

b2) 12...b5!? (D) was Dautov’s more enterprising approach.

11 a3 Bxc3 12 Bxc3 Qc7 Black just needs one more move (e.g., castling) and the central square d5 will be under his total control.

White to play This move typically prevents c4, while here it also hints at an advance with ...a5 and ...b4. 13 Bd3 (or 13 Bb3 Be4! 14 Rhe1 Bd5 =, when the strike on f7 is incorrect) 13...0-0 (exchanging on d3 is inadvisable on general grounds as it gives White a tempo to redeploy his rook to the kingside: 13...Bxd3?! 14 Rxd3 0-0 15 Rg3 Nd5 16 Bd2 with an attack) 14 Bxf5 exf5 15 Qf3 (it might seem that Black loses a pawn on either f5 or c6; but let’s see

13 d5! (D)

116

Let me point out Black’s strategy. On the queenside he has two pawns against three, but his pieces are exerting pressure on this side of the board. On the kingside he has four vs three, which gives his king added security. This majority may be of use later. 21 a4 Nb6 22 Be5 22 a5 is inappropriate as the pawn could perish: 22...Na4 23 Be5 Qxa5 24 g4 (forced because after 24 Bxb7 Rxc2! 25 Rxc2 Bxc2 –/+ the weakness of White’s back rank reveals itself) 24...Bd3! with good play for Black.

Black to play As is known, other things being equal, bishops prefer space and open diagonals.

22...Qe7 23 b3 Qg5 24 Bb2 Nd5 25 g3 (D) 13...cxd5 14 Bxd5 14 Bxg7?? Rg8 –+. 14...0-0 15 Bf3 Rac8 16 0-0 Nb6 Here come the Caro-Kann outlines. In that opening a similar pawn-structure may arise after an undermining ...c5. 17 Rac1 Rfd8 18 Rfd1 Or 18 Be5 Qc4 19 Rfe1 Qxe2 20 Rxe2 Rd7 21 b3 f6 22 Bb2 Nd5, with approximate equality.

Black to play

18...Rxd1+ 19 Rxd1 h6 (D)

25...Qg6

There is no point in 19...Na4 because after 20 Be5 Qxc2 21 Qxc2 Bxc2 22 Bxb7 Bxd1 23 Bxc8 all Black can dream about is a draw.

25...Nc3 physically prevents the advance of White’s c-pawn, but it doesn’t suffice for full equality. 26 Qe1 b6 (the knight is untouchable because of the attack on the c1-rook) 27 h4 Qf6 28 Kh2 (a prophylactic measure against ...Ne2+ and ...Qxb2) gives White the initiative. 25...b6 doesn’t equalize either: 26 h4 Qg6?! (26...Qd8 +=) 27 Bh5! Nf4 (27...Qh7? 28 Qa6 +–) 28 Qa6 +/–. 25...Nf4!? is a further possibility. 26 Qb5 The obvious advance 26 c4 must be considered, but it doesn’t attain its goal: 26...Nf4 27 Qd2 Nh3+ 28 Kh1 (28 Kg2 Ng5 gives Black the initiative) 28...Be4 (the exchange of the lightsquared bishops would considerably ease Black’s game because then h1, g2 and f3 will become holes in White’s kingside while the pressure on the h1-a8 diagonal will be no more) 29 Bg2 (29 Qe3? is bad because of the tactical nuance

White to play 20 Rc1 The outrageous-looking 20 h4!? is interesting, with possible ideas of h5 and g4. However, not everyone is eager to weaken their king with moves like this. 20...Nc4 117

transferring his queen to g6, Dautov combines pressure on the c2-pawn with piece-play on the kingside. He also activated his knight and bishop in such a way that they fought effectively against White’s two bishops. With his pressure on the light squares, Dautov sought an exchange of lightsquared bishops, which would have been especially painful for White after he played g3.

29...Rd8! 30 Bxe4? Qxe4+ 31 Qxe4 Nxf2+ 32 Kg2 Nxe4 –/+) 29...Ng5 =. 26 Qe5!? is a further possibility. 26...Rxc2 Now the game is becoming drawish. 27 Rxc2 Bxc2 28 Qxb7 Qf5

In our next game we examine a subtle plan by which White seeks to obtain the bishop-pair in a situation where they are truly an advantage. After studying this subject I have gained the impression that both players must handle some intricate tactical nuances with great accuracy. It is not easy for White to transform his slight initiative into anything tangible. Moreover, an inaccuracy can result in hardships for White.

Or 28...Be4 29 Bxe4 (the computer-like 29 Bd1 += is unlikely to be chosen by a human) 29...Qxe4 30 Qb8+ Kh7 31 Qe5 =. 29 Qc8+ Kh7 30 Bxd5 exd5 ½-½

Conclusions

Game 28

In this line, an exchange of knights sometimes brings the white d-pawn to e5, where it impedes Black’s light-squared bishop. However, with this structure Black can even hope to seize the initiative. Mass exchanges on the d-file can lead to equality (Rogers-Wahls), while Black can seek counterchances after an exchange of minor pieces on d5 (Sai Krishna-S.Kasparov), even in a drylooking position with opposite-coloured bishops.

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 c6 5 Bc4 Bf5 6 Nf3 Nf6 7 Bd2 e6 8 Qe2 Bb4 9 0-0-0 Nbd7 10 a3 Bxc3 11 Bxc3 Qc7 12 Ne5 Nxe5 13 dxe5 Nd5 14 Bd2 (D)

In the lines without an exchange of knights on e5, the pawn that remains on d4 hinders the white dark-squared bishop (which sits on c3 following Black’s ...Bxc3 exchange). Black has to reckon with the d5 and c4 advances, but it is not easy for White to engineer either of these moves in a satisfactory way. The resource ...b5 looks like a true fighter’s move, taking control of c4 and bringing up the possibility of attacks on opposite wings. It is especially pointed after White has played a3, as this provides a ‘hook’ to help Black open lines by ...b4 (Z.Almasi-Dautov).

Black to play

Ki. Georgiev – Henrichs Recklinghausen 1998

Of course, White is not obliged to capture on d5; he has the right to keep both bishops alive. Who will benefit from this?

The alteration to the pawn-structure that occurred in Hjartarson-Cu.Hansen appeals to me. The shift of the c6-pawn to b5 opens the c-file for Black’s rooks, while the pawn itself is ready to team up with its neighbour on the a-file and make further advances when appropriate. If Black can establish a firm grip on the light squares, then White’s bishop-pair will count for very little.

14...0-0-0 15 g4 Bg6 16 f4 White intends to trap the bishop by simply playing f5. Black must react somehow. 16...h5

In our main game we see some clear ‘Caro-Kann’ motifs. White gets a queenside majority, but Black limits its mobility by exerting pressure on the cfile. Both players acted almost ideally. By

Natural and strong: the h8-rook is included in the fight immediately without leaving its home square. 118

17 h3 17 g5? is absolutely anti-positional because it provides Black with a strongpoint on f5 and leaves the g6-bishop totally secure. The kingside is then ‘deadened’, and the dark-squared bishop on d2 becomes bad. 17...Qb6 Sooner or later the h-file will be opened; Black just needs to choose the best moment. In principle, it is acceptable to do so now; 17...hxg4 18 hxg4 Rxh1 19 Rxh1 Qb6 20 Rf1 Qd4 (D) and then:

White to play 18 Rhf1 18 Rdf1?! is a tempo-loss in comparison with the examples given just above. 18...hxg4 19 hxg4 Rxh1 20 Rxh1 Qd4 = Luther-Jackelen, Bundesliga 1993/4. In the previous note, the black queen had only reached b6 at the equivalent moment. 18...hxg4 19 hxg4 (D)

White to play a) 21 c3 weakens the light-square complex and is unpromising in view of the forced line 21...Qe4 22 f5 exf5 23 gxf5 Bxf5 24 Qxe4 Bxe4 25 Rxf7 Rh8, when an approximately equal ending emerges, Golubović-Wahls, Berne 1995. b) It is also a bit too early to push 21 f5?! exf5 22 gxf5 in view of 22...Bh5!, distracting the white queen from its duties. 23 Qxh5 Qxc4 gives Black the initiative.

Black to play This position is important for the evaluation of the whole variation. Faced with the menace of the f5 advance, Black must create counterplay as a matter of some urgency.

c) 21 Bb3 Re8 (the rook lies in ambush, in preparation for the opening of the e-file) 22 Qh2 (22 f5 exf5 23 gxf5 Rxe5 =+) 22...Be4 23 Re1 Rd8 24 Qe2 (Morović-Wahls, Cienfuegos 1996) and here Black should play 24...Bg6 25 Qf3 Rh8 (Black is managing to create sufficient counterplay) 26 f5 exf5 27 gxf5 Rh2 28 Ba5 Bh5. All the black pieces are active, and the situation is unclear.

19...Qd4

We now return to 17...Qb6 (D):

b) 19...Qc5 has the point that the queen is eyeing c2. Then:

Let’s consider other continuations: a) 19...Rh3?! is weak because it allows the breakthrough 20 f5! exf5 21 e6!. White’s attack is extremely dangerous; e.g., 21...f4 22 Rxf4! Nxf4 23 Bxf4 Rxd1+ 24 Qxd1 Qd8 25 exf7 +/–.

b1) 20 f5?! is an over-ambitious approach: 20...exf5 21 e6 (one must constantly bear in mind that with the rook on d1, 21 gxf5? doesn’t work due to 21...Bh5 –/+) 21...f4! (White’s activity has 119

22...Qxe2 23 Bxe2 Rxd8 (D)

turned against him, thanks to Black’s accurate response) 22 exf7 Bxf7 23 Bb4 Nxb4 24 Bxf7 (T.Ernst-Westerinen, Gausdal 1992) and after the unsophisticated 24...Rxd1+ 25 Qxd1 (or 25 Rxd1 Na6 =+) 25...Nd5 =+ Black has a healthy extra pawn. White’s expansion in the centre just ended up costing him material. b2) 20 Bb3 (D).

White to play Black has given up an exchange in return for two pawns; this looks reasonable. The material balance is not upset too much, and Black just has to coordinate his forces. 24 Bf3?! Black to play

White can keep an initiative by 24 c4! Nb6 25 e6!. This pawn was weak anyway, so it makes sense for White to give it up on his own terms; the black pawn on g7 will now drop. After 25...fxe6 26 Rg1 Be8 27 Rxd8+ Kxd8 28 Rxg7 += White’s initiative might not reap material rewards, but Black will have to tread carefully.

20...Rh3 (hinting at ...Rxb3) 21 Rf3 Rdh8 (Black never gives his opponent any respite; even though Black’s threats are not deadly, they distract White from his own aggressive ideas; now ...Rh2 is in prospect) 22 Rdf1 Rh2 23 Qe1 = Psakhis-Wahls, Baden-Baden 1992. Now 23...Re8!? is interesting, lying in ambush: 24 f5 exf5 25 gxf5 Bh5. The situation is absolutely unclear. Black’s pieces grip the board like octopus tentacles.

24...Ne3 25 Rxd8+ Kxd8 26 Rh1 (D)

We now return to 19...Qd4 (D):

Black to play 26...Kd7?! White to play

After 26...Nc4!? Georgiev might have regretted not having the ‘painful tooth’ (the e5-pawn) extracted earlier. Now the intrusion on the eighth rank by 27 Rh8+? Kd7 28 Rb8 Nxe5 29 Rxb7+ Kc8 30 Rb3 f4 leaves White a little worse; the setup g7-f7-f4 is extremely effective. Even after 27 Re1 Ke7 it is clear that White is not the one playing for the win.

20 f5 20 Ba5? Nxf4 –/+. 20...exf5 21 Bg5 Qxg4 22 Bxd8 22 Qxg4 fxg4 23 Bxd8 Rxd8 (23...Kxd8!?) is unclear.

27 Rh4 Ke6 28 Ra4 Ng4? 120

burst into the enemy camp. Meanwhile, the black queen seeks to cause White discomfort from squares like d4 or c5.

I know how difficult it is to play faultlessly and don’t wish to be too critical of the players. I shall just state that unfortunately the logical thread has been broken, and the game now ends abruptly in just five moves.

In some cases, Black will place a rook ‘in ambush’ (Morović-Wahls and Psakhis-Wahls), preparing to gain counterplay along the open lines that are created when White executes his central breakthrough. Using military terminology, it reminds me of a cannon placement inside the fortress gates. If the attacking army storms the gate – it gets blasted head-on!

Here 28...a6!? looks best: 29 Rb4 f4 30 Rxb7 (30 Rxf4 Kxe5 is unclear) 30...Bxc2. Of course, it is unpleasant for Black to give up the pawn in the middle of his queenside phalanx, but White has his own problems because it is not clear how best to halt Black’s kingside advance.

In our main game, Black gave up an exchange in very reasonable fashion. After that the logical thread was lost as both players committed inaccuracies.

29 Bxg4 fxg4 30 Rxg4 Kxe5 31 Kd2 Kf6 32 Rb4 b5 33 a4 (D)

6 Ne5 Our next game is a rarity: a game in a world championship match featuring the Scandinavian Defence! And despite the eventual result of the game, the opening was a success for Black. Garry Kasparov chose the plan of putting his knight on e5 and supporting it on that square by playing f4. We shall also look at a variety of plans for both sides in the lines after 6 Ne5.

Black to play Black is worse, but could have put up stern resistance. Instead...

Game 29

33...a5??

G. Kasparov – Anand

33...a6 is best met by 34 a5! +/–, fixing the weak pawn on a6, which White plans to win by playing Rh4-h8-a8. Meanwhile the a5-pawn is absolutely invulnerable, as it is on a dark square and the black king is far away. Instead 34 axb5?! axb5 35 c4 bxc4 36 Rxc4 Ke5 37 Rxc6 Be4 += offers White far less.

PCA World Ch (game 14), New York 1995 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 d4 Nf6 5 Nf3 c6 6 Ne5

34 axb5 1-0

We already know this manoeuvre from Chapter 1; the only difference is that there the black queen was on d6.

Alas, the white pawn is unstoppable after 34...axb4 (34...cxb5 35 Rxb5 a4 36 Ra5 +– is also hopeless for Black) 35 b6 +–.

In the following miniature, White put his knight on e5 one move later, and things went quickly awry: 6 Bd2 Bf5 7 Ne5 Nbd7 (D).

Conclusions White’s main idea in this line is either to trap the f5-bishop or to carry out a breakthrough in the centre and on the kingside. Once 16 f4 is met by 16...h5 it is inevitable that the h-file will be opened, allowing a black rook (or both of them) to 121

a) 7 Bd3 (the idea of exchanging bishops does not seem very appropriate here; why simplify?) 7...Bxd3 8 Qxd3 Nbd7 9 Nc4 Qa6 10 Qe2 (NayerLaznička, Erevan 2014) and here the simple 10...e6 equalizes. b) White may include 7 Bc4 e6 before playing 8 g4 (D) (there is some logic to White’s choice: now the black bishop has no choice and can’t go to e6; the slower 8 Bb3 allows Black time to attack the centralized knight by 8...Nbd7 9 Nc4 Qd8 10 Qe2 Be7 11 Bf4 Nb6 12 Nxb6 axb6, when in Hamdouchi-Kotronias, Erevan Olympiad 1996 Black’s game was preferable – his pawns and pieces are arranged more harmoniously).

White to play 8 g4 Be6 (the situation is already unclear) 9 f4 00-0 10 Qf3?! Nxe5 11 dxe5?? (Sutovsky has mixed up something major, and reaches a hopeless position by move 11 as White; 11 fxe5 Nxg4 =+) 11...Rxd2! 12 exf6 (alas 12 Kxd2 fails to 12...Bd5 13 Qe3 Ne4+ followed by ...Nxc3 and ...Bxh1) 12...Rxc2 13 Bd3 Rxb2 14 Rc1 exf6 15 0-0 h5 16 g5 fxg5 17 Nd1 Bc5+ 18 Kh1 g4 0-1 SutovskyVan Wely, Istanbul Olympiad 2012. This game reminds us of a simple truth: even White can’t play whatever he wants; one must always be careful. We now return to 6 Ne5 (D):

Black to play 8...Bg6 9 h4 (this looks imposing, but Black has everything in order and can systematically find good roles for each of his pieces) 9...Nbd7 10 Nxd7 Nxd7 11 h5 Be4 (forced but sufficient) 12 0-0 Bd5 13 Nxd5 cxd5 14 Bd3 Bd6 15 Bd2 Qd8 with approximately equal chances, FedorovNisipeanu, European Team Ch, Batumi 1999. The bishop-pair is an undeniable asset, but White has considerably weakened his king’s cover whereas Black’s structure is intact.

Black to play

c) The direct 7 g4 (D) seems the most dangerous for Black.

Black must now decide where to put his queen’s bishop. 6...Be6!? The bishop often hides on this square, ‘under the cover’ of the enemy knight. I view 6...Bf5 as somewhat risky. Why provoke an advance of White’s pawns? The white knight on e5 is already in a good position to support a broad kingside advance. He can push forward in a couple of different ways:

Black to play 122

He must now decide where to put the bishop: c1) 7...Bg6?! has scored disastrously for Black after 8 h4! (now the play takes on a forcing character, and no matter how Black struggles, he seems unable to set up a viable defence; 8 g5?! provides Black with the f5- and h5-squares and after 8...Nd5 9 Bd2 Nxc3 10 Bxc3 Qd5 11 f3, Zhang Pengxiang-Yu Ruiyuan, Tianjin 2013, the prosaic 11...e6 = leaves Black with no problems) 8...Nbd7 9 Nc4 Qc7 10 h5 Be4 (by following a narrow winding path, Black manages to balance on the edge of the abyss but, of course, we don’t even mention equality) 11 Nxe4 Nxe4 12 Qf3 Nd6 13 Bf4 (D) and now either way the knight moves there is a lamentable outcome:

White to play The only distinction from our main game (with the immediate 6...Be6) is that here the white pawn is on g4. Who benefits from this difference? The verdict from practice is not so clear; White has a wide choice but in all lines Black gets a more or less normal position: c21) 8 Nc4 Qc7 9 g5 Nd5 10 Ne4 Nd7 11 Na3?! (an attempt to be creative, presumably, but this looks odd given that the knight was not under attack, and it is not a surprise that White soon ran into problems) 11...Bf5 (11...g6 =+ is a good and logical alternative) 12 Ng3 Bg6 13 c4 Nb4 14 f4 e6 15 Qf3 0-0-0 16 Bd2 Qb6 17 Bc3 h6 18 gxh6 Rxh6 –/+ Pikula-S.Sarić, Kragujevac 2010. The knight placement on the edge of the board (a3) can’t be considered sound.

Black to play c11) 13...Nf6 14 h6 Nd5 15 Be5 +/– (I don’t think Black should go for such positions and I hope you agree with me) 15...f6 16 hxg7 Bxg7 17 Bg3 0-00 18 0-0-0 b5 19 Nxd6+ exd6 20 Bxb5 +– Macieja-Tomczak, Polish Ch, Opole 2007.

c22) 8 Bd2 Qb6 9 f4 Qxd4 (not everyone would dare to play this greedy move, and it may well be a poor decision; 9...g6 is more prudent) 10 Qe2 Nbd7 11 Nf3 Qd6 (11...Qb6!?) 12 f5 Bd5 (Pavlović-Šolak, Ljubljana 2003) and now 13 g5! gives White more than enough compensation for the pawn.

c12) 13...Nb6 14 Ne3 (besides this move White has a wide choice; for example, 14 Ne5 Nd5 15 Bg3 {15 Nxf7!?} 15...Qa5+ 16 c3 e6 17 Be2 Be7, Gallagher-Turpanov, Bled Olympiad 2002, 18 0-0 +/–; the bishop-pair and spatial plus speak in White’s favour) 14...Na4?! (when you are well behind in development, it is rarely a good idea to seek activity; better is 14...e6 15 0-0-0 h6, somehow holding on) 15 0-0-0 Qb6? 16 b3 Nc3 17 Bxd6! Nxd1 (17...exd6 is even worse in view of 18 Nc4, when White wins after 18...Qb4 19 a3 or 18...Nxa2+ 19 Kb2 Qa6 20 Nxd6+) 18 Bc5 Qa5 19 Nxd1 Qxa2 20 Nc3 with an overwhelming position for White, Vlassov-Malakhov, Moscow 1996.

c23) 8 Bg2 Nbd7 9 Nxd7 Bxd7 10 g5 Nd5 11 0-0 Nxc3 12 bxc3 e6 13 Rb1 Qc7 14 c4 Bd6 15 Qh5 0-0 16 c5 (D) was played in Macieja-Papaioannou, Achaia 2013.

c2) The central retreat 7...Be6 (D) is more solid.

123

to equality. Of course, the two bishops are a significant factor, but Black has no weaknesses and his pieces are perfectly mobilized. b) 7 Bc4 (D) seems to me rather harmless as each exchange is welcome for Black, especially when it means an end to any worries about his light-squared bishop being hunted down by the white pawns. Black to play White has a space advantage and his rook might in some cases manoeuvre via the third rank (Rb3h3/g3). The d7-bishop is still out of the picture but at least there is no immediate catastrophe facing Black. We now return to 6...Be6 (D): Black to play 7...Bxc4 8 Nxc4 Qd8 (the typical manoeuvre 8...Qa6!? is interesting, especially after the exchange of light-squared bishops; e.g., 9 b3 e6 10 0-0 Nbd7 11 Bf4 Nb6 12 Nd6+ Bxd6 13 Bxd6 Qa5 14 Qf3 Rd8 15 Be5 Nbd7 16 Bg3 0-0 is about equal, A.Sokolov-Chabanon, French Team Ch 2002; Black happily finishes his development having exchanged two minor pieces) 9 Bg5 e6 (9...Nbd7!? also looks logical, planning, if necessary, to capture the f6-knight) 10 Bxf6 gxf6 11 0-0 Na6 12 Re1 Bg7 13 Qh5 0-0 14 Rad1 b5 15 Ne3 f5 16 Ne2 Qd6 17 c3 Nc7 was played in IllescasAzmaiparashvili, Madrid 1996. There are positive things about Black’s position. His bishop controls the dark-square complex and his pawns cover the light squares. In the future the g-file may become useful.

White to play When I was young it seemed artificial to put the bishop in front of a central pawn, but after many years playing the Scandinavian Defence I have got used to such a development. Black’s plan is as follows: the bishop on e6 controls d5 and c4, thus discouraging White from developing his bishop to c4, and making a successful d5 pawn-break highly unlikely. The f5-square will be guarded by the g6pawn and the other bishop is developed to g7.

We now return to 7 Bd3 (D):

7 Bd3 Avoiding exchanges. Other continuations: a) 7 Nc4 is a manoeuvre with which we are already familiar, but it causes Black few problems here. 7...Bxc4 (Black can also choose 7...Qd8, preserving his bishop) 8 Bxc4 e6 9 0-0 Be7 10 Qd3 Nbd7 11 Ne4 Rd8 12 Bf4 (Rabiega-Beliavsky, Austrian Team Ch 1996/7) 12...Nxe4 13 Qxe4 Nf6 is close

Black to play 124

7...Nbd7 7...g6 8 0-0 Bg7 is another move-order. Then: a) 9 Ne2!? is interesting but hardly promises an advantage, as White relaxes his grip on the central squares. 9...00 10 c4 Nbd7 11 Bd2 (11 Nf4 looks more consistent with the knight manoeuvre to e2, but after 11...Nxe5 12 dxe5 {12 Nxe6? Nxd3 13 Nxf8 Nxc1 –/+} 12...Qxe5 13 Re1 Qd6 14 Nxe6 fxe6 the position is dynamically balanced) 11...Qb6 12 Nxd7 Bxd7 13 Qc2 Rfe8 14 Rad1 e5 (the predictable response to White’s slow plan) 15 dxe5 Rxe5 16 Bc3 Re7, E.Hossain-Kotronias, Hamburg 2001. Following the exchange of the centralized e5-knight and the annihilation of the d4-pawn, the position is absolutely equal.

Black to play 10...Bf5!? A weighty decision. 11 Bc4?!

b) 9 f4 Bf5 10 Bc4 0-0 11 g4 takes up an aggressive stance, but is it any good? His own king is exposed. After 11...Bc8 12 Be3 Nbd7 13 Be2 Nxe5 (13...Qc7!?) 14 fxe5 Nd5 15 Nxd5 Qxd5 16 Bf3 Qa5 17 Qe2 Be6 18 c4 Rad8 19 b3 the game is roughly equal, T.Heinatz-Lau, Bundesliga 1999/00. Now that the knights are exchanged, Black has more air to breathe. White has more space, but Black has no weaknesses.

After 11 Bxf5 gxf5 12 Qd3 e6 13 b3 White stands better. He can try the idea a4 and Ba3, or simply develop the bishop to d2. 11...e6 12 Be2 h5 It turns out that on f5 the bishop feels quite comfortable. 13 Be3 Rd8 14 Bg1 0-0 15 Bf3 Nd5

We now return to 7...Nbd7 (D):

I like this manoeuvre but it is usually appropriate only if it forces the exchange. 15...c5!? is interesting, although the opening of the h1-a8 diagonal is welcome to the f3-bishop. 16 Nxd5 It looks tempting to avoid the exchange, intending to play c4, but there are tactical problems: 16 Ne2? Nxf4! 17 Nxf4 Nxe5 –/+ or 16 Ne4 Nxf4 =+. 16...exd5 17 Bf2 Qc7 18 Rc1 (D)

White to play 8 f4 Moves like 8 Nxd7 don’t deserve detailed consideration if White to seeking to gain any significant advantage. In some positions I won’t even mention options like this; my implicit comment in those instances is: “there is no point exchanging the centralized knight because almost every exchange makes Black’s life easier”. 8...g6 9 0-0 Bg7 10 Kh1 (D)

Black to play 125

Black’s position looks good. The e4-square is a nice post for a bishop or a knight, while the e5knight isn’t so secure because of the possibility of ...f6.

28 Nc6 Rd7 =+. 28...hxg4 29 Nxg4 Bg7 Or 29...Ne4!?, with the point 30 Rc7 b5 31 Rxa7? Qb6 –/+.

18...f6 Houdini recommends an ‘inhuman’ exchange sacrifice: 18...Bh6!? 19 Bg3 Nxe5 20 dxe5 h4 21 Bxh4 Bxf4 22 Bxd8 Rxd8 23 Rb1 Qxe5 24 g3 Be3 =+. This would be hard for a carbon-based player to venture.

30 Rc7 Ne4

19 Nd3 Rfe8 20 b3 Nb6

31 Rxa7 would again be met by 31...b5!, planning ...Qb6, with the better chances for Black.

30...Be4 31 Bxe4 Nxe4 32 f5 leaves White with the initiative. 31 Ne3 (D)

Usually the knight is poorly placed on this square if its prospects are stifled by a b3-pawn, but the knight was already passive on d7 (facing the d4pawn), so it begins a journey along the route b6c8-d6-e4. 21 a4 Nc8 22 c4 Qf7 23 a5 Bf8 23...Nd6 24 a6 b6 25 Nb4 gives White the initiative. 24 cxd5 cxd5 25 Bh4 Nd6 26 a6 b6?! 27 Ne5!? (D)

Black to play 31...Bh3? After 31...Rc8 32 Rxc8 (32 Rxa7? Nc3 33 Qd2 Qxe3 –/+) 32...Rxc8 33 Nxf5 gxf5 a dynamic balance is maintained. Each player has some pluses to boast of. 32 Rg1

Black to play

All the white pieces are very active. Black is especially bothered by the rooks on the seventh rank and the g-file.

Wow! There is no attack in the offing. But Garry is right, as usual. Other continuations would have left him in a slightly worse position.

32 Bg4 Bxg4 33 Qxg4 Qxg4 34 Nxg4 +/– is also possible; the c7-rook is the lynchpin of White’s game.

27...Qe6!

32...g5

After 27...fxe5 matters are unclear: 28 fxe5 Ne4 29 Bxd8 Rxd8 30 g4 hxg4 31 Bxg4 Bxg4 32 Qxg4 Nf2+ 33 Rxf2 Qxf2 34 Qxg6+ Bg7 35 Rc7 Qf1+ 36 Qg1 and now Black must force a draw by 36...Qf3+ 37 Qg2 Qd1+ because after 36...Qxg1+?! 37 Kxg1 he is not the one who is playing for a win.

32...Rd7!?. 33 Bg4 33 fxg5!? is even stronger: 33...fxg5 34 Rxa7 +/– (not 34 Bxg5?? Nf2#). 33...Bxg4 34 Qxg4 Qxg4 35 Rxg4 Nd6 36 Bf2 Nb5 37 Rb7 Re4 38 f5 Rxg4

28 g4! 126

examples after 7...Bg6?! 8 h4!. The other retreat 7...Be6 is more solid, when the question is how useful the extra move g4 really is. In MaciejaPapaioannou we saw White seize the initiative after exchanging knights on d7, while PavlovićŠolak featured White sacrificing his d-pawn (which Black didn’t have to accept) for solid compensation. Overall, I feel Black shouldn’t push his luck by allowing White to grab space with gain of time.

After 38...Rxd4? 39 Rxd4 Nxd4 40 Rxa7 Nxb3 41 Rb7 Nc5 42 Rxb6 White has a considerable advantage. 39 Nxg4 Rc8 40 Rd7 (D) 40 Kg2!?; 40 h4!?.

That brings us to 6...Be6, as played by Anand in the main game. Now if White plays 7 Bc4, he only achieves simplifications and approximate equality (A.Sokolov-Chabanon and IllescasAzmaiparashvili). Black to play

The manoeuvre 7 Nc4 is not too dangerous either. Black can either retreat his queen or simply exchange on c4 (Rabiega-Beliavsky). Both options have acceptable consequences.

The situation has been significantly simplified. White has some advantage due to his piece activity. The f5-pawn also cramps Black’s game.

In our main game, Garry chose 7 Bd3, putting neither piece on c4, and so avoiding exchanges.

40...Rc2?? This time-trouble blunder ruined the game. Correct was 40...Rc3 41 Ne3 Rxb3 42 Kg2 += (instead, 42 Nxd5 Ra3 43 Ne7+ Kh7 44 d5 Rxa6 45 d6 Ra1+ 46 Kg2 Rd1 is not so clear).

Bringing the knight from c3 to e2 yields no special dividends if Black responds adequately; in E.Hossain-Kotronias we saw a successful ...e5, which is logical since White has lessened his control of the centre. In essence, in positions of the Caro-Kann-type (as well as in many others), White’s claim to an advantage is generally based on the presence of the d4-pawn in the centre, whereas Black has no pawns in this important part of the board. Thus, the annihilation of the d4pawn usually leads to equality.

41 Rxd5 1-0 Now Black’s game is hopeless – a shame after such a good start.

Conclusions

In our main game, Anand had a pleasant, weakness-free position for a long time. Garry obviously felt somewhat uncomfortable because of the weak e4-square. However, with skilful piece-play he managed to confuse his opponent and escape from his somewhat dubious position. In particular the white rook on the seventh rank caused the black king many problems.

Since my childhood I have been a fan of my namesake, the 13th World Champion. So in our main game everything was perfect for my purposes. Firstly, Anand improved the reputation of the Scandinavian Defence (Black’s position was even preferable), but then Garry proved more resourceful and managed to win the game. Sutovsky-Van Wely featured an opening disaster for White. Sutovsky made a few careless moves and was almost instantly beaten.

Rare Plans for White To end this chapter, I wish to alert you to the problems presented by a very different idea for White, namely the attempt to take advantage of

6...Bf5 seems a little risky to me. After 7 g4 (which is the most testing reply), the retreat to g6 leads to serious problems; we saw some sad 127

but now it is just a target) 18 Qc3 Bg4 19 Nf4 Rac8 gave Black the initiative in BerchtenbreiterNisipeanu, Bad Wiessee 2013.

the queen’s position on a5 by means of the aggressive advance b4-b5. We examine two distinct ways to do so. In Game 31, White first fianchettoes his king’s bishop and only then advances his b-pawn. Clearly the b7pawn is in his sights.

b) 5...Qc7 (D) looks sturdier, but the queen might sometimes be subject to a tempo-gaining attack by Nd5.

But first we shall see what happens when White opts for 4 Rb1!?, directly preparing to advance the b-pawn before deciding how to develop the rest of his pieces. Game 30 Charbonneau – Hungaski USA League 2010 White to play

1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 Rb1!? (D) For instance:

b1) 6 Bc4 Nf6 7 Nge2 is too modest a development to promise much of an advantage. The knight is usually more active on f3. 7...Bf5 (after 7...b5!? 8 Bb3 e6 Black’s position is at least not worse; later on he can choose ...a5 or ...a6, depending on what White does) 8 d3 e6 9 b5 Bd6 10 bxc6 bxc6 11 h3 0-0 12 0-0 Nbd7 and Black stood better thanks to her superior control of the centre in Skripchenko-Bollengier, French Women’s Ch, Nîmes 2014.

Black to play This is the most original line for White: the bpawn will rush forward immediately. How should Black react to such a direct plan?

b2) 6 b5 (D).

4...Nc6 This is the move for those who like piece-play. The main alternative is 4...c6. On the one hand, this provides a retreat for the queen, but on the other hand it may provide a ‘hook’ for White’s b4-b5 advance. We can only determine which is more significant by analysing some lines. So, 5 b4, and then:

Black to play

a) 5...Qd8 6 g3 Nf6 7 Bg2 g6 8 Nge2 Bg7 9 0-0 00 10 b5 cxb5 11 Rxb5 Nc6 12 d4 Ne8 (the knight is redeployed to d6, where it supports the b7-pawn) 13 Ba3 Nd6 14 Rd5 (an exotic manoeuvre; meanwhile Black consolidates his grip on c4) 14...Na5 15 Bxd6 exd6 16 Ne4 Nc4 17 Qd3 Qc7 (a drawback of White’s set-up is the backward c2pawn; if it stood on c4, shoulder-to-shoulder with the d4-pawn, it would be a force to reckon with,

6...Nf6 (it is desirable to keep the b-file closed, but after 6...c5 7 Nd5 Black may regret that he didn’t drop the queen back to d8 immediately) 7 Nf3 (7 h3 g6 8 Nf3 Bg7 9 Bd3 0-0 10 0-0 cxb5 11 Nxb5 Qd8 12 Re1 Nc6 13 a4 a5?! 14 Ba3 += C.Bauer-Bonnaud, Le Port Marly 2012; White has a solid advanced post on b5) 7...e6 8 g3 Be7 9 Bg2 0-0 10 0-0 Bd7 (this looks safe but what future now lies ahead for the b8-knight? White is 128

The smoke has cleared and we can draw our first conclusions. The doubled c-pawns are by no means as weak as it might seem to an inexperienced player. The c-file is closed, so frontal pressure is impossible. The exposed position of the b4-pawn raises hopes for an exchange by ...c5. Besides, with dark-squared bishops the fact that the front pawn (c6) is situated on a light square is also encouraging for Black.

not obliged to capture on c6) 11 a4 cxb5 12 axb5 Rd8 13 Nh4 Bc8 14 Qe2 Nbd7 15 Nf5, EhlvestDe Guzman, Las Vegas 2002. Actually this is not too terrible for Black, but visually I don’t like his position. The b5-pawn performs as a ‘breakwater’, holding back Black’s a- and b-pawns. The faraway g2-bishop also exerts pressure on Black’s queenside. We now return to 4...Nc6 (D):

12 d3 Nd5 13 Ne4 e5 14 a3 a6 15 Bb2 f6 16 0-0 Be7 17 Rfe1 Rhe8 18 d4?! 18 g3 is more prudent, still not revealing White’s intentions while continuing to improve the pieces’ positions (Kg2, etc.). 18...exd4 19 Bxd4 Bf8

White to play

I don’t know why White allowed and Black avoided the uncomplicated 19...Nxb4!; after 20 axb4 Rxd4 21 Nc5 Rd5 22 Re6 Re5 23 Re1 Rxe6 24 Rxe6 Kd8 White must be very cautious, as his king is far away from the battlefield, though objectively it appears to be about equal.

Planless play brought the higher-rated player a rapid defeat in Galego-Hernando Rodrigo, Dos Hermanas 2002: 5 Bb5

20 c3 Re6 21 Nd2 Rde8 22 Rxe6 Rxe6 23 Kf1 Nb6 24 c4?!

5 Nf3

Now this is in vain, as now the black rook is active.

Nf6 6 b4 Qb6 7 d4?! (combining b4 and d4 is probably not the best idea) 7...Bg4 8 f3 Bf5 9 g4?! (the pseudo-active 9 d5? doesn’t work because of the standard reply 9...0-0-0 10 Bxc6 bxc6 –+ and the d-pawn falls) 9...Bg6 10 Be3 0-0-0 11 Nge2 e5! 12 Bxc6 Qxc6 13 0-0 exd4 14 Bxd4 Qc4 15 Qd2 Bxb4 –+.

24...Rd6 25 Be3 Rd3 26 Ra1 Rc3 27 c5 Nd5 28 Ke2 a5! (D)

5...Nf6 6 Bb5 Bg4 7 b4 Qb6 8 h3 Bxf3 9 Qxf3 00-0 Black operates in quite a simple way; all his moves are logical and unsophisticated. 10 Bxc6 Qxc6 11 Qxc6 bxc6 (D) White to play Now the isolated and doubled pawns appear in rather a good light! 29 Ne4 Rb3 30 bxa5 Nxe3 31 fxe3 f5 32 Rf1 g6 This is still no disaster for White, though he must continue precisely. White to play

33 Nd2? 129

Game 31

33 g4! Bh6 34 g5 Bg7 35 Nf6 offers White counterplay.

Yudasin – Bartholomew

33...Rxa3 34 e4 Bh6 35 Nc4?! Rc3 36 Ne5 Rxc5?! 37 Nf7 Bg7 38 a6 Ra5 39 exf5 gxf5 40 Ng5 h6 41 Ne6??

Philadelphia 2008 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qa5 4 g3 (D)

This must have been a time-trouble blunder. White has chances of saving the game after 41 Nf3. 41...Re5+ 42 Kd3 Rxe6 0-1 The passed a-pawn is stopped by the king.

Conclusions After 4 Rb1, Black has two basic options.

Black to play

4...Nc6 is based on ‘active defence’. Black seeks piece-play and relies on counterthreats to parry White’s queenside aspirations. Since the black queen is not provided with an easy retreat after White’s b4 advance (which is inevitable sooner or later), I can recommend this option only to players with nerves of steel. On the other hand, White will also have to calculate some concrete variations. Playing by ‘general considerations’ may lead to disappointment (Galego-Hernando Rodrigo).

4...c6 We shall examine two other options, both featuring ...Nc6 rather than ...c6: a) 4...Nc6 leads to lively piece-play, but I think White has the better chances here. 5 Bg2 Nf6 6 h3 Bf5 7 d3 e5 8 Bd2 (lying in ambush; White does nothing too committal for the time being, and just develops his pieces and keeps a ‘coiled-spring’ position) 8...Qc5 9 Nf3 Be7 10 0-0 0-0 11 Re1 Rfe8 (D) and now:

In our main game, Black equalized rather easily and in the ending Black even took over the initiative. However, please do not form the impression that 4...Nc6 is a simple solution to all Black’s problems in this line; please examine the critical lines carefully. A more solid option is 4...c6, which has a CaroKann look about it. It provides a safe retreat for the black queen, but also offers a target for White’s b-pawn advance; for instance, the b-file might be opened by b4-b5xc6. After analysing these lines I got the impression that in general White exerts some pressure, but that this is very much a line where whoever plays the better chess will win!

White to play a1) After 12 a3 Bf8 (Rozentalis-Tomczak, Warsaw rapid 2009) 13 g4 White’s chances are definitely preferable. His position is flexible, and his pawns can push aside enemy pieces on both flanks.

In our next game, the b-pawn advance is prepared in more leisurely fashion: White fianchettoes his king’s bishop first of all. Once again, Black’s principal choice is between set-ups based on ...c6 and ones with ...Nc6.

a2) It is strange that White avoided the natural 12 b4!, inviting Black to open the b-file. It is obviously unfavourable for Black to capture the b130

White can also combine the b4 advance with placing the knight on e2, which has the benefit of leaving the long diagonal open: 6 Nge2 Bf5 7 b4!? (exploiting the unprotected b7-pawn) 7...Qc7 8 Rb1 e5 9 d3 Bd6 10 0-0 0-0 11 f4 (this looks like a good idea, as the f1-rook casts a hungry glance at the f5-bishop) 11...exf4 12 Bxf4 Bxf4 13 Rxf4 Bg6 14 b5 (your computer may not be too worried here, but from a human point of view White enjoys a strong initiative: his rooks are hyperactive, the g2-bishop is impressive, while the b5-pawn acts like a wrecking ball against Black’s underdeveloped queenside) 14...Nbd7 15 bxc6 bxc6 16 Nd4 += J.Friedel-Dzhakaev, Pardubice 2011.

pawn, as he loses a central pawn in return: 12...Nxb4?? loses to 13 Rxe5 +–, while after 12...Qxb4 13 Nxe5 +/– the g2-bishop and the b1rook will make Black’s life on the queenside unbearable. However, retreating by 12...Qd6 also results in the loss of a pawn: 13 b5 Nd4 14 Nxe5 +/–. b) 4...Nf6 5 Bg2 Bg4 6 Nf3 Nc6 7 h3 Bh5 8 0-0 e5 (D).

6...Bf5 (D)

White to play 9 d3 Bb4 10 Bd2 0-0-0 11 g4 Bg6 12 Nh4 e4 (this central break looks reasonable given the opposition of the queen and the rook) 13 Nxg6 (13 a3!? Bxc3 14 Bxc3 offers White the initiative) 13...hxg6 14 g5 Bxc3?! (after 14...Nh7! 15 Qg4+ Kb8 16 a3 Bxc3 17 Bxc3 Qxg5 it is difficult to give an unequivocal evaluation; Black’s pieces, particularly his rooks, are rather active) 15 bxc3 Nd7 (the capture on e4 entails either the loss of the h3-pawn or problems on the d-file) 16 d4 gave White an auspicious position in KülaotsBartholomew, Las Vegas 2014, but after 16...Rh4 Black has some counterplay.

White to play 7 0-0 In Godena-Predojević, Rijeka 2010 White chose the immediate 7 b4!? but then didn’t push the pawn to b5, keeping it as a constant threat. Was this a good and creative plan, or was he showing excessive respect to his high-rated opponent? Let’s delve. 7...Qc7 8 0-0 e6 9 Rb1 Be7 10 d3 0-0 11 Re1 (11 b5 could well be considered here: 11...a6 12 b6 with slightly better chances for White, though it is difficult to predict the fate of the breakaway pawn) 11...Nbd7 12 Bf4 Bd6 13 Bxd6 Qxd6 14 Nd2 Rfd8 15 Nc4 Qc7 16 Qd2 Rac8 17 a4 Nb6 18 Ne3 Bg6 19 a5 Nbd5 20 Nc4 Ne7 left Black is cramped and lacking counterplay, and advancing his b-pawn would weaken the long light-square diagonal. White could now have chosen 21 Na4 +=.

5 Bg2 Nf6 (D)

7...e6 8 h3 Be7 9 Nh4 Bg6 10 Nxg6 hxg6 11 a3 Nbd7 12 d3 0-0 13 Rb1 Qc7 14 b4 a6 15 Re1 Rfe8 16 Qe2 Rac8 17 Nd1 (D)

White to play 6 Nf3 131

to point out lines like this after the game, working with a computer, but it is far more difficult to stay calm at the board when facing two terrible bishops. 27 Qf3 Nxc4 Now both the bishop and the queen attack f6, promising trouble for Black. 28 Bxf6 gxf6 29 bxc5 Qc6? Black to play

This was probably a time-trouble mistake; now the end is really close.

White regroups in order to continue his queenside expansion in the most beneficial circumstances. Black must take action before White consolidates.

The right line is 29...Qxa3! 30 c6 Rd6 31 Qxa3 Nxa3 32 Rb7 Nb5 33 Rxd5 Rxc6 34 Rdd7 +/–. Black has serious problems but the fight continues.

17...b5!

30 Rxd5 Nd2

This may not promise equality, but at least it disrupts White’s plans.

30...Rxd5 31 Qxd5 Qxd5 32 Bxd5 Nxa3 33 Rb7 Rf8 34 c6 Nb5 35 f4 +–. Black is paralysed and the white king is ready, if necessary, to set out for the queenside.

18 c4 c5! 19 Ne3 Red8 19...Nb6!?.

31 Qd3 Nxb1 32 Rxd7 Re1+ 33 Kh2 Qxc5 34 Qb3 Re6 35 Qxb1 Qxf2 1-0

20 Bb2 Nb6 21 Ba1 Qd7 22 Red1 bxc4 23 dxc4 Qa4 24 Nd5!? Sharply played, but I can’t say it was necessary. 24 Rxd8+ Rxd8 25 Be5 keeps the initiative.

Conclusions You can see how Black might suffer in the ‘active line’ with ...Nc6. White gained the advantage in both Rozentalis-Tomczak and KülaotsBartholomew.

24...exd5 25 Qxe7 Rd7 Perhaps 25...Nxc4 is more accurate. After 26 Bxf6 Rd7! 27 Qe2 gxf6 28 Rxd5 Nxa3 it is highly probable that all the pawns will be exchanged on the queenside, leading to a draw.

However, White can obtain interesting play against the ...c6 plan as well. In GodenaPredojević, White didn’t hurry to push the pawn to b5, but brought its colleague to a5.

26 Qe2 (D)

In our main game, Yudasin obtained the bishoppair and then prepared queenside play. He got the upper hand in a hand-to-hand fight and the centre opened, allowing his bishops to sweep the whole board with their fire.

Conclusions on 3...Qa5 Placing the queen on a5 seems to me less solid than the retreat to d6. White has several plans: putting his bishop on d2, creating threats of the knight leap Nd5; pressure against Black’s lightsquared bishop with the advances g4, f4 and so on;

Black to play 26...Re8? Objectively, Black should prefer 26...dxc4 27 bxc5 Rxc5 with unclear play. Of course, it is easy 132

e5! 8 d5 Nb4 (a very secure place for the knight after White has played a4) 9 Bg5?! Bf5 10 Bb3 h6 11 Bxf6 Qxf6 12 Ng3 led to comfortable equality in Naiditsch-Danielsen, Reykjavik 2014. In fact, after 12...Bg6 Black’s position could even be preferred due to his bishop-pair and the white b3bishop being hampered by its own d5-pawn.

while the idea of pushing the b-pawn is also quite disturbing for Black. 4: 3...Qd8 One doesn’t have to be a grandmaster to notice that this retreat leaves all the black pieces on their initial positions, while White has already developed one knight and now it’s his turn to move again.

b) 5 Bg5 h6 (this looks provocative: Black is behind in development and spends time inviting White to double his pawns) 6 Bh4 c6 7 Bc4 Bf5 (D).

Due to the relative rarity of this line, I shall devote just one main game to it. That said, it could hardly be a higher-level encounter: the world’s numbers 1 and 2, playing on top board in an Olympiad!

Game 32 Caruana – Carlsen Tromsø Olympiad 2014 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd8 (D) White to play 8 Bxf6!? (judging that this is a better moment to take – this decision appeals to me) 8...exf6 9 Qe2+ Be7 10 0-0-0 (perhaps castling kingside would be more promising; of course, if Naiditsch had lost having gone kingside, the commentators might have suggested queenside as a better option...) 10...0-0 11 Nf3 Bb4 12 g4 Bh7 13 Bd3 Bxd3 14 Qxd3 Qa5 was roughly equal in Naiditsch-Djukić, Istanbul Olympiad 2012. After 15 a3 Bxc3 16 Qxc3 Qxc3 17 bxc3 Nd7 neither side is in much danger of losing.

White to play This retreat looks rather modest. On the positive side, White is most unlikely to gain further time by attacking the black queen with his minor pieces. But logically, the queen must be more passive here than on a5 or d6, and it will take Black additional time to complete his development. But when a move is played by Carlsen – and against one of his main rivals – it must be taken seriously...

We now return to 5 Nf3 (D):

Before moving on, I shall mention the surprising fact that in 2700+ chess Black quite often wins in games where 3...Qd8 was played.

Black to play

4 d4 Nf6 5 Nf3

5...Bg4

Other possible continuations:

5...c6

a) 5 Bc4 a6 6 a4 Nc6 7 Nge2 133

6 Bc4 b5 is an unconventional approach. There is a temptation, on principle, to put a ‘?!’ on such early aggression by Black while still underdeveloped, but while it is hard to recommend the idea, it may not be so terrible. 7 Bb3 (7 Bd3!? +=) 7...e6 8 Bg5 Be7 9 a4 b4 10 Bxf6 gxf6 11 Ne2 Na6 12 Qd2 h5 was played in Shirov-Dizdarević, Calvia Olympiad 2004. Such exotic play (pawns on h5 and b4, the knight on a6, the bishop still on c8) places great demands on Black’s resources. After 13 0-0 Nc7 14 Nf4 Bb7 15 Rfd1 += White plans c4, a5, etc.

a1) 9 g3 Nbd7 10 Bg2 Bd6 11 0-0 0-0 12 a3 (I don’t know why Carlsen didn’t play the simple 12 Ne4 Nxe4 13 Qxe4, when White has the usual long-term plus due to the bishop-pair – with his powerful technique he could no doubt have put it to good use) 12...Qc7 13 Rd1 (White is waiting for Black to play ...e5 and isn’t seeking to take drastic action of his own for the time being) 13...Rad8 14 b3 Rfe8 15 Bb2 (it would be nice for White to put his knight on f3, controlling e5, and advance the pawn to c4; but these are dreams) 15...a6 16 Ne2 e5! (precisely when the knight has left the centre!) 17 c4 exd4 18 Nxd4 Be5 19 Qc2 c5 (I am not sure about this move as the diagonal is opened for the g2-bishop; but was there any better option?) 20 Nf3 (maybe 20 Ne2 is more precise, intending Nc3/f4-d5) 20...Bxb2 21 Qxb2 = Carlsen-Djukić, Tromsø Olympiad 2014. White no longer has the bishop-pair, while Black has a strongpoint on d4.

6 h3 Bxf3 7 Qxf3 c6 (D)

a2) 9 Be2 Nbd7 10 0-0 Bd6 11 Bg5 Qc7 12 Ne4 Bh2+ 13 Kh1 Bf4?! (Kariakin-Iotov, Tromsø Olympiad 2014) and here White missed the elegant 14 Qa3! c5 (other options are even worse: 14...0-0-0? 15 Bxf4 Qxf4 16 Nd6+ Kb8 17 Nxf7 +–; 14...Bxg5?? 15 Nd6+ +–; 14...Kd8 +/–) 15 Bxf6 Nxf6 16 Nxf6+ gxf6 17 Qa4+ +/–.

White to play This position has arisen rather naturally after 3...Qd8. The game looks similar to the Caro-Kann, but White has secured the bishop-pair rather easily, and it is hard for Black either to exchange one of them off, or else to bring about a structure where he could claim that a knight might be as effective as a bishop. This is one of the reasons why I am sceptical about 3...Qd8.

b) 8 Be3 e6 9 0-0-0 Bb4 (D) and then:

8 Ne2 White must protect his d-pawn one way or another, but is such an early retreat justified? Other options: a) 8 Qd3 is a prosaic choice, but has been used at elite level. 8...e6 (D) and then: White to play b1) The natural 10 Ne4 can be rejected because of the simplifying operation in the centre: 10...Nxe4 11 Qxe4 Qd5 (a fork! The queen attacks a2 and e4 at the same time, forcing White to exchange) 12 Qxd5 cxd5 =. b2) 10 Nb1!? is an exotic way to avoid the exchange. Then: b21) 10...Nbd7 11 g4 Nd5 12 c4 Nxe3 13 fxe3 Qf6 14 Qg3 Qe7 15 h4 0-0 16 Bd3 failed to give

White to play 134

White an advantage, even with the queens still on the board, in A.Vovk-Tratar, Trieste 2014. Now the central counterstrike 16...e5 would balance the chances. Black should concentrate on the e-file (...Rfe8), seeking to bring the queen to e3.

also doesn’t promise equality: 11 0-0 0-0 12 Rd1 e5 13 Qb3 Qc7 14 dxe5 Nxe5 15 Bf4 Nfd7 16 a4 (this makes sense: the pawn intends to squeeze the enemy and, if given a chance, to run all the way to a6, making the long-diagonal pressure critical) 16...Nc5 17 Qb4 Ne6 18 Be3 a5 19 Qb3 Nd7 20 Nc3 Ndc5 21 Qc4 Na6, Fedorchuk-Danielsen, Kolkata 2014. Despite the harmonious placement of the black pieces, the initiative is entirely in White’s hands. Here is a sample variation: 22 Ne4 Bxb2 23 Rab1 Bg7 24 Bb6 Qe7 25 Bxa5 +=.

b22) 10...Qd5 (D) is a typical idea in such structures.

9 g4 Qd5 10 Bg2 Nbd7 11 Qg3 Qc4 12 Qb3 Qxb3 13 axb3 (D)

White to play In case of an exchange on d5 Black has a pick of pawn-structures (...Nxd5, ...exd5, ...cxd5) each with differing potential benefits, depending on the specifics of the position. Here 11 Qxd5?! is rendered most harmless by 11...Nxd5 =, as the e3bishop will be exchanged. However, White can evade the exchange with 11 Qg3! since 11...Qxa2? loses to 12 Qxg7. Black has nothing better than 11...0-0 12 c4 +=, when White has achieved everything he could have wished for with his knight retreat.

Black to play I still can’t see any defects in the white position except for the slight weakening caused by the move g4. I would prefer the pawn to be on a dark square – g3. 13...Bd6 14 c4

We now return to 8 Ne2 (D):

Again I would be happier placing the pawns on dark squares (d4, c3, b2 and b4). 14...a6 15 Be3 0-0-0 16 0-0-0 Rhe8 17 Ng3 Nf8 18 Bf3 The point of this manoeuvre is not very clear to me. 18 g5 looks simpler: 18...N6d7 19 h4 +=. 18...Ng6 19 h4 Bf4 There goes the bishop-pair.

Black to play

20 h5 Bxe3+ 21 fxe3 Ne7 22 e4?!

8...e6

22 g5!?. I guess it’s my obsession: if you have a light-squared bishop, put the pawns on dark squares.

I would be happy to say: “A central strategy is a safeguard against all troubles!” If only it were true...

22...h6 23 e5 Nh7 24 Ne4 Rf8 25 Nd6+ Kc7 26 Bg2

A fianchetto with 8...Nbd7 9 g3 g6 10 Bg2 Bg7 135

White’s initiative has ground to a halt, while Black has an extra pawn. Furthermore, the g4pawn is going to fall.

See the note to White’s 18th move. 26...Ng5 = 27 Rhf1 f6 28 Kc2 fxe5 29 dxe5 Nc8 30 c5 Ne7 (D)

40 Ra3 b6 41 Ra7 bxc5 42 Ra6+ Kc7 43 bxc5 Nd7 44 Ra7+ Kc6 45 g5 Nxc5 46 Nf7 d4 47 Ne5+ Kd5 48 Nd7 d3+ 49 Kc1 Nxd7 50 Rxd7+ Ke4 0-1 Excellent work by the World Champion. This is how White can lose at the 2800+ level. There is no threat of ‘draw death’ for chess...

Conclusions White to play

In Naiditsch-Danielsen, White played Bc4 and Nge2, and was hit with the vigorous and effective 7...e5!.

Now Black can put his knight on d5. An exchange may result in the appearance of a protected passed pawn (after ...exd5).

With opposite-side castling, the doubled pawns following ...exf6 gave the black king a solid fortress in Naiditsch-Djukić. It offered Black full scope to create counterplay on the other flank – see 14...Qa5.

31 b4?! Nd5 32 Bxd5 cxd5 This came as a surprise to me, but 32...exd5?! is indeed weaker since 33 Rf5 += gives White a strongpoint on f5.

It can be risky for Black to overestimate his resources and play ambitiously on both sides of the board. In Shirov-Dizdarević, White played in the centre to good effect – see 15 Rfd1, when he had ideas of c4 and a5.

33 b5?! This plan (not necessarily now, but in general) is the reason why I was sceptical about ...cxd5. 33...axb5 34 Nxb5+ Kc6 35 Nd6 Nf3?!

I don’t understand why in Carlsen-Djukić White failed to preserve his bishop-pair. As a result Magnus didn’t achieve anything special in the opening.

35...Rf3!. 36 b4 Ra8 37 Ra1?!

Meanwhile, Black didn’t achieve absolute equality by putting his bishop on g7 in FedorchukDanielsen. White kept a standard initiative with two bishops in an open position.

37 Nf5!?. 37...Rxa1 38 Rxa1 Nxe5 39 Ra7 Rb8 (D)

In our main game, Carlsen manoeuvred skilfully in the ending, firstly depriving his opponent of the bishop-pair and then exploiting White’s pawnweaknesses. Frankly speaking, I was surprised by Caruana’s feeble play, losing what had looked such a pleasant endgame... Overall, I feel that while the variation 3...Qd8 is legitimate, it is somewhat more passive than 3...Qd6 or 3...Qa5.

White to play

136

3 Nf3 Bg4

5: 2...Qxd5 without 3 Nc3

Game 33

1 e4 d5 2 exd5 Qxd5 (D)

Trygstad – S. Kasparov Leros 2010 1 e4 d5 2 exd5 Qxd5 3 Nf3 (D)

White to play If White doesn’t feel like engaging in a theoretical debate in the main lines, he doesn’t have to bring the knight to c3 immediately. However, this should not be viewed as arbitrary theoryavoidance, as there is some strategic substance to these ideas. In earlier chapters we have sometimes seen White moving his knight from c3 in order to advance his c-pawn, so as to unite his pawns on c4 and d4, with a view to further central advances. That leads to the idea of advancing the c-pawn before putting the knight on c3, or at least keeping that option open, so as to give Black additional problems to deal with.

3...Bg4 4 Be2

We shall deal with three lines in this chapter:

4 Nc3

3 Nf3 Bg4

is harmless because Black doesn’t experience any problems after the immediate exchange of queens: 4...Qe6+ 5 Qe2 (5 Be2? Bxf3 6 gxf3 Nc6 =+) 5...Qxe2+ 6 Bxe2 Nc6 7 h3 Bh5 (7...Bf5!?) 8 g4 Bg6 9 Bb5 (this seems unpleasant but in fact the creation of doubled c-pawns shouldn’t worry Black; in return he gets the bishop-pair, while even the doubled pawns themselves are useful, as they successfully restrain White’s queenside majority) 9...e6 10 Ne5 Ne7 11 d3 0-0-0 (even more precise is 11...f6 12 Nxc6 Nxc6 13 Bxc6+ bxc6 14 Be3 Kf7 =) 12 Bxc6 Nxc6 13 Nxg6? (after 13 Nxc6 bxc6 14 Be3 c5 15 b3 {so as not to allow ...c4} 15...Kb7 = the king goes to c6, ‘gluing’ the weak pawns together) 13...hxg6 =+ Laubert-S.Kasparov, Pardubice 2012. Black then exerted long-term pressure on the backward hpawn.

Black to play This is the most flexible move-order. White retains the possibility of putting the d-pawn on either d4 or d3. There are also lines where the pawn remains on d2 for a long time. But White’s main plan is to put his pawns on d4 and c4, and only then developing the knight to c3.

3 Nf3 Nf6 3 d4 The last of these doesn’t work out very well for tactical reasons; indeed, when we come to examine it in more detail, you will notice that I feel sufficiently certain of this conclusion to give 3 d4 a ‘?!’ marking. In that section we also take a brief look at 3 c4?!. 3 Nf3 offers Black less of a target for his counterplay; if there is a specific drawback to this move, it is that it allows a pin by 3...Bg4. If Black doesn’t wish to get involved in the cut-and-thrust after that move (and the possible loss of the bishop-pair), then he can play the more relaxed 3...Nf6, though this move isn’t wholly problemfree. This line is also significant because it can arise via 1 e4 d5 2 exd5 Nf6 3 Nf3 (a refined move-order that seeks to avoid 3 d4 Bg4) 3...Qxd5.

4...Nc6 5 d4 White can determine the bishop’s intentions first, and only then play in the centre. We examine this 137

but he can even hope for better: 7...Bxf3 (an attempt to keep the game complex and so maintain winning chances) 8 Bxf3 Qxe5 9 Nd2 Bc5 10 Qe2? (10 Bxc6 Bxe3 11 Bxb7+ Kb8 12 fxe3 Qxe3+ 13 Qe2 Qxe2+ 14 Kxe2 Kxb7 = PokojowczykCu.Hansen, Esbjerg 1981) 10...Bxe3 11 Qxe3 Qxe3+ (11...Qxb2!? –/+) 12 fxe3 Ne5, KostourosS.Kasparov, Vrachati 2011. An unpleasant fight for a draw lies ahead of White, thanks to his isolated pawn in the centre.

idea (via both 5 h3 and 5 0-0 0-0-0 6 h3) in Games 35 and 36. 5...0-0-0 (D)

b) The aggressive-looking 7 c4 leads to a much richer struggle. However, Black’s resources are wholly adequate. The queen check 7...Qa5+ forces 8 Bd2 (D); I doubt that this tempo-loss is very welcome for White.

White to play Black combines ‘work with development and pressure on the played this position as White, feeling under some pressure, as critical decision. Playing considerations’ can prove fatal.

pleasure’ (i.e. d-pawn). I have and remember each move is a by ‘general

White now has a choice between 6 Be3 and 6 c4. 6 c4 Black to play

Forcing the queen to choose a square before putting the bishop on e3.

Now:

6 Be3

b1) 8...Qa6!? is an interesting resource. 9 d5 (it may seem that the simple 9 c5 is winning, but Black then damages White’s pawn-structure by 9...Bxf3 10 gxf3 and follows up with 10...b5 11 cxb6 Qxb6, when the situation is absolutely unclear) 9...Bxf3 10 Bxf3 Nd4 11 b3 f5 (White’s decision to avoid complications on move 9 has left Black with sufficient play) 12 0-0 and both players have attacking chances against the enemy king, Külaots-Kovalenko, Iasi 2014.

is the less ambitious option and generally leads to more solid positions. White’s modest stance invites a vigorous reaction in the form of 6...e5 (D).

b2) 8...Bb4 9 d5 Bxf3 10 Bxf3 Nd4 (D) and then:

White to play Then: a) 7 dxe5 is both unpopular and unpromising. Black can play in very straightforward fashion for a draw,

White to play 138

b21) 11 Nc3 is the most common move, although most of the high-level games with this move come from rapidplay events. 11...Qa6 12 b3 (12 Be2

After the text-move (7 Be3), putting the bishop on this square only when the black queen has moved to f5, it seems that White will achieve his ideal set-up and gain the advantage. For example, in reply to 7...e5?! there follows 8 d5 += with comfortable play. But suddenly tactics come to Black’s aid...

Nf6 13 a3 Rhe8 is OK for Black; if White risks 14 axb4?! Qxa1 15 Qxa1 Nc2+ 16 Kf1 Nxa1 17 Bd1 c6 then it is Black who gets the initiative, Bologan-Grafl, Warsaw rapid 2009) 12...Qg6 (now ...Nc2+ is threatened; in general, the d4knight is worth a lot and it is highly unpleasant to tolerate such a piece in one’s camp) 13 0-0 Nf6 14 a3 Bd6 15 Be2 (15 b4 is met by 15...e4) 15...e4 16 Rc1 Qf5

7...Bxf3! The quieter 7...Nf6 is possible, but far less promising; Black is still playing for equality then, which he may not obtain given that White has two good pawns in the centre. After 8 Nc3 (8 0-0 transposes to note ‘b322’ to Black’s 4th move in Game 37) 8...e5 9 d5 e4 10 Nd4 Nxd4 11 Bxd4 (D) I like the look of White’s position.

(in such complicated positions, accurate play can be expected only from a computer; 16...Kb8!?) 17 f4?! (17 Be3 can be met by 17...Nf3+ 18 gxf3 exf3 19 Bxf3 Qh3 with a draw after 20 Re1 Bxh2+ 21 Kh1 Bg3+ 22 Kg1 Bh2+ 23 Kh1 =) 17...exf3 (17...Bxa3!?) 18 Bxf3 Nxf3+ (18...Rhe8!?) 19 Rxf3 Qh5 20 Bf4 Bxa3 21 Rb1 Bb4 22 Qd4 Nd7! 23 Kh1 (Movsesian-Grafl, Bundesliga 2005/6) 23...g5!? 24 Bg3 Qg6 with a complicated situation that is hard to evaluate. b22) 11 Be2 Bxd2+ 12 Nxd2 Qb6 13 b3 Qg6 (I like this redeployment of the queen to the kingside, but there were many other attractive options, such as 13...Kb8, 13...Nf6 and 13...f5) 14 0-0 f5 15 Nf3 Nxf3+ 16 Bxf3 Nf6 left Black at least no worse in Tassopoulos-S.Kasparov, Iraklion 2011. With opposite-side castling, one must hurry to attack the king.

Black to play He has no weaknesses, predominates in the centre and is already eyeing Black’s castled position. See how Delchev got the advantage in two fairly recent games:

We now return to 6 c4 (D):

a) 11...a6 12 0-0 h5 13 f3 exf3 14 Bxf3 Rh6 (Delchev-L.Milov, Bad Wiessee 2014) and after the prophylactic 15 Qd2 it seems that the black pieces are stuck on the kingside. The half-open ffile denies Black the chance to create an effective attack: 15...Rg6 16 Be2 +/–.

6...Qf5 7 Be3

b) 11...Bxe2 12 Qxe2 Bb4 13 0-0 Rhe8 (DelchevV.Spasov, Sunny Beach 2014) and here 14 Qc2 += plans pressure on the e4-pawn while threatening to bring the queen into the attack on the queenside.

7 0-0 Nxd4 8 Nxd4 Bxe2 9 Qxe2 Rxd4 is unclear.

8 Bxf3 (D)

Black to play

7 Nh4? is futile: 7...Qe4 8 f3 Qxd4 9 fxg4 (9 Qxd4 Nxd4 –/+) 9...Qf6 –/+. 139

Hoogeveen 2014. It turns out that 16 Qxg7? is impossible because of 16...Bf6. b) The modest 13 Qxa6 is more reasonable: 13...bxa6 14 Bxa7 e6 = Meribanov-S.Kasparov, Minsk rapid 2013. We now return to 9 Bxd4 (D):

Black to play At first it looks like Black was crazy to exchange on f3, as Bg4 and Bxc6 are threatened. But suddenly... 8...Nxd4! It is hard to see a logical reason why this ought to work; White hasn’t broken any precepts and has been developing in a healthy way. The tactical basis relies on pins on the e- and d-files.

9...Qe6+ 10 Be2 (D)

But what about White’s Bg4 threat?

Not 10 Kf1?

9 Bxd4

Qxc4+.

Black to play

The deadly-looking 9 Bg4 actually has unclear consequences after the forced 9...Nc2+ 10 Qxc2 Qxg4 11 0-0 (D).

Black to play We have reached a key position. In the resulting positions, Black is generally a pawn up, but White has more or less sufficient compensation. Black to play

10...Qe4

Here a lot of paths lead to unclear positions. Nevertheless, I’ll propose 11...Qg6 12 Qa4 Qa6, which is enough to equalize, ‘cynically’ seeking an exchange. This may be particularly appropriate against a younger player, or indeed anyone who is less comfortable in an endgame than in a sharp middlegame. Then:

Now White has no decent way to protect the bishop. In Game 34 we shall examine 10...c5!?. 11 0-0 Qxd4 12 Qa4 e6 12...Qxb2? is too risky: 13 Qxa7 e6 14 Bf3 Nf6?! 15 Nc3! Qxc3? (the only reply that keeps the struggle going is 15...Qb6 16 Qa8+ Kd7 17 Qa4+, but White has a very strong attack) 16 Qxb7+ Kd7

a) 13 Qc2 e6 14 b4? (creative but ... why?) 14...Bxb4 15 Qb2 Be7 –+ J.Kramer-S.Kasparov, 140

17 Bc6+ Ke7 18 Qxc7+ Nd7 19 Rad1 g5 20 Bxd7 Bg7 21 Be8+ 1-0 V.Meijers-Antoniewski, Martigny 2005. 13 Nc3 Bd6 14 Nb5 14 Bf3 Qf4 15 Rfd1 (Bruzon-Kurajica, La Laguna 2010) 15...a6!? is unclear. 14...Qe5 (D)

Black to play 22...e5?! 22...Rc7 keeps the balance: 23 c5 dxc5 24 bxc5 Qb2 25 R1c2 (25 c6 b6 =) 25...Qa1+ 26 Rc1 Qb2 27 R1c2 Qa1+ 28 Rc1 =. 23 c5 dxc5 24 bxc5 += Qc7 25 Ra3 25 Bxb7?! Qa5! has curious consequences: 26 Bc6 (26 Qxa5?? Rd1+) 26...Qxa4 27 Bxa4 Rd3 28 Rxd3 Rxd3 29 Rb1+ Kc8 =. Not 25 c6? Rd4 =+.

White to play A ‘geometrical’ defence. The queen escapes from the attack while in turn threatening ...Qxh2# and attacking the e2-bishop. This denies White time to take on a7.

25...a6

15 Nxd6+ cxd6 16 Bf3 Kb8

26 c6 Rd4 (D)

25...a5!?.

A problematic position. Which is more significant: Black’s extra pawn or White’s attack? Right now Black has no weaknesses. I feel it should be equal as at some point Black will need to give back the pawn in order to simplify. 17 Rad1 17 Rfe1 Qc5 18 b4 Qc7! 19 Rac1 Ne7 also leaves White with sufficient compensation, d’AmorePrié, European Union Ch, Arvier 2007. White to play

17...Nf6 18 Rd3 Qc5 18...Ne4!? is interesting: 19 b4 (19 Ra3 a6 20 b4 d5 21 b5 Rd6 is unclear – then 22 Rb1? can’t be played due to 22...Nd2) 19...d5 20 c5 h5 21 c6?! bxc6 22 Qxc6 Rd6 23 Qc1 Rc8 =+. Black’s centralized grouping looks impressive, doesn’t it?

White has acted skilfully so far, but as the game approaches the time-control, mutual inaccuracies start to creep in.

19 Rfd1 Rd7

27 Qb3! +=.

19...h5 20 Rb3 d5 21 cxd5 Nxd5 =.

27...e4?!

20 Rc3 Rhd8 21 b4 Qb6 22 Rdc1 (D)

27...Rd2 28 Qc4 (28 Qb1 e4 29 Rb3 b5 30 a4 exf3 31 axb5 Ng4 32 bxa6+ Ka8 –+) 28...e4 29 Be2 R8d4 30 Qc3 Qxc6 31 Qxc6 bxc6 32 Bf1 =+.

27 Qc2?

141

28 Rb3! b5 29 Be2 R8d5

cxd8Q+ Rxd8 43 Qb5+ Kf8 44 Qxf5+ Kg8 45 a8Q 1-0

Simpler is 29...Rd2 30 Qc5 R8d5, when White should avoid 31 Qf8+?! Ka7 –/+. 30 a4?! Rd2?

Conclusions

The computer suggest the ‘emotionless’ capture of the pawn by 30...Rxa4 31 Bxb5 axb5 32 Rxb5+ Rxb5 33 Qxa4, but this position seemed dangerous to me. With seconds left, it was difficult to appreciate that after 33...Qxc6! Black wins anyway.

If White plays too conservatively and tries to liquidate the central tension, then he may face problems in the ending (Kostouros-S.Kasparov). In the sharper lines after 6 Be3 e5 7 c4 Qa5+ 8 Bd2, White must even be quite accurate, as otherwise he runs the risk of being worse (Bologan-Grafl). If both sides play well, then a race to attack can ensue. No wonder – this is standard in positions with opposite-side castling. All other things being equal, both players should avoid weakening their king’s pawn-cover. Advancing them without a good reason can provide a ‘hook’ for the opponent to open files (Movsesian-Grafl).

31 Qb1? 31 Qc4 is better. 31...Rxe2 32 axb5 (D)

In our main game we acquainted ourselves with one of the sharpest variations of the Scandinavian Defence, which is overall a relatively calm opening. As you can see, the chances are approximately equal and the battle is rich in tactical ideas. On move 12, I can’t recommend the capture on b2 because it invites a swift attack (V.MeijersAntoniewski).

Black to play The moment of truth.

In the comments to the main game you can see how ‘active’ defence should be carried out. For some time both players operated rather logically, but with time-trouble came some mistakes by both sides. The advantage passed from one to the other, but it was I who made the last mistake. However, we can state that in the line 10...Qe4 Black’s chances are not inferior.

32...Ng4?? The right way was 32...axb5 33 Rxb5+ Kc8 (not, of course, 33...Rxb5?? 34 Qxb5+, when the e2rook perishes) 34 Rb7 (who wouldn’t fear such an attack when under time-pressure? White’s powerful grouping – 2R+Q+P – bursts into the king’s apartment; 34 Rxd5 Nxd5 35 Qa1 Qb6 –+), when after the cool 34...Qf4 White can’t catch the lonely black monarch, while the white king is also endangered. 35 Rf1 e3 36 Rxf7 and now 36...Qxf2+! is the only winning move.

A further way (on move 10) to capture the d4bishop can be observed in the next game.

33 bxa6+ Kc8 34 Rg3

Game 34

This is sufficient to win, but a cleaner path to victory is 34 Rb8+! Qxb8 35 Qb7+! Qxb7 36 cxb7++ Kd7 37 b8Q +–.

Sharma Dinesh – S. Kasparov Llucmajor 2013

34...Kd8 35 Rxg4 f5 36 Rg3 Red2 37 Rgc3 Ke8 38 g3 Qa7 39 Rf1 Qd4 40 a7 Rd8 41 c7 Qxc3 42

1 e4 d5 2 exd5 Qxd5 3 Nf3 Bg4 4 Be2 Nc6 5 d4 0-0-0 6 c4 Qf5 7 Be3 Bxf3 8 Bxf3 Nxd4 9 Bxd4 Qe6+ 10 Be2 c5!?(D) 142

Another acceptable option.

13 Na3 13 0-0 should lead to a draw by perpetual check after 13...Qxe2 14 Qa8+ Kc7 15 Qa5+ Kd7!. 13...e6 Or 13...d3 14 Qa8+ Kc7 15 Nb5+ Kc6 (avoiding the repetition by 15...Qxb5? is a bad idea: 16 Qxd8+ Kxd8 17 cxb5 dxe2 18 Kxe2 +/–) 16 Na7+ Kc7, with a draw. 14 Qa8+

White to play

14 Nb5!?.

Now Black will get a pawn on d4. On the one hand, this pawn may prove strong as it is ready to move to d3 at any point; on the other hand, the lack of a c-pawn leaves the black king less well defended.

14...Qb8 That is why the queen is needed on the b8-h2 diagonal.

11 Qa4

15 Qxb8+

White has tried some other moves, but without much success:

Or 15 Qa5 d3 (15...Ne7!?) 16 Bf3 d2+ 17 Ke2 Ne7 18 Rhd1 (Baron-Kurajica, Rijeka 2010) 18...Nc6 with unclear play.

a) 11 Nc3 cxd4 12 Nd5 Qd6 (a good sidestep in positions like this: it prepares ...e6, while the queen can work on the b8-h2 and f8-a3 diagonals) 13 Qxd4 e6 14 Qxa7 exd5 15 c5 Qxc5 16 Qxc5+ Bxc5 17 Rc1 b6 18 b4 (Jurković-Nevednichy, Bizovac 2008) and now Black can win a pawn by 18...Kb7 19 bxc5 Rc8 =+.

15...Kxb8 16 Nb5 Bb4+ 17 Kf1 d3 (D)

b) 11 Qc2 cxd4 12 0-0 Nf6 gave Black the initiative in L.Milov-Pitl, Barcelona 2009. 11...cxd4 12 Qxa7 Qe5 (D) This is correct. As has been noted before, the queen should stay on the dark squares (e5, d6), as it is then useful both in defence and attack.

White to play ½-½

12...d3 is weaker: 13 Nc3 dxe2 14 Qa8+ Kd7 15 Qxb7+ Ke8 16 Nd5 with unclear consequences.

Here we agreed on a draw and there is a sad comment in my database: “I have been doing everything right, but evaluated the position as approximately equal”. The struggle could have continued as follows: 18 Bf3 Nf6 19 a3 Be7 and now 20 Rd1 e5 21 Nc3 Rd4 =+ or 20 Re1 Ne8! (alas, I didn’t find this move during the game) 21 g3 Nd6 22 Nxd6 Rxd6 =+.

White to play 143

Conclusions

Black must decide whether to retreat or exchange:

If White forces exchanges, this may lead to an ending which is better for Black (JurkovićNevednichy).

a) It is a shame to give up the bishop by 6...Bxf3 without a truly concrete follow-up. Black gets some compensation in the form of central control, but it seems insufficient. 7 Bxf3 Qd7 and then:

In the main game, I also got slightly better chances but... didn’t understand all the benefits of the situation and agreed to a draw.

a1) 8 g3?! and here: a11) Not 8...Qxh3?? 9 Bg4+ +–.

In the line with 8...Nxd4! Black enjoys ample counterplay, with reasonable hopes of creating winning chances. In general it makes sense for White to use this variation if he likes to attack and is not afraid to sacrifice.

a12) 8...e5 9 d3 f5 10 Nc3 (10 Re1 Bd6) 10...Nd4 11 Bg2 Nf6 with a promising position for Black, Sitnikov-S.Kasparov, Donetsk 2011.

But given that White’s pawn on d4 quickly becomes a target, it is natural to ask if White should not delay this move and seek to prepare the ground better for this advance. From this springs the idea of forcing the g4-bishop to declare its intentions by playing an early h3. I shall present two games from my own experience, both against quite high-rated opponents. The results are different but in both encounters Black had little to complain about until well into the endgame.

a2) 8 c3 (D) prepares d4, which, if allowed, would give him a comfortable long-term advantage thanks to his bishop-pair.

a13) 8...h5 is probably even stronger; e.g., 9 Bxh5?? Nf6 –+, 9 Bg2 h4 10 g4 f5 11 gxf5 Nf6 12 Qf3?! Rh5 –/+ or 9 h4 g5 10 hxg5 h4 11 g4 e5 12 Nc3 Nge7 =+.

Game 35 Bogdanovich – S. Kasparov Donetsk 2010 1 e4 d5 2 exd5 Qxd5 3 Nf3 Bg4 4 Be2 Nc6 Black to play

White is not obliged to advance his pawn to d4 immediately; he can instead ‘check the bishop’s documents’.

So: a21) To my mind, 8...e5

5 h3 doesn’t equalize: 9 Re1 Nf6 (9...Nge7 10 b4 Ng6 11 Na3 +=) 10 b4 (White could grab the e-pawn by 10 Bxc6 Qxc6 11 Rxe5 Bd6, but Black has compensation) 10...Bd6 11 d3 Qf5 12 Qa4 Qxd3 13 b5 Nb8 14 Be3 gave White a strong attack in Vachier-Lagrave – Papaioannou, European Team Ch, Warsaw 2013.

Another move-order, with similar goals, is 5 0-0 0-0-0 6 h3 (D).

a22) 8...Ne5 (the knight both attacks the bishop and glances at d3) 9 d4 (White can’t preserve the bishop because after 9 Be2?! Nd3 it is impossible to tolerate the d3-knight and he will soon have to play Bxd3) 9...Nxf3+ 10 Qxf3 Nf6 11 Nd2 Qd5 12 Qe2 e6 13 Nf3 Bd6 14 Re1 Qh5 15 c4 += Short-Salov, Dos Hermanas 1997.

Black to play 144

b) I think 6...Bh5 (D) is best, maintaining the tension.

5...Bxf3 In this version the retreat 5...Bh5 doesn’t equalize. White has time to put both pawns in the centre as he has saved a tempo on castling, compared to the previous note. 6 d4 0-00 7 c4 (the pawn reaches c4 before the knight gets to c3!) 7...Qd6 8 d5 Bxf3 9 Bxf3 Ne5 (Gaponenko-Jakubiec, Women’s Bundesliga 2010/11) and now White can act in gambit style: 10 Nc3 Nxc4 (10...e6?! is even worse due to 11 00 Nxf3+ 12 Qxf3 Nf6 13 Bg5 +/–) 11 0-0 Nf6 12 Qa4 (the f3-bishop and the d5-pawn squeeze the enemy forces; both ...e6 and ...c6 are risky, but how can Black develop?) 12...Qa6 13 Qxa6 bxa6 14 Rd1 +=.

White to play Let’s look at several replies for White: b1) 7 d4?! (one way or another, the pawn is given away) 7...Bxf3 8 Bxf3 Qxd4 is unclear.

6 Bxf3 Qe6+ (D)

b2) 7 Nc3 Qd7 8 b4 Nf6 and now 9 b5?! is, as normal, parried by an exchanging operation: 9...Bxf3 10 Bxf3 (not 10 bxc6?, when the sated bishop joyfully returns by 10...Bxc6 –/+) 10...Nd4 11 a4 Qf5! (White already experiences problems!) 12 d3 Nxf3+ 13 Qxf3 Qxf3 14 gxf3 Nd5, B.Smith-Korley, USA League 2014. Please look at the ruins of White’s kingside. Further comments are superfluous.

6...Qe5+ comes to the same thing.

b3) 7 d3 e6 8 Be3 Nf6 9 Nbd2 Rg8 10 d4 (White’s twostep pawn advance is not pointless, but it’s still a tempo-loss) 10...g5 11 g4! (otherwise Black would have played ...g4 himself) 11...Bg6 12 c4 Qd7 13 Bxg5?! (13 Ne5 is better) 13...Be7 14 Be3 (Palac-Prié, French Team Ch 2009) and after 14...Ne4! the extra pawn brings White no pleasure; he is too busy trying not to get mated.

White to play Exactly! With the king in the centre this is an additional resource. In return for gaining the bishop-pair, White must acquiesce to a rather awkward exchange of queens. 7 Qe2

We now return to 5 h3 (D):

7 Kf1?! (this looks artificial) 7...0-0-0 8 Nc3 g6 9 d3 Bh6 10 g3 Nd4 11 Bg2 Bxc1 12 Rxc1 Nf6 =+ Moutousis-Nikolaidis, Greek Team Ch, Athens 2001. 7 Be2 Nd4 =. 7...Qxe2+ 8 Bxe2 Nd4 (D) In Game 36 we take a look at 8...0-0-0!?.

Black to play

145

I alluded to this idea in my note to Black’s 8th move. The rook move seems more accurate than 11 d3, even though the latter has been played in some high-level games: a) 11...Bc5?! (I suspect that developing the bishop to this square is a bit suspect) 12 Re1 f6 13 Nd2 Ne7 14 Nc4 Ndc6 (so what was the point of ...Nd4?) 15 g3 g5 16 Kg2 h5 17 Nd2! Bb6 18 Ne4 += (a strong transfer of the knight: from this square it gives Black plenty to worry about, and the ...f5 advance is hard to engineer due to the loss of the g5-pawn; in its turn the d1-bishop ‘ties down’ the h8-rook to the h5-pawn) 18...Rdf8 (but now the pressure on the d-pawn is reduced) 19 c3 g4 20 Rh1 (20 hxg4 hxg4 21 Bxg4+? f5 22 Bh3 Rxh3 23 Kxh3 fxe4 =+) 20...f5 21 Ng5 Nd8 22 Bd2 Ng6 23 hxg4 hxg4 24 Bb3 f4 25 Rxh8 Rxh8 (KuzubovDanielsen, Reykjavik 2012) and after 26 Ne4 f3+ 27 Kg1 +/– White is obviously better. Black’s kingside aggression is deadlocked; without queens and with only one rook available it is difficult to give mate. And in the future White’s bishops and superior pawn-structure will come to the fore.

White to play The knight’s leap to the centre looks the most natural and logical, but don’t rush to make an evaluation, as there are some nuances here. Either way, White will hide his bishop from exchange. So should we really open the e-file for the white rook, drive the bishop to the queenside and place the knight on a square where it provokes a tempogaining advance by the white c-pawn? Let’s see. 9 Bd1 e5 Otherwise White will simply gain space by playing c3 and d4. 10 0-0

b) 11...g6 (D) and then:

Several years later I got to play this position as White: 10 c3 Ne6 11 0-0 Nf4?! 12 d4 Nd3 13 dxe5 (slightly more precise is 13 Ba4+ c6 14 dxe5 0-0-0 15 Be3 +=; since Black has played ...c6 he will not be able to meet Bxa7 with ...b6, trapping the bishop) 13...0-0-0 14 Be3 Nxe5 15 Nd2 Nf6 16 Nf3 Nc4 17 Bc1 Bc5 18 Bb3 += S.Kasparov-Z.Varga, Erbil 2013. The bishops ensure a long-term initiative. 10...0-0-0 (D) White to play b1) The sharp 12 f4 is too good to be true. After 12...Bh6 the bishops are exchanged, and White’s hopes for a tangible advantage vanish with them. 13 Na3 Bxf4 14 Bxf4 exf4 15 Rxf4 f5 16 Rf1 Nf6 17 c3 (17 g4 Rhe8 gives Black the initiative) 17...Nc6 18 Bc2 Nd5 = Beikert-Prié, French League 1996/7. b2) 12 Re1 Bg7 13 Nd2 (after 13 f4 exf4 14 Bxf4 Ne6 15 Bc1 Ne7 Black’s piece-play denies White time to finish his development in comfort) 13...Ne7 14 c3 Ne6 15 Nf3 Nc6 16 Bc2 Rd6 17 Ng5 Ncd8 18

White to play 11 Re1 146

Nf3 Nc6 19 Nd2 Rd7 20 Ne4 f5 21 Ng5 Nc5 = Nakamura-Tiviakov, Wijk aan Zee 2010. White got nothing special out of the opening.

15...Nce7?! Too subtle. 15...Nge7 is more natural because 16 b5 is not dangerous: 16...Na5 17 Ba3 Nd5 =.

We now return to 11 Re1 (D):

16 Re2 Nd5 17 g3 Nge7 18 Na3 Rhe8 18...Nf5. 19 Bb2 Nf5 20 Rae1 += My young opponent performs almost perfectly. By means of dainty manoeuvres he has managed to consolidate his forces. A freeing advance like f4 or d4 is on the agenda. I should note that playing c4 is pointless as it weakens b4. 20...Nd6 21 Bb3 c6 22 Nc2 Nc7 (D)

Black to play 11...Nc6 11...f6!? is interesting (but I didn’t want to weaken the light-square complex given that I had no light-squared bishop): 12 d3 Ne7 13 Nd2 h5 14 a4 Nd5 15 c3 Ne6 16 g3 Nc5 (16...h4?! 17 Bg4 Re8 18 Nf3 +=) 17 Bc2 Ne7 (the knights’ gyrations and the pressure upon the c2- and d3pawns make it hard for White to develop) 18 Re3 Nd5 19 Rf3 Ne6 20 h4 g5 gave Black counterplay in Sherzer-C.Horvath, Hungarian Team Ch 1996.

White to play

12 c3 g6 13 Bc2 Bh6

Black must get ready for the opening of the centre and patiently await developments.

The idea is clear, I presume. The advance of the dpawn will allow the exchange of the c1-bishop, depriving White of the bishop-pair – his only real asset. Meanwhile, Black’s pressure on the d2pawn hampers White’s development.

23 d4 exd4 24 Nxd4 Rxe2 24...f5!?. 25 Rxe2 f5 26 c4

14 b4 f6 15 a4 (D)

26 Re7 looks dangerous but Black has enough resources to neutralize White’s activity: 26...Rd7 27 Rxd7 Kxd7 28 Bg8 (evidence of the bishop’s proverbial superiority over a knight: the bishop ‘flies’ from one flank to the other in an instant; however, the black pieces also have a harmonious disposition and are in time on the kingside) 28...Ke7 29 c4 (here and later the h7-pawn is poisoned: 29 Bxh7? Kf7 –/+) 29...Bg7 30 c5 (30 Kg2?! Kf8 31 c5 Ndb5! =+) 30...Nde8 31 Bc4 Ne6! =.

Black to play

26...Ne4 27 c5 Nd5 28 Bxd5 Rxd5 29 g4 (D)

Please note how the Ukrainian master playing White solves the problem of his queenside development while preserving his bishops. 147

34 Kg2. 34...Ke6 35 Kg2 Kd5 36 Kf3 Not 36 a5? Ke4, when White is suddenly defending rather an unpleasant ending. 36...Bc1 37 Kg4 Kc4 38 Be1 a5 (D)

Black to play With the exchange of the light-squared bishop, White has lost his trump-card and the chances are now fully equal. By this point I was already thinking in terms of taking over the initiative. 29...Kd7?! White to play

This can hardly be called a mistake but there were stronger options. For instance, the ‘computer-like’ 29...Bc1!? 30 Ba1 a5 31 gxf5 gxf5 32 Nxf5 Rxf5 33 Rxe4 axb4 34 Rxb4 Ba3 gives Black the initiative. Also good is 29...Bg5!, taking control of e7 and escaping from the knight’s attack; e.g., 30 gxf5 gxf5 31 Nxf5 Nxc5 32 bxc5 Rxf5. 29...a5!? also looks attractive.

Now Black’s active pieces and White’s ragged pawns fully compensate for the pawn. 39 bxa5 Kxc5 40 f4 Kd5 41 f5 c5?! After 41...Bb2! Black’s chances are at least no worse: 42 Bh4 (42 Kg5?! c5 and now 43 Kh6? Ke5 –/+ or 43 f6?! Ke6 =+) 42...c5 43 Kf3 c4 44 Ke2 Ke5 45 f6 Ke6 46 Ke3 Be5 47 Ke4 Bb2 =.

30 gxf5 (D)

42 f6 Ke6 43 Bc3 Bh6? 43...Ba3 is sufficient to hold the draw: 44 Kf4 Bb4 45 Bxb4 cxb4 46 Ke4 Kxf6 47 Kd4 Kg5 48 Kc4 Kh4 49 Kxb4 Kxh3 50 Kb5 Kg3 51 Kb6 h5 52 Kxb7 h4 53 a6 h3 54 a7 h2 55 a8Q h1Q+ with perpetual check, as the white queen is poorly placed. 44 Kf3 Kf5 45 Bb2 45 Ke2 +/–.

Black to play 30...gxf5?!

45...Bg5

And here 30...Bc1! was already necessary if Black is not to be fighting just for equality: 31 fxg6 (31 Bxc1 Rxd4 32 fxg6 hxg6 gives Black the initiative) 31...Bxb2 32 Rxb2 hxg6 with compensation for Black.

Black could repeat with 45...Ke6 46 Bc3 Kf5, but White can, as already noted, vary with 47 Ke2 +/–. The remainder is more or less clear without comments. 46 f7 Bh6 47 Ke2 Ke4 48 Bf6 Bf8 49 Kd2 Kd5 50 Kc3 Ke6 51 Bh4 Kd5 52 Bf6 Ke6 53 Bg5 Kd5 54 Be3 Bg7+ 55 Kb3 Bf8 56 Kc3 Bg7+ 57 Kd3 h5 58 Bc1 Bf8 59 Bg5 c4+ 60 Kc3 Kc5 61 Be3+ Kd5 62 Bc1 Bg7+ 63 Kb4 Ke6 64 Kxc4

31 Nxf5 Rxf5 32 Rxe4 Rf4 The right decision; it should be enough for a draw. 33 Rxf4 Bxf4 34 Bc3 148

In our main game, White got somewhat the better position, having completed his development while preserving the bishop-pair. However, it was unclear how he should then make progress. In the ending I missed an opportunity to take over the initiative, and further inaccuracies even led to a loss.

Kxf7 65 Kc5 Bc3 66 Kb6 Ke6 67 Kxb7 Bxa5 68 Be3 Kd7 69 Bb6 Bxb6 70 Kxb6 1-0 An almost perfect game from White’s side.

Conclusions Sometimes White prefers the cautious d3 advance so as to continue with quiet development and avoid committal play until later. However, such sluggish tactics give Black time to launch a kingside attack (Palac-Prié).

I used another move-order in our next game, though the difference can be seen only with a microscope.

After 5 0-0 0-0-0 6 h3 I believe it is undesirable to exchange bishop for knight with 6...Bxf3 as then, with best play, White can get some advantage in a calm position (Short-Salov) or sacrifice a pawn to launch a strong attack (Vachier-Lagrave – Papaioannou). That’s why it is more reasonable to preserve the bishop with 6...Bh5; if appropriate Black can still capture on f3 if a better opportunity presents itself. For instance, if the white pawn runs up to b5, the knight will joyfully leap to the centre (...Nd4) attacking f3, b5 and so on (B.Smith-Korley).

Panchanathan – S. Kasparov

Game 36

Ranshofen 2012 1 e4 d5 2 exd5 Qxd5 3 Nf3 Bg4 4 Be2 Nc6 5 h3 Bxf3 6 Bxf3 Qe6+ 7 Qe2 Qxe2+ 8 Bxe2 0-0-0!? (D)

It turns out that with the 5 h3 move-order, matters are reversed: the retreat to h5 is not so good because White has enough time to occupy the centre with his c- and d-pawns (GaponenkoJakubiec). So Black should exchange on f3, which immediately moves the game into a complicated ending (or maybe one should talk here of a queenless middlegame). A long positional struggle lies ahead. Personally, I find it rich enough that I am happy to play it with both White and Black!

White to play In contrast to the previous example, Black tries to save time by not putting his knight on d4 just yet. 9 c3 White has a wide choice here:

Black must manoeuvre knights with some care because if he allows the centre to open it will generally benefit White (S.Kasparov-Z.Varga).

a) Timid play with 9 d3 allows Black to seize the initiative with natural developing moves: 9...e5 10 Nd2 f5 11 Nc4 Nf6 12 Bd2 e4! (simple and strong) 13 0-0-0 Bc5 14 Ne3 g6 15 Rhe1 Rhe8 16 g3 Nd4 =+ J.RamirezMunoz Pantoja, Barbera del Valles 2014. I think of positions like this (from Black’s point of view) as ‘muscular’. All his pieces interact harmoniously.

After capturing on f3 it is reasonable for Black to bring his sole remaining bishop to h6. Thus it will exert pressure on the d2-pawn and if this pawn advances, an exchange of bishops on c1 will deprive White of his bishop-pair. This idea has been used by Prié and Tiviakov, and it gave Black a very decent position against a very formidable adversary inNakamura-Tiviakov.

b) One can hardly recommend 9 b3

Developing the bishop to c5 makes less sense, as the piece is useless there (Kuzubov-Danielsen).

g6 10 Bb2 Nd4 (Black’s next move will be ...Bg7, threatening a knight check followed by capturing 149

Bd3 Nc5 22 Bc2 e5 23 Ne2 Nc6 24 b4 Ne6 25 a3 (all the pawns are put on dark squares, denying Black the chance to create strongpoints) 25...Be7 26 Bb3 Rxd1 27 Rxd1 Rd8 28 Rd2 Bg5 29 Rxd8 Ncxd8 30 h4 (see the previous comment) 30...Be7 31 Bd2 += S.Kasparov-Klerides, Triesen 2014. White has a slight but comfortable advantage. It is not so easy to win positions like this, but it is far more unpleasant for Black to defend them, without any counterplay.

on b2) 11 Bxd4 (after such an exchange White has not a vestige of an advantage) 11...Rxd4 12 Na3 Bh6 13 Nc4 Nf6 14 c3 Rdd8 15 Bf3 (LamoureuxPrié, Nice 1994) and after the precise 15...e5! 16 0-0-0 (16 Nxe5?? loses to 16...Bxd2+) 16...e4 Black stands better. c) 9 0-0 Nf6 10 c3 e6 (D).

We now return to 9 c3 (D):

White to play 11 d3 (this was my choice when I was on White’s side of the board, seeking a minimal advantage; in reply to the natural 11 d4 I was disconcerted by 11...e5! 12 dxe5 Nxe5, when the knight might penetrate to d3) 11...h6 12 Nd2 Ne5 13 d4 (now this advance has fewer drawbacks) 13...Ng6 14 Nf3 (the knight has run from the queenside to the kingside, placing e5 under control) 14...Bd6 15 g3 (denying the black pieces access to f4; note that there is no chance for Black to trade knight for bishop, and in general it is favourable to put the pawns on dark squares when one has a ‘single’ light-squared bishop) 15...c5 16 Kg2 cxd4 17 Nxd4 (17 cxd4?! leaves White with an isolated pawn and provides the black knights with a strongpoint on d5) 17...Ne7 18 Rd1 a6 19 Bd2 (D).

Black to play 9...e5 Black can also play 9...g6. Then: a) The straightforward 10 d4 is, of course, met by 10...Bh6 11 Na3 Bxc1 12 Rxc1 e6 (the sharp 12...e5 13 dxe5 Nxe5 equalizes even faster; one bishop is not the same as two!) 13 Nc4 h5 14 h4 Nf6 15 b4 and White stands slightly better, but no more than that, Lemmers-Geirnaert, Belgian Team Ch 2009. b) 10 Na3 Bh6 11 Nc4 Nf6 12 b4 Nd7 13 Bb2 Nb6 (13...e5 is more accurate – but then why not play this on move 9 rather than make manoeuvres like ...Nf6d7?) 14 Nxb6+ axb6 15 0-0-0 Bg7 (again 15...e5 is logical, hindering White’s d4 advance) 16 b5 Na7 17 Rhe1 Kb8 18 a4 += C.Garcia-Mancebo, Cullera 2001. One can state that Black did not play very successfully in this game, as he kept ignoring the strong and natural ...e5 advance. 10 d3 (D)

Black to play Such manoeuvres can make sense if you enjoy the bishop-pair – you should keep them safe; the bishop is redeployed to e1 after the rook has moved to the queenside. 19...Ne4 20 Be1 Kc7 21 150

But after the text-move, it seems that White has a comfortable position given that he still has both bishops. However... 14...Bxg5! This may seem an odd way to fight against the bishop-pair, but it works well for concrete reasons. 15 Bxg5 f4

Or at once 10...g6!?.

Usually two bishops are considerably stronger than two knights but there are situations when the value of these pieces is altered, and this is one such case. Right now, the simple threat is ...h6 followed by ...g5.

11 Nd2

16 Bh4

11 0-0?!

Black has the preferable position after 16 Bg4 h6 17 Bxd7+ Kxd7 18 Bxf4 exf4 =+.

Black to play 10...f5

is unsatisfactory for White: 11...Nf6 12 Na3?! e4 (or 12...Bxa3) 13 Rd1 Bxa3 14 bxa3 h6 (14...Rhe8!?) 15 Bb2 Rhe8 gave Black the initiative in Doettling-Kalod, Paget Parish 2001. White’s weakened queenside structure is significant.

16...h6 17 f3 Nge7 (D)

11...g6 12 Nf3 The knight has managed to reach f3 before Black had time to ‘pin’ it with ...Bh6. However, this idea is still reasonable. 12...Bh6 13 Ng5 Rd7 (D) White to play Look closer to the position. You are no doubt aware that knights like strongpoints, and g3, f5 and e3 are all excellent candidates. Bishops need space and open diagonals. At the moment the light-squared bishop has little scope. How can it be freed? The d4 advance is most unlikely, while the manoeuvre Bd1-b3/a4 is also impeded by the pressure on the d3-pawn. 18 g4 fxg3 White to play 18...Rf8 is also not bad, simply ignoring the pawn’s advance.

During the game I wasn’t sure about my ‘wellbeing’ but now I realize that Black is OK.

19 Bxg3 Rf8 14 Bd2 Another idea is also good: 19...Nf5 20 Bf2 Nce7 21 Rg1 g5 with the initiative.

Play on the wing doesn’t reach the goal either: 14 h4 Nf6 15 h5?! gxh5 16 Bxh5 Rg8 =+.

20 0-0-0 Nd5 21 h4 Nf4 151

Here the knight is so strong that one way or another White has to give up one of his bishops. 22 Bxf4 Rxf4 23 Rdg1 Rd6 24 Rg4 Rdf6 25 Bd1 Kd7 26 Re1 (D)

Black to play The chances are approximately equal but I felt that Black wasn’t taking any risks at all by playing on. The white king will be diverted by the passed hpawn, granting Black time to cause havoc on the other side of the board; even if things somehow go wrong, Black should be able to eliminate the last white pawn.

Black to play 26...Rxg4?! Black shouldn’t reconnect the isolated h- and fpawns, but, alas, I didn’t see anything better. A preferable course is 26...h5! 27 Rxf4 Rxf4 28 Re4 Ne7 29 d4 Nf5 30 Rxe5 Nxh4 =+. This gives Black better chances than in the game continuation.

37...Nd5 38 Be4 h4 39 Kf3 Nb6 This weird-looking manoeuvre is in fact highly venomous. In particular, ...Na4 is threatened and in general Black will attack on the dark squares. 40 c4?!

27 fxg4 Rf4 28 Kd2

40 b3 is stronger but requires quite lengthy calculations. After 40...Nd5 41 Bxd5 Kxd5 42 Kg4 White finds salvation in the pawn ending: 42...e4 43 c4+ Kd4 44 dxe4 Kxe4 45 Kxh4 Kd4 46 Kg4 Kc3 47 Kf5 Kb2 48 Ke6 Kxa2 49 Kd7 Kxb3 50 c5! a5 51 Kxc7 a4 =.

After 28 Re4 Rxe4 (28...Rf2 29 Ba4 offers White counterplay) 29 dxe4 one can agree on a draw. The bishop is blunted by the e4- and g4-pawns, but he has a good set-up on the queenside (b2 and c3) and there is little scope for the black king to break through.

40...Kc5 41 Bf5?! Na4

28...Kd6 29 Ke3 Ne7 30 Bf3 h5 31 gxh5 Nf5+ 31...gxh5 =.

The last pawn has to move to a light square, which makes White’s position indefensible.

32 Ke2 gxh5 33 Rg1

42 b3

Obviously, the Indian grandmaster was still striving for victory.

42 Ke4 Nxb2 43 Kxe5 h3 –+. 42...Nc3 43 a3 Kd4! 44 b4 (D)

33...Rxh4 34 Bxb7 Perhaps 34 Rg6+ is more precise: 34...Ke7 35 Bxb7 Rh2+ 36 Rg2 Rxg2+ 37 Bxg2 =. In the game continuation, the black pieces end up better placed. 34...Ne7 35 Be4 Rh2+ 36 Rg2 Rxg2+ 37 Bxg2 (D)

Black to play 152

44...Nd1 is even more accurate: 45 Ke2 Ne3 46 Be6 Nc2 47 b5 Nxa3 –+.

emerged where the knight was better than the bishop, even in play on two flanks. The moment of truth came after the manoeuvre ...Nd5-b6, which saddled White with queenside weaknesses.

45 Ke2 h3 46 Bxh3 Nc1+ 47 Kd2

3 Nf3 Nf6

47 Kf3 Nxd3 –+.

Black is not obliged to get embroiled in a fight for the d4-pawn and can confine himself to unhurried development. But there is a negative: White can then establish pawns on c4 and d4. However, if Black is happy to play for a draw, then this may not be such a problem.

44...Na2

47...Nxd3 48 c5 e4 49 Bf1 Ne5 Black’s queenside pawns are ideally situated on dark squares, which are inaccessible to the bishop, but also far enough from the phalanx of white pawns. Thus White gets no chance to draw by eliminating all the black pawns.

Game 37

Not 49...e3+? 50 Kc2 =+.

Grishchuk – Feygin

50 a4 Nc4+ 51 Ke2

European Clubs Cup, Ohrid 2009

51 Kc1 Ne3 52 Be2 Nd5 –+.

1 e4 d5 2 exd5 Qxd5 3 Nf3 Nf6 (D)

51...Ne3 52 Bh3 Nd5 0-1 Now the c5-pawn falls and all the others will follow. Besides, the white king can’t attack the passed e-pawn.

Conclusions From time to time I have to fight ‘on the other side of the barricades’, handling these positions with White (S.Kasparov-Klerides). White mustn’t play too passively, as this can lead to trouble (J.Ramirez-Munoz Pantoja).

White to play This is a generally useful developing move, but plays little role in the fight for d4. If Black now launches an assault on d4, this move will be almost a loss of a tempo.

There is no point in White seeking a position with opposite-coloured bishops (Lamoureux-Prié). I would like to warn Black not to ignore the natural advance ...e5 without good reason (C.Garcia-Mancebo).

4 d4 I shall generally ignore moves like 4 Nc3 (settling down in front of the pawn) as it leads, at the worst for Black, to structures we have already discussed in previous chapters.

A weakening of White’s queenside structure can prove a serious matter, as the black knight can then seek out strongpoints (Doettling-Kalod).

4...c6

In our main game, White found time for the rare manoeuvre Nd2-f3-g5, the point being to preserve the bishop-pair. He succeeded, but at a high price, as it turned out Black could set up a favourable ‘two knights vs two bishops’ scenario. To save a piece, White had to weaken his kingside pawns, and as a result the knights got strongpoints. After a manoeuvring struggle, an interesting ending

Let’s examine other continuations: a) 4...Bf5 5 c4 Qd8 6 Qb3 Qc8 does not promise Black an easy life. It is hard to believe Black can equalize with ...Qxd5-d8-c8 if White doesn’t also waste 153

time. 7 Nc3 e6 8 Bf4 Be7 9 Nh4 Bg6 10 Nxg6 hxg6 (Van den Doel-Tiviakov, Dieren 2006) 11 Rd1 +=. b) 4...Bg4 is a popular and important alternative. After 5 Be2 (D) Black has a wide choice:

White to play b21) 12 Bg3 Be7 13 f4 (a weighty decision {pawns don’t go back!}; with the light-squared bishop exchanged I am suspicious of this move: the g3-bishop is blunted but it can be exchanged via h4) 13...g6 14 Bh4 (White must hurry, since once the knight leaves f6, it will be difficult to exchange off the bad bishop) 14...0-0 15 Rad1 Re8 16 Qf3 Qc7 17 Qb3 Nd5 18 Bxe7 Rxe7 19 Nxd7 Rxd7 20 Ne5 Rd6 with approximate equality, Swinkels-Motylev, European Clubs Cup, Ohrid 2009. Having exchanged three pieces, Black can sigh with relief.

Black to play b1) 5...e6 looks a bit passive. 6 h3 Bh5 7 c4 Qd7 (7...Qd8 8 Qb3 Qc8 9 Nc3 Be7 10 Be3 Nbd7 11 g4 Bg6, Korneev-Tiviakov, FIDE World Cup rapid, Khanty-Mansiisk 2005, 12 Nh4 Be4 13 Nxe4 Nxe4 14 Qc2 += is definitely somewhat better for White) 8 Nc3 (8 g4!? Bg6 9 Ne5 +=) 8...Bxf3 (the optimistic 8...Nc6 is well met by 9 g4 Bg6 10 d5 exd5 11 cxd5 Nb4 12 Ne5 +=, when the b4-knight feels uneasy on foreign soil) 9 Bxf3 Nc6 10 Be3 0-0-0 (Rozentalis-Krakops, Tallinn 1998) and now the far-from-obvious 11 Qe2! could have consolidated White’s plus. He has the bishop-pair and the prospect of attacking in the centre and on the queenside. The key point is that the d-pawn is poisoned: 11...Nxd4?? 12 Bxd4 Qxd4 13 Bxb7+! Kxb7 14 Rd1 +– and Black suffers material losses.

b22) 12 Rfe1!? is probably more accurate. Black cannot play the natural developing move 12...Be7?? because the bishop lacks the protection of the far-away queen, and White wins by 13 Nxf7 Kxf7 14 Qxe6+ +–. Black should choose 12...Nb6 13 Ne3 Nfd5 14 Bg3 +=. b3) 5...Nc6 (D) looks similar to the ideas we examined in the previous section, but it is less effective here since Black’s move ...Nf6 is not the most relevant in this context.

b2) 5...Qa5+ 6 Nbd2 c6 7 0-0 Nbd7 8 Nc4 (I feel the knight on c4 ensures comfortable play for White, as he keeps a grip on e5 without taking on great responsibilities – in contrast to the more critical play with c4; this is an ideal type of strategy when facing a stronger player) 8...Qc7 9 Nfe5 (it is probably better to include 9 h3!? Bh5 before 10 Nfe5 +=) 9...Bxe2 10 Qxe2 e6 11 Bf4 Qc8 (D) and then:

White to play Now: b31) 6 h3 Bxf3 7 Bxf3 Qd7 (7...Qxd4?? loses to 8 Bxc6+) 8 c3 0-0-0 9 Qb3 Nd5 10 0-0 e5 (Black takes a radical approach, but let’s not forget that White’s bishops will be glad about the opening of 154

the centre) 11 Rd1! exd4 12 cxd4 (an isolated pawn is created but White intensifies the pressure on the blockading square d5) 12...Kb8 13 Nc3 Ndb4?! (alas, Black can’t establish his knight on d5: 13...Nce7? 14 Bf4 Ka8 15 Nxd5 Nxd5 16 Bxc7! Qxc7 17 Rac1 Qd7 18 Bxd5 +–; 13...Nb6 is relatively best) 14 Qa4! Be7 (Adhiban-Akshat, Sharjah 2014) 15 Be3 +– threatening a3.

5 Be2 Bg4 In case of 5...g6 6 0-0 Bg7 7 Re1 0-0 8 c3 White has a solid and slightly better position; e.g., 8...Nbd7 9 Nbd2 Nb6 10 Ne5 Bf5 11 Ndc4 Qd8 (Slingerland-Tiviakov, Leiden 2010) 12 a4 +=. 6 0-0 Or:

b32) 6 c4 (D) and then:

a) 6 c4 Qa5+ 7 Nc3 e6 8 0-0 Be7 9 h3 Bh5 10 a3 Nbd7 (D) and then:

Black to play b321) 6...Qh5 7 Be3 (this is more flexible than the straightforward 7 d5) 7...0-0-0 8 Nbd2 e5 9 d5 (now that the possibility of ...e6 has disappeared, the pawn is safe on d5) 9...Nd4 10 Nxd4 exd4 11 Bxd4 Bd6 12 Bxg4+ Nxg4 13 h3 (MovsesianDamaso, Cappelle la Grande 1998) 13...Rhe8+ 14 Kf1 +/–. Black’s counterplay doesn’t compensate for the pawn deficit.

White to play a1) 11 Be3 0-0 12 b4 Qc7 13 Qb3 Rac8 14 Rac1 Qb8 15 Rfd1 a6 16 a4 Bd6 17 a5 gave White the initiative in Esserman-Kritz, Pawtucket 2008. Black lacks the possibility of ...b5 as it would be parried by the en passant capture axb6. a2) 11 Bf4 is even more challenging: 11...0-0 12 Re1 Rfe8 13 g4 Bg6 14 Bf1 h6 15 Bg2 Rac8 16 b4 Qd8 17 Qb3 a6 18 Rad1 b5 19 Ne5 += BaklanBaratosi, Mamaia 2013. After the pawn has left b7 the g2-bishop is unencumbered.

b322) In case of 6...Qf5 the most precise move-order is 7 0-0 (removing the king from all possible checks and pins) 7...0-00 8 Be3 e5 9 d5 e4 10 Nd4 Nxd4 11 Bxd4 Bd6 12 Nc3 Rhe8 (Piorun-Cioara, Arad 2014) and after 13 Re1 White stands better; e.g., 13...Bxe2 14 Qxe2 +=. The e4-pawn is separated from its colleagues while the c4-d5 pair is more secure.

b) 6 h3 Bh5 7 0-0 e6 8 c4 Qd8 9 Nc3 Be7 10 Bf4 0-0 11 a3 (D) and then:

We now return to 4...c6 (D):

Black to play b1) 11...a5 stalls White’s queenside offensive. However, White is free to become active in other

White to play 155

parts of the board: 12 Re1 (12 g4! Bg6 13 Ne5 += is more critical) 12...Re8 13 Ne5 (I hope Oleg Korneev won’t be offended if I venture to criticize such a strategy; other things being equal, it is inadvisable for White to decrease the number of pieces on the board) 13...Bxe2 14 Rxe2 Na6 15 Rd2 Nd7 16 Ne4 Qc8 17 Qc2 Nf8 18 c5 Ng6 (another pair of pieces will leave the board) 19 Nxg6 (19 Bg3!?) 19...hxg6 20 Bd6 Nc7 (Black mustn’t capture on d6) 21 Bxe7 Rxe7 22 Nd6 Qb8 23 Rb1 Ne8 += Korneev-Tiviakov, FIDE World Cup rapid, Khanty-Mansiisk 2005. After the knight retreats to c4, White’s position remains a little better, but Black’s defence is eased by the exchanges.

Black to play 10...Bxf3 Is this necessary? The prophylactic 10...Qc7 (in case of ...Nbd7 being met by Bf4) is good enough.

b2) 11...Nbd7 (D) took control of e5 in another game in the same match:

Now White’s bishop-pair will be a long-term asset. 11 Bxf3 Rd8 12 Rfd1 Nbd7 13 Rac1 13 Bf4 Qb4 +=. 13...Qc7 14 Qe2 Re8 15 g3 Rad8 16 Bg2 a6 17 h3 Bd6 18 a3 e5 Black seeks to avoid a gradual suffocation and aims for play in the centre. 19 c5 Bf8 20 b4 (D)

White to play 12 Ne5 Bxe2 13 Qxe2 Re8 (13...Nxe5 doesn’t equalize: 14 dxe5 Nd7 +=) 14 Rad1 Qb6 (14...Nf8!? seeks to consolidate the kingside) 15 Rfe1 (with White’s major pieces concentrated on the e-file, Black must be wary of the Nxf7 sacrifice) 15...Qa6 16 g4!? Nf8 (I like the fact that Tiviakov refrains from pawn advances that would create ‘hooks’ for White’s kingside advance) 17 g5 N6d7 18 Nxd7 Nxd7 19 d5!, KorneevTiviakov, FIDE World Cup, Khanty-Mansiisk 2005. White has finally executed his breakthrough, but it isn’t so deadly because Black has no weaknesses. After 19...Bf8 20 d6 White has the initiative but the struggle is double-edged. White’s king position is compromised, while the passed d6-pawn is solidly blockaded and it could even become a weakness in the ending.

Black to play White has arranged his pawns perfectly: they are placed on the opposite-coloured squares from his unopposed bishop, the idea being that if the darksquared bishops are exchanged, he will be left with a good bishop. 20...exd4 21 Rxd4 Ne5 22 Rcd1 Rxd4 23 Rxd4 Qc8 24 g4 h6 25 f4 Ng6 26 f5 Ne5 27 Qf2

6...e6 7 c4 Qd6 8 Nc3 Be7 9 Be3 0-0 10 Qd2 (D)

However, the line of pawns on h3, g4 and f5 (light squares) is less appropriate for the reasons described in the previous note. This is especially 156

so because Black can set up a sturdy barrier on the d8-h4 diagonal.

Now the threat of Bh6 is always in the air, so it is dangerous for the black king to move far away.

27...Rd8

51...Ke8 52 Bb3 Ke7 53 Bg8 Nf3+ 54 Ke3 Nfe5 55 Ke4 Kf8 56 Bb3 Ke7 57 Be3 Ng4 58 Bf4 Nge5 59 Be6 Ke8 60 Be3 Ke7 61 Ba7 Kd6 62 Bf2 Ke7 63 Bg3 Ke8 64 Kd4 Nf3+ 65 Kd3 Nfe5+ 66 Kc3 Ke7 67 Bf2 Kd6 68 Be3 Ke7 69 Kb3 Ng4 70 Bg1 Nge5 71 a4 bxa4+

27...Be7!?. 28 Qd2 Be7 29 Bf2 Rd7 30 Qe2 Qc7 31 Rxd7 Nfxd7 32 Bg3 Bf6?! Better is 32...a5. 33 Ne4 Qd8 34 Nd6 +/–

The reduction in the number of pawns is welcome to Black. However, an additional threat appears: a king-march with Kxa4-a5-a6-b7.

Earlier this square was guarded by the black bishop, but it is available now.

72 Kxa4 Nd3?! 73 Ka5 73 Bxd7! wins: 73...Kxd7 74 Bd4 (now Bxf6 is threatened) 74...Ke7 75 Ka5 Kf8 76 Bc5+ Ke8 77 Kb6 +–.

34...b6 35 cxb6 Qxb6+ 36 Bf2 Qb8 37 Ne4 (D)

73...N7e5 74 Bb3 Kd6 75 Be3 Ng4 76 Bd2 Nde5 77 Kb6 Nd7+ 78 Kb7 Nde5 79 Bf4 (D)

Black to play Black’s chances of a successful defence are limited. It is difficult to protect the queenside weaknesses while fighting against the bishop-pair. Black to play 37...Qb5 38 Nxf6+ Nxf6 39 Qxb5 Zugzwang. The prolonged bishop dances have tortured the opponent to death.

39 Qe3 is interesting, not allowing Black to reconnect his isolated a- and c-pawns.

79...Ke7 39...axb5 40 Bg3 Nfd7 41 Kf2 f6 42 Ke3 Kf7 43 Kd4 Ke7 44 Be4 Kd6 45 Bc2 Ke7 46 Bd1 Kd8 47 h4 Ke7 48 Bf4 Kd8 49 g5 hxg5 50 hxg5 Ke7 51 g6 (D)

In case of 79...Nf2 80 Bxe5+ Kxe5 (80...fxe5 81 f6 +–) 81 Kxc6 the knight can’t stop the passed bpawn: 81...Kxf5 82 b5 Nd3 83 b6 +–. 80 Kc7 Nf2 81 Bxe5 fxe5 82 Kxc6 Nd3 83 b5 e4 84 b6 e3 85 Bd1 Ne5+ 86 Kd5 Nd7 87 b7 Kf6 88 Kd6 Nb8 89 Be2 1-0

Conclusions Personally, I don’t like Black’s sluggish plan with ...Nf6; it feels like an ‘appeasement strategy’. Black to play 157

Please excuse me for giving such a categorical marking, but this move truly is unpromising.

4...Bf5 is dubious, as it allows White to make strategic gains without weakening his kingside (Van den Doel-Tiviakov).

3 c4?!

If Black reverts to the ‘aggressive’ plan with 4...Bg4 and 5...Nc6, then he suffers because ...Nf6 doesn’t fit in with the plan of putting pressure on the d4-pawn; White gets slightly the better chances (Piorun-Cioara and Movsesian-Damaso). The modest 5...e6 also doesn’t equalize; White typically gains the bishop-pair and a strong centre (Rozentalis-Krakops). I also don’t believe in 5...Qa5+; White stands a bit better then (SwinkelsMotylev).

also has serious drawbacks as the black queen moves away with check and can deny White his goal of setting up pawns on d4 and c4. The simple 3...Qe4+ is good enough: after 4 Qe2 (4 Be2? Qxg2; 4 Ne2?! Qxc4) 4...Qxe2+ 5 Nxe2 (Mellado Trivino-Monell Camarasa, Manresa 2012) 5...Nc6 it is White who feels uneasy. For example, 6 d4?! doesn’t work due to 6...Nb4 =+.

Instead 4...c6 5 Be2 Bg4 looks safe but dooms Black to defending a slightly worse position. In the line 1 e4 d5 2 exd5 Qxd5 3 Nc3 White could only dream of getting the pawn to c4 and the knight to c3 without losing time. This set-up intensifies his control over the centre and queenside. So why does Black allow it? In our main game Black unsuccessfully fought against the bishop-pair in a long and unpleasant ending. Black to play

On the other hand, I suggest you pay special attention to the games of the match KorneevTiviakov, played in the Khanty-Mansiisk World Cup. In matches, players use different tactics from those they apply in open tournaments. One doesn’t always need to play for victory; it is enough to make draws as Black and exert pressure with White. That’s what Sergei did. Black was under pressure all the time without equalizing the chances. Tiviakov avoided pawn-weaknesses and played extremely solidly. Although White definitely had the initiative, Korneev didn’t win any of these games.

3...e5!? If Black wants to avoid immediate simplifications, he can attack the pawn with 3...Nc6, when 4 Nf3 Bg4 5 Be2 transposes to the line we discussed in Games 33 and 34. White has an independent option, viz. 4 Be3 e5 5 Nc3 Bb4 6 Qd2?! (6 Ne2), and then: a) 6...Bxc3 7 bxc3 (Meertens-S.Kasparov, Prague 2011) 7...Nf6 8 c4 Qd6 (precisely here!) 9 d5 Ne7 gives Black ideal play due to White’s damaged pawn-structure.

3 d4

b) 6...Qa5 is also not bad, when Black can handle the position in gambit style: 7 dxe5 Nge7 8 a3 Bf5 9 0-0-0 0-0 10 Bd3 (10 axb4? Nxb4 –+) 10...Rfd8 with the initiative for Black, Dražić-Tiviakov, Trieste 2007. Who could say, looking at this position, that the Scandinavian Defence is a passive opening?

Black has a wider choice if White plays d4 before developing any other pieces. Besides the plans we have already considered, after 3 d4 Black can also strike immediately in the centre by 3...e5.

Game 38

4 dxe5

de Firmian – Granda

It is difficult for White to count on any advantage whatever he plays.

Amsterdam 1996

4 Nf3 (D) is the alternative.

1 e4 d5 2 exd5 Qxd5 3 d4?! (D) 158

equal. The advantages and disadvantages balance each other out. b2) 11...Bf5 12 0-0-0 0-0 13 Re1 (13 Bc4 looks more reasonable) 13...Rfe8 14 f4 exf3 15 gxf3 Re6 (D).

Black to play This might have some value when facing a much stronger opponent, as a silent hint at a draw. Then: a) 4...exd4 leads to very dry positions after 5 Qxd4 Qxd4 6 Nxd4 =. White’s extra tempo means little when there are no queens and the pawn-structures are symmetrical.

White to play This is definite asymmetry; Black has blocked the e5-pawn, which now obstructs its own bishop. Moreover, the good blockading square f5 is available now thanks to the pawn’s shift from g2 to f3. 16 Kb1 Rc6 17 Rc1 h6 18 Be2 Rd8 19 Rhd1 Nb6 20 b3 Rd5 (forcing another pawn to move to a dark square, which is undesirable for the d2bishop) 21 f4 Rdc5 (I would call this the ‘ramming’ of the light squares; the c2-pawn can be protected only through an alteration of the pawn-structure) 22 Bd3 Bxd3 23 cxd3 Rxc1+ 24 Bxc1 Nd5 =+ Pazos-Granda, Erevan Olympiad 1996. Black has the better prospects, as his knight is definitely stronger than the bishop. A long tiring defence lies ahead of White.

b) 4...Nc6 leads to a richer struggle. After 5 Nc3 Bb4 6 Bd2 Bxc3 7 Bxc3 e4 the structures are asymmetrical, and we have a fight between a bishop and knight. Clearly if Black wants to play for a win, this is the right path. The risk is not high; the aim is to play around White’s darksquared bishop and start a ‘light-square symphony’. 8 Ne5 Nxe5 9 dxe5 Ne7 10 Qxd5 Nxd5 11 Bd2 (D) and then:

We now return to 4 dxe5 (D):

Black to play b1) 11...0-0 12 c4 Ne7 13 0-0-0 Re8 14 Bc3 (this set-up, with a pawn on c4, and a bishop on c3, is logical; it is unclear how Black will cope on the kingside without his dark-squared bishop) 14...Ng6 15 Rg1 h5 16 Be2 Nf4 17 Rde1 Bg4 (pinpointing White’s inaccuracies) 18 Bf1 (18 Bxg4?! Nd3+ 19 Kd2 Nxe1 20 Rxe1 hxg4 =+) 18...Bf5 19 h3 h4 20 g3 hxg3 21 fxg3 Ne6 22 Re3 Nc5 23 g4 Bh7 24 Bg2 = Ahmadinia-Tiviakov, Mashhad 2010. The Iranian player has operated quite accurately and the situation is approximately

Black to play Black now has a pleasant choice. 4...Qxd1+ 159

This pawn sacrifice keeps the game interesting, and is the way to try to punish White’s 3rd move.

b) 6 Nf3 Bg4 7 Bf4 Nge7 8 Bb5 0-0-0+ gave Black the initiative in Orel-Marinsek, Bled 1993.

4...Qxe5+ looks insipid, but any advantage White gets is little more than symbolic: an extra tempo or so in a position where not a great deal is happening. 5 Be2 Bg4 and then:

c) 6 Bf4 Bc5 7 Nf3 Nge7 8 Nbd2 promises White nothing good; e.g., 8...Bg4 (the simple 8...Bxf2 =+ is promising) 9 Bg3 0-0-0 10 Bd3 Nf5 11 h3 Nxg3 12 fxg3 Bxf3+ (Black also stands slightly better after 12...Be6) 13 Nxf3 Rhe8 14 Kd2 Nxe5 15 Nxe5 Rxe5 ½-½ Movsesian-Istratescu, French Team Ch 2003. It looks like Black was happy with a draw, since he passed up many opportunities to strive for an advantage.

a) The obvious-looking 6 Nf3? is bad because after 6...Bxf3 7 gxf3 Nc6 =+ White’s doubled pawns are as ugly as they look. b) 6 Be3 Bxe2 (6...Qxb2 7 Bd4 Qb4+ 8 c3 Qe7 9 Nd2 has unclear consequences, despite Black’s considerable lack of development) 7 Qxe2 Nc6 8 Nf3 Qe6 (R.Dimitrov-Rusev, Sunny Beach 2014) 9 Nc3 +=.

6...Bf5 7 c3?! 0-0-0+ 8 Ke1 8 Nd2 may be vigorously met by 8...g5!, or 8...f6!?, with the initiative.

c) If White doesn’t wish to give up the b2-pawn, he can play 6 Nd2 Bxe2 7 Nxe2 Nc6 8 Nf3 Qd6 9 Bf4 Qxd1+ 10 Rxd1 Bb4+ 11 c3 Ba5 12 0-0 Nge7 13 b4 Bb6 14 a4 a5 15 b5 Nd8 16 Ned4 0-0 17 c4 += Berelowitsch-Saljova, Hlohovec 1996. It would be rather difficult for White to win, but the initiative is definitely on his side and Black lacks counterplay.

8...f6 (D)

5 Kxd1 Nc6 (D) This seems to be the most accurate, though 5...Bc5 is also interesting: 6 f4 Nc6 7 Nf3 Bf5 8 Bd3 0-00 9 Ke2 Bxd3+ 10 cxd3 f6 with compensation, Dochev-V.Panbukchian, Plovdiv 1994.

White to play Black has a huge development advantage, which is more than sufficient compensation for a pawn. Please bear in mind that White has lost his castling rights. 9 Bb5 Not, of course, 9 exf6? Nxf6 10 Bb5 Bc5 –+. 9...fxe5 9...Be4!? 10 Nf3 a6 gives Black compensation. 10 Bxc6 bxc6 11 fxe5 Bc5 12 Nf3 Nf6! (D)

White to play

This impressive move could easily have been overlooked when calculating this position from some moves earlier.

6 f4 Other moves are no better: a) 6 Bb5 Bg4+ 7 f3 0-0-0+ 8 Bd2 Bd7 9 Bxc6 Bxc6, Linker-Bandza, Wiesbaden 1993. If anyone is in danger, it is not Black. 160

Or: 19 Kxh4 Ne3 20 Kg5 Rhg8+ 21 Kf4 Rd3 –+; 19 Rxh4 Rd1 –+. 19...Bh7 20 Nf3 20 Na3 Rhg8 21 Rhg1 Nxe5+ also leaves White in desperate trouble. 20...Rhg8 21 Rxh5 Also hopeless is 21 Kh4 Nf2 22 Kxh5 Be4 –+. White to play

21...Nf6+ 22 Rg5

13 Bg5

22 Kh4 Nxh5 23 Kxh5 Rd1 –+.

13 exf6? loses to 13...Rhe8+.

22...Ne4+ (D)

13...h6 14 Bh4 14 Bxf6 gxf6 15 Nbd2 fxe5 =+. 14...g5! 15 Bf2?! Capturing on f6 is always a mistake: 15 exf6?? Rhe8+ 16 Kf1 Bd3#. White also gets no joy from 15 Bg3 Ng4 16 Na3 Ne3 –/+. 15...Bxf2+ 16 Kxf2 Ng4+ 17 Kg3 17 Ke1 Rhe8 is not a position that White will survive.

White to play Now Black has a material advantage to go with his positional dominance.

17...h5! (D)

23 Kh4 Nxg5 24 Nxg5 Rd1 25 e6 25 Nxh7 Rh1#. 25...Rg1 25...Bxb1 –+. 26 Nd2 Rxa1 27 Ndf3 Be4 28 Nd4 28 Nxe4 allows 28...Rh1+ 29 Nh2 Rxh2#. 28...c5 29 Nb3 Rg1 0-1

White to play White’s extra pawn is a woefully inadequate reward for such suffering.

Conclusions

18 h4

In the play after 3 d4?! e5!? 4 Nf3, please investigate in which cases a knight in combination with a bishop is preferable to two bishops (PazosGranda). It is by no means easy for Black to seek winning chances in these lines, but there is scope for creative play in the ending.

18 Nxg5 doesn’t help either: 18...h4+ 19 Kf4 Rhf8! –+. 18...gxh4+ 19 Nxh4

161

Before delving into the material, we should note that while White’s main plan is to play d4 and a well-timed c4, he can begin with either 3 d4 or 3 Nf3. The latter is often motivated by the desire to avoid the sharp gambit line 3 d4 Bg4; this is understandable, even though I feel that this line is objectively dubious for Black. Note that after 3 Nf3, Black can reply 3...Qxd5, but this transposes to the line 2...Qxd5 3 Nf3 Nf6, which we examined in Game 37, and did not find especially satisfactory for Black.

In the main line Black sacrifices a pawn by meeting 4 dxe5 with 4...Qxd1+. He enjoys an enduring initiative and can sometimes restore the material balance (e.g.,Linker-Bandza). In our main game, the excellent 14...g5! led to serious pressure on White’s position. Conclusions on White’s Rare Third Moves By omitting 3 Nc3, White is seeking to set up his pawns on d4 and c4. I urge Black not to ignore this plan; he should disrupt his opponent’s schemes by countering vigorously in the centre. From White’s viewpoint, he should avoid 3 d4?!, but 3 Nf3 is very much a viable and interesting way to meet the Scandinavian.

Game 39 Topalov – Kamsky Wijk aan Zee 2006

6: 2...Nf6

1 e4 d5 2 exd5 (D)

1 e4 d5 2 exd5 Nf6 (D)

Black to play

White to play This continuation is much less common than the capture with the queen on d5. I feel that the placement of the knight on d5 gives White the possibility of carrying out the c4 advance, which we witnessed him struggling to achieve in earlier chapters, without a hitch. And while the queen is vulnerable on d5, it also exerts pressure on White in a way that a knight on this same square doesn’t quite manage.

2...Nf6 2...c6?! is too exotic. The pawn sacrifice can’t be correct as White hasn’t created any weaknesses yet. 3 dxc6 Nxc6 4 Nf3 e5 5 d3 Bc5 6 Be2 Nf6 7 0-0 h6 8 Nbd2 0-0 9 c3 leads to normal play, except that Black is a pawn down. 9...Re8 10 Ne4 Bb6 11 Nxf6+ Qxf6 12 Be3 Bc7 13 Nd2 +/– LupulescuBakalarz, Rijeka 2010.

While there are some sharp tactical possibilities after 2...Nf6, in many of the lines we shall examine the precise move-order doesn’t play a major role, so I shall divide the material into three games based on where Black develops his bishops: Game 39: ...Bf5 Game 40 : ...Bg4 Game 41 : ...Bg7 Game 39 also features some miscellaneous options for Black in the first few moves, including the unjustified gambit 2...c6?! and the provocative idea of an early ...Nc6.

3 Nf3 In this game we shall mainly examine Black’s plan of development with ...Bf5; in this case, it doesn’t matter so much whether White plays Nf3 or d4 first. But we should also mention that Black occasionally seeks rapid central counterplay by putting his knight on c6. However, it is hard to 162

recommend putting knights on c6 and d5 without any ‘pawn-cover’ as they will be forced back by the enemy pawns, which should result in White’s superiority. There are a few possible tricks, so it might make some practical sense to try it in blitz games. 3 d4 Nxd5 4 Be2 Nc6?! (D).

White to play Now White has a pleasant choice. 5 Bd3 A completely straightforward approach. Here are a couple of other options, the second of which may be the most testing reply:

White to play Even in this relatively favourable form, Black’s idea is unpromising; clearly it would be even less justified if White had played Nf3 instead of Be2. 5 c4 Nb6 6 d5 (there is no need for a sophisticated response; other things being equal, the space-grab must be profitable) 6...Ne5 7 Qd4 Ng6 8 Nf3 e6 9 Nc3 c5 10 dxc6 Qxd4 11 Nxd4 bxc6 12 0-0 (the c6-pawn is not especially tasty: 12 Nxc6 Bb7 13 Nd4 Rc8 14 b3 Bxg2 15 Rg1 +=) 12...Bb7 13 Be3 0-0-0?! (13...e5 is a better try) 14 c5 Na8 (an oddlooking manoeuvre, but 14...Bxc5 15 Nxe6 fxe6 {15...Bxe3?? 16 Nxd8 +–} 16 Bxc5 +/– is unfavourable for Black, while White’s pawnstructure is even better after the exchange in the centre: 14...Nd5 15 Nxd5 Rxd5 16 b4 +=) 15 b4 Nc7 16 Rfd1 f5 17 Bc4 +/– S.Kasparov-Starostits, Cutro 2005. Black’s problems are clear: the b7bishop, the c6-pawn...

a) 5 Be2 is not just modest, but verging on passive. Nonconfrontational development like this allows Black to continue playing in ‘standard mode’ and poses few problems. 5...e6 6 0-0 Be7 7 a3 0-0 8 c4 Nf6 9 Nc3 Ne4!? (guided by the principle that each exchange is welcome for the side with less space) 10 Be3 Nxc3 11 bxc3 Nd7 12 Qb3 b6 13 Rfd1 Be4 14 a4 a5 with an acceptable position for Black, Ribli-Beim, Austrian Team Ch 2004/5. b) The immediate attack on the bishop by 5 Nh4 (D) appeals to me.

White has other options on his third move, but they are less critical, and we won’t examine them in detail here. 3 Bb5+ seeks to frustrate Black’s smooth recovery of the d5-pawn, but also disorganizes White’s own pieces. 3 c4

Black to play

can be met with the very sound gambit 3...c6 (when White’s best option is to transpose to a Caro-Kann Panov Attack by 4 d4 cxd5) or the more ambitious and riskier Icelandic Gambit, 3...e6.

b1) 5...Be4?! (wasting time) 6 Nd2 Bg6 7 g3 c6 8 Bg2 e6 9 Nxg6 hxg6 10 Nf3 gave White a modest but very clear plus in Tkachev-Rausis, Hastings 1997/8; he has the bishop-pair for ‘free’. b2) After 5...Bg6 White can achieve a comfortable two-bishops advantage: 6 Nd2 e6 7 Nxg6 hxg6 8 Nf3 Nd7 9 g3 c6 10 Bg2 += Rowson-Arkell,

3...Nxd5 4 d4 Bf5 (D) 163

1997) and after 19 Rxd8 Qxd8 20 g3 += the black knight feels uncomfortable, as it lacks a good ‘circuit’.

Edinburgh (1) 1996. This is like a Caro-Kann variation where White would have played Nh3/e2f4 in order to capture the bishop by Nxg6. White has prospects of torturing Black for a long time, and a plan of queenside expansion with c4, etc. Please note that the knight is much better placed on f3 to support this plan than it is on c3 in the similar structures in the 2...Qxd5 lines.

7 0-0 Nc6 8 c4 (D)

b3) If the bishop chooses the more central retreat 5...Be6, it will come under fire from White’s pawn advances: 6 Be2 g6 7 c4 Nb6 8 d5 Bc8 9 Qd4 f6 10 Nf3 (the difference in piece-activity is palpable; a simple tempo-count is about 6-3 in White’s favour) 10...Bg7 11 Nc3 e5 12 dxe6 Bxe6 13 Qe4 Qe7 14 Qxb7 Bxc4 15 0-0 +/– ZviagintsevMellado, Barbera del Valles 1996.

Black to play Topalov understands what he has gained as a result of the ‘undesirable’ exchange of bishops: a very pleasant position with central domination.

5...Bxd3 6 Qxd3 (D)

8...Nb6 8...Nf6 doesn’t equalize either: 9 Nc3 Be7 10 Bf4 0-0 11 Rad1 Qd7 12 Ne5 Qe8 13 Qf3 Bd6 (Mikhalchishin-Ubilava, Rostov-na-Donu 1980) and now 14 Nb5! +/– looks strong, eyeing c7 and d6. More specifically, Nxc6 is threatened. Black to play

9 Nc3 Be7 10 Bf4 g5? (D)

The drawback of White’s approach is that it exchanges like pieces (bishop for bishop rather than knight for bishop), which eases Black’s cramp without making a strategic gain. On the positive side, it removes the active black bishop and prepares further rapid development. White grabs space and keeps the better chances.

Wow, so abruptly! If some weak player had the black pieces, then such a rash decision would not be so surprising, but when it is made by a player with the stature and experience of Gata Kamsky, we wonder what drove him to this idea. Presumably he missed something. On the other hand, the ‘normal’ 10...0-0 11 Rad1 Bf6 12 Ne4 +/– gives Black an unpleasant position, so maybe desperate measures already seemed relatively reasonable.

6...e6 With 6...c6 7 0-0 e6 Black sets up a Caro-Kann structure, which gives him a slightly passive, but solid position. 8 c4 Nf6 9 Nc3 Be7 10 b3 0-0 11 Bb2 Nbd7 12 Rad1 Qc7 13 Rfe1 Rfe8 14 Ne5 Nxe5 (I am not sure this is correct; now the white pawn on e5 will spearhead White’s pressure; the patient 14...Rad8 is more prudent – then 15 f4 might have seemed unpleasant but I don’t see anything worse than a slight initiative for White after 15...Nf8 16 Qf3) 15 dxe5 += Red8 16 Qg3 Ne8 17 Qe3 Rxd1 18 Rxd1 Rd8 (Kotronias-Tzermiadianos, Ikaria

White to play 164

11 Bg3!

17 b4?!

White finds the most challenging reply, sacrificing his d-pawn. Perhaps Kamsky had reckoned on 11 Be3 g4 12 Nd2 +=.

There was an instant and elegant win: 17 Be5! f6 18 Ng5 Nf5 (18...fxg5 19 Nh6#) 19 Nh6+ +–. 17...Nd5?

11...g4 12 Ne5 Nxd4 (D)

Could the players already have been in timetrouble?

Other continuations are also unsatisfactory: 12...Qxd4 13 Qe2 Qc5 14 Ne4 Qb4 15 b3 +/–; 12...Nxe5 13 Bxe5 f6 14 Bg3 Qd7 15 Nb5 +/–.

18 bxc5?! Again 18 Be5! would have finished the game: 18...Nf5 19 Nh6+ Nxh6 20 Qg3+ +–. 18...Nf5 19 Qf3 Rc8 20 Bd6 Nxd6 20...Bxd6 21 Rxd5 +–; better is 20...Rc6. 21 cxd6 Bh4 22 d7 Rc6 22...Rc7 23 Rxd5 exd5 24 Nef6+ Bxf6 25 Nxf6+ Kg7 26 Ne8+ Rxe8 27 dxe8Q Qxe8 28 Qg3+ +–; 22...Qxd7 23 Rxd5 +–.

White to play

23 Ne5 Rc7 24 Qg4+ Kh8 25 Nd6 1-0

13 c5!? Topalov entices the bishop to c5 to prevent Black from reinforcing his d4-knight with ...c5. He will then seek to attack and win the knight.

Conclusions If Black is allowed to exchange pieces without any concessions (such as White gaining the bishop-pair, or rapidly setting up a strong centre), then he can expect to gain equality (Ribli-Beim).

A more natural option is 13 Rad1 c5 14 Ne2 Nf5 15 Qe4 +/–; Black is so far behind in development that the extra pawn brings him no joy. 13...Bxc5 14 Rad1 0-0 15 Ne4 Be7 16 Nxg4 +/–

I like 5 Nh4, since after 5...Bg6, White can get the bishop-pair under excellent circumstances. The bishop’s retreat to e6 invites trouble from another direction as it provokes White to play aggressively in the centre (Zviagintsev-Mellado).

Less critical is 16 Qxd4 Qxd4 17 Rxd4 f5 +=, when an intense struggle still lies ahead. 16...c5? (D) 16...f5? also loses: 17 Nh6+ Kg7 18 Be5+ Kxh6 19 Qh3+ Kg6 20 Rxd4 +–. The sly 16...f6 enables Black to put up resistance; e.g., 17 Qxd4 Qxd4 18 Rxd4 f5 +/–.

If White simply offers an exchange of bishops on d3, then if Black plays the restraining ...c6, there can arise positions akin to lines of the Caro-Kann, with White having a pawn on e5, which are in his favour, and where he can play for a win with little risk of losing (Kotronias-Tzermiadianos). However, putting the knight on c6 is no better. In our main game, Kamsky played in a somewhat risky manner. Topalov drove Black into a critical situation with a vigorous pawn sacrifice. Kamsky retreated his knight to b6 when it was attacked by c4. Heading for the kingside (...Nf6) doesn’t promise equality either (MikhalchishinUbilava). By continuing with natural development, White gets a considerable advantage.

White to play 165

b1) 5 c4 is met by 5...e6, when Black’s gambit play generally yields compensation.

In our next game we move on to the more aggressive development plan with ...Bg4. This includes the notorious gambit line 3 d4 Bg4.

b2) 5 Bb5+

Game 40

Nbd7 6 c4 is a line on which I won’t go into great details, but just show a game where our main ‘consultant’ demonstrated a good option for White: 6...e6 7 dxe6 Bxe6 8 d5 Bf5 9 Nc3 Bb4 10 Ne2 00 (D).

Svidler – Dreev Russian Ch, Elista 1997 1 e4 d5 2 exd5 Nf6 3 Nf3 After 3 d4, Black can play 3...Nxd5 4 Nf3 Bg4, transposing to our main game. However, there is a more ambitious option in 3...Bg4 (D).

White to play 11 Bxd7 (Tiviakov gets rid of his ‘problematic’ bishop; if the knight were allowed to move to, say, c5, the bishop would remain ‘offside’) 11...Nxd7 12 0-0 += Tiviakov-Damaso, Caldas da Rainha 2014. Black has some compensation for the pawn (two bishops, etc.) but I am not sure that it is sufficient.

White to play This can lead to sharp gambit play, and is one of the major reasons why some players like to play 2...Nf6. White has both a simple option and a ‘show me’ approach:

b3) 5 g4 (D) is principled and strong.

a) 4 Nf3 challenges Black to prove that he has gained anything from his risky move-order: a1) 4...Nxd5 transposes to the main game. a2) 4...Qxd5 is the standard try. It transposes to note ‘b’ to Black’s 4th move in Game 37, a line where Black’s early ...Nf6 didn’t fit in well with sharp counterplay against the d4-pawn, so Black tended to end up a little passive. b) 4 f3 Bf5 (D) and here:

Black to play 5...Bg6 6 c4 e6 7 Nc3 exd5 8 g5 (the knight is kicked away from the centre with tempo) 8...Nh5?! (8...Nfd7 is preferable) 9 f4! (it turns out that Black’s problems are not restricted to the d5-point; the knight on h5 may perish as well: f5 and Be2 are unpleasant threats) 9...Nc6 10 Be2 dxc4 11 Bxh5 Bxh5 12 Qxh5 Nxd4 13 Kf2 +/– I.PopovLaylo, Quezon City 2014. Black doesn’t have enough compensation for the knight.

White to play 166

3...Nxd5 4 d4 Bg4 (D)

5...Bh5 The meek 5...Bxf3 demonstrates Black’s willingness to fight just for a draw, as the bishop-pair enables White to exert long-term pressure with minimal risk. Of course, that is assuming White plays precisely, which your opponents may very well not! 6 Qxf3 c6 7 Be2 g6 8 0-0 Bg7 9 Rd1 (saving a tempo on c3 as this pawn will probably want to move directly to c4 later on) 9...0-0 10 c4 Nc7 11 Qe4 f5?! (on general considerations I wouldn’t want to make such a weakening move; there might be a point if Black could follow up with ...e5, but this is hard to engineer) 12 Qe3 Kh8 13 Nc3 Nd7 14 b3 Qe8 15 f4 += Carlsen-Short, Wijk aan Zee 2010. White suppresses ...e5, and has a significant edge.

White to play This thrust, as usual, is rather provocative, as it invites White to chase the bishop with his pawns. 5 h3

6 c4 Nb6 7 Nc3 (D)

The modest 5 Be2 is also not bad: 5...e6 6 0-0 Be7 7 Ne5!? (in general, exchanges play into Black’s hands but White gains some useful space in the centre with this move) 7...Bxe2 8 Qxe2 0-0 9 Rd1 Nd7 10 c4 N5f6 (10...N5b6 can be met by 11 Nf3! +=, keeping as many pieces as possible on the board) 11 Bf4 c6 (with the knight on f6, it is unfavourable to exchange on e5: 11...Nxe5 12 dxe5 Nd7 13 Nc3 +=) 12 Nc3 Re8 13 Rd3 Nf8 14 Rad1 += (it is quite difficult to blast through Black’s fortifications, but the concentration of white pieces in the centre promises a long-term initiative) 14...Qa5 15 h4 Rad8 16 Qf3 (16 h5!? prevents ...Ng6 and intends Rg3) 16...Ng6 17 Bg3 Bd6?! 18 h5 Bxe5 (Alekseev-Gareev, Kochin 2004) and now 19 b4! is an effective ‘stab’: 19...Qc7 (or 19...Qxb4 20 Bxe5 Nxe5 21 dxe5 Rxd3 22 Rxd3 +/– and the black knight dies, as there is no place to step back) 20 dxe5 Rxd3 21 Rxd3 Nxe5 22 Qe3 Nfd7 23 h6 +/– with strong pressure; at some point Rxd7 may follow.

Black to play 7...e5?! You sometimes see this thrust in the Classical Caro-Kann (especially when White plays Nh3). Black seeks to sow disorder in the enemy army, but White turns out to have a strong and vigorous response. 7...e6 is more cautious, but then the typical 8 g4 Bg6 9 Ne5 gives White a standard advantage:

We now return to 5 h3 (D):

a) 9...N8d7 10 Nxg6 hxg6 11 Qe2 (11 a4!?) 11...Bb4 12 Bg2 c5 (Conquest-Ekström, European Team Ch, Leon 2001) 13 d5 +=. b) 9...Nc6!? is more challenging, but has an obvious drawback: doubled pawns. 10 Nxc6 bxc6 11 Bg2 Qd7 12 Qf3 0-0-0 13 c5 (13 0-0!?) 13...Nd5 14 0-0 Be7 15 Nxd5 cxd5 16 Bf4 Qc6 17 b4 += Hraček-

Black to play 167

Cerveny, Czech Team Ch 2008/9. White’s better pawn-structure (the advanced post on c5 guarantees an attack) enables him to start active play.

Better is 14...Bxc5 15 Re1+ Kf8 (15...Be7 offers White a choice between 16 Bxh6 0-0 17 Be3 and 16 f6! gxf6 17 Bxh6 with a strong attack) 16 Be3 Na6 17 Nb3 +/–. White’s compensation for a pawn is more than sufficient, but the struggle would have continued.

We now return to 7...e5?! (D):

15 g5! Effective and efficient. Svidler must have calculated this breakthrough from some moves back because other moves are considerably weaker. 15...0-0 If Black accepts with 15...hxg5, there is the stunning 16 Ne6! Qa5 (16...fxe6? 17 Qh5+ g6 18 fxg6 Bxc5+ 19 Kh1 +–) 17 Nxg7+ Kf8 18 Ne6+ fxe6 19 fxe6+ with an attack. 15...Bxg5 16 Bxg5 Qxg5 doesn’t work either, since after 17 Ne4 Qe3+ 18 Kh1, it turns out that in this position without weaknesses, Black will inevitably lose his queen in the centre of the board: 18...0-0 19 Nc2 +–.

White to play 8 g4! Svidler’s concrete reaction instantly brings Black to the verge of disaster. 8 dxe5?! is obviously weaker due to 8...Qxd1+ 9 Nxd1 Nc6 with compensation.

16 g6! fxg6 17 Ne6 (D)

8...exd4 9 Nxd4 Bg6 10 Bg2 c6 11 0-0 11 Qe2+!? is probably even stronger: 11...Be7 12 Bg5 h6 13 Bxe7 Qxe7 14 Qxe7+ Kxe7 15 0-0-0 +/– with a big development advantage. 11...Be7 12 f4 h6 13 f5 Bh7 14 c5!? (D)

Black to play 17...Qc8 18 Nxf8 18 Nd5! cxd5 19 Nxf8 Nxf8 20 Qxd5+ +/– reminds Black of his queenside problems (b7). Black to play

18...Bxc5+ 19 Kh1 Bxf8 20 fxg6 Bxg6

Black’s problems are obvious. The h7-bishop doesn’t participate in the fight, his queenside is undeveloped, while his king is still in the centre. White sacrifices a pawn to deny his opponent time to castle.

It seems that Black has escaped the worst, as material remains balanced. However, the attack continues.

14...N6d7

Capturing the bishop is even worse; here is a summary of the key lines: 21...gxh6 22 Qg4 Kg7 (22...Kh7 23 Rf6 +–; 22...Qe8 23 Rae1 Ne5 24

21 Bxh6! Ne5

168

Rxe5 +–) 23 Be4 Qe8 24 Rg1 Ne5 25 Bxg6 Nxg6 26 Rae1 Qf7 27 Re6 +–.

Game 37), a line where Black has spent time on the ‘unnecessary’ move ...Nf6.

22 Bf4 Nbd7 23 Ne4 Qe8 24 Qb3+ Qf7 (D)

Let’s move on to 3 Nf3 Nxd5 4 d4 Bg4.

All roads lead to Rome now. White has both an attack and a material advantage.

Even the ‘non-confrontational’ 5 Be2 is quite promising. White establishes his knight on e5 and develops an initiative around it (Alekseev-Gareev). But 5 h3 looks even more logical, asking the bishop to declare its intentions. The exchange 5...Bxf3 6 Qxf3 is unpromising for Black because White gets comfortable play with the two bishops (see Carlsen-Short). The retreat 5...Bh5 results in more complicated play but here White gains the advantage in a typical manner: placing the knight on e5 and advancing his pawns. After 6 c4 Nb6 7 Nc3, one possibility is 7...e6 8 g4 Bg6 9 Ne5, but Black fails to equalize with either 9...N8d7 (since a central break may follow; see Conquest-Ekström) or 9...Nc6 (his structure is weakened; see HračekCerveny). In the main game, the central counterstrike 7...e5?! had even graver consequences. Svidler excluded Black’s lightsquared bishop from the battle by a speedy advance on the kingside. White’s non-stop threats to the black king led in the end to material gains.

White to play 25 Qg3 25 Qxb7 +–. 25...Bxe4 26 Bxe4 Nf6 27 Bg2 Nh5 28 Qg5 Nd3 29 Be5 Qe7? A mistake, but there was no escape anyway.

Now we shall acquaint ourselves with the pattern of the struggle when Black fianchettoes with ...g6.

30 Qxh5 Qxe5 30...Nxe5 31 Rae1 Re8 32 Be4 +–. 31 Qf7+ Kh8 32 Rf5 Nf2+ 33 Kg1 Nxh3+ 34 Kf1 1-0

Game 41 Svidler – Berend

There are no more checks.

European Clubs Cup, Vilnius 1995 1 e4 d5 2 exd5 Nf6 3 d4 Nxd5 4 Nf3 g6 (D)

Conclusions In the ...Bg4 lines, it makes a difference if White starts with 3 d4 or 3 Nf3. 3 d4 Bg4 can lead to rather peculiar positions after 4 f3. Black’s gambit offers lively play, but isn’t 100% correct. The compensation proved not quite sufficient in Tiviakov-Damaso, while I.PopovLaylo showed how White could play aggressively on the kingside. If White simply interposes the knight with 4 Nf3 then Black can, if he wishes, capture on d5 with the queen rather than the knight. This transposes to 2...Qxd5 3 Nf3 Nf6 4 d4 Bg4 (see the notes to

White to play 5 c4

169

The text-move (5 c4) is simpler and more aggressive. Can grabbing the centre (with tempo!) be bad?

The most natural move. 5 Bc4 might appear harmless as now the knight is not obliged to retreat, but it is not so easy for Black to equalize. Note the similarity with some lines of the Alekhine Defence. Now:

5...Nb6 (D)

a) 5...c6 6 0-0 Bg7 7 h3 0-0 8 Re1 Nd7 9 Nbd2 a5 (the idea is to limit the white bishop’s possibilities on the queenside) 10 a4 N7b6 11 Bf1 Bf5 12 c3 (White sets up a solid b2-c3-d4 structure; by inserting ...a5 and a4 Black has given his knight the b4-square in case White plays c4: 12 c4 Nb4) 12...Qc7 13 Ne4 Rad8 14 Bg5 Rfe8 15 Qc1 Nc8 16 Ng3 += Wang Hao-Nepomniashchy, St Petersburg rapid 2012.

White to play Clearly, the knight now has less central influence than it did on d5.

b) 5...Bg7 6 0-0 0-0 7 Re1 (D) and then:

6 Nc3 White relies on simple development, and probably rightly so. However, 6 a4!? is interesting. White wants to determine Black’s intentions on the queenside while the knight is unable to return to d5. White’s future plans may include the move c5, in which case he would ideally like ...Nd5 not to be the reply. If Black allows a5, the knight will have to retreat again. Thus:

Black to play b1) 7...Nb6 8 Bb3 c5!? 9 c3 (exchanging queens by 9 dxc5 Qxd1 10 Rxd1 N6d7 offers White no more than approximate equality) 9...cxd4 10 Nxd4 e5 (10...Nc6!?) 11 Nb5 Bd7 12 a4 Na6 13 Be3 Qe7 14 Qd6 += A.Rotshtein-Maliutin, Minsk 1993. White has the initiative on the queenside.

a) 6...Bg7 7 a5 N6d7 8 Nc3 0-0 9 Be2 Nc6 10 d5 Nb4 (White appears better due to his space advantage; on the other hand, Black has no weaknesses and plans to undermine White’s centre by ...b6 and ...c5) 11 0-0 Nc5 12 Be3 b6 13 Nd4 += Swiercz-B.Savchenko, European Ch, Budva 2009.

b2) 7...Bg4

b) 6...a5 (D), stopping the aggression, is perhaps more reasonable.

(playing both ...Bg4 and ...g6 indicates a willingness to grant White the bishop-pair) 8 c3 e6 9 h3 Bxf3 10 Qxf3 c6 11 a4 a5 12 Na3 Nd7 13 Bd2 N5b6 14 Bb3 e5 15 Nc2 exd4 16 cxd4 (an isolated pawn is usually undesirable, but in this case White has strong pressure on the blockading square d5; 16 Nxd4 Nc5 allows Black counterplay) 16...Nf6 17 Bg5 Nbd5 18 Re5 Qb6 19 Rae1 gave White the initiative in Adams-Pechenkin, Edmonton 2009. White’s rooks are unusually but effectively placed.

White to play 170

(14...e6 15 Bf4 +/– leaves material balanced but White much better positionally) 15 Qxe7 Be6 16 Qxd8 Rfxd8 17 Bf4 += de FirmianAzmaiparashvili, Erevan Olympiad 1996. White has an extra pawn.

So the question is how the a-pawns’ moves affect matters after White pushes his c-pawn. 7 Nc3 Bg7 8 c5 Nd5 9 Bc4 Nb4 10 0-0 0-0 11 h3 b6 12 cxb6 cxb6 13 Bg5 Ba6 14 Nb5! (it is unfavourable to trade the light-squared bishop, and it is advisable to keep control over the blockading square d5) 14...Nd7 15 Qb3 Nf6 16 Ne5 Nfd5 17 Rfe1 Bb7 (17...f6? 18 Nc6 +/–) 18 Bd2 e6 19 Rac1 with acceptable play for Black, Ni Hua-A.Muzychuk, Wijk aan Zee 2010.

b) 7...Nd5 appears to give Black a strongpoint on d5, but White’s idea is to start immediate and strong pressure on the d5-knight. 8 Bc4 and then: b1) The attempt to consolidate with 8...c6 doesn’t equalize. 9 0-0 (D) and here:

We now return to 6 Nc3 (D):

Black to play

Black to play 6...Bg4?!

b11) 9...Be6

This invites a vigorous continuation that gives White a powerful initiative.

10 Qb3 Nxc3 11 bxc3 Bxc4 12 Qxc4 Nd7 13 Rb1 b6 14 cxb6 axb6 15 Qxc6 0-0 (15...Rc8 16 Qb5 Rxc3 17 Bg5 0-0 18 Qb4 += attacks c3 and e7) 16 Qc4 += Nayer-Zurek, Pardubice 2004.

6...Bg7 is much more standard. The best attempt to create an initiative is then 7 c5 (D), planning to bombard the knight if it returns to d5:

b12) 9...0-0 10 Re1 e6 (10...Nxc3 11 bxc3 b5 12 Bb3 a5 13 Bg5 Ra7 14 a3! {once again, we see that the bishop should remain on this diagonal} 14...Bf6 15 Bh6 Bg7 16 Qd2 +/– Landa-Seegers, Senden 2011; there is a huge difference in pieceactivity) 11 Bg5 Nxc3 12 bxc3 Qa5 13 Qd2 Nd7 14 Bh6 += Landa-Vishniakova, Sochi 2014. The c8-bishop doesn’t decorate Black’s position. b2) After 8...Nxc3 9 bxc3 the blockade has been lifted, while the damage to White’s structure plays no major role. White’s rapid development and pressure on the e-file lead to direct threats against the black king. Black might seek counterplay with ...b6 and ...bxc5, but this might even open the b-file to White’s advantage. 9...0-0 (D) and here:

Black to play a) 7...N6d7 is flexible but looks sluggish. 8 Bc4 0-0 9 0-0 c6 10 Re1 Nf6 11 h3 (White’s huge development advantage indicates problems for Black) 11...b5 12 Bb3 a5 13 a3 (making sure the bishop can stay on the a2-g8 diagonal, where it exerts unpleasant pressure on d5 and f7) 13...Na6 14 Qe2 Nc7 171

White to play

White to play

b21) 10 Qe2 Nd7 11 0-0 c6 (passive defence turns out to be very ineffective; for better or worse, Black should try the queenside break with ...b6 at some point) 12 Re1 (typical frontal pressure; it is strategically unpleasant for Black to have to play ...e6, entombing the c8-bishop) 12...Bf6? 13 Bh6 Qa5 (13...Re8?? 14 Bxf7+ Kxf7 15 Qe6#) 14 Bxf8 Nxf8 +– Kamsky-Abdullaev, Baku 2010. White simply has an extra exchange.

15 Ne5 (again best: the queen intends to leap to f3 with threats on f7 and a8) 15...c6 (the most resilient; 15...Bb7? 16 Qb3 +–) 16 d5 (seeking immediate gains before Black can develop his queenside; 16 Qf3 cxd4 17 Rad1 +=) 16...Bb7?! (better is 16...Qc7 17 dxc6 e6 {17...Nxc6? 18 Nxc6 Qxc6 19 Bd5 +–} 18 Qf3 +=) 17 Rb1 +/– (Black hasn’t made a lot of mistakes but by move 17 his position is cheerless; this suggests that the variation as a whole is a little dubious) 17...Qc8? (S.Kasparov-Zasypkin, Fagernes 2012; 17...cxd5 18 Rxb7 dxc4 19 Qxd8 Rxd8 20 Rxe7 +/–) and now I could have won by 18 d6 exd6 19 Nxf7 Rxf7 20 Bxf7 Kxf7 21 Qxd6 +–. Although material is nominally equal, it is as if White has an extra rook.

b22) 10 0-0 b6!? 11 Re1 (D) and then:

We now return to 6...Bg4?! (D):

Black to play b221) 11...Bb7 12 Bg5 (an almost universal reply in this position) 12...Bf6 13 Bxf6 exf6 14 d5 gave White the initiative in Nataf-Galego, Andorra 1999. b222) 11...Nd7 12 Bg5 Nf6 13 Ne5 +/– Zhang Zhong-West, Sydney 1999.

White to play 7 c5 Nd5 8 Qb3 Nxc3 9 bxc3 Bxf3 10 gxf3 b6 11 Bc4 e6 12 0-0

b223) 11...Nc6 12 Bg5 h6 13 Bh4 Bf6 14 Bxf6 exf6 15 cxb6 axb6 leaves White for preference due to Black’s doubled f-pawns, Vorobiov-Nureev, Moscow Ch 2002.

White’s splintered pawns look ugly, but dynamic factors are much more important here. White’s development advantage and bishop-pair define his advantage.

b224) 11...bxc5 may not be the best continuation, but at least it forces White to play accurately to keep the initiative. 12 Bg5 Bf6 13 Bh6 (stronger than the prosaic 13 Bxf6 exf6 14 dxc5 +=) 13...Bg7 14 Bxg7 Kxg7 (D).

12...Bg7 It seems that Black is about to castle. However... 172

I prefer the natural 6 Nc3. In our main game Black chose 6...Bg4?!, with rather deplorable consequences. Svidler replied with great energy, undaunted by the damage to his pawn-structure.

13 Qa4+! (D)

Black should prefer the more flexible 6...Bg7. It makes sense to continue chasing the wandering knight with 7 c5 Nd5 and then putting pressure on d5 in order to open lines. (Instead retreating the knight to d7 keeps a flexible structure but leaves Black too far behind in development – see de Firmian-Azmaiparashvili.) White should trust in his development advantage, and seek to break through (maybe at a cost in material or structure) before Black consolidates his forces (S.KasparovZasypkin). In Kamsky-Abdullaev, Black was severely punished for his passive play.

Black to play 13...Nd7 13...Qd7 is elegantly met by 14 cxb6! Qxa4 (14...cxb6 15 Bb5 +–) 15 b7 +/–.

Attempts to reinforce the d5-knight with ...e6 also fail to achieve equality; White should then avoid exchanging on d5 as ...exd5 would liberate the c8bishop (Landa-Vishniakova). The exchange of light-squared bishops helps Black but still doesn’t ensure an easy life (Nayer-Zurek).

14 Bb5 a6 15 Bc6 b5 16 Qa5 Rb8 16...Ra7? 17 a4 +/–; the most resilient course is 16...0-0 17 Bxa8 Qxa8 18 Qxc7 Qxf3 19 Qg3 +/–. 17 Bf4 e5?? 17...Rc8 is better.

Conclusions on 2...Nf6

18 dxe5 g5 19 Bg3 0-0 20 Rad1 1-0

I feel that 2...Nf6 is somewhat weaker than 2...Qxd5. White gains the advantage in every line we have considered. Even a simple pawn advance in the centre is sufficient for an advantage in many lines.

Conclusions Black has an uneasy time even after 5 Bc4, although this move doesn’t look very dangerous. White builds a ‘breakwater’ against the g7-bishop, with pawns on b2, c3 and d4. Typically he plays Re1 and Bg5, exerting unpleasant pressure on the e7-pawn. It is also hard for Black to find an effective role for his queen’s bishop. If it moves to g4, then after h3 it has little choice but to take on f3, since ...Bh5 is no good with a black pawn on g6 (Adams-Pechenkin). But if the bishop stays on c8, then after White plays h3, it is unclear where to put this piece. Usually White keeps his bishop on the a2-g8 diagonal, where it exerts pressure on the f7-pawn. However, Bf1 may also make sense (Wang Hao-Nepomniashchy).

Black’s light-squared bishop problems finding employment.

has

enduring

The bishop’s thrust to g4 can be parried by the natural h3. If it retreats, White’s kingside pawns may pursue the bishop: g4 and so on. Meanwhile, the exchange by ...Bxf3 gives White the bishoppair with all that it entails. If the bishop develops to f5, then ‘pestering’ moves like Nh4 can cause discomfort. None of these problems with the queen’s bishop are solved by the fianchetto with ...Bg7. 7: Rare Second Moves for White

After 5 c4 Nb6, White can play 6 a4!?, when it is probably best for Black to halt the expansion by ...a5, with an interesting fight, as in Ni HuaA.Muzychuk.

1 e4 d5 (D)

173

differences are mostly in Black’s favour: the white king is more exposed, while only Black still has his castling rights. a) 5 Be3 Nd7 6 Nd2 Bc5 7 Bxc5 Nxc5 8 Bc4 Nf6 9 f3 Ke7 10 Ne2 a5 11 c3 Bd7 12 Nc1 Rhd8 13 Kc2 (Bielinski-S.Kasparov, Guben 2013) and after 13...b5 Black stands slightly better. b) In reply to 5 Nd2 Black can bring the bishop directly to a good diagonal by 5...Bc5 6 f3 a5 7 Nc4 f6 8 Be3 Na6 9 c3 (this would make more sense if there were a black knight on c6, but on the other hand, the king needs a shelter on c2) 9...Be6 10 Bxc5 Nxc5 11 Kc2 Ne7 12 Ne3 a4 13 Bc4 (D).

White to play So far in this book, we have examined only 2 exd5. While this is without any doubt the main and most critical reply, there are alternatives, to which we devote our two final games in this book.Game 42 deals with moves that defend the e4-pawn: the reasonably healthy 2 Nc3 and the compromising 2 d3?!. In Game 43, we take a look at 2 e5?!, when Black gets an improved French or Caro-Kann, depending on how you look at it. In the notes, we also mention 2 d4?!, which fails to defend the pawn at all! Game 42 Dinić – S. Kasparov

Black to play

Paracin 2013

13...Bd7!? (no, I wasn’t drunk when carrying out the manoeuvre ...Be6-d7-e6; the point is to keep the bad enemy bishop alive {the f3- and e4-pawns are fixed on light squares} and to kick it back by ...b5) 14 a3 (with the pawn-structure now altered, the exchange of bishops seemed acceptable to me) 14...Be6 (the a4-pawn fixes White’s queenside, securing a long-term initiative for Black) 15 Ne2 Bxc4 16 Nxc4 Kf7 gave Black the initiative in J.Novak-S.Kasparov, Prague 2013. If White attempts to free himself by pushing the bpawn, this will only create isolated a- and c-pawns.

1 e4 d5 2 Nc3 White wishes to avoid the beaten tracks. While this isn’t such a bad move, it doesn’t give Black much in the way of real problems to solve. The weird 2 d3?! (D) is very rare, and allows Black at least equality by the simplest possible means:

We now return to 2 Nc3 (D):

Black to play 2...dxe4 (the most natural reply; 2...e6 and 2...e5 are possible too) 3 dxe4 Qxd1+ 4 Kxd1 e5. The position is close to symmetrical, but the Black to play 174

As briefly mentioned in the Introduction, in contrast to the similar Two Knights Variation of the Caro-Kann (1 e4 c6 2 Nc3 d5 3 Nf3) here the situation is significantly better for Black. In fact, he basically has an extra tempo as ...c6 is not very useful in these circumstances.

b) 3...Nc6!? is a worthy option; as usual, the knight’s placement on c6 provides active play. 4 Bb5?! Qd5! 5 Qe2 Bf5 (this tactical skirmish favours Black) 6 Ng3 Qxg2 7 Qe5 (PlaskettSpeelman, Catalan Bay 2003) 7...Bd7 8 Qxc7 Nf6 9 Qxb7 Rd8 =+.

2...d4

c) 3...Nd7 (D) is good, with a kind of tempo-up Caro-Kann.

Black grabs space. I believe this is the most ambitious reply. In many lines, a kind of reversed King’s Indian structure emerges. Certainly, White has an extra tempo, but it is not really felt with the centre closed. The assessment is likely to hover somewhere around ‘dynamic balance’, which is not a bad deal for Black by any means. If you are satisfied with equality and prefer to play with a more open centre, it is more than reasonable to seek an ‘improved Caro-Kann’ by 2...dxe4 3 Nxe4 (D).

White to play Then it is natural for White to put his pawn on d4 (when the analogy with the Caro-Kann is strengthened), but he might also try to profit from placing the pawn on d3. While that leads to a different type of game, it doesn’t provide any more chances for an advantage: c1) 4 Bc4 e6 (it is natural to establish a ‘breakwater’ on f7 and e6 against the c4-bishop; this makes accidents on f7 less likely) 5 Nf3 Ngf6 6 d3 (Demkovich-S.Kasparov, Donetsk 2010; White can also strive for the initiative with 6 Nxf6+ Nxf6 7 d4 a6 8 0-0 Be7 9 Ne5 c5 10 Be3 {10 dxc5!?} 10...cxd4 11 Qxd4 Qxd4 12 Bxd4 b5 13 Bb3 Bb7, but this is about equal, MeštrovićMalakhov, Croatian Team Ch, Tucepi 1996) and now 6...Nb6!? is interesting, with full equality after 7 Bb3 (7 Qe2 Nxc4 8 dxc4 Nxe4 9 Qxe4 =) 7...Nxe4 8 dxe4 Qxd1+ 9 Kxd1 =.

Black to play Now Black has a wide choice of viable continuations, including simple active piece development: a) A particularly interesting idea from a practical viewpoint is 3...Qd5!?. Then White’s best reply is without doubt 4 Nc3. I hope you recognize this position! Yes, we have transposed to the main line of the Scandinavian, as covered in Chapters 1-4 of this book. Of course, White has chances of keeping an edge here, so objectively this is less satisfactory than Black’s other third-move options, but on the other hand, with his choice of 2nd move White has already shown that he prefers to avoid these main lines. A final attempt to deviate by 4 Qf3?! is dubious: 4...Nc6 5 Ne2 f5!? 6 Qh5+? (6 Ng5 Qxf3 7 Nxf3 e5 8 d4 e4 =) 6...g6 7 N4c3 Qc5 (7...Qf7!?) 8 Qh4 e5 =+ Kenworthy-Prié, Paris-London 1994.

c2) 4 d4 (D) also leads to equality:

Black to play 175

We now return to 3...e5 (D):

4...Ngf6 5 Ng5 (5 Ng3 e6 6 Nf3 c5 saves a tempo: ...c5 instead of ...c6-c5; after 7 Be2 b6 8 00 Bb7 9 c3 Be7 10 h3 0-0 11 Be3 Nd5 Black’s position is already better, Sharma-S.Kasparov, Bhubaneswar 2010) 5...h6 6 N5f3 e6 7 Bc4 c5 (Black’s gains are evident again: the move ...c6 is absolutely unnecessary in this structure, and the pawn moves directly to c5) 8 Ne2 (DeljouS.Kasparov, Ahvaz 2007) and now after the simple 8...Nb6 9 Bb5+ (where else to go?) 9...Bd7 10 Bxd7+ Qxd7 it is full equality. And no wonder: in almost any opening (or open position) an extra tempo will significantly influence the game’s course.

White to play 4 d3

3 Nce2 (D)

4 Ng3!? is interesting. White intends to develop the bishop actively to c4 or b5, before the pawn moves to d3. Black should frustrate this plan. 4...Be6 (taking control of c4) 5 Be2 (White seeks a strategically-desirable exchange of bishops with Bg4) 5...Nf6 6 Nf3 Nfd7 7 d3 h6 8 0-0 g6 9 c4 a5, Kluger-Van Wely, Vlissingen 2001. White has problems finding full employment for his lightsquared bishop and the g3-knight. Furthermore, Black has a space advantage thanks to the d4pawn. Black to play

4...c5 5 g3 Nc6 6 Bg2 Bd6 7 Nf3 Nge7 8 0-0 f6 9 Nd2 Be6 (D)

Now Black can choose between two logical continuations: 3...e5 and 3...c5. The move-order is not of great importance though, as Black normally ends up with his pawns on e5, d4 and c5, though there are some subtleties. One way or another, Black should obstruct the active deployment of White’s light-squared bishop. 3...e5 A typical method for Black in the struggle against the King’s Indian set-up was demonstrated in the following game: 3...c5 4 g3 (such an early fianchetto with the centre already blocked provides a ‘hook’ for the enemy infantry) 4...Nc6 5 d3 h5!? (an interesting treatment: with this sharp flank advance, Black causes a weakening of White’s kingside, as he quite reasonably doesn’t want to allow ...h4) 6 h4 (now the f4 advance will entail an irreparable weakening of the g4-square) 6...Nf6 7 Bg5 c4!? 8 Bh3 (Black stands better in all variations; e.g., 8 Bxf6 exf6 9 dxc4 Qa5+ 10 Qd2 Bb4 =+) 8...Qa5+ 9 Kf1 cxd3 10 cxd3 Ng4 11 Nf3 e5 =+ S.Novikov-Ponkratov, Moscow 2010.

White to play The position is a kind of reversed King’s Indian. Of course, White has an extra tempo which, however, doesn’t play a big role in closed structures. Moreover, White’s early commitment has allowed Black to place his pieces ideally for this structure – in a regular Sämisch he would have needed to negotiate some tactical problems. In particular, the fact that Black has not yet castled can often turn out to be in his favour.

176

10 c4 g5

25 Ne1 Qa5 26 Nf3

It’s a pleasure to play such a position with Black, isn’t it? In fact, he enjoys the initiative from the very beginning.

Still with a hint at Nxg5, on occasion. 26...Qxd2+ 27 Nxd2 Rfb8 28 Bc2 Be8 29 Ne2 bxc4 30 Nxc4 Nb4 31 Bb1

11 f4 Qd7 12 Nf3 h6 13 Bd2 Bg4 14 f5 a6 15 Ne1 Bh5 16 Bf3 Bf7 (D)

31 Rfc1 a5 =+. 31...Bb5 32 Rc1 a5 33 a3 Bxc4 34 Rxc4 Na6 Although objectively the game should end in a draw, White’s task is unpleasant. He has to defend all the time without counterplay. 35 Ba2 Kf8 36 Rac1 Nc8 37 Ng1? A time-trouble mistake. 37...Nd6 38 Ra4 (D)

White to play An exchange of the light-squared wouldn’t be beneficial for Black.

bishops

17 Nc1 h5 18 Nc2 h4 19 g4 b5 20 b3 Bc7 On the other hand, it is desirable to exchange the dark-squared bishop as then the b2- and c3squares will be potentially weak, and may be used as entry-points by the black pieces. Besides, while the d2-bishop remains alive, White can pin his hopes on a piece sacrifice on g5 (Ne1-f3xg5).

Black to play 38...c4

21 Qe1 Qd8 22 Kf2 Ba5 (D) This is decisive. 39 Rxa5 Or: 39 bxc4 Rb2+ –+; 39 dxc4 Nc5 –+. 39...Nc7? Time-trouble. 39...cxd3 40 Kf3 Rb5 41 Ra4 Nxe4! would have won. However, it is difficult to venture such a line under time-pressure. 42 Kxe4 is met by 42...d2 –+ and then 43...Nc5+. White to play

40 Rxa8 Rxa8

The idée fixe is carried out.

40...Nxa8 =+.

23 Bd1 Bxd2 24 Qxd2 0-0

41 bxc4 Rxa3 42 Bb1 Na6 43 c5

After the dark-squared bishops have left the board it is more difficult for White to make a destructive sacrifice on the kingside, which is why castling is more justified now.

43 Ke2 Nb7 –/+. 43...Nb5 –/+ 177

The pawn sacrifice 2 d4?! (D) is highly dubious, and is again most likely to occur in blitz games, though this type of gambit does have its own band of fanatical supporters.

White even has an extra pawn, but from the strategic point of view his position is bad due to his awkwardly placed minor pieces. If only the bishop could reach c4. If, if, if... 44 Ne2 Nc3 45 c6 Ke7 46 c7? 46 Ke1 is more resilient: 46...Nc7 47 Kd2 N7b5 – /+. 46...Nxc7 47 Bc2 N7b5 48 Ke1 Kd6 49 Kd2 Ra2 50 Re1? Na3 51 Nxc3 Rxc2+ 0-1

Conclusions Black to play

I wouldn’t call the position after 2 d3?! dxe4 3 dxe4 Qxd1+ 4 Kxd1 e5 ‘deadly drawish’ as Black has a slight advantage already, and in the ending one can outplay an inexperienced opponent based on micro-nuances (and a strong adversary wouldn’t play such a passive variation!). You saw how things went in the game J.Novak-S.Kasparov.

By the way, this position can also arise from another move-order: 1 d4 d5 2 e4?!. It is logical for Black to accept the sacrifice: the strongest reply is of course 2...dxe4 (though transpositions to the French {2...e6} or Caro-Kann {2...c6} are possible too). 3 Nc3 Nf6 (3...e5 is another idea) 4 f3 (this is the notorious Blackmar-Diemer Gambit) and now:

Black can meet 2 Nc3 with either 2...dxe4, with an improved Caro-Kann, or 2...d4, with a kind of reversed King’s Indian. Both options are fully satisfactory, as we saw.

a) The peaceful 4...e3 is not so ambitious but possible – and a safe option for blitz games. 5 Bxe3 Nbd7 6 Bc4 Nb6 7 Bd3 c6 8 Qd2 e6 9 Nh3 Nbd5 was about equal in Kokholm-Agrest, Copenhagen 2007. Of course, moves like ...c6 are optional; Black could play ...a6 and ...b5, etc.

In our main game, I just acted by analogy with the King’s Indian Defence, obtaining slightly better chances thanks to my space advantage. Our final game features the uninspiring 2 e5?!, with a note on the reckless 2 d4?!.

b) 4...exf3 5 Nxf3 (D).

Game 43 Frink – S. Kasparov Arad 2013 1 e4 d5 2 e5?! This move is rarely played and, if you encounter it at all, it is most likely to be in a blitz or ‘bullet’ (one-minute) game. I feel it is appropriate to mark this move with a ‘?!’ as White gets an inferior version of the Advance Caro-Kann (1 e4 c6 2 d4 d5 3 e5). Black has a tempo in hand because he has no need to play the now-pointless move ...c6. It seems that White is fighting for equality from the very beginning.

Black to play 5...e6 (5...g6!? is a sturdy alternative; e.g., 6 Bc4 Bg7 7 0-0 0-0) 6 Bc4 a6 7 Bg5 Be7 8 Qe2 b5 9 Bd3 Bb7 (the black pieces have a harmonious disposition, and he still has the extra pawn) 10 00-0 Nbd7 11 Rhf1 Nd5 (a liquidation strategy in action; with a material advantage it is quite effective) 12 Nxd5 Bxg5+ 13 Ne3 (Naole-Rogers, 178

Kona 1998) 13...0-0 =+. Note that there is surprisingly extensive literature on the BlackmarDiemer Gambit.

In an Advance French, the bishop would still be on c8. In an Advance Caro-Kann, White would have an extra developing move. Either way, it is a substantial difference!

We now return to 2 e5?! (D):

6 Bd3 Nge7 7 Ne2 h5 8 b4 cxd4 9 cxd4 Bxd3 10 Qxd3 Nf5 11 0-0 Be7 12 h3 Qd7 Nothing dramatic has happened, but Black has arranged his forces conveniently and his position is marginally better. 13 Be3 a5 13...0-0 might be preferable. 14 b5 Na7 15 Nbc3 a4!? (D)

Black to play 2...Bf5 Black can also play 2...c5, at once controlling d4, where White will generally want to put his d-pawn. Note that Black carries out this advance in one move (rather than by ...c6-c5, as in the CaroKann). In the following examples White failed to equalize after 3 f4: a) 3...Nc6 4 Nf3 Bg4 5 c3 e6 6 Bb5 (here, unlike in the French or the Caro-Kann, White doesn’t even have enough resources to place his pawn on d4: 6 d4? cxd4 7 cxd4 Bxf3 –/+) 6...Ne7 7 0-0 a6 8 Ba4 Qc7 9 d3 Nf5 =+ P.Dimitrov-G.Szabo, Sunny Beach 2009.

White to play Preventing White from supporting his b-pawn by playing a4.

b) 3...Bf5 4 Nf3 e6 5 c4 dxc4 6 Bxc4 Nc6 7 Nc3 a6 =+ P.Dimitrov-Milanović, Zadar 2009. Everything is simple and logical. The d-pawn is backward, and d4 is safely controlled by Black.

16 Ra2

The text-move (2...Bf5) is like an improved French Defence, as Black has developed his ‘problem child’ light-squared bishop actively outside the pawn-chain.

16...Nc8 17 Bc1 Nb6 18 f4?!

16 Nxa4 is a bit worse. After 16...Qxb5 =+ the a3pawn is weaker than the b7-pawn.

As is often the case, putting the pawn on this square only blunts White’s own bishop and leaves weaknesses. After 18 Ng3 problems remain, but they are not so appreciable.

3 a3 e6 4 d4 c5 5 c3 Nc6 (D)

18...h4 –/+ Of course, Black shouldn’t let White create a pawn-phalanx on g4, f4, e5 and d4. 19 Kh2 Rc8 (D) As you can see, the black pieces have a lot of strongpoints (g3, f5, c4), unlike the enemy forces. White’s pawns are fixed on dark squares, hindering the c1-bishop. White to play 179

Conclusions on Rare Second Moves for White I imagine you are convinced that if White avoids 2 exd5 he can expect no advantage, and sometimes even fails to keep the position equal! As a matter of fact, this is one reason I prefer the Scandinavian over the Caro-Kann. A chess-player who starts the game with 1 e4 is ‘sentenced’ to wandering in the Scandinavian forests. Conclusion White to play

I hope that your time studying this book has been rewarding. Now you are aware not only of many concrete variations but also a whole range of typical pawn-structures and have a better understanding of Scandinavian plans.

20 g4 This weakens the kingside. 20...hxg3+ 21 Nxg3 Nxg3 22 Qxg3 g6 23 Rg2 Rc4 24 Qd3 Qc8 25 Rf3 Kd8

You have no doubt gained many insights into the hidden potential of the queen in the opening and middlegame. It turns out that this piece can work miracles in the hands of a skilful chess-player, even when some of the other forces are still waiting to be developed.

From a strategic point of view, Black’s position is very auspicious. He controls the strongpoints h5, f5 and c4, while White has many weaknesses. 26 Bb2 Qd7 27 Nd1 Rc8

What conclusions can be drawn? Here I shall express my opinion (which may differ from yours) as clearly as possible.

Already Black could start to show off with lines like 27...Qxb5!? 28 Ne3 Rb4 –+. 28 Bc3? Nc4 29 Nb2

The best line for Black to choose is 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd6.

It is infeasible to protect both weaknesses (a3 and b5).

Instead:

29...Qxb5 (D)

3...Qa5 is not so ‘fundamental’ but also acceptable. 3...Qd8 is possible but a little more passive. It is more difficult to equalize then. 2...Nf6 definitely doesn’t lead to equality. Of course, within almost all of these lines, there is scope for a stronger player to outplay a weaker one, provided that the choice of individual subvariation is made carefully. That’s the current state of affairs as it seems to me. Nevertheless, theory is constantly evolving. Who knows, maybe in 2020 1...d5 will be the most popular reaction to 1 e4. Or conversely, the Scandinavian Defence might have fallen completely out of favour.

White to play The remainder is simple. 30 Nxc4 dxc4 31 Qd1 Qd5 32 Rfg3 b5 33 Ba5+ Kd7 34 Qb1 Rh4 35 f5 gxf5 36 Rg7 Rch8 37 R2g3 Re4 38 Qb2 Rf8 39 Bb4 Rxd4 40 Bc3 Rd1 41 Rg2 Kc6 42 Qe2 Rd3 43 Bb4 Bxb4 44 axb4 Qf3 45 R7g3 Qxe2 46 Rxe2 Kd5 0-1

I wish you success with the Scandinavian! If you need inspiration, just recall our heroes’ names (and of course their games): Tiviakov, Kovalenko, Prié... 180

Exercises The following ten exercises are designed to test how well you have understood the Scandinavian and the typical positions and structures to which it leads. In each case, be sure to assess the position as well as choosing a move. Take the decision as seriously as you would in an important over-theboard game. Points are awarded for correct assessment as well as the correct choice of move, and in some cases there are bonus points for seeing important additional points.

Black to play 2) Paleologu – S. Kasparov Davos 2014 Choose your move: a) 21...Ng4 b) 21...Nd5 c) 21...Ng8

White to play answer

1) S. Kasparov – Lubbe Neustadt an der Weinstrasse 2009 What should White play? a) 11 Nxe4 b) 11 h5 c) 11 Ne2

answer

Black to play 3) Sitnikov – S. Kasparov Donetsk 2011 Make your choice: a) 27...Rxd4 b) 27...Nh3+ c) 27...Nxg4

answer 181

White to play

Black to play

4) S. Kasparov – Hapke

6) S. Kasparov – Smerdon

Prague 2011

Haarlem 2006

What would you play as White?

The piece sacrifice 8...Bxh3 is...

a) 14 Nf3

a) interesting and has unclear consequences.

b) 14 Nb5

b) a correct and winning decision.

c) 14 Qxe4

c) a losing mistake.

answer

answer

White to play

White to play

5) S. Kasparov – Fasser

7) Potapov – B. Savchenko

Ybbs 2008

Khanty-Mansiisk 2013

What should White do?

The queen sacrifice 10 Nxe5 is...

a) 19 a4

a) correct and offers lively piece-play.

b) 19 Ne6

b) a mistake which gives Black the advantage.

c) 19 Nc6

c) possible, but other continuations are not worse.

answer

answer 182

Black to play

Black to play

8) Vachier-Lagrave – Papaioannou

10) R. Byrne – Rogoff

European Team Ch, Warsaw 2013

USA Ch, Pasadena 1978

Choose the best move for Black:

What should Black play?

a) 12...Qxd3

a) 12...Na5

b) 12...Ne7

b) 12...Nd4

c) 12...e4

c) 12...Bxf3

answer

answer Answers

In these solutions, bold text is used only for the moves played in the game.

White to play 9) Alekseev – N. Mamedov Turkish Team Ch, Kocaeli 2014 How would you continue as White? White to play

a) 17 Bd4

1) S. Kasparov – Lubbe

b) 17 d6

Neustadt an der Weinstrasse 2009

c) 17 Ba5

White has a big advantage (1 point). answer 183

a) In the game the languid 11 Nxe4 (0 points) was met by 11...Bxe4 12 c3 Be7 13 Qe2 with just an initiative for White. b) After seeing the correct 11 h5! (2 points) one just needs to calculate the variations: b1) 11...Nxg3? 12 hxg6 Nxh1 13 gxf7+ Kd7 (13...Kxf7 14 Qe2 +–) 14 Qe2 Bd6 15 Bxe6+ Kc7 16 Bg5 Qf8 17 0-0-0 Nxf2 18 Qxf2 with an overwhelming position and material equality. Black to play

b2) 11...Bf5 12 Nxf5 exf5 13 Ne5 +/–.

3) Sitnikov – S. Kasparov

c) 11 Ne2?! (0 points) is equal.

Donetsk 2011

next exercise

You are right if you think Black has a won position (1 point). Currently White’s most dangerous piece is his knight. Therefore Black should destroy it: 27...Rxd4! 27...Nxg4 is also good, as 28 fxg4 Rxd4 –+ (2 points) is just a transposition. 27...Nh3+ (0 points) is considerably weaker. Then 28 Bxh3 Qxh3 is unclear.

Black to play

28 Qxd4 Nxg4 (2 points)

2) Paleologu – S. Kasparov

Destroying the king’s cover at any cost.

Davos 2014 Black has the advantage (1 point).

29 Rfe1

a) 21...Ng4! –/+ (1 point). After 22 Qe4 f6 23 exf6+ Black went wrong in the game with 23...exf6? =. The f-file should have been kept open for the rook: 23...Nxf6! (+2 points) 24 Qc2 (24 Qxe7+? Rf7 25 Qe5 Rd5 26 Qf4 Ng4 –+) 24...Ng4 25 Nf4 e5 –/+.

29 fxg4 is hopeless: 29...f3 30 Qe5 (30 Rf2 Qh2+ 31 Kf1 fxg2+ 32 Ke2 Re8+ –+) 30...Bxe5 31 Bxe5+ Kc8 –+. In the game 29...Nxf3+?! was played and the struggle dragged on. The winning line is 29...Qh2+ 30 Kf1 Qg3 31 fxg4 f3 32 Bh1 Qh3+ 33 Kf2 Bg3+ 34 Ke3 Qh2 –+.

b) 21...Nd5?! (0 points) is a little weaker. 22 Qf3 =.

next exercise

c) 21...Ng8?! (0 points) is solid (and equal) but more passive. next exercise

184

b) 19 Ne6 (2 points) is the most effective, as now Black is devoid of defensive resources. 19...Qc8 (19...fxe6 20 dxe6 Qc8 21 exd7 Nxd7 22 c5+ Kh8 23 Ba6 Qxa6 24 Rxd7 +–) 20 Nxg7 Kxg7 21 Rxe7 Rxe7 22 Qc3 +–. If you got this far in your calculations, you receive a bonus (+1 point). c) 19 Nc6 Bxc6 20 dxc6 Qc8 +/– (1 point). next exercise White to play 4) S. Kasparov – Hapke Prague 2011 White’s position is obviously better (1 point) but he must be accurate with his move-order. The author failed, I regret to say. a) If you chose 14 Nf3!, then you have earned 3 points. After 14...Qxe2 (14...Qf5 15 Ng5 Qg6 16 f4 +–) 15 Rxe2 +/– the black knight doesn’t reach d6 in time and so Ng5 will give Black problems protecting the f7-pawn.

6) S. Kasparov – Smerdon

b) 14 Nb5 Qxe2 15 Rxe2 a6 16 Nc3 += (1 point).

Haarlem 2006

c) 14 Qxe4 is good but not best (1 point). 14...Nxe4 15 Re1 Nd6 16 Rc2 +=.

The piece sacrifice 8...Bxh3? is the wrong decision (8...Bd7 is better); Black is on the verge of defeat (2 points). After 9 gxh3 Qxh3 White played 10 Re1?!, but stronger was 10 Ng5! (+1 point) 10...Qf5 11 Bf3 +–, when the white king is under no real threat.

Black to play

next exercise

next exercise

White to play 5) S. Kasparov – Fasser Ybbs 2008 White to play

Sometimes one has to choose between several good options, and it is important not to settle for a second-best move.

7) Potapov – B. Savchenko Khanty-Mansiisk 2013

a) 19 a4 +/– (1 point).

a) The sacrifice 10 Nxe5!? is quite correct (2 points) and suitable for chess-players who like 185

complicated play. After 10...Bxd1 11 Nxf7 Qd7 12 Rxd1 White had sufficient compensation for the queen. b) Since it is not a mistake, this answer gets 0 points. c) This is a reasonable answer if you had in mind the most solid move, 10 Bf4!? (2 points). After 10...Bxf3 11 Bxf3 Nxf3+ 12 Qxf3 e5 13 dxe6 Qxe6 14 Bg5 += White has a long-term comfortable advantage.

White to play

next exercise

9) Alekseev – N. Mamedov Turkish Team Ch, Kocaeli 2014 The natural 17 Bd4! (2 points) places Black in a difficult position (1 point). After 17...Qd6 (17...Qa6 18 Qc3 +/–) 18 f3 Re7 19 Qa5 Re2 (19...a6 20 c4 +–) 20 c4 +– White’s strong attack and positional advantage ensured victory.

Black to play

Other lines are weaker: 17 d6 (1 point) 17...Ne5 18 Bd5 Rxd6 19 Bxe4 Nxe4 gives White the advantage but the struggle continues. 17 Ba5? gets 0 points because 17...Qxb2 is unclear.

8) Vachier-Lagrave – Papaioannou

next exercise

European Team Ch, Warsaw 2013 Black’s situation is unenviable (1 point). He must hinder White’s attack at any cost. a) In the game the Greek grandmaster preferred to win a pawn by 12...Qxd3? (0 points) but after 13 b5 Nb8 14 Be3 Qf5 15 Nd2 +– he couldn’t defuse White’s pressure. 15...Bc5 16 Ne4 Bxe3 17 Rxe3 h5 18 Qxa7 Nd5 19 Nc5 Nxe3 20 Qxb7# (1-0). b) 12...Ne7 (0 points) doesn’t help: 13 Qxa7 c6 14 Be3 +–.

Black to play 10) R. Byrne – Rogoff

c) The pawn sacrifice 12...e4! (2 points) is the least of the evils, cutting off the terrible enemy bishop. After 13 Bxe4 Nxe4 14 dxe4 Qe5 15 f4 Qe6, intending ...g5, or 13 dxe4 Rhe8 (e.g., 14 Nd2?! Qe5) he at least has the initiative.

USA Ch, Pasadena 1978 a) After 12...Na5? (0 points) 13 Be2 +/– Black shouldn’t capture on c5 because of b4. b) 12...Nd4?! (0 points) is a poor move-order as it allows White to advance aggressively on the kingside: 13 g4! Nxf3+ 14 Qxf3 Bg6 and now rather than 15 b4?, many roads would have led to a significant advantage for White; for instance, 15 d6 cxd6 16 cxd6 Bxd6 17 Rd1 +/– or 15 c6 bxc6 16 dxc6 +/–.

next exercise

186

c) Black should get rid of the bishop: 12...Bxf3! (2 points) 13 Qxf3 Nd4 14 Bxd4 exd4 15 d6 Ne5 with unclear consequences (+1 point).

5...c6

Results (your level):

Now (after 5...c6):

26-31 – You understand the Scandinavian Defence better than the author. I am really pleased that you have spared your time for my humble work.

6 Ne5 – Chapter 1

21-25 – Your chess level is 2300-2450, and this book has hopefully improved your opening knowledge.

Detailed Index

10-20 – You are at good club level or a strong junior. Your outlook is widening whenever you study chess. Keep at it!

2 exd5

Less than 10 – If you are fairly new to chess, then don’t despair. But if you have been grinding away for a long time... then maybe the Scandinavian is not for you.

2 d3?!

Index of Variations

Now (after 2 exd5):

Chapter Guide

A: 2...Nf6

1 e4 d5

B: 2...Qxd5

2 exd5

2...c6?!

Other moves – Chapter 7

A)2...Nf6

2...Qxd5

3 d4

2...Nf6 – Chapter 6

3 Bb5+

3 Nc3

3 c4

Other moves – Chapter 5

3 Nf3:

3...Qd6

a) 3...Qxd5 – see 2...Qxd5 3 Nf3 Nf6

3...Qa5 – Chapter 3

b) 3...Nxd5 4 d4 – see 3 d4 Nxd5 4 Nf3

3...Qd8 – Chapter 4

3...Nxd5

4 d4

3...Bg4

Other moves – Chapter 2

4 Nf3

4...Nf6

4 Be2 Nc6?!

4...c6?! – Chapter 1

Now (after 4 Nf3):

5 Nf3

4...g6

Other moves – Chapter 2

4...Bf5

Other moves – Chapter 2

Other moves – Chapter 2

1 e4 d5

2 d4?!

2 e5?! 2 Nc3

187

4...Bg4

7...Bxf3

B)2...Qxd5

7...Nf6

Now:

8 Bxf3 Nxd4

B1: 3 Nf3

9 Bxd4

B2: 3 Nc3

9 Bg4

3 c4?!

9...Qe6+

3 d4?!

10 Be2 10 Kf1?

B1)3 Nf3 Bg4

Now (after 10 Be2):

3...Nf6

10...Qe4

4 Be2

10...c5

4 Nc3 4...Nc6

B2)

5 d4

3 Nc3

5 0-0 0-0-0

Now:

6 h3

B21: 3...Qa5

5 h3:

B22: 3...Qd6

a)5...Bh5

3...Qd8

b) 5...Bxf3 6 Bxf3 Qe6+:

B21)

b1)7 Kf1?!

3...Qa5

b2)7 Be2

4 d4

b3) 7 Qe2 Qxe2+ 8 Bxe2:

4 Rb1

b31)8...Nd4

4 g3

b32)8...0-0-0

4...Nf6

5...0-0-0

5 Nf3

6 c4

5 Bd2:

6 Be3

a) 5...c6:

6...Qf5

a1)6 Bc4

7 Be3

a2)6 Bd3

7 Nh4?

b)5...Bg4

7 0-0

5 Bc4: 188

a)5...Nc6

14 Bxd5

b) 5...c6 6 Qe2

14 Bd2

c)5...Bg4 5...c6

B22)

6 Bc4

3...Qd6

6 Ne5

4 d4

6 Bd2 Bf5

4 Nf3 Nf6

7 Ne5

5 g3

6...Bf5

4 Bc4 Nf6:

7 Bd2

a)5 Nge2

7 Ne5

b)5 d3

7...e6

4...Nf6

8 Qe2

4...c6?!

8 Nd5

5 Nf3

8...Bb4

5 g3

8...Bxc2??

5 Bd3

8...Be7

5 Nge2

9 0-0-0

5 Bc4

9 a3 Nbd7 10 0-0-0 – see 9 0-0-0 Nbd7

5 Nb5

10 a3

5...c6

9 Ne5

5...a6

9...Nbd7

5...Bg4:

10 a3

a)6 Be2

10 Kb1

b) 6 h3:

10 Ne5

b1)6...Bxf3

10...Bxc3

b2) 6...Bh5:

11 Bxc3 Qc7

b21) 7 Be2

12 Ne5 Nxe5

b22) 7 g4 Bg6:

12...b5

b221) 8 Bg2

13 dxe5 Nd5

b222) 8 Ne5

Now:

5...g6: 189

a) 6 h3

8 a4:

b) 6 g3

a) 8...g6

c) 6 Bc4

b) 8...Nb6

d) 6 Ne5

8...Nb6

e) 6 Nb5:

9 Bf4

e1) 6...Qd8

9 Ne5

e2) 6...Qb6

9...Qd8

6 Ne5

9...Qd7:

6 Be3

a) 10 h3

6 Be2

b) 10 Ne5

6 Bd3

c) 10 Be5

6 h3

d) 10 Nxb6

6 g3:

10 Be5

a) 6...Bf5

10 0-0-0?

b) 6...Bg4

10 Ne5

6...Nbd7

10 h3

6...Be6

10...Bg4

7 Nc4

10...Nbd5

7 f4

10...Be6

7 Bf4:

11 Qg3

a) 7...Nxe5

11 Qf4

b) 7...Nd5:

11...h5

b1) 8 Nxf7??

11...Be6

b2) 8 Nxd7

Now (after 11...h5):

b3) 8 Bg3

12 h4

b4) 8 Nxd5 Qxd5:

12 f3

b41) 9 Be2

12 h3

b42) 9 Nf3 7...Qc7 8 Qf3 8 d5?! 190

Index of Main Games

HUNGASKI – Charbonneau 30

Numbers refer to games. When a player’s name appears in bold, that player had White. Otherwise the FIRST-NAMED PLAYER had White.

IVANCHUK – Kovalenko 16 ; Tiviakov 11

ALAVI – Kasparov, S. 24

KAMSKY – Topalov 39

ANAND – Kasparov, G. 29

KASPAROV, G. – Anand 29

; Tiviakov 10

KASPAROV, S. – Alavi 24

; Van Wely 26

; Bogdanovich 35

BACROT – Tiviakov 20

; Dann 6

BARTHOLOMEW – Yudasin 31

; Dinić 42

BEREND – Svidler 41

; Frink 43

BOGDANOVICH – Kasparov, S. 35

; Harutjunyan 7

BOLOGAN – Tiviakov 3

; Lupulescu 25

CARLSEN – Caruana 32

; Neiksans 19

CARUANA – Carlsen 32

; Panchanathan 36

; Leko 21

; Sharma Dinesh 34

; Tiviakov 13

; Shukh 5

CHARBONNEAU – Hungaski 30

; Trygstad 33

DANN – Kasparov, S. 6

; Van den Heever 14

DAUTOV – Vogt 27

; Vorobiov 18

DE FIRMIAN – Granda 38

KOTRONIAS – Tiviakov 2

DINIĆ – Kasparov, S. 42

KOVALENKO – Ivanchuk 16

DRAŽIĆ – Oleksienko 9

; Laznička 15

DREEV – Svidler 40

LAZNIČKA – Kovalenko 15

FEYGIN – Grishchuk 37

LEKO – Caruana 21

FRINK – Kasparov, S. 43

LUPULESCU – Kasparov, S. 25

GEORGIEV, KI. – Henrichs 28

NEIKSANS – Kasparov, S. 19

GRANDA – de Firmian 38

NISIPEANU – Radjabov 8

GRISHCHUK – Feygin 37

OLEKSIENKO – Dražić 9

; Tiviakov 4

PANCHANATHAN – Kasparov, S. 36

HARUTJUNYAN – Kasparov, S. 7

RADJABOV – Nisipeanun 8

HENRICHS – Georgiev, Ki. 28

SHARMA DINESH – Kasparov, S. 34 191

SHUKH – Kasparov, S. 5 SVIDLER – Berend 41 ; Dreev40 ; Tiviakov 17 SWIERCZ – Tiviakov 1 TIVIAKOV – Anand 10 ; Bacrot 20 ; Bologan 3 ; Caruana 13 ; Grishchuk 4 ; Ivanchuk 11 ; Kotronias 2 ; Svidler 17 ; Swiercz 1 ; Van Kampen 22 ; Vovk, A. 23 ; Wan Yunguo 12 TOPALOV – Kamsky 39 TRYGSTAD – Kasparov, S. 33 VAN DEN HEEVER – Kasparov, S. 14 VAN KAMPEN – Tiviakov 22 VAN WELY – Anand 26 VOGT – Dautov 27 VOROBIOV – Kasparov, S. 18 VOVK, A. – Tiviakov 23 WAN YUNGUO – Tiviakov 12 YUDASIN – Bartholomew 31

192