Larry Kaufman - Sabotage the Grünfeld_small

Sabotage the Griinfeld

Larry Kaufman

Sabotage the Griinfeld A

Cutting-Edge Repertoire for White Based on 3 .f3

New In Chess 2014

© 2 0 1 4 New In Chess Published by New In Chess, Alkmaar, The Netherlands www.newinchess.com All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission from the publisher. All photos: New In Chess Archives. Cover design: Volken Beck Supervisor: Peter Boel Proofreading: Rene Olthof Production: Anton Schermer Have you found any errors in this book?

Please send your remarks to [email protected]. We will collect all relevant

corrections on the Errata page of our website www.newinchess.com and implement them in a possible next edition. ISBN: 9 7 8-90-569 1 -440-0

Contents

Explanation of Symbols .

6

Introduction. .

7

.

.

.

11

The Role of Komodo

Chapter 1 History of the f3 Anti-Grunfeld ............................... 1 3

Chapter 2 Third Move Offshoots

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

49

Chapter 3 The N eo-Grunfeld ......................................... l 1 Chapter 4 The Samisch with ...cl -c5 .................................. 11l

Chapter S The Samisch without ...cl -c5 ................................ 141

Exercises

.

................................................. 1 6 3

Solutions ................................................. 1 6 9

Index of Variations .

1 77

Index of Names.

1 83

Game List. . .

1 85

Bibliography .

1 87

5

Explanation of Symbols

The chess board with its coordinates:

8 7 6 5

;!;

White stands slightly better

:j:

Black stands slightly better

±

White stands better

+

Black stands better

+-

White has a decisive advantage

-+

Black has a decisive advantage

=

4

3 2 a

b

c

d

e

f

g

h

balanced position

00

unclear position

�

compensation for the material

>

strong (sufficient)

<

weak (insufficient)

�

better is

:::;

weaker is good move

!!

excellent move bad move

D

White to move

??

blunder interesting move

•

Black to move

!?

� II

King

?!

dubious move

Queen

D

only move

/::,.

with the idea

-

attack

Rook

� Bishop ttJ

Knight

t

initiative

tt

lead in development

�

counterplay

#

mate

corr.

6

correspondence

Introduction This book is about the position arising after the opening moves l .d4 tt::l f6 2 .c4 g6 3 .f3 . It is sometimes called 'Alekhine's Anti-Grunfeld' due to the critical role played in its early days by then World Champion Alexander Alekhine, and to the motivation of White's 3 .f3 move to avoid a proper Grunfeld Indian Defense, which would arise after the natural 3 . tt::l c 3 d S . Although a full White repertoire is part of this book, it also in­ cludes coverage of major non-repertoire lines so that a Black repertoire can also be derived from it. Explanations are aimed at club level players, but the actual analysis is intended to be suitable even for grandmasters. The idea of the weird-looking 3 .f3 (weird because the g I -knight usually goes there) is to provoke 3 . . . dS by threatening to play 4.e4, at a time when White's knight is not yet on c3. Then the Exchange Variation of the Grunfeld is more effec­ tive since Black cannot swap off his attacked knight on dS . The counter-argument is that White has paid a price for this, in the sense that f2-f3 is not as useful a move in general as tt::l f3 would have been. However it is still quite useful, as it goes well with queenside castling and a kingside attack similar to the Yugoslav Attack against the Dragon Sicilian or the Samisch King's Indian, into which play often transposes. The move 3 .f3 does have some other drawbacks. Instead of playing Grunfeld-style (3 . . . dS) , Black can target the dark squares weakened by f2-f3 with moves like 3 ... e S , 3 . . . cS , and 3 . . . tt::l c 6, the move chosen for the Black repertoire in The Kaufman Repertoire For Black and White (KRBW) , my last book. Black can also just choose the King's Indian, content with the fact that White is virtually forced to choose the Samisch Variation against it, which may not be his best option. This book might seem an odd choice for an author who just enthusiastically rec­ ommended the Grunfeld for Black in KRBW But I am always keen to try to prove White's advantage in chess, and when I realized the strength of the move 3 .f3 I felt that it deserved a book of its own. I expected this to be the only book in the current century devoted to this position, but a similar book by grandmaster Svetushkin came out first. While we agree more often than not, I point out the many places where I could not agree with him after doing further analysis. Although we are both grandmasters, Svetushkin is considerably higher rated than I am, but on the other hand I probably have much better computer hardware and software than he is likely to have used. These days this is very important. 7

S a b o t a g e t h e G r ii n fe l d

So what is my verdict on 3 .f3 from a theoretical standpoint? In my view, it is ex­ tremely difficult to demonstrate any meaningful advantage against the Griinfeld, but the 3 .f3 d5 neo-Griinfeld does seem to give White his normal opening edge. I think this is also true of the alternative third moves other than 3 . . . �g7 (or 3 . . . d6 first) , the King's Indian. Against the King's Indian, I don't think that the Samisch is White's best option, but I do think it suffices for a normal opening edge. So White is giving up something against the King's Indian to get something against the Griinfeld. Since the Griinfeld has a much better theoretical reputation than the King's Indian, White has more latitude as to what to play in the second case. To summarize, 3 .f3 seems to give White his normal slight plus whatever Black plays, whereas the usual 3 .tlJc3 fails to do so against 3 . . . d 5 ! . If all this is true, then 3 .f3 may simply be ' the best move' ! So the move 3 .f3 is suitable to use at every opportu­ nity, but is especially appropriate against opponents known to prefer the Griinfeld over the King's Indian. The 3 .f3 Anti-Griinfeld is by no means new, though it is newly popular. It was in­ troduced in 1 9 2 9 in games by Flohr and Nimzowitsch, but was quickly taken up by World Champion Alexander Alekhine in his title defense against Bogoljubow, and he played it many times over several years thereafter. It was named 'Alekhine's Anti-Gruenfeld Attack' in one book with the same name written by Alan 1. Watson in 1 99 6 . It was also played by many other famous players of that era, including Fine, Yudovich, and Bondarevsky. However the majority of these early games trans­ posed to the Samisch, and in the majority of games featuring 3 . . . d5 White refrained from queenside castling, which is now considered essential in this varia­ tion. Still, some of the early games do feature modern lines, and we trace their development in the historical section of this book. This book attempts to cover all the reasonably important variations arising after 3.f3 , but with the Samisch King's Indian being given restricted coverage to avoid duplication with Schandorffs excellent coverage of it in his recent White repertoire book. However, he only covers lines with 6 .�e3 so I fully cover lines without it, as well as some improvements I have found on his lines. This is not strictly a reper­ toire book, but I do make clear which lines I recommend for a repertoire for White, as well as which ones are playable for Black (i.e. limit White to no more than a 'par' opening plus) . Virtually everything in this book has been checked by the two strongest engines at the time of writing, Houdini 3 and (at a later stage) 4, and Komodo, for at least 1 5 minutes per position, usually more. Both of these engines are much stronger than the corresponding versions that were used for my last book, so the quality of analy­ sis is that much higher. This analysis is done using the 'IDeA' feature of 'Aquarium' so that hundreds of positions can be scheduled for analysis overnight (one core per position) . With an eight core and a twelve core computer devoted to this task, you can see how it was possible to analyze deeply the many thousands of positions in this book in a reasonable time. Most writers just use the engines to check as they 8

In tr o d u c t i o n

write, s o many positions will have less than a minute's scrutiny. M y method insures that the quality of the analysis will be top-notch, at least to the extent that the com­ puters can approach the truth. I use my own judgment as a grandmaster to decide which engine to believe when they disagree, as well as to identify the occasional in­ stances where they both get things wrong, primarily in endgames and in severely blocked positions. Because I am a co-author of Komada (together with the late Don Dailey) , I am usually able to explain in words why it evaluates a position as it does. I tend to favor Komodo's analysis over Houdini both because I better under­ stand where the scores are coming from, and because I believe that Komodo's eval­ uations are on average a bit more realistic in human terms. At the time of writing, Houdini is the stronger engine at blitz levels, while tests at levels averaging a couple minutes per move generally favor Komada. Based on this trend, I believe that at lev­ els like 1 5 minutes per move as used for this book, Komada is likely stronger than Houdini, but no one tests at such long levels so I can't be certain of this. But regard­ less of which engine is objectively stronger, Komada seems to 'like' the white side of most of the recommended lines in this book more than Houdini, correctly so in my opinion as the lines in question do score well for White in human practice. As with my previous books, my choice of recommended lines is primarily ob­ jective, and hence suitable even for the elite grandmasters, but since the text is aimed at ordinary club players, I hope that this book will appeal to a wide range of players, including even grandmasters. Another feature of the previous book which I retained for this one is that all refer­ ences to material values are based on my own scale, which is pawn = 1, knight = 3 . 5 , single bishop = 3 . 5 , two bishops together = 7 . 5 (i.e. half a pawn bonus for the pair) , rook = 5 . 2 5 , and queen = 1 0. This is far more accurate and reliable than the traditional 1 -3 - 3 - 5 -9 count. Most chess opening books focus on master I grandmaster games, with analysis by engines of the moves. In this book, I work partly from databases which include large numbers of games played by computers. Consequently I focus more on the analysis and less on the actual games, since the analysis done at 1 5 minutes per move will be of much higher quality than almost all of the actual games, which are played at much faster time limits than this. I make an effort to cite relevant games in the variations, but I 'm not fanatical about it, I don't always try to track down which computer-recommended move is the actual novelty in a sideline, especially since many novelties are played in computer games anyway. Frankly, since most opening analysis used in games comes from computers, I don't think it's terribly interesting to know which grandmaster happened to get the first chance to use a computer­ inspired novelty. In at least one case, an elite game featured a novelty followed for many moves by a variation I had already written up for this book! Working on this book has been somewhat of a novel experience for me. In my previous books, which covered the full range of openings, I had to cut off analysis at some reasonable depth in order to be able to write such books. This time, I was 9

S a b o t a g e t h e G rii n fe ] d

able to analyze all the way into the endgame in many lines, since the range of cov­ erage was so much narrower. I hope that I have come fairly close to the truth in at least most lines. The usual result of such deep analysis of a good White opening is that White should reach an endgame where he is the only one with winning chances, although Black should be able to hold. That is indeed the result shown here for many of the best Black defenses. I also want to point out that I have been playing this 3 .f3 line in tournaments since late 2 0 1 2 myself (mostly as White, once as Black) , and so I have some first­ hand experience with the lines. So far I believe I have gotten a clear (maybe win­ ning) advantage out of the opening in every White game, though I make too many mistakes at the board at age 65 so my actual results aren't that wonderful. For those who are not familiar with me, my tournament successes now span more than half a century, from second place in the Maryland Junior Championship in 1 96 1 to the World Open Senior Championship in 2 0 1 3 . I earned the grandmaster title by winning the World Senior Championship in 2 0 0 8 . I won the American Open Championship in 1 96 6 , became an international master in 1 9 8 0 , played in four U.S. Championships and two Student Olympiads, and have been Maryland champion eight times, including as recently as 2 0 1 2 . My son Raymond is an inter­ national master. This is my third chess book. I have been involved with computer chess (off and on) since 1 9 6 7 , when I worked on 'MacHack' , the first computer to compete in human tournaments. More recently I worked on Rybka and now Komodo. I would like to thank IM Eric Kislik for his help in analyzing many of the lines in this book, the late Don Dailey, my Komodo partner, for this engine without which the book just wouldn't be nearly as good, and of course all the players who played the games featured herein. The book includes relevant grandmaster games right up to February 2 0 1 4. L arry K aufman P otomac MD, USA F ebruary

10

2 0 14

The Role of Komodo Because of my great faith in the Komada chess engine, I have decided to introduce a new feature in this book. Traditionally, most variations end with a symbolic as­ sessment like ;!;. While retaining these symbols, in key positions in the analytical section of the book I also give Komodo's evaluation (in pawns) after a deep think (usually 1 S minutes or more) . That way I am free to express my own opinion ei­ ther in symbols or words while letting the reader see the unbiased opinion of a very deep computer analysis by what I believe to be the best analytical engine avail­ able. That there is some objective basis for this opinion was demonstrated just be­ fore my deadline on this book when Komada won the 'TCEC' internet tournament, which is widely considered to be the real (if unofficial) world software champion­ ship, ahead of all major rivals including Houdini and Stockfish. Most regrettably, my partner in the Komada project died just 9 days before Komodo's TCEC tri­ umph. It's hard to say what Komodo's rating would be in human terms, but there is no doubt that it starts with a ' 3 ' and that it would be hundreds of points above even World Champion Magnus Carlsen. Normally computer evaluations are translated to symbols with the t evaluation covering the range 0 . 2 S to 0 . 7 S pawns. I think this is a bit too high, my guideline for that symbol is about 0 . 2 0 to 0 . 6 0 . But the symbolic evaluation shown often dis­ agrees with the one implied by the last sentence, because it reflects my opinion. I try to consider how easy it is to play the position as well as its objective merit. So for example even with a score like + 0 . 8 0 or more I may give just;!; if the position is tricky and double-edged. On the other side I may give a + 0 . 0 7 position a t if it's not excessively drawish and if White is clearly (if only slightly) for choice. The computer's evaluation is probably the better guide for a correspondence player, but I would like to think that my own opinion has some added value for practical over-the-board players at least. Occasionally I totally disagree with Komada, but these instances are quite rare. One more note about Komodo's evaluations. After deeply analyzing all serious opening lines in chess with Komada, my conclusion is that White's initial advan­ tage is about 0 . 1 S pawns. This is not enough to win, but should be enough to re­ tain some advantage into the endgame, so Black will at least have to be careful to se­ cure the draw. Consequently positions evaluated as well above +0. 1 S are deemed (by Komada at least) to be desirable for White, and those well below + 0 . 1 S (which I call 'par') are acceptable for Black. With 3 .f3 White nearly always achieves 'par ' , the only exceptions being rare lines that just haven't been analyzed very deeply. All the major third move replies to 3 .f3 usually lead to this par score with best play, but in many cases Black must tread a very narrow path to reach this score.

�

I also use the known feature of the 'tip' in this book, indicated with the sign:

11

Chapter 1

History of the f3 Anti-Grunfeld This section is a chronological review of games played in the last century that intro­ duced new opening ideas in this variation. By playing through these games, you will learn how to meet all sorts of unusual moves Black may try, as well as some un­ orthodox white tries, although often I have omitted white ideas that have been found wanting. You will also acquire the foundation to better understand and remember the modern ideas found in the theoretical section, as well as a better feel for how to play typical positions. All the analysis here has been checked by Komo­ do, though not always to the depth used in the theoretical section. Once we get past the point where a game has any theoretical value, I don't insist on really deep checking, since any mistakes won't affect your opening choices. The first two games featuring l .d4 tll f6 2 .c4 g6 3 .f3 were played in 1 92 9 by two of the best players of the time on the white side, namely Salo Flohr and Aaron Nimzowitsch. I start off this chapter with the Nimzowitsch victory over Saviely Tartakower, also a top player back then. As happens more often than not, Black sim ply allowed transposition to the Samisch King's Indian, but the game does illustrate a theme I emphasize in the book, namely that with a pawn on f3 it is usually good to play tll h3 if the knight cannot be profitably taken. See Game 1 . 1 . Studying these old games might be a waste of your time if you are a grandmaster, but for amateur players there is often more to learn from the old games than from the latest battle between titans, because the top players back then often made strategic mistakes similar to what many of today's amateurs will do. Instead of a wild battle with no real theme you are more likely to see how the stronger player exploits the mistakes of the weaker one. One point really struck me when annotating these games of the 1 93 0s. It seemed to me that the superstar players played at a much lower level than today's stars, perhaps at the level of an ordinary grandmaster today, while most of the play­ ers of that time who are not famous today were likely not even of what we would now call master strength. This is partly due to unfamiliarity with what are now considered standard plans and ideas, but also to missing more tactics. Probably this is simply a reflection of the fact that today's stars play thousands and thousands of games on the internet to hone their tactical skills, while the top players back then rarely had much chance to play other players of their own level except in the occa­ sional tournament. Perhaps this opinion will be somewhat controversial, but it agrees with my own experience as someone who played in many tournaments from the early 1 9 60s to today. Today's players are just much stronger, for many reasons. 13

S a b o t a g e t h e Gr ii n fe l d

Anyway, the next three games in my survey are all between World Champion Alexander Alekhine as White and his challenger Efim Bogoljubow. The first two are from their title match in 1 9 2 9 , the third is from a 1 9 3 1 tournament which Alekhine won by a huge margin. The first (Game 1 . 2) transposed to an unusual Samisch line where Black chose . . . c7 -c6 and . . . d7 -dS rather than . . . d7 -d6. This is a known idea in the King's Indian, a favorite of the late grandmaster Alex Wojtkiewicz, but it usually aims to meet eS by . . . tl'ie4, which is obviously impos­ sible in the Samisch. The other two games featured the neo-Grunfeld move 3 . . . dS . For the most part I skipped over games that transposed to the Samisch in this section, with a couple exceptions. In the first of these two games (Game 1 . 3 ) we see the first use o f the defense i n which Black develops his queen's knight before castling, which is not recommended though it was tried in one recent elite game. Alekhine won a very nice game, although he missed multiple opportuni­ ties to win the game along the way. In the last of these games against Boga (Game 1 .4) , Alekhine played 9.dS ? in the position where everyone castles queenside nowadays, and followed this error by four more in the next few moves ( ! ) , arriv­ ing at a dead lost position which the engines rate at about the equivalent of a rook up for Boga. Despite this, Alekhine miraculously drew the game ! It's hard to imagine this set of events happening in an elite game today, although I'm sure someone will be able to point out a similar case in modern play. This game should be a lesson that this system is not without risk for White; a couple mis­ takes and White can be dead lost. In the next game we see the first use of the system with 8.f4, which is still topi­ cal as it was played in a very recent elite game (see theoretical section) . Black responded poorly and soon fell victim to a standard h-file attack. These early games are very good at showing how to punish routine or inaccurate play by Black. This game (Game 1 . 5) was rather one-sided. In Game 1 . 6 we have a 1 9 3 3 battle between two of the titans of American chess at the time, Fine and Dake, both of whom I met in the 1 9 7 Os. It featured the first use of 8 . . . eS (as in Anand-Gelfand from their recent title match) , but White played the natural but weak 1 O . a4? instead of Anand's 1 O .h4. Black missed the proper way to punish this move, giving White the edge, with the advantage tilting back and forth after multiple errors on both sides, with Fine finally winning brilliantly. Next we have the first use of the move 3 . . . e6, which Maxime Vachier-Lagrave has brought into the limelight this year. I was truly shocked to see that this move dated back to 1 934! Naturally both sides made dubious choices in the next few moves, with White emerging on top and winning a nice miniature. See Game 1 . 7 . The next battle, again to my amazement, featured Black switching to the Benko Gambit with 3 . . . cS 4.dS bS , long before Benko ever played his Gambit! ! However it's not so good here because in the real Benko Gambit Black would meet an early f3 by . . . e7-e6 rather than by . . . g 7-g6. In the game White declined the pawn and Black should have equalized, but he blundered twice on moves 2 0 and 2 1 . Black (Opocensky) was famous for his opening innovations, but perhaps his practical playing strength was not on a par with his originality. See Game 1 . 8 . 14

C h a p t e r 1: H i s t o ry of t h e f3 A n t i - G rii nfeld

Next we see the first use of 8 .�bS against the premature knight development of Game 1 . 3 . On moves 1 5 and 1 6 White missed the chance for advantage with the thematic d4-d5 ! , something which I doubt a modern master would miss. The way he played allowed Black easy equality and a quick draw. Black was one of the titans of the day, Grigory Levenfis h. See Game 1 . 9 . In our next game Black essays a very rare gambit with 4 . . .c 6 , which shouldn't give full compensation for the pawn but does offer some practical chances. White declined the gambit and still obtained some advantage. Later Black equalized but White again built up a winning advantage, only to throw the game away with missed wins on moves33 and3 4. See Game 1 . 1 0 . Next we come to another example of the line from Game 1 . 9 , with the new move 9 . . . e S . This could have led to a fascinating piece sacrifice (analyzed in theory section) but instead Black just conceded a modest plus to White, which White went on to exploit for a win. The game featured two heavyweights of 1 948 , Bondarevsky vs. Aronin. Curious that Lev Aronin was a top player then, while Levon Aronian is a top player now! See Game 1 . 1 1 . Now we come to two games with the Enevoldsen brothers behind the white pieces, introducing a long variation that remained popular into the current century. Black plays 9 . . . eS 1 O.dS tlJd4, getting a nice outpost, which White proceeds to un­ dermine. In Game 1 . 1 2 Jens missed several chances for a modest plus, then one chance for a repetition draw, and quickly went down to defeat. In Game 1 . 1 3 Harald had the more dangerous attack until a very understandable error on move 2 3 turned things around dramatically and Black went on to win, although missing a nice mating combination on move 2 6 . This subvariation should favor White, but it is very tricky and the price of a mistake by either player is high. Our next game takes place after my own playing career had begun, in 1 964, be­ tween two of the best American players of that time, Larry Evans and Donald Byrne. Black played the original but dubious 9 . . . aS ? ! , and after the unnecessary 1 O . . . e6? he was already losing. Later Black missed a chance to complicate, but then went down to a one-sided defeat. The game is a good illustration of the power of the h-file at­ tack if Black fails to find sufficient counterplay. See Game 1 . 1 4. Next comes our first example of the Benko Gambit (with f2-f3 and . . . g 7 -g6) Ac­ cepted. I think that White did not adopt the best plan in this variation and Black missed some chances to equalize, but eventually he blundered and lost. See Game 1 . 1 5 for both the actual game and the recommended line for White. Game 1 . 1 6 features a return to the 8.f4 attack (which seems to have been the main line in the last century) , this time with the now usual response 8 . . . t2Jc6 , fol­ lowed by 9 . . . tlJaS when the knight is attacked. The game went back and forth be­ tween equality and white advantage until Black made a fatal error on move 2 8 . The game and notes show the importance of exchanging off Black's dark-squared bishop when possible. In Game 1 . 1 7 Black tries to improve on the previous game by retreating his knight back home when it is attacked, but I believe this was actually a worse move. However White missed the only way to punish the move ( 1 O . a4!) and Black equal15

S a b o t a g e t h e G riinfeld

ized easily. The queens were traded into a virtually equal endgame, but somehow Black was outplayed and went down to defeat. Next we see the first example of meeting the then-popular 7 . . . CiJc6 by the 'safe' 8 .dS rather than the complex 8 .�bS . White wins a pawn in the game and goes on to victory after multiple errors on both sides, but the notes show that Black could have regained the pawn with only a moderately inferior game. This 7 . . . CiJc6 line is a bit dubious but not clearly refuted. See Game 1 . 1 8 . In Game 1 . 1 9 White tries to improve over Game 1 . 1 6 on move 1 1 , but I think his new move was worse than the old one, and Black equalized. However Black al­ most immediately went astray on move 1 3 and found himself in a losing position, which White went on to win fairly easily. In our next game Black simply plays to prevent d4-d5 by 7 . . . e6. This allowed White to obtain a dominant center and space advantage, although at the price of the bishop pair. But White missed the way to consolidate his plus and Black equal­ ized. Later White was fortunate that he could still draw the game. See Game l . 2 0 . I n Game l . 2 1 we have a return to the Benko Gambit Accepted, this time with White adopting a different plan than the one I recommended in Game 1 . 1 S . His choice also seems sufficient for at least a small plus, and although both players missed a nice equalizing trick on move 1 4, White went on to win a fairly nice and instructive game. Next we have a return to the 8 .f4 fS variation of Game 1 . 5 , but this time with a much better reply to the check on move 9 . It seems that Black could have equalized by keeping queens on the board on move 1 S . I think Black's queen is more active than White's in this line and that White would have trouble playing actively with the black queen present. As the game went he was able to keep improving his posi­ tion and won a nice and instructive endgame. See Game l. 2 2 . Now we come to the first game featuring the initial 1 4 moves of the recent Anand-Gelfand game. However White went astray with 1 S .g4?, after which Black missed first a chance for advantage and then an exchange sac for equality. Instead Black chose a line giving him two pawns and the bishop pair for a knight, which was not enough, and he lost. See Game 1 . 23 . Our final game in this section features the first use of 9 . . .fS , a still-topical varia­ tion. White played the sharp 1 O .h4, which I analyze to a draw in the theoretical section. In this game though, Black chose to suffer for a pawn, a poor decision in my opinion, and he lost rather badly. See Game 1 . 24. This brings us to the present century. We have seen examples of all major variations (excluding King's Indian transpositions) except for the very recent 9 . . . 1/Wd6 and for 3 . . CiJc6 , for which see the theoretical section.

16

C h a p t e r I: H i s t ory of t h e f3 A n t i - G rii nfeld

Game 1 . 1 (ES 7) Nimzowitsch, Aaron Tartakower, Saviely Karlsbad 1 929 (2 1 ) 1 .d4 tll f6 2.c4 g6 3.f3 This was one of the first two games played with this move according to the database. 3... �g7 4.e4 d6 5.tll c 3 0-0 6.�e3 tll b d7

Aaron N imzowitsch

7.tll h 3! With a pawn on f3 , f2 is usually the best home for the king's knight if it can reach that square. I think 6 . . . tll b d7 is strategically refuted by this reply. 7... e5 S.d5 as 9.tllf 2 b6 1 0.'i¥d2;!; tll c5 1 1 .�gS �d7 1 2.g4 �cs 1 3.h4

play really that low then? Apparently Tartakower judged that winning the bishop pair was worth having his king­ side pawn structure ruined and his king attacked, but if so that is very bad judg­ ment. Black should try l 3 . . . c6, though White's attack should prevail. 1 4.hS gxh5 1 5 .�xf6 �xf6 1 6.l::rx hS �g7 1 7.tll h 1 I 7 .O-O-O �es 1 s .�d3 f6 1 9 .l:!h2 @gs 20 . .ld.dh l �g6 2 1 .�c2+ - ; White has a strategically won game. 1 7...f6 1 S.�h2 h6 1 9.tll g 3 @h7 20.�e2 I:igS 21 .@f2 2 1 . 0-0-0 looks even better. 21 ...l::l h S 22.l::r h 4 ?! �es 23 . .l:rg1 �fS 24.@g2?! tll b 7 25.tll h S �96 26.f4 tll d S?

1 3...@hS? This is ridiculous, moving toward the attack. Tartakower was one of the great players of the era; was the standard of

17

S a b o ta g e t h e G r ii n feld

27.�f3? 2 7 .gs fxg5 28 .fxg S +- . 27...Ci:Jf7? 2a.Ci:Je2 2 8 . g S ! fxg5 2 9.fS +- . 2 a...�e7 29.\t!h1 @ga 30.Ci:Jeg 3 @fa 31 .Ci:Jfs .:g:ga 32.�d2 :t:Ica 33..l::!. h 2 @ea 34.b3 @da 35.a3 :t:Iaa 36.�c1

White is already clearly better here, but after the next move the game is de­ cided. 36 ...�fa?? 37.Ci:Jh4 �h7 3a.Ci:Jxf6+�ha 39.Ci:Jxga �xga 40.gS exf4 41 .gxh6 �h7 42.�xf4 �xh6 43.�f6+ @ca 44.Ci:JfS �xf5 45.exfS @b7 46.�g6 l::!. h a 47.�xh 7 .l:!.xh7 4a. .l:i.g6 @ca 49.f6 .l::!. h a so.�g4+ @da 51 .�e6 @ea 52.�xf7 + @xf7 53.�hxh6 1-0 If someone told me this was a recent game, I would guess the players to be rated around 2 00 0 . But they were among the top five at the time!

7.e5 7.cxd5 ! cxd5 8.e5 is a more accurate move order to reach the game without allowing the gambit mentioned on move 8. 7...Ci:Jfd7 7 . . . Ci:Je8 8.cxd5 cxd5 9 .h4 is slightly better for White. a.cxd5 cxd5 After 8 . . . Ci:Jb6 ! 9.dxc6 Ci:Jxc6;\; Black has some but not full compensation for the pawn. 9.Ci:Jxd5 �as+ 1 O.Ci:Jc3 Ci:Jc6

1 1 .�c4?! l l.a3 ! prepares b4 and prevents . . . Ci:Jb6 ,

(E70) Game 1.2 Alekhine, Alexander Bogolj ubow, Efim Germany/Netherlands Wch m 1 929 ( 1 5) 1 .d4 Ci:Jf6 2.c4 g6 3.f3 �g7 4.e4 0-0 5.Ci:Jc3 c6 6.�e3 d5 18

leaving Black little compensation for the pawn. 1 1 ... Ci:Jb6 1 2.�b3 �e6 1 3.Ci:Je2?! 1 3 .�xe6! fxe6 1 4.Ci:Je2 still gives White an edge. 1 3...Ci:Jc4 1 4.�xc4 �xc4 1 5.0-0 .l::!.f da 1 6.f4

Chapter

1:

H i s t o ry o f t h e f3 An t i - G rii n f e l d

29 ...tt:JxaS= 30.'ii' b 1 ? 3 0 .'fia2=. 30...tt:Jc4 31 Jk1 .l::!.a 8 32.dS .l:!.xd5 33.ld.xdS 'fixd5 34.tt:Jc3 'iY'd2 35.tt:JxbS

16 ....ld.acS 1 6 . . . f6 , and with the bishop pair, a bad enemy bishop, and pressure on the backward d-pawn, Black has full com­ pensation for his pawn. 1 7.a3 1 7 .b3 .te6 1 8 .a3 ;!; . 1 7... 'iY'b6 1 8.'fid2 tt:Ja5 1 9 . .ld.ae1 'iY'c6 20 ..if2 b6 21 .'iY'e3 f5?! 2 1 . . . .ie6 2 2 . .ld.d l tlJc4 2 3 .'ii'c l 'ii'b 7=. 22 ..ih4 .ld.d7 23.b4 23 . .i::!d 1 .if8 24.b4;!;. 23 ....txe2 23 . . . tt:Jb3 24 . .l::t d l e6=. 24.tt:Jxe2 tt:Jc4 25.'ii' b 3 b5 26 ..l:!.d1 e6 27..tf2 2 7 .'ii'f3 t. 27....ifS 28. .l::!.d 3 as

35 ...tt:Jxa3 A) 3 5 ...'ii'xf4? 36.g3 °iVg4 3 7.'iYb3+-; B) 35 . . . .ld.c8 3 6. ld.c3 Wixf4 3 7 .'ii'c l 'ii'xc l + 3 8..l:!.xc 1 ld.b8 would slightly favor Black. 36.tt:Jxa3 .l::!.xa3 37.g3 .l::!.a 2 38. .ld.f1 a'.b2 39.'ii'c 1 'fib4 40.'fic6 'ii' b 3 41 .'ii'a S 'ii'c4 42.'ii' d S 'ii'c6 43 . .l:!.d 1 'ii'f3 44. .l::!.f 1 'ii'c6 1/2-1/2 (D7 0 ) Game 1 .3 Alekhine, Alexander Bogolj ubow, Efim Germany/Netherlands Wch m 1 929 ( 1 7) 1 .d4 tt:Jf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tt:Jb6 6 ..te3 .ig7 7.tt:Jc3 tt:Jc6?!

29.bxaS With 2 9 . .ld.fd l ! axb4 3 0 .dS l:i'.xdS 3 1 ..ld.xdS exdS 3 2 .axb4 .l::!. a 8 3 3 .tlJd4;!; White returns his extra pawn, retaining the better pawn structure. 19

S a b o t a g e t h e G r iinfeld

Castling is far more common here, and probably better, but the knight move was played in at least one recent elite game so it's not 1 0 0% clear. 8.d5 8 .�bS ! ? is the greedier move, but it en­ tails more risk due to the following piece sacrifice: 8 . . . 0-0 9 .tll g e2 and now: A) 9 . . . tll a s 1 0 .b3 a6 1 1 .�d3 tllc 6 1 2 .�c2 tll b 4 1 3 . 0 - 0 ( 1 3 .�b l i) 1 3 . . . c6 1 4.l:i'.c l ( 1 4.�b l i) 1 4 . . . tll xc2 1 5 . .l:txc2 aS (Borwell-Miillner, corr semifinal Wch- 1 9 1 99 2 /94) 1 6 .a4i; B) 9 . . . eS 1 0 .dS tll a S ! ? ( 1 0 . . . tll e 7 Bondarevsky-Aronin. Moscow ch-URS 1 948 ; 1 l .tlla 4i) 1 1 .�xb6 axb6 l 2 .b4 c6 1 3 .�d3 bS 1 4.bxaS b4 and Black gets decent if not full compensation for the sacrificed knight, see the theory sec­ tion for details. 8...tll e5

9.�d4 9.f4! tllg 4 1 0 .�bs + �d7 1 I .�d4 �xd4 l 2 .�xd4 0 - 0 (Speelman­ Lputian, Hastings 1 9 8 6 / 8 7 ) 1 3 .�xd7 �xd7 l 4.h3 tll f6 1 5 .tll g e2 c6 1 6.a4i. 9...f6? 9 . . . 0-0 1 0.f4 and now: A) 1 O . . . tll e d7 1 1 .�xg7 ..t>xg 7 1 2 .�d4+. Alekhine claimed a big edge here based on the attack with h2-h4, but. . . 1 2 . . . tllf6 1 3 .tllf3 e6 1 4.dxe6 �xe6 1 5 .�e2i. 20

Analysis diagram

White has no attack, just a more ad­ vanced majority and slightly better placed knights; B) 1 O . . . �g4?! 1 1 .�e2 �xe2 1 2 . tll g xe2 tll g 4 1 3 . �xg 7 ..t>xg 7 tll f6 1 5 . 0- 0 - 0 cs 1 4.�d4+ (Laurentius-Alexander, Warsaw Olym­ piad 1 9 3 5 ) 1 6.�d3 ± ; C ) 1 0 . . . tll g 4 1 1 .�xg7 'it>xg 7 1 2 .�d4+ tll f6 transposes. 1 0.f4 1 O .a4! aS l 1 .�b3 0-0 1 2 .tll b S (by threatening tll x c7 White provokes . . . 'it'h8 , which makes h2-h4 even more effective) 1 2 . . . 'it>h8 1 3 .h4±. 1 o...tll f 7 1 1 .a4 Now it's less effective. 1 1 ... e5! 1 2 .dxeG �xe6 1 3.a5 tll d 7 14.aG b6 1 5.�b5 �e7 1 6.tll g e2 c5 1 6 . . . 0-0 1 7 . O-Oi.

1 7.�f2

C h ap t er 1: H i s t o ry of t h e f3 An t i - G rii nfeld

1 7 .fS ! gxfS 1 8 .exfS �xfS 1 9 .tll d S �d6 20.�xf6 �xf6 2 1 .0-0 �es 2 2 . l:rxfS 0-0 2 3 .�xd7 l:!.ad8 24.�c6 Wh8 2 5 .�b7 +- . Alekhine missed this very difficult win of a piece for a pawn. 1 7... 0-0-0

In case of 1 7 . . . 0-0 1 8 .tll d S �xdS ( 1 8 . . . �d6 l 9 . �a4+-) 1 9 .�xdS l:!.ad8 2 0 .0-0i White has a solid bishop pair advantage plus more space. 1 8.�a4 Alekhine missed the difficult winning line 1 8 .tll d S ! �d6 l 9 .�a4 �xdS 20.l::r d l �c7 2 1 .l::r x dS tll d 6 2 2 .tll c 3 tll f8 2 3 . .l:!.d3 Wb8 24. tll d S �f7 2 5 .�c6+- . 1 8...f5 1 9.e5 g5! 20.�c4 After 2 0 .g3 tll fxeS ! 2 1 .fxeS tll x eS Black gets enough for the piece. 20... tt:Jdxe5 20 . . . tllb 8 2 1 .0-0+- . 21 .�xe6+ �xe6 22.fxe5 tt:Jxe5 23.0-0

23 ...�c4? Better was 2 3 . . . tll d 3 24.'iYbs l:rd7 and with two pawns, center control, and a super knight outpost, Black has almost enough for the sacrificed knight. 24.b4!! �xb4 Or 24 . . . cxb4 2 5 .tll b S ! �xe2 (25 . . . Wb8 2 6 .�xb 6 +- ) 2 6 .tll x a 7 + Wb8 2 7 . tll c 6 + +- Wc7 28 . .l:Iac l tll c 4 2 9 .tll xd8 l:!.xd8 3 0.�xb6+ and White wins. 25.�c2? 2 5 .�xb4! cxb4 2 6 .tll b S @b8 2 7 .tll ed4 f4 2 8 . l:rad l �f6 2 9 .l:rfe l +- . Black's three scattered pawns are no match for the piece here, as White has numerous threats. 25 ...tll d 3 26JHb1 ? 2 6 .l::rfd l i. 26 ... �c4 27.l::ra4 �e6 2 7 . . . �f? 2 8 .tll b S±. 28.tll b 5

28...WbS? After 2 8 . . . tll xf2 2 9. Wxf2 Wb8 3 0 . .l:rd 1 �c8 3 1 .tll e c3 l:!.hf8 3 2 .Wg l White has a slight edge. He is up half a pawn in material, but otherwise the position is balanced. 29.tll e d4 �e4 2 9 . . . �xd4 3 0 .�xd4 l:!.xd4 3 1 ..l:!.xd4 �e3 + 3 2 . Wfl +- . 30.tll c3 �ea 31 .�xd3 cxd4 21

S a b o t a g e t h e Grii nf e l d

Black resigned, a s after h e takes the bishop 3 5 . "iW c6 is mate in 7 , as is also 3 5 . .l:f.xb6+.

32 ..2.xd4? 3 2 .°i:Wf3 ! °i:Wf7 3 3 . .l::rxb6+ axb6 34.a7+ "i:Wxa7 35 . .l::!.xa7 @xa7 3 6 .tllb S + @a6 3 7 . tll c 7 + @a7 3 8 .�a3 + @b7 3 9 .tlle 6+-. 32 ... °i:We6? 33.°i:Wf3 °i:Wf7 34.�xb6!

Game 1 .4 (D7 0 ) Alekhine, Alexander Bogolj ubow, Efim Bled 1 93 1 (22) 1 .d4 tllf6 2.c4 g6 3.f3 d5 4.cxd5 tll x d5 5.e4 tll b 6 6.tll c3 �g7 7.�e3 0-0 8.'i!Wd2 tll c6

9.d5? Only castling offers a chance for advan­ tage, while the old 9 . .l:i.d 1 ?! may suffice for equality. 9 ...tll e 5 1 0.�g5? The lesser evil was I 0 . .l:i.d I . 1 o ... c6 1 1 ..l:rd 1 ? cxd5 1 2.exd5 �f5 1 3.g4 �d7 14.d6? f6 1 5 ..2.h6?

Alexander Alekhine

1 5 ...�c6 22

C h a p t e r I: Hi s t o ry of t h e f3 An t i - G ri infeld

1 5 ... .ixh6 1 6.�xh6 exd6 1 7 .�e2 �c6 l 8 .�f4 �e7-+ ; with a pawn less and a crushing deficit in development and king safety, White could reasonably re­ sign. 1 6 ..ixg7 @xg7 1 7.�f4 exd6 1 8. .ie2 �e7 This is the same position as in the previ­ ous note after l 5 . . . .ixh6, except for the position of the black king, which is not important. 1 9.@f2 White's position is still resignable, and yet somehow he doesn't lose, although Black was among the world's best play­ ers then. 1 9... g5 20.�d4 f5 21 .h4? fxg4 22.hxg5 �xg5

27...tt:Jd5? 27 . . . tt::l c 8+. 28.@g3 tt:Jxf4 29.a'.xf4 lixf4 30.@xf4 .l:f.xd6 31 .@xe5 l:i.d5+ 32.�eG h5 33.�xg4 hxg4 34.�xg4+ @h5 35.l:i.g3 .ld.d2 36.tt::l f3 l::rx b2 37.tt::l d 4= 112-1/2 White's vastly superior king position makes the draw easy despite the pawn deficit. What an amazing save!

Game 1 .S (D7 0) Rodi, Ludwig Helling, Karl Bad Pyrmont ch-GER 1 93 3 (5) 1 .d4 tt::lf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tt::l b 6 6.tt::l c3 .ig7 7..ie3 0-0 8.f4 The computers rate this as about a rook up! 23.tt::l e4 �f4? 2 3 . . . .ixe4 24.�xe4 hS 2 5 . l:i.xd6 l:i.ac8-+. 24.tt:JxdG �xd4+? 24 . . . �gS 2 5 .tt::lh 3 'i:!:Yf6 26 . .l::!.h g l hS 2 7 . f4 .!d.ad8-+ . 25.�xd4 @g6 26 ..lah4 liad8? 2 6 . . . .ixf3 2 7. tt::l xf3 hS is still clearly better for Black. 27.f4 Black still has the upper hand here, but with his next move he definitely lets his advantage slip.

8 .. .f5?1 As shown in the theoretical section, the unknown move 8 . . . cS! is quite promis­ ing, while 8 . . . tt::l c 6 is normal and suffi­ cient for equality with precise follow-up. 23

S a b o t a g e t h e G r iinfeld

9.'YWb3+ @h8?! The immediate 9 ... e6 is better as on h8 the king is a target for h4-h 5 . 1 0.e5 e 6

1 1 .tllf3 l l .h4 directly is probably more accu­ rate. 1 1 ...tll c6 1 2.h4t a5 1 2 . . . hS looks like the lesser evil. 1 3.h5 gxh5 14Jhh5 l 4.a3 looks even better, to prevent . . . a5-a4 and . . . a4-a3 . 1 4 ... 'YWeS 1 5.�h3 a4 1 6.'YWd1 a3 1 7.b3 tll b4

1 8.�d3?! 1 8. �f2. ! ±. Why make the bishop be a hostage to the knight?. 1 8... �d7?! 1 8 . . . h6t. 1 9.g4± �c6 20.gxf5 tll x d3+ 21 .'YWxd3 �xf5 22.@f2 �xf3 23.Wxf3 24

23 ...J::l d S 23 . . . 'YWc6+ 24.dS tll x dS 2 5 .tll x dS 'YWxdS + 2 6 .'YWxdS exdS 27 . .l:!.ah l h6 2 8 . �g4 l:laf8 2 9 . e 6 +-. 24.!iah 1 +- h5 25.tll e 2 'YWg6 26.�g1 'YWf7 27.tll g 3 l:lxe5 28.tll e4? 2 8 . @e2 �eds 2 9 .fS l::rxfS 3 0 .tll xfS exfS 3 1 . .&!. g s tll ds 3 2 .l::r xfS tll f6 3 3 .!ihxhS + @g8 34 . .l:!.hgS+- . 28.....thG 29.@e2

29 ...�xf4?? Mistakes like this give me the general impression that the standard of play was a lot lower then than now. The players who finished below 5 0% in major tournaments back then were probably below what we would call master level ( 2 20 0) today in the USA. Kasparov remarked in the 1 9 8 0s that even the grandmaster who finished last in a certain tournament (Ljubojevic) was stronger than Capablanca had been half

C h a p t er 1: Hi s t ory of t h e f3 A n t i - Grii nfeld

a century earlier. Although not a diplo­ matic thing to say, it was probably true. 29 . . . �fs 3 0 .tt:Jg3 ± . 30.�xf4 �f5 Maybe Black missed 3 0 . . . �xf4? 3 1 .dxeS l::ixd3 3 2 . �xhS + with mate next when he took on f4. 31 .�e3? 3 1 .ti:Jg 3 +- . 3 1 ...l::rf8 32.tt:JgS �g6 33.l:igh1 @g8 34.l::rg 1 @ha 35.tt:Jf3 �f6 36.lagS 1::1.x gS 37.�xgS �f7 38.�f6+ 1-0

1 5.tt:Jge2 1 s . .!d.a4;£.

Game 1 .6 (D 7 0 ) Fine, Reuben Dake, Arthur William Detroit 1 9 3 3 ( I 0) 1 .d4 tt:Jf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tt:Jb6 6.�e3 �g7 7.tt:Jc3 o-o 8.�d2 es 9.d5 c6

Reuben Fine

1 5 ... tt:Ja6= 1 6.0-0 tt:Jc5 1 7.�xcS?! �xc5+ 1 8.@h1 .ld.d8 1 9.�gS 1 0.a4? Only 1 O .h4 promises advantage, as shown in the theoretical section, while 1 0 . .lld 1 is playable. 1 0...cxd5 1 1 .exd5 �h4+? With 1 l . . . tt:Ja6 ! + Black prepares . . . tt:Jb4 while showing that 1 2 .aS was not a threat due to 1 2 . . . tt:Jc4 1 3 .�xc4 �h4+ . 1 2.�f2? 1 2 .�f2 ! ± . 1 2 ...�b4 1 3.aS tt:Jc4 1 4.�xc4 vt¥xc4

1 9...f6? 25

S a b o t a g e t h e G riinfeld

1 9 ... .id7+ with the bishop pair versus two knights and a fluid position. 20.'t!Vh4 g5 21 .'t!Va4 .if5? The pawn should go there as now the g7 -bishop is pretty bad. 22JUd 1 �ac8 23.t2Jg3 .ig6 24.t2Jge4;!; 't!Vc4 25.'t!Va3? 2 5 .'t!Vxc4 l:lxc4 2 6 .a6;l;. 25 ....ifS 26.d6?

26... @g7? With 26 .. .fS 2 7 .tLlxgS l:rc6 2 8 .l:lac l l:lcxd6+ Black regains the pawn with a dominating position. 27.�d5? Jl.f7? 2 7 . . . .ixe4 2 8 .fxe4 l:!.c6 wins a pawn. 28..l:!.d2 .l::!.c 6 28 .. .fS ! 2 9 .tLlxgS .ig8+. 29..l::l.a d1 ? 2 9 .g4! to prevent 2 9 . . .fS . 29 ....l::!.a 6? 29 .. .fS ! 3 0 .tLlxgS .l:!.cxd6+. 30.'t!Va1 b6 31 .'t!Vc1 ;t .a:xa5?

32.t2Jxf6!± 26

Although this was not a well played game up to here, the finish is very nice! 32 ...@xf6? 33.t2Je4+ @g7 34. .l:lc2 't!Va4 35.'t!Vxg5+ .ig6 36 ..l:!.c7+ @g8 3 6 . . . J:[d7 3 7 .b3+- . 37.'t!Vxg6+ 1-0 On a personal note, I met both of the players in this 8 0 year old game about 40 years later.

Game 1.7 (E60) Yudovich, Mikhail Freiman, Sergey Leningrad ch-URS 1 934 1 .d4 t2Jf6 2.c4 g6 3.f3 e6 Amazingly, this move, which Vachier­ Lagrave has brought back from obscu­ rity, dates back to 1 934! 4.t2Jc3 4.e4 d5 5 .cxd5 exd5 6 .tLlc3 is better, see the theoretical section. 4 ... d5

-�.t��[t Ii i .,., . ., ,. . .,. . ·8� • • • � .8. 8� ·8·8� .:: �'iV�i. .:: 5 ..ig5 5 .cxd5 t2Jxd5 ! (5 . . . exd5 ? ! 6 . e4 trans­ poses to the recommended line for White) 6.e4 t2Jxc3 7 .bxc3 c5 is an Ex­ change Grunfeld where . . . e7 -e6 looks more helpful to Black (restraining d4-d5) than f2-f3 does to White. 5 ... �e7 6.'t!Vd2 In the event of 6 .e4 0-0 7 .�h6 !:!.e8 8 .e5 tLlh5 9.f4 c5oo it's hard to say who has the better chances.

C h a p t e r 1: H i s t o ry of t h e f3 An t i- G riin feld

6 ...t2Jc6 It's usually wrong to block the c-pawn like this when the cl-pawn is advanced two squares, because neighboring pawns that are not part of the castle should work together. 7.e3 t2Jg8?! 8.�f4 �h4+ 9.�g3 �xg3+ 1 0.hxg3;t

White has space while Black has a bad bishop. 1 0 ... 'fHdG 1 1 .0-0-0 t2Jge7 1 2.g4 b6 1 3 .l2Jh3?! 1 3 .e4 a6 1 4.l2Jh3 +- . 1 3 ...fG? 14.l2Jf4?! 1 4.e4 dxc4 1 5 .�xc4 l2Ja5 1 6.�d3 tbec6 1 7 . tbf4+-; the threat of dS is too strong. 14 ...�b7? 1 5.l2Jb5+�d7 1 6.t2Jxe6 .l::!.c8 1 7.tbf4 dxc4 1 8.�xc4 t2Ja5 1 9.�e2 1 9 .�d3 ! +- . 1 9 ... gS 20.l2Jh5 0-0 21 .e4 c5 22.l2Jc3 °fHe6 23.d5 'fHd6 24.f4 h6 25.e5

Black resigned, because if 2 5 . . . fxeS 2 6.fxgS gives White a crushing attack. This was a rather one-sided game.

Game 1.8 (E60) Engels, Ludwig Opocensky, Karel Bad Nauheim 1 935 (5) 1 .d4 t2Jf6 2.c4 g6 3.f3 c5 4.d5 b5?! Amazing, a version of the Benko Gam­ bit before Benko ever played his gam­ bit! 5.e4

The problem for Black is that f2-f3 (on move 4 or 5 ) against the Benko Gambit is best met by . . . e7-e6, not by . . . g7-g6, which is already on the board. 5 ...d6 6.t2Jc3?! 6.cxbS ! a6 7 .l2Ja3 �g7 8 .l2Je2 0-0 9 .tbc3 l2Jbd7 1 0 .�e2;t.

Analysis diagram

27

S a b o t ag e t h e G rii n f e l d

White has a good version of the Benko Gambit here. 6 ... bxc4 6 . . . b4!=; White should not allow his knight to be kicked thus in the Benko. 7.�xc4 �g7 8.tLlge2 t2lbd7 9.0-0 t2lb6 After 9 . . . 0-0= Black is fine. White really should have grabbed the pawn when he could. Black should not invite the check by moving his knight before castling. 1 o.�b5+ it.d 7 1 1 .a4 1 1 .�d3 ;!;; the black bishop on d7 is misplaced, preventing the knights from using this vital square. 1 1 ... 0-0 1 2.f4!? Perhaps too ambitious. 1 2 ...�xb5 1 3.axb5 e6 14.dxeG fxe6 1 5 .�d3 �e7= 1 6.�h3 d5 1 7.e5 t2ifd7 1 8.b3 g5 1 9.it.d2 gxf4 1 9 . . . c4 2 0 .tLld4 tLlcS 2 1 .tLlc6 �e8 2 2 .bxc4 t2lxc4 2 3 .it.e3 ± . 20.tLlxf4

22.�g5+- �b4 23.tLlxfS tLixfS? 24.�fG 2s.nab1 ncs b2 26.it.xg7 �d4+ 27.Wh1 Wxg7 28.eG! tLlgG? 29 ..l::If 7+ 1-0 Black was a rather famous player I open­ ing innovator, so I'm surprised to see so many tactical mistakes by him.

Game 1 .9 (D7 0) Goglidze, Viktor Levenfish, Grigory Moscow 1 93 5 (7) 1 .d4 tLlfG 2.c4 g6 3.f3 d5 4.cxd5 t2ixd5 5.e4 tLlbG 6.�e3 it.g7 7.t2lc3 tLicG 8.�b5 0-0 9.tLlge2

9... t2ia5 9 . . . es 1 O.dS tLlaS is an interesting piece sacrifice analyzed in the theoretical section. 1 O.b3 a6 1 1 .�d3 tLicG 1 2.�c2 t2lb4 1 3Jk1 t2ixc2+ 1 4.�xc2 a5

20...c4?? 20 . . . it.xeS 2 1 .tLlxe6 1:1.xfl + 2 2 . l:rxfl lle8 2 3 .�g4+ h8 24.tLlgS tLlf6=. 21 .tLlxeG± cxb3?? 2 1 . . .�xfl + 2 2 .l::!: xfl it.xeS 2 3 .bxc4 kte8 24.cxdS tLlxdS 2 5 .tLlxdS �xe6 2 6 .�d3 ± and White is a healthy pawn up. 28

C h a p t e r 1: H i s t o ry of t h e f3 An t i - G ri i n fe l d

1 5 .0-0 1 5 .dS ! a4 1 6 . 0-0 transposes to the note to move 1 6 while avoiding the note to Black's 1 S th move. 1 5 ...a4 1 5 . . . c6 1 6 . .l:!.fd l a4 1 7 .dS cxdS l 8.bxa4;l;. 1 6.tt:Jxa4? 1 6 .dS ! axb3 1 7 .axb3 �d7 (after 1 7 . . . c6 l 8.dxc6 bxc6 1 9 . .l:!.fd l �d7 2 0 .�hl tbc8 2 1 .tba4 tbd6 2 2 .tbd4 the pressure on c6 and on the d-file, and the option of tbcS give White a clear plus despite Black's bishop pair) 1 8 . .l:!.fd 1 tbc8 l 9 .1/Wd2 tbd6 2 0 .�d4 f6 2 l .tlJf4;t_ White's huge edge in space and piece placement easily outweigh Black's bish­ ops. 1 6 ...tt:Jxa4 1 7.bxa4 c6 1 8.�fd1 'ifas 1 9.tt:Jc3 �e6

20.dS?! Better was 2 0 .l:rb l b6 2 1 .tbe2 �d7 2 2 .'i¥d2 1/Wxa4 23 . .l:!xb6 '1Wxa2 24.I:l.b7 when White's rook on the 7 th and su­ perior pawn structure slightly outweigh the bishop pair. 20 ... cxdS 21 .tt:JxdS �xd5 22.f!xd5 1/Wxa4 23.1/Wxa4 l:l'.xa4 Draw agreed in this totally equal end­ game. Black in this game was one of the world's best players.

Game 1.10 (D7 0) Engels, Ludwig Richter, Kurt Aachen ch-GER 1 935 (4) 1 .d4 tt:Jf6 2.c4 g6 3.f3 d5 4.cxd5 c6

This is an interesting gambit. Probably after S .dxc6 tbxc6 6.e3 eS 7 .�bS Black doesn't get full compensation for the pawn. 5.e4 cxd5 6.e5 tt:Jfd7 7.tt:Jc3 7 .�d3 ! tbc6 8.tbe2 e6 9 .�e3 ;!; looks like a rather good version of the French for White. 7...tt:Jb6 After 7 . . . tbc6 ! 8 .tlJxdS tbdb8 9 .�c4 �e6 1 O .tbe3 �xc4 l l . tbxc4 b S 1 2 .tba3 '1Wxd4 Black i s slightly more comfortable. 8.�d3 tt:Jc6 9.tt:Jge2 �g7 1 0.0-0 0-0 1 1 .f4 �g4 1 2.h3

29

S a b o t a g e t h e G rii n f e l d

1 2 ...�f5?! Normal would be to take the knight, since this is a rather closed position. Apparently Black provoked h2-h3 be­ fore playing . . . . �f5 to avoid the rook maneuver :I.-f3 -h3 . 1 3 .�xf5 gxf5 1 4.b3 @h8 1 5 .�e3 ld:g8 1 6.tll g 3?! tll - c l -d3 looks better. 1 6 ...�xe5? 1 6 . . . e6 1 7 .�f2 �f8=. 1 7.tll xf5;t �g7 1 8.WVh5 l 8 .tllx g 7 l:txg7 1 9.fS t . 1 8 ...WVf8

1 9.tll e 2? l 9 . g4t. 1 9...tll d 7= 20.g4 tll f6 21 .WVh4 tll e4 22.Wh2 l::rc8 23 ..l::!.a c1 e6 24.tll x g7 WVxg7 25.f5 e5? 26.�h6 WVf6 27.WVxf6+ tt:Jxf6 28.�e3? After 2 8 .dxeS tllx e5 2 9.@g2;\; White's better pawn structure and very strong bishop give him the edge. 28..Jke8= 29.�f2 h5 30.�h4 tll e4 31 .tll g 3 hxg4? 32.tll xe4 dxe4 33.d5? 3 3 .�f6+ @h7 34.hxg4 .l:!.xg4 35 . .i::l c 3 @g8 3 6 .dS tll e 7 3 7 . .i::l h 3 @f8 3 8 . .l::!.d l +- . 33 ...tll d 4 34.d6?? 34.�f6+ @h7 3 5 .hxg4 .!::l.xg4 36 . .l::!. c 3 tll xf5 3 7 .l:[xfS @g6 38 . .l::!.xeS +-. 30

34 ... tllf3+ 35 ..l:!.xf3 gxf3 36.d7 .l:!.a8 37. .l::!.c4? @g7 38. .l::!.xe4 f6 White resigned, as after the coming exchange of f3 for d7 he will be down the exchange for nothing.

Game 1.1 1 (D7 0) Bondarevsky, Igor Aronin, Lev Moscow ch-URS 1 948 ( 1 8) 1 .d4 tllf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tll b 6 6.tll c3 �g7 7.�e3 tll c6 8.�b5 0-0 9.tll g e2 e5 1 0.d5 tt:Je7

This concedes White some advantage, whereas 1 O . . tll a 5 is a speculative piece sacrifice. I have analyzed this idea deeply in Chapter 3 , in Game 3 . 1 be­ tween Khismatullin and Yandemirov, after Black's 9th move. 1 1 .0-0 1 1 .�cS tll d 7 1 2 .�f2 c6 1 3 .dxc6 tll x c6 1 4. 0-0;\; appears to be a slight refine­ ment over the game as the bishop is brought to a less vulnerable square with no loss of time. 1 1 ...c6 1 2.dxc6 tt:Jxc6 1 3.�c5 .l:!.e8 1 4.tll d 5 1 4.WVe l �e6 1 5 .WVf2;t looks a bit better, as Black's queen is less active than White's. White will keep a lead in development. .

C h a p t er 1: H i s t ory o f t h e

f3

An t i - Gri i n fe l d

14...�fS 1 5.�e3 �d7? 1 5 . . . �e6 1 6 .tl:lec3 tl:ld7 1 7 .�e2 a6 1 8 .�d3 - White has moderately supe­ rior development and is slightly better after, say, 1 8 . . . .ld.c8 . 1 6.tl:lec3 tl:lxd5 1 7.tl:lxd5 �g7 18 ..l:Ic1 .l:Ic8 1 9.�d2 a6

20.�a4 2 0 .�b6 ! �h4 2 1 .�a4± was better, to avoid the useful 20 . . . bS.

Igor Bondarevsky

20 ... b5 21 .�b3 �e6? Black should not have missed the chance to play 2 1 . . . tl:ld4!, even though it doesn't equalize. 22.�g5! f6 23.�e3 @h8 24. .l:Ifd 1 ±

24...f5?? After 24 . . . tl:ld4 2 5 . .l:Ixc8 �xc8 2 6 .�xd4 �xdS 2 7 .�xdS exd4 2 8 . f4 �cs 2 9 .�xd4 �xd4+ 3 0 . l:lxd4;\; the opposite-colored bishops give Black fair drawing chances despite being a pawn down. 25.tl:lb6 �xb3 26.axb3 �xd2 27. .ld.xd2 f4 28.�f2? 2 8 . tl:lxc8 ! fxe 3 2 9 .l::!. d 7 ( 2 9 . .l:Id6 tl:ld4±) 2 9 . . . l::!. xc8 3 0 . .l::!.d 6 was win­ ning for White. 28 ... .!::!.c dS 29 . .l::i.x dS

29 ....l:IxdS? 2 9 . . . tl:lxd8 would avoid the loss of a piece, but still it loses a pawn: 3 0 .l::!.c 8 .l:If8 3 1 .l:la8 tl:lc6 3 2 . �xa6 tl:lb4 3 3 .l:laS tl:ld3 34.@fl tl:lxf2 3 S .@xf2 �b8 3 6. tl:lds �f8 3 7 .b4+- . 30 ..!::!.xc6 .l:f.d1 + 31 .�e1 .l:Ixe1 + 32.@f2 .lab1 33..!kS+ �f8 34. .ld.xfS+ @g7 35 . .ld.eS .l:Ixb2+ 36.@e1 .ld.b1 + 37.@d2 i::tx b3 31

S a b o t a g e t h e G r ii n f e l d

38.ld'.xe5 ld'.b2+ 39.\t>d3 l:rxg2 40.ld'.e7+ 'itif6?? 40 . . . 'it>h6 4 1 .eS J::l. f2 42.'itie4 ld.xh2 43 .'itixf4+- . 1-0 41 .tll d 5+ 'itig5 42 ..ld.xh7 Black must lose his rook to delay mate. The opponents were both top Soviet players, especially Bondarevsky who was a candidate for the World Cham­ pionship.

Game 1 . 1 2 (D7 0) Enevoldsen, Jens Heinicke, Herbert Oldenburg 1 949 (3) 1 .d4 tll f 6 2.c4 g 6 3 .f3 d5 4.cxd5 tll x d5 5.e4 tll b6 6 .tll c 3 �g7 7.jie3 0-0 8.'iYd2 tll c 6 9.0-0-0 e5 1 O.d5 tll d 4 1 1 .f4 c5 1 2.fxe5 �g4 1 3.l::r e 1 .ixe5 1 4.h3 'iVh4

This line is still being played, but I think White is significantly better. 15.'itib1 1 5 .�d3 also favors White: 1 5 . . . ld'.ac8 1 6.�gS 'iYhS 1 7 .'itib l .id7 1 8 .ld.fl f6 1 9.�h6:t . 1 5 .. .f6?! 1 s . . . fs 1 6.�f2 'iVhs 1 7 .hxg4 'iYxh l 1 8 .gxfS 'iVhS 1 9.tll f3 tll xf3 2 0 .gxf3 'iVh2 2 1 .�d3 c4 2 2 . .ic2 �ae8 2 3 . .l:!d l gxfS 24.exfS :t . 32

Analysis diagram

White is down 1,4 pawn and has a bad pawn structure, but the dangerous passed d-pawn combined with threats to the black king will give him the edge. 1 6.!k1 1 6.tll b s �hs 1 7 .�f2 .ig3 1 8 .�xg3 'iYxg3 1 9 .tllxd4 cxd4 2 0 .tll e 2 .ixe2 2 1 .�xe2 'iYes 2 2 .'iYd3 tll d 7 2 3 . .l::!. c l .l:i.ac8 2 4 . .l::rhd l tll c S 2 5 .'iYxd4 tllxe4 2 6 .l::rxc8 l::rxc8 2 7 .'iYxeS fxeS 2 8 . .ig4.

Analysis diagram

White should be slightly better in this endgame, with rook and bishop vs. rook and knight and the more advanced passer. 1 6 ... a6 1 7.�f2 1 7 . .id3 l::ra c8 1 8 .tll g e2 tll x e2 1 9 .tll x e2 .ixe2 2 0 .�xe2 'iYxe4+ 2 1 . .id3 'l!Ha4 2 2 .�xcS .if4 23 . .ie3 .ixe3 24.'iYxe3 'tWb4 2 5 .a3 'iYd6 2 6 .'iYe6+ 'iYxe6 2 7 .dxe6 l::rfe8 2 8 .�e4:t .

C h a p t er

Analysis diagram

The rook and bishop will be stronger than the rook and knight here. 1 7...'tWh5 1 8.tt:Jce2 1 8 .�d3 ! fS 1 9 .tll g e2 �xe2 2 0.�xe2 tll xe2 2 1 .tll x e2 tll d 7 2 2 .�xcS tllxc5 2 3 ..lhcS 'tWh4 24.exfS 'tWe4+ 2 5 .'tWc2 'i¥xf5 2 6 . .ld.d l .

1:

H i s tory of t h e f3 An t i - Gri i n fe l d

21 .tllf 3? 2 1 ..l:i.e l �f4 2 2.'tWc3 �es 2 3 .'tWd2 leads to a draw by repetition. 21 ... tll a4 22.�e2?? A) 2 2 .tll h4 'tWd7 :+ ; B ) 2 2 . l:Ie l �xb2 2 3 . .ld.xe4 VWxe4 2 4.�d3 'tWxdS 2 5 .tll e 3 'tWc6 2 6.�c2 �a3 +. 22 ...�xb2 23.g4 'tWd7 24.�xc5 0-1 �xc1 25.�xc1 tt:Jxc5

Game 1 . 1 3 (D70) Enevoldsen, Harald Bolbochan, Julio Dubrovnik Olympiad 1 95 0 (7) 1 .d4 tllf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tll b 6 6.tll c3 �g7 7.�e3 0-0 8."i¥d2 tt:Jc6 9.0-0-0 e5 1 O.d5 tll d 4 1 1 .f4 �g4 1 2.�e1 c5

Analysis diagram

Black has only partial compensation for the pawn. 1 8...�f5 1 9.tll xd4 �xe4+ 20.tll c2 'iif5

1 3.h3 1 3 .fxe5 �xe5 l 4.h3 is the more accu­ rate move order. 1 3 ...�d7 With 1 3 . . . exf4 1 4.�xf4 �d7 1 5 .tll f3 l:le8= Black retains an outpost on d4. 1 4.fxe5 �xe5 1 5.tll f3 tt:Jxf3 1 6.gxf3 'tWe7 If 1 6 . . . .ld.e8 1 7 .�d3 .i::r c 8 1 8 .@b l ;l; and with ideas of f3 -f4 and h3 -h4, White must be better. 1 7.h4 33

S a b o t a g e t h e G rii n fel d

1 7... h5 1 7 . . . fS was relatively best, but 1 8 .hS or 1 8 .�gS give White a clear advantage. 1 8.�d3 tt:Ja4 1 9.tt:Jxa4 �xa4 20.f4 c4 21 .fxe5 cxd3 22.�d4 l::t ac8+

23.@b1 ? 2 3 .�c3 ! bS 24.'�b l �c2 + 2 S .Wa l aS 2 6.ld.hg l t. White's attack is stronger than Black's. 23 ...�c2+ 24.@a1 1d.c4 25.�c3 .l:!.a4 26J1e3?? 26 . .l:!.h2 �cs + .

34

2 6 ... b5?? After 2 6 . . . �cS ! ! there is no reasonable defense to the threat of 2 7 . . Jha 2 + with mate to follow soon. 27..l:!.xd3 �xd3 28.�xd3 b4-+ 29.b3 After 2 9 .�d2 �xeS Black is just up the exchange. 29 ... bxc3 30.bxa4 �xe5 31 .�c2 .l:!.b8 32.�h3 �b4 33.�d3 �xe4 34.�xc3 �xc3+ 35 . .l::!.xc3 ld'.xh4 36.l'.ld3 .l:!.b4 37.dG J::r b 8 38..l:k3 @ta 39Jk7 h4 40.d7 1:1.da 0-1 (D70) Game 1 . 14 Evans, Larry Melvyn Byrne, Donald New York ch-USA 1 963/64 ( 1 1 ) 1 .d4 tlJfG 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tt:Jb6 6.tt:Jc3 �g7 7.�e3 0-0 8.�d2 tt:Jc6 9.0-0-0 a5?!

This is reasonable, but doesn't seem to be satisfactory. 1 0.h4 1 o .@b l a4 1 1 .tLlbS tlJaS 1 2 .tlJa3 t was a fancy way to neutralize Black's coun­ terplay. 1 0...eG? This is just too slow. After 1 O . . . a4 1 1 .hS tlJaS 1 2 .hxg6 fxg6 1 3 .�gS 'iYd6 1 4.Wb l tlJbc4 l S . .ixc4+ tlJxc4 1 6.�c l t White is for choice with his strong center and open h-file

C h a p t er 1: H i s t ory of t h e f3 An t i - Gri i n fe l d

in return for the bishop pair, but at least Black has counterplay. 1 1 .h5 a4 1 2.g4+White is already winning due to the threat of 1 3 .hxg6 and 1 4.WVh2 . 1 2 ... a3 1 3 .b3 WVe7 14.hxg6 fxg6 1 5.WVh2 �h8 1 6.f4 .!:lf7 1 7.tt:lf3 �b4

Larry Evans

1 8.@c2 After 1 8. l::!'. d 3 ! +- Black has no attack and no defense against 1 9 .tll g S . 1 8... l::ra S 1 8 . . . e s ! 1 9 .dxeS �xg4 2 0 . .ixb6 .ixf3 2 1 .�c4 �xd l + 2 2 .tll xd l @f8 2 3 . .ie3 and White should still win, but at least this line is messy. 1 9.e5+- WVe7 20.tt:lg5 .ld.g7 21 .tll ce4 tt:Jd5 22.tllfG+ @f8 23.tll x d5 .ld.xd5 24.tt:lxh7+ .ld.xh7 25.WVxh7 WVxh7 26 . .ld.xh7 �g7 27.�g2 tt:lb4+ 28.@b1 'it>g8 29.�xd5 tt:Jxd5 30. .l::!. h 3 tt:lc3+ 31 .@c2 tt:Jxd 1 32.@xd 1

With the exchange and a pawn for the bishop pair (plus 2 1/4 pawns) , the win is almost in the bag. 32 ....ifS 33 ..id2 �d7 34.b4 b5 35 ..l::!.xa3 �c6 36.@e2 �d5 37.�a7 c6 38.a3 �g7 39.�e1 .ih6 40.@e3 �b3 41 .�h4 �d5 42.�fG �f8 43.@12 �h6 44.@g3 �f8 45.I;lc7 �e4 46.@h4 �d5 47.f5 1 -0

Game I.I S (A5 7) Zaitsev, Alexander Adorjan, Andras Polanica Zdroj I 9 7 I ( 3) 1 .d4 tllf6 2.c4 g6 3.f3 c5 4.d5 d6 4 . . . bS 5 .e4 d6 6 . cxbS transposes to the game while avoiding 4 . . . d6 5 . tll c 3 , which prevents . . . b7-b5 . 5.e4 b5 6.cxb5 a6

35

S a b o t a g e t h e G riinfeld

7.bxa6 7 .tll a 3 ! ilg7 8 .tll e 2 0-0 9 .tll c 3 . White scores 6 7 % in 69 games from here in the database. It's a bad Benko for Black because White will keep a piece on bS to block the usual pressure down the b-file, for example: 9 . . . tll b d7 1 0.ile2 tll e 8 1 1 .0-0 tll c 7 1 2 .�c2 ;t . This is per­ haps even better than the usual 1 2 .�g S . I n neither case does Black have much to show for the pawn. 7... �g7 8.ilb5+ tll bd7 8 . . . tll fd7 ! 9 .tll e 2 ilxa6= and Black seems to have normal Benko Gambit compensation here. 9.tll e 2 �a5+ 1 0.tll bc3 �xa6 1 1 .�xa6 �xa6 1 2.0-0 0-0

1 3.ile3 1 3 .ilgs ;t - in the Benoni it's usually better to play ilgS to provoke . . . h7-h6 and then retreat to e3 than to play the bishop directly to e3. Once you have met . . . c7-c5 by d4-d5 , the bishop on e3 is attacking a strong pawn on cS ; it's better to at least have one target on h6 that is not pawn-protected. 1 3 ... llfbS 14.�c2 tll e5 1 5.b3 tll e8 1 6.r.!.ad1 tll c 7 1 7.h3 1 7 .a4 to prevent Black's next move might be best, though Black has good compensation. 1 7...tll b 5 1 8.tll x b5 .!::rxb5 1 9.�g5 llb7 20.f4 36

20...h 6 O r 2 0 . . . tll d 7 2 1 .ilxe7 �xa2 2 2 .�xa2 .l::rx a2 2 3 .tll c l .l:i.a6= and Black has full typical Benko Gambit compensation. 21 .ilh4 g5?! 22.fxe5 gxh4 23.exd6 exd6 24. .l:i.d2 �ba7

25.e5?! After 2 5 .tll c l t Black's compensation looks thin in view of his own weak pawns. 25 ... �xe5 26.�e4 �ca 27.�xh4 'i!Yd8? 2 7 . . . .!:rxa2 28 . .!:rxa2 .l:!xa2=. 28.�xh6 l:rxa2??

C h ap t er

I:

H i s t o ry of t h e f3 Ant i - G rii nfeld

29 ..l::!.xa2?! 29 . .l:rf5 ! +- . 29....l:!.xa2 30. .ld.f5 .ld.xe2 31 ..l::!. g 5+ 'iWxg5 32.'iWxg5+ Wf8 33.h4+­ .l::i. b 2 34.h5 .l:!.xb3 35.h6 l::t b 1 + 36.wf2 wea 37.'iWh4 3 7 .h7 .id4+ 3 8 .We2 .l:i.h l 3 9 .Wd3+-. 37....l:tb2+ 38.We3 .l::i. b3+ 39.f3 .l::!. h 2 45 ..l:Ixf6++-

Now White has a winning advantage. The rest needs no comment. 45 ... @g7 46 . .l::!. g 6+ @f7 47.@g3 ld.b2 4a.nc6 .l:!.xb3 49.'it>f4 .l::r b 4+ 50.@g5 tll e 4+ 51 .@h6 .l::!. xa4 52.tll c4 h4 53.tll e 5+ @e7 54.l::f. e 6+ @da 55.f6 tll xf6 56 . .l:i.xf6 .i::t a 3 57.tll f3 a5 5a.@g5 @e7 59.ld.f4 .l:Ib3 60.@g4 nb4 61 .J::l'. x b4 axb4 62.tll d 2 @f6 63.@xh4 @g6 64.'it>g4 @h6 65.tll b3 @g6 66.tll a 5 @h6 67.h4 @g6 6a.h5+ @h6 69.@h4 @h7 70.@g5 @g7 71 .h6+ @h7 72.@h5 @ga 73.'it>g6 @ha 74.tll c 6 b3 75.tt:Jda b2 76.tll f 7+ @ga 77.h7+ 1 -0

Game 1 . 1 8 (D 7 0 ) Bjarnason, Saevar Svenn, Matti Stockholm 1 97 8 ( 6) 1 .d4 tllf6 2.c4 g6 3.f3 d5 4.cxd5 tll x d5 5.e4 tll b 6 6.tll c3 �g7 7.jie3 tll c 6

C h a p t e r I : H i s to ry of t h e f3 A n t i - G ri i n fe l d

8.d5 This is the safer route to a plus, 8 .�bS being the alternative. a ... tll e 5 9.f4 tll g 4 1 o.�b5+ c6 1 0 . . . �d7 1 1 .�d4 �xd4 1 2 .'i1'xd4 0-0 1 3 .�xd7 'i1'xd7 1 4.h3 tll f6 1 S .a4t. This is often a good move against a knight on b6. The reply . . . a7-a5 makes . . . c7-c6 unplayable. 1 1 .dxc6 0-0 1 2.'iVxdS .l::!.x d8 1 3.c7t Black will have to struggle to regain the pawn. 1 3 ... �fS 1 4..itc5

1 4...�d7? Better was 1 4 . . . �xc3+ 1 5 .bxc3 �d7 1 6 .a4 .l::!.fc8 1 7 .aS �xbS 1 8 .axb6 axb6 1 9 ..l:ha8 l:f.xa8 20.�xe7 fS 2 1 .tll e 2 fxe4 2 2 .tll d4 �d3 23 .@d2t , but the 'bomb' on c7 gives White a plus. 1 5.0-0-0+�xb5 1 6.tll x b5 �xb2+ 1 7.@b1 tll a4 1 8.�xe7

1 8...tllf2? But also after 1 8 . . . .l:!.fe8 1 9 .l::!. d8 l:texd8 2 0 . cxd8'Y!!V + .l::!. x d8 2 1 .�xd8 tll f2 2 2 .tll x a7 tllxh l 2 3 .tll b S tll f 2 24.eS tll g 4 2 5 .tll f3 +- , with a healthy extra pawn and better piece placement, White should win. 1 9.�xfS? 1 9 .tll h 3 ! tll x d l 2 0 . l:!'.xd l .l:!.fc8 2 1 . .l:!.d8+ @g7 2 2 .tll d 6 J:lxc7 2 3 .tll e 8+ @g8 24.tll xc7 + .l:!.xd8 2S . .itxd8 +­ and White is a clear knight ahead. 1 9 ... tll xd 1 20.tll h 3

20...tll e 3?? He should have played 2 0 . . . tll dc3 + 2 1 .tllx c3 �xc3 2 2 .�d6 �as 2 3 . .ld.c l l::t c 8 24.eS±. White is a pawn up, but the win is not yet certain. 21 .�e7 .l:!.c8 22.ne1 tt:lc4 23Jid1 @g7 24. .:bi.dS tll a b6 25.e5 f5 26.�ta+ 1 -0 It's mate in three. 41

S a b o t a g e t h e G r iinfeld

(D7 0) Game 1 . 1 9 Gheorghiu, Florin Jansa, Vlastimil Warsaw Zonal 1 9 79 (4) 1 .d4 tllf6 2.c4 g6 3.f3 dS 4.cxdS tt:JxdS 5.e4 tll b 6 6.tll c3 �g7 7.�e3 0-0 8.f4 tll c6 9.dS tll a S 1 0.�d4

1 5 ...cs After 1 5 . . . tllx eS 1 6.�e2 tll ec4 1 7 .d8 1 3 .�dl tll b 6 1 4.b3 tll x dS l S .cxdS tll e 7 1 6.f4 +0 .4 1 . White's space advantage together with Black's loss of castling rights define White's advantage. 1 2.tll d 5 0-0?! 1 2 . . .@d8 + 0.48 . 1 3.f4 Wk'e4?! 1 3 . . . Wk'd4 1 4.e3 + 0 . 8 0 . c7 falls. 14.tllf2 ( + 1 . 2 7 ) 14 ...Wk'xc4 1 5.e3 'l?Ha4

1 6.1le2?! 1 6.b3 ! Wk'a3 1 7 .tllxc7 .ld.b8 1 8 .1le2 tll g 7 1 9 .tll e 4+- ; White wins a pawn while keeping the superior position. 1 6 ...�e6 1 7.b3 Wk'a5 1 8.'i:Vxa5 tll xa5 1 9.tll xc7 .ld.ac8 20.tll b 5 d5 21 .flxh5 gxh5 22.tll d 4? 2 2 .@d2;!;. 22 ...tll c 6= 23.@d2 tll xd4 24.exd4 ilf5 25.l::r h c1 h6 26.h4 hxg5 27.fxg5 f6 28.gxf6 .l::!.xc1 29.l::rxc1 nxf6 30.l:rg1 + @f7 31 .@e3 l!e6+ 32.@f4 1lg6 33.@f3 1/2-112 (E60) Game 2.2 Ganguly, Surya Shekhar Gupta, Abhijeet Kavala 2 0 1 2 ( 6) 1 .d4 tllf6 2.c4 g6 3.f3 tll c6 This was my choice for Black in KRBW; it makes f2-f3 look a bit silly. 4.d5 4.tll c 3 is probably the second best move: 4 . . . dS . So we get a Grunfeld with each side having made a questionable move (3.f3 and 3 . . . tll c 6) . S .cxdS tll x dS 6 .e4 tllx c3 7. bxc3

Analysis diagram

7 . . . 1lg7 ( 7 . . . es 8 .ilbs 1ld7 9 .tll e 2 �g7 1 O .�e3 0-0 1 1 .dS ;!; + 0 . 1 8 . This is similar to the 7 . . . �g7 line) 8 .h4 hS 9 .�d3 eS 1 0.dS tll b 8 1 I .nb 1 tll d 7 S2

Chapter

2:

T h i r d Move O ffs h o o ts

1 2 .tll e 2 0-0 1 3 .�e3 �f6 1 4.g3 �e7 1 5 .0-0

Analysis diagram

Analysis diagram

White's space advantage gives him the edge, as there are still many pieces on the board. +0. 2 9 .

+0. 1 8 . White has better development and more space for free. 4 ...tt:Je5 5.e4 d6 6.tll c3 �g7 6 . . . c6 7 .f4 tll e d7 8 .tll h 3 cxdS 9.cxdS 'lWb6 1 O .tll f2 �g7 l l . .te2 0-0 1 2 .0-0 tbcS 1 3 .�c2;!; + 0 . 3 7 (Svetushkin) .

Surya Shekhar Ganguly

7.f4 7 . .te2N may be the best move, post­ poning the choice between tll f3 and tll h 3 for a move to avoid the possibility of Black playing . . . tbcS before castling. 7 . . . c6 (7 . . . 0-0 8 .f4 tll e d7 9 .tll f3 trans­ poses to the next note, while the vol­ untary retreat 7 . . . tll e d7 is met by 8 . tll h 3 ) 8 . f4 tll e d7 9 . tll h 3 cxdS 1 O.cxdS 0-0 l I .tll f2 b S 1 2 . 0 - 0 b4 l 3 .tll a4 �as 1 4.a3 tll b 6 1 5 .axb4 �xb4 1 6 .�d2 �d4 1 7 .�c3 �xd l 1 8 .�fxd l ;!; .

7...tt:Jed7

8.tll h 3 53

S a b o t a g e t h e G rii n fe l d

8 .tL'if3 and now: A) 8 . . . 0-0 9 .�e2 (Svetushkin men­ tions only the inferior 9 .�d3 as leading to equality) 9 . . . tLlcS 1 0 . e S tLife4 ( 1 0 . . . ttJfd7 l 1 .�e3 t) l l .tL'ixe4 tLixe4 1 2 .�e3 c6 1 3 .0-0 �fS 1 4.°iYb3 'i¥c7 1 S . �ad l llfd8 l 6 .�d4t + 0 . 2 2 . White's space advantage is obvious, while Black's compensation for it, the bishop pair, may disappear shortly; B) 8 . . . c6 9 .�e3 °iYaS 1 0 .°iYd2 tLlcS 1 1 .l:Id l t + 0 . 2 6 ; C ) 8 . . . tLlcS ! ? 9 . e S + 0 . 2 2 . 8 0-0 A) 8 . . . eS (or 8 . . . e6 9.dxe6 transposing. I failed to capture in a similar position against Shabalov recently and I lost the game) 9.dxe6 fxe6 1 O.�e2 (Svetushkin thinks 1 O.�d3 to prepare f4-fS is even better, but anyway Black can avoid this by waiting for ite2 before playing . . . el -eS) 1 0 . . . 0-0 1 1 . 0 - 0 b6 1 2 .�e3 itb7 ( 1 2 . . . tLlcS 1 3 .°iYc2 aS 1 4.�f3 1:!'.b8 l S .i::rae l tLlfd7 1 6.tLlgS �b7 l 7 .b3 'i¥e7 1 8 .a3 eS 1 9.fS h6 20.tLlh3 (20.b4!t) 20 . . . gxfS 2 1 .exfS e4 22.itxcS itxc3 23 .'i¥xc3 bxcS was Grischuk-Navara, Rhodes 20 1 3 . Now White played 24.tL'ifl and eventually won, but 24.�hS ! +0.70 was much stronger) 1 3 .itf3 l:Ib8 1 4.°iYc2 'i¥e7 l S . .l:Iae l t

B) 8 ... c6 9.�e2 (Svetushkin prefers 9 .�e3 but 9 . . . °iV aS 1 0 . tLlf2 tLlcS l l .itd2 seems an unnecessary conces­ sion although White is still for choice) 9 . . . 0-0 1 0 .tL'if2 cxdS 1 1 .cxdS bS ! ? 1 2 . 0-0 (Svetushkin considers only ac­ cepting the gambit, which he prefers to avoid by 9 .ite3 . But it seems that 1 1 . . . bS can just be ignored) 1 2 . . .b4 1 3 .tLla4 °iYaS 1 4.a3 tL'ib6 1 S .axb4 'i¥xb4 1 6.�d2 °iYd4 1 7 .�c3 °iYxd l 1 8 . .!d:fxd l t

...

Analysis diagram

+ 0.44; 54

Analysis diagram

+ 0 . 2 9 . White has an advantage in piece placement and space. 9.�e2 ttJcS A) 9 . . . l:re8 1 0 .�e3 e6 l l .dxe6 �xe6 l 2 .tL'if2t + 0 . 2 9 . White's space advan­ tage is significant; B) 9 . . . e6

� Don't fail to capture on e6 when legal in this line.

1 O.dxe6 fxe6 1 1 .0-0t; see the note to Black's eighth move; C) 9 . . . c 6 l O . tLlf2 ( 1 0 .ite3 cxdS 1 1 .cxdS bS 1 2 .a3 a6 +0. 1 0 , Wallace­ Gormally, 4NCL 2 0 1 3 / 1 4) 1 0 . . . °iYb6 1 1 .0-0 as 1 2 .°iYc2 tLics 1 3 .�e3 'i¥b4 1 4.eS (the surprising 1 4.�f3 is also good) 1 4 . . . dxeS l S .fxeS tLlfd7 1 6.e6 fxe6 1 7 .tL'ig4 exdS 1 8 . cxdS Wh8 1 9 . .i::!. ad 1 t. White's large advantage in development is more significant than Black's extra pawn.

Chapter

2:

T h i rd Move O f f s h o o t s

A) 14 ... b5 1 5 .�e3 tZ'icxe4 1 6.tZ'icxe4 tt:Jxe4 1 7 .�xe4 f5 1 8 .�d4 fxe4 1 9 .�xg 7 @xg7 20 . .l:le l �f5 2 1..l:!. c l 'iVe7 2 2 . g4 e3 23 .'iVd4+ @g8 24.gxf5 exf2 + 2 5 .'iVxf2 'iVf6 2 6 .fxg6;!;

Analysis diagram

1 9 . . . a4 2 0 .a3 + 0.44; D) 9 . . . tZ'ib6 1 0.tZ'if2 e6 1 1 .0-0 exd5 1 2 .cxd5 c6 1 3 . dxc6 bxc6 After 1 4.'iYc2 this was eventually drawn in Postny­ Nyback, Germany Bundesliga 2 0 1 3 / 1 4. Instead 1 4.�e3 + 0 . 2 9 gives an edge. 1 0.tt:Jf2 e6 1 0 . . . ki.e8 1 1 . 0 - 0 ( 1 1 .�e3 + 0 . 1 8) 1 1 . . .e6 ( 1 1 . . .e 5 ? ! 1 2 .f5 gxf5 1 3 . exf5 e4 ( 1 3 . . . �xf5 1 4.tZ'ig4 �xg4 1 5 .�xg4 e4 1 6.�f5 ;!;) 1 4.tZ'ih l + 0 . 2 9 . The knight belongs on g3 here) 1 2 .�f3 exd5 1 3 .cxd5 ;!; transposes to the note to move 1 2 . 1 1 .0-0 exd5 1 2.cxd5

Analysis diagram

+ 0 . 3 7 . White's extra passed pawn on the kingside is clearly worth more than Black's extra backward pawn on the queenside; B) 1 4 . . . b6 1 5 .'iYc2;!;; C) 14 ... a5 1 5 . .l::!. e l tt:Jfd7 1 6.�e3 b6.

Analysis diagram

1 2 ...c6 1 2 . . . !Ie8 1 3 .�f3 h5 ( 1 3 . . . a5 1 4 . .l::!. e l b 6 (+0 .44) 1 5 .�e3 ;!; i s similar to the note to 1 3 . . . h5 1 4.h3 a6) 1 4.h3 and now:

I called this position equal in KRBW, but as grandmaster Sam Shankland pointed out to me, this is wrong: 1 7 .�d4;!; (+0 . 3 7 ) 1 7 . . . �h6 1 8 .g3 �a6 1 9 .h4 + 0 . 6 0 . Svetushkin prefers 1 9 .�e3 . White has a large space advantage, and Black lacks counterplay. 1 3.dxc6 bxc6 55

S a b o t a g e t h e G r ii n fe l d

14.�e3?! ,-{\. Don't overlook 'sham sacrifices'

\d,) on e4.

Better was 1 4.�f3 ! and now: A) 1 4 . . . �b8 1 S .�e3 �xb2 1 6.eS lUdS 1 7 .tUxdS cxdS 1 8 .�xdS lUe6 1 9 .�xd6 �xd6 2 0 .exd6 lUd4 2 1 .1:!.ac l ld:xa2 2 2 . !::r fd l lla4 2 3 .�h l lld8 24.ti:Je4i + 0 . S 6 . White's advanced passed pawn gives him the edge; B) 1 4 . . . �a6 1 S .1:!.e l 1:!.b8 1 6.�e3 .l:Ixb2 1 7 .est + 0 . 3 7 ; C) 1 4. . . �e7 1 S .�e3 d S 1 6.eS lUfd7 1 7 .1:!.e l lUe6 1 8.llcl �a6 ( 1 8 . . . �b7 l 9 .lUa4i +0.44) l 9 .lUe2 llfc8 20.b3i.

Analysis diagram

Svetushkin: 'Black has a bad version of hanging pawns.' The computer would say: 'White has fewer pawn islands , and more active bishops and rooks.' + 0 . 64. 14 ...t2Jfxe4 1 5.t2Jfxe4 56

Draw agreed. Play might continue 1 S . . . tUxe4 1 6. tUxe4 .l:Ie8 1 7 .ii.f3 dS 1 8 .l::t e l (-0.04) l 8 . . . dxe4 l 9.�xd8 llxd8 2 0 .�xe4 �e6 2 1 .llad l l::rx d l 2 2 .l:i.xd l �xb2 23 .ii.xc6 llc8 24.�d7 .l::!.d 8 2S . .l::!. b l l::l'.x d7 26 . .l::!.x b2 and the position is a dead even draw. (E60) Game 2.3 Vitiugov, Nikita Vachier-Lagrave, �axiine Gibraltar 20 1 3 ( 1 0 . 1 ) 1 .d4 t2Jf6 2.c4 g 6 3.f3 e6!?N

The idea of this clever novelty is to wait for lUc3 before playing . . . d7-dS so that the knight on dS will be able to ex­ change itself when attacked by e2-e4, or else to recapture with a pawn on dS . The path to a white edge is very narrow. 4.e4 4.lUc3 dS and now: A) S .cxdS tUxdS 6 .e4 tUxc3 7 .bxc3 cS 8 .�bS + (8.�c4 is inferior for

C h a p t er

White to the normal �c4 Exchange Griinfeld, as the move . . . e7 -e6 is more useful than f2-f3 here) 8 . . . tt:ld7 ? ! (8 . . . �d7 ! 9 .�xd7+ �xd7 1 0 .tt:le2 tbc6 1 1 .�e3 cxd4 1 2 .cxd4 �g 7=) 9 . tt:le2 cxd4 1 0 . �xd4 f6 1 1 . eS ( + 0 .29) l 1 . . . @f7 1 2 .exf6 �xf6 1 3 .�f4 es 1 4.�c4+ �e6 1 S .�g3 �xc4 1 6 .�xc4+;!; was KasimdzhanovNaumann , Germany Bundesliga 2 0 1 2 I 1 3. White won; B) S .�gS ! h6 (S . . . �g7 6.e4 dxe4 7 .fxe4 h6 8 .�h4 g S 9 .�g3 0 - 0 1 O.�e2;\;) 6.�h4 ( 6 .�xf6 �xf6 7 .e3 + 0 . 1 8 ; White has good compensation for the bishop pair) 6 . . . �e7 (6 . . . �g7 7. e4 dxe4 8 .fxe4 transposes to the pre­ vious note) 7 .�d2 tt:lbd7 8 . �g3 (8.cxdS exdS 9 .e 3 ;\;) 8 ... c6 9 .e4 dxe4?! (9 . . . 0-0 1 0 .eS tt:lhs 1 1 .cxdS + 0 . 2 2 ) 1 O.fxe4 �b4 1 1 .�d3 eS 1 2 .a3 �as 1 3 . d S cxdS 1 4. cxdS tt:lcS 1 S . b4 tt:lxd3 + 1 6.�xd3 �b6 1 7 .tt:lf3 0-0 1 8 .tt:lxeS aS 1 9 .bS a4 2 0 .tt:lc4±

2:

T h ir d Move Offs h o o t s

Analysis diagram

2 9 . . . �xd6? (29 . . . �g l 3 0 .tt:laS �b6 3 1 .tt:lxb7 l:rb8 3 2 .tt:lcS �xcS 33 . .l::i.x cS fS 34.@a2 fxe4 3 S .d7 .l:!.bd8 3 6 . .l:!.c7 ;\; + 0 . 7 1 . The passed pawns should regain the exchange, keeping a pawn plus, though Black has fair drawing chances) 3 0 . tt:lxd6 +- .l::i. c 3 3 1 .@b2 .ld. b 3 + 3 2 .@a2 .l::!. d 8 3 3 . .l:!.d4 hS ? 34.�h4 .l:id7 3 S .�f6 @h7 3 6 .eS Black resigned, Nakamura-Safarli, Troms0 2 0 1 3 . 4 ... d5 4 ... cS S .dS (S .tt:le2 !?) S . . . d6 6 .tt:lc3 transposes to lines covered under 3 . . . c S . 5.e5?! I think S .cxdS is better than the game move. See Svidler-Vachier-Lagrave. If S . tbc3 dxe4 6 .fxe4 and now: A) 6 . . . cs 7 .ds i.g7 8 .tt:lf3 exdS 9.exdS 0-0 1 O.�e2 .l:!.e8 1 1 .0-0 �fS

Analysis diagram

+ 1 . 1 0 . White dominates the center and has half a pawn more. 2 0 . . . �cS 2 1 . 0-0-0 �g4 22 . .l:rd2 (22 . .l:!.dfl ! ± + 0 . 9 8 . The pressure on the f-file aug­ ments White's plus) 22 . . . �e7 2 3 .d6 �e6 24.tt:lds tt:lxdS 2 S .�xdS .l::!. a c8 2 6 .'iitb l �xdS 27 . .l:!.xdS �e6 28 . .l::!.hd l i.xdS 2 9 . .l:rxdS

Analysis diagram

1 2 .�gS ( 1 2 .�d3 ! �xd3 tt:lbd7 1 4.�f4;t + 0 . 2 7 )

1 3 .�xd3 1 2 . . . h6= 57

S a b o t a g e t h e G r iin f e l d

1 3 . .ih4 tt:lbd7 1 4.tt:lbs 'ffe 7 l S . l:!'.e l 'ffe 3+ 1 6.�f2 'f/f4 1 7 . .ig3 "ffe 3+ l 8 . .if2 'fif4 1 9 .�g3 'ffe 3 + 20.�f2 1/2- 1/2 Laznicka-Vachier-Lagrave, Warsaw Ech-tt 2 0 1 3 ; B) 6 . . . eS= 7 .ds �cs (Yang-Mikha­ levski, Santa Clara 2 0 1 4) + 0 . l S , but here I disagree with Komodo. By anal­ ogy to other similar openings, I think Black has full equality. 5 ...tt:lh5 6.f4? 6 .�e3 ! cS 7.cxdS exdS 8 . .ibS + and now:

Analysis diagram

A) 8 . . . .id7 ! 9 .�xd7 + 'fixd7 1 0 .tt:le2 tt:lc6 l l .tt:lbc3 cxd4 l 2 ..ixd4 tt:lg7 1 3 .tt:lf4 .ib4 ( 1 3 ... 0-0-0 0 . 0 0 - that's a lot of zeroes! - may be better and per­ fectly OK for Black) 1 4.0-0 .ixc3 1 S .bxc3 0-0 1 6 . .if2 (+0.07) 1 6 . . . tt:le7 1 7 .'fid2 tt:le6 1 8 .c4 l:rad8 1 9 .�ad l tt:lxf4 2 0 .'ffxf4 'ff e6 2 1 .cS and White is slightly better due to potential pressure on b7 as well as dS ; B) 8 . . . tt:lc6 9 .tll e 2 cxd4 (if 9 . . . a6 1 0 . .ixc6+ bxc6 1 1 .dxcS 'fih4+ 1 2 .g3 'fih3 1 3 .'ii a 4 �d7 1 4 . .id4 'ff e 6 1 S .tt:ld2 �h6 1 6.f4;!; White has a 'half-pawn' material plus for nothing) 1 O.tt:lxd4 .id7 1 1 .0-0 �cs l 2 .f4 a6 1 3 .�xc6 bxc6 1 4.'fid2 �a7 1 S .tt:lb3 �xe3 + 1 6 .'ffx e3 ;!; ; White has knight versus bad bishop with firm control over the dark squares. 58

6 ... 'fih4+ 7.g3 tt:lxg3 8.tt:lf3 "fih5 9.hxg3 'ffx h1 __.n The modern trend is to sacrifice

ta) the exchange rather freely, but sometimes (as here) I think this is overdone. 1 0.tt:lc3

White apparently judged that he would have enough compensation for the lost material with his large space and devel­ opment advantage, but this appears to be false. Since 6.f4? appears to be un­ sound, I'll give only brief notes to the rest of the game as this is an opening book. 1 0 ... h5? 1 0 . . . dxc4! l l .Wf2 �d7 (- 1 . 2 7) 1 2 .'ffe 2 �c6 1 3 . .ig2 "fihS 1 4.'fixc4 .ie7 l S .dS exdS 1 6 .tt:lxdS �xdS 1 7 .'fixdS tt:lc6 1 8 .�e3 'f/fs 1 9 .�cl 0-0-+ .

Analysis diagram

Chapter

2:

T h i r d M o v e O ffsh o o t s

White has some space and development advantage, but that doesn't even come close to offsetting 2 1,4 pawns. 1 1 .�e3? 1 1 .cxdS ! exdS 1 2 .'iit> f2 h4 1 3 .�bS+ c6 1 4. �xh l hxg 3 + 1 5 . 'iit> g l l:rxh l + 1 6.'iit>xh l cxbS 1 7 .ti:JxdS t2ia6 1 8 .ti:Jf6+ 'iit>e 7 1 9 .'iit> g 2;!;.

N ikita Vitiugov

22.'iit>c 2± b5 23.�b3 b4 24.axb4 �xb4 25.tt:Je4 tt:Jd5 Analysis diagram

After capturing the pawn on g 3 , White will be down the bishop pair but will have tremendous compensation in cen­ ter, space, and development. 1 1 ... h4 1 2.gxh4 c6? l 2 . . . �e7 + - 0 . 8 9 . 1 3.'iit> d 2 26.�d2? 2 6 .'iit>b l ;!;. 26 ...�xd2 27.tt:Jxd2 Ith8 28.ti:Jf1 �e1 28 . . . �a l -+ . 29.�xd5 exd5 30. .!d.xd7 'iit>xd7 31 .�g4+ 'iit> c 7 32.�xg6 �f2+ 3 2 . . . 'iit>b 6 -+ . 33.tt:Jd2 l:[h2? 34.�d6+= 'iit> b 7 35.�b4+ @ca 36.�fa+ 112-1/2 1 3 ... !:txh4? l 3 . . . �e7+. 14.tt:Jxh4 �xh4 1 5.�f3= �d7 1 6.�d3 tt:Ja6 1 7.l:[h1 �d8?! 1 8.a3;t dxc4 1 9.�xc4 �a5? 20.l:[h7 0-0-0 21 .�xf7 tt:Jc7

(E60) Game 2.4 Svidler, Peter Vachier-Lagrave, �axiille Paris/St Petersburg 2 0 1 3 (5) 1 .d4 tt:Jf6 2.c4 g6 3.f3 e6!? 59

S a b o t a g e t h e Gr iinf e l d

This recent idea of Vachier-Lagrave deserves serious consideration. 4.e4 d5

4 ... cS S .dS would just transpose to 3 . . . cS 4.ds e6 S .e4. 5.cxd5 Probably best, but S .eS tZ'ihS 6.�e3 is a playable alternative. See Vitiugov­ Vachier-Lagrave for analysis of this. 5 ...exd5 6.t2Jc3 dxe4 Or 6 . . . �b4 ( 6 . . . cS 7 .�gS;l; +0.60) 7 .es tZ'ihS 8 .g3 0-0 9 .tZ'ie2 l:!.e8 1 0.�g2 cS l l .�e3 tZ'ic6 1 2 .f4 �g4 1 3 . 0-0;l;; White has a space advantage and pressure on dS and hence a modest edge. + 0 . 2 9 . 7.fxe4 �b4 8.�g5 h6 9.kxf6 'iYxfG 1 O.t2Jf3 �g4 1 1 .�b5+ White provokes l l . . c6 because otherwise . . . tZ'ic6 will be annoying. 1 1 ... c6 1 2.�e2 l 2 . �c4! ? + 0 . 2 9 . 1 2 ...t2Jd7 1 3.0-0 'iYe7 .

14.'iYc1 1 4.tZ'id2 ! ? �e6 1 S .a3 �as 1 6.tZ'ic4 + 0 . 1 7 , Flear-Vachier-Lagrave, Gibraltar 2 0 1 4; 1 4.'iYc2 0-0 1 S .h3 �e6 1 6 .kd3 .l:i.ad8 1 7 .a3 �as 1 8 .eS �g7 1 9 .b4 �c7 2 0 .tZ'ie4 �ds 2 1 .llae l as 2 2 .kc4;l;

Analysis diagram

+O . 1 8 . White has a central and space advantage, while Black is losing his only trump, the bishop pair. 1 4... 0-0-0 1 5.'YWf4 f5 1 6.tZ'ib5 1 6.exfS ! � Take stuill

this case postponing � the capture had adverse consequences as will be seen at move 1 9. In

1 6 . . . gxfS 1 7 .tZ'ibS cxbS 1 8 . .l::I.ac l + tZ'icS 1 9 .a3 'iYe4 2 0 .'iYxe4 fxe4 2 l .axb4 exf3 2 2 .l:!.xcS+ Wb8 2 3 . gxf3 kh3 24 . .l::r d l l:l'.hg8+ 2 s .wf2 l:!.g2+ 2 6.We3;t.

Analysis diagram

White will keep an extra pawn. Black has fair drawing chances after 2 6 . . . �e6 2 7 .�xbS + 0 . 3 7 . 60

Chap ter

16 ...cxb5 17..l::!.a c1 + tt:Jc5 1 S.exf5 'iYxe2 1 9.'iVxg4

1 9 ...gxf5 l 9 . . . �e3 + ! 2 0 .@h l gxfS 2 1 .'iYxfS + @b8 2 2 .dxcS .ld.hf8 2 3 .�hS a'.c8 . Black will regain the pawn on cS with near-equality: 24.h3 0 . 0 0 . 20.'iVxf5+ @bS 21 .dxc5?! 2 1 .'iYf4+ @a8 2 2 . dxcS ( + 0 . 5 2 ) 2 2 . . . �hg8 2 3 .g3;\;. This shows why Black should have inserted 1 9 . . . 'iY e3 + earlier, as now it will be difficult for him to regain the lost pawn. 21 ... a'.hfS 2 l . . . �e3 + 2 2 .@h l transposes to the note to move 1 9 . 22.'iYe5+?! 2 2 .'iYc2 'iVxc2 23 . .l::!.xc2 ;\; was a better way to trade queens than in the game, as now the second rank is defended. 22 ...'iVxe5 23.tt:Jxe5 1::1'.feS 24.c6 �d2 25.�f5?! �d6 26.tt:Jd7+ @cs

2:

Thi r d Move O f fs h o o t s

27.tt:Jf6? 2 7 .cxb7+ Wxd7 2 8 . .l::!.f7 + a'.e7 2 9 .l'.:!.f6 �b8 3 0 . .l::!. f8 .l:Ie8 3 1 . .ld.f? + @e6 3 2 . lah7 - with two pawns including a dangerous one on the seventh rank for the bishop, White has decent drawing chances.

Maxime Vachier-Lagrave

27... a'.e5 2S.a'.xe5 �xe5 29.tt:Je4 .l:Ixb2 30.cxb7+ @xb7 31 .tt:Jc5+ was 32 ..l:Id1 �c7 33.a'.e1 a5-+

With his rook on the 7th rank, bishop vs. knight with opposing majorities, and the threat against a2 , Black is win­ ning. 34.�e6 �c2 35.tt:Jd3 h5 36 ..&!.h6 .l::!.x a2 37Jbh5 l:Id2 3S.tt:Jc5 a4 61

S a b o t a g e t h e G rii n fe l d

39.l:i.h6 �d6 40..l:!.hB+ @a7 41 .@f1 l:k6 0-1 Black's advanced passed pawns will cost White material.

Game 2.5 (E60) Socko, Bartosz Leniart, Arkadiusz Poland tt 2 0 1 2 ( 3 . 2) 1 .d4 t2Jf6 2.c4 g6 3.f3 c5 4.d5 jLg7 5.t2Jc3 0-0 Black should really play S . . . d6 first, be­ cause here White could play 6.tbh 3 , which i s usually a good move in the Samisch if jLxh3 is not possible. 6.e4 d6 7.jLg5

This move scores fairly well here, but with . . . c7-cS d4-dS not yet played it does not score so well, because Black can play 6 . . . a6, planning a later . . . c7-cS d4-dS b7-bS in the style of the Benko Gambit. In the game position . . . a7-a6 is almost always met by a2-a4, ruling out any . . . b7-bS Benko Gambit ideas. 7 .tbge2 transposes to the main line of the Samisch with 6.tbge2 cS 7 .dS . I give the 7 .jLgs move here for those readers who prefer to meet the Samisch by 6 .jLe3 cS 7 .tbge2. 7... h6 .-f\ It usually pays to lose a tempo

� with �gS -e3 if it provokes . . . h7-h6. 62

7 . . . e6 8.'ifd2 exdS 9.cxdS (9.tbxdS ! ? �e6 1 O.tbe2 tbc6 1 1 . 0-0-0;\; +0.26. The d6-pawn is weak and the pin annoying) 9 . . . �e8 1 o .t2Jge2 t2Jbd7 1 l .tbg3 a6 1 2.�e2 bS 1 3 .0-0 V/iic 7 1 4. .ld.ac l ( 1 4.a4 + 0 . 2 9 ) 1 4 . . . b4 ( 1 4 . . . t2Jb6 + 0 . 1 S ) l S .tbdl aS 1 6.jLh6 �h8 1 7 .tbe3 was Dreev-Khairullin, St. Petersburg 20 1 1 . White should be better, as Black lacks any obvious play. +0.44. 8.�e3 a6 8 . . . e6 9 .V//id 2 exdS 1 0 .cxdS and now: A) 1 O . . . t2Jbd7 1 l .tbh3 (again a good move as the knight can't be captured by a bishop) 1 l . . . tbh7 1 2 .�e2 tbeS 1 3 .tbf2 hS 1 4. 0-0 fS l S . �ae 1 . This was Mamedyarov-Vachier-Lagrave, Dresden Olympiad 2 0 0 8 . +O . S 2 . White looks better; . . . f7-fS and . . . h7-hS don't go so well together; B) 1 0 . . . h S .

Analysis diagram

This seems better than the game, be­ cause Black saves the tempo . . . 'itih7 . However because the white knight is not yet on g3 , he can bring it to c l in­ stead, making 1 O . . . hS look pointless: l l .tbge2 tbbd7 1 2 .tbcl �e8 1 3 .�e2 tbeS 1 4.0-0 a6 1 S .a4 t2Jh7 1 6.�b l (this plan of preparing b2-b4 is often good against Benoni formations) 1 6 . . . fS 1 7 . b4 b6 1 8 .bxcS bxcS 1 9 .�h6;\;.

C h a p t er

9.'ii'd 2 'it>h7 1 O.tt:lge2 e6 1 O . . . b5 1 1 .cxb5 tt:lbd7 1 2 .tt:lg3 tt:le8 1 3 .a4 tt:lc7 1 4.'ii'd l ;!; looks like a rather good version for White of the Benko Gambit. 1 1 .a4 exd5 1 2.cxdS tt:lbd7 1 3.tt:lg3;!;

1 3 ...l::r bS 1 3 . . . h5 1 4.�e2 h4 1 5 . tt:lfl .ld.e8 1 6.�g5 h3 1 7 .g4;!;. The pawn on h3 looks rather lonely, and g3 will be a pretty good square for the knight. 1 4.�e2 tt:le8 1 5.0-0 tt:lc7 1 6.f4 b5

1 7.eS! dxe5 1 8.fS This is the pawn sacrifice plan made fa­ mous by Hans Kmoch's book Pawn Power in Chess more than half a century ago. He called it 'Sealer - Sweeper' . The computer has no difficulty in seeing that White has the advantage despite the pawn deficit.

2:

T h ir d Move O ffsh o o t s

18 ...c4 1 9.d6 tt:le8 20.axbS axb5

21 ...tf3 2 1 . 1::!'. a 7 ! tt:ldf6 2 2 . .ld.xf7 ! gxf5 ( 2 2 . . . .l::!: xf7 2 3 .fxg 6 + 'it>xg6 24.'ii'c 2 + +-) 2 3 .1::!'. xf8 �xf8 24.tt:lxf5 ..txf5 2 5 . .ld.xf5 �xd6 2 6 .'ii'c 2 Wh8 2 7 .'iYc l ;!;. White has the safer king. the better pawn structure, and more active pieces in return for the half-pawn mate­ rial deficit, clearly more than enough. 21 ... b4? 2 1 . . .tt:ldf6 2 2 .fxg6+ fxg6 23 .l:ra7;!;. 22.tt:lce4± tt:lef6? 23 . .l:!.a7 tt:lg8?

24.'i¥c2 24.fxg6+ fxg6 25 . .l::!.c l +- . 24...'it>ha 2s.�e2 2 5 . fxg6 fxg6 2 6 .'ii'x c4+-. 25 ... l:rb7 26.l:rxb7 �xb7 27.jlxc4 tt:lgf6 28.tt:lxf6 �xf6 29.'ii'd 2 2 9.fxg6 'ii'x g6 3 0 .'ii'd 2 +-. 29...e4 30.fxg6 'ii'x g6 31 .'ii'xb4 ..tc6 63

S a b o ta g e t h e G r iin feld

32.tllf S? 3 2 .b3 fs 3 3 .'iVas +- . 32 ... l:tbS 33.'iVaS �xb2 34.�dS+ @h7 3S.tll h 4 �xd6 36.�xf7t l:tb1 + 37.�f1 ?? �bS 3 7 . . . �e6-+ . 38.'iVeS tll e S 39 . .l:f.xg7+ @xg7 40.tllfS+ @f6 41 . 'iYxbS 'iYd 1 42.�a6+ @xfS 43.h3 hS 44.�cs+ tll d 7 4S.'iYc4 tt:ies 46.'iVcS+ tll d 7 47.'iVc4

47... �d3?? 47 . . . tll f6-+. 48.'iVf7+ tllf6 49.g4+ hxg4 SO.hxg4+ @es S1 .'iYe7+ @dS S2.'iVf7+ @es S3.'iYe7+ @dS S4.'iVf7+ @es SS.'iVe7+ 112-112

Game 2.6 (E60) Anand, Viswanathan Gelfand, Boris Moscow Wch m 20 1 2 (8) 1 .d4 tllf6 2.c4 g6 3.f3 cs 64

The advantage of this move over the King's Indian is that White must ad­ vance, since tll e 2 is not legal. The disad­ vantage is that White has not commit­ ted either to �e3 or to tll g e2 yet. 4.dS d6 S.e4 �g7 6.tlle 2 The idea is to play this knight to c3 , solving its development problem since neither g3 nor c l is ideal for this piece. The other knight then becomes the problem piece, but it can go to a3 and later to c2, or wait for the c l -bishop to develop and then go to d2. For the rou­ tine 6 .tll c 3 see the previous game. Note that this idea of tll - e2-c3 is only reason­ able when the moves . . . c7-c5 d4-d5 are already on the board, as otherwise d4 may need protection and the suppres­ sion of . . . b7-b5 by tll a 3 may be irrele­ vant. The jury is still out as to whether this tll e 2-c3 in the game is better than the standard tllb c3 , but I think it is. 6 ... 0-0

7.tll e c3!? For the ordinary 7 .tllb c3 see the 'Samisch with . . . c7 -cs ' chapter. 7...tll h SN A) The usual move is 7 . . . e6 8 .�gS (provoking . . . h7-h6 before placing the bishop on e3 is often desirable in King's Indian and Benoni lines, so that �d2 will be with tempo) 8 . . . h6 9 . �e3 exdS I O.cxdS a6 l l .a4 tll b d7 l 2 .�e2 tllh 7

C h a p t er

2:

T h ir d Move O ffs h o o t s

1 3 . 0-0 �e8 1 4.°YWd2 hS (Gil Capape­ Cortes Bueno, Spain 1 99 2 ) 1 5 .tt:la3t

out by Svetushkin, appears to render 8.g4 harmless) 1 1 .cxdS and now:

Analysis diagram

Analysis diagram

+ 0 . 5 6 . This looks like a rather ideal Benoni for White. Black can do nothing on the queenside while the pawn on h5 makes . . . f7-f5 less desirable; B) 7 . . . tt:la6 ! ? 8 . .2.e2 tt:lc7 9 . �e3 a6 1 0.a4 .ld.b8 1 1 . 0-0 �d7 1 2 .tt:la3 ( 1 2 .aS also favors White) 1 2 ... tt:lfe8 1 3 .°YWd2 e6 1 4.aS bS 1 5 .axb6 l:Ixb6 1 6 .Sfb l °YWb8 1 7 .l::!. a 2 .ld.b4 (Laznicka-Polzin, Germany Bundesliga 2006/07) l 8.dxe6 fxe6 1 9 .tt:ldl .l:!.f7 2 0.:!d.ba l .ld.b3 2 1 .tt:lb l °YWb6 2 2 .tt:lbc3 t.

A l ) l 1 . . .tt:lfd 7 ( 1 1 . . .h S l 2 .h 3 ::t ; l 1 . . .tt:lbd7 1 2 .tt:ld2 'YWe 7 1 3 . 0 -0::t) 1 2 .h4 ( 1 2 . 0-0 fs 1 3 .gxfS gxfS 1 4.°YWd2 tt:Jes 1 5 .�gS 'YWc7 1 6 .�h l :t; White aims to play a2-a4 and tt:la3) 1 2 . . . fS ? ( 1 2 . . . tt:Jes 1 3 .h S g s 1 4.h6 �f6 1 5 .°YWd2 �h8 1 6.a4 tt:la6 1 7 .tt:la3 tt:lb4 1 8 . '>t>fl ;!:; and White has a nice space advantage though the blocked position makes a draw likely) 1 3 . gxfS gxfS 1 4.'iWd2 b5 ( 1 4 . . . a6 1 5 .�gS 'YWe8 1 6.a'.g 1 �h8 1 7 . .2.h6::t ; Black must ei­ ther sacrifice the exchange or suffer from a weak king and weak pawns)

Analysis diagram

Black has two weak pawns on half-open files. 8.�g5?! A) Critical is 8.g4! ? tt:lf6 9.�e2 e6 1 O.�e3 exd5 (1 O ... a6! l l .a4 tt:le8= and Black plans . . . f7-f5 . This defense, pointed

Viswanathan Anand

65

S a b o t a g e t h e G r iin f e l d

White plans to expel the outposted knight while retaining a nice space ad­ vantage; C2) 8 . . .itd4 appears to favor Black, but White will eventually dislodge the bishop by tLla3-c2 and then castle: 9.ith6 .l:!.e8 1 0.g4 tLlf6 l l .�d2 e6 1 2 .tLla3

� look for moves like this to expel ·

Analysis diagram

I S .ith6 (or I S .tlJxbS ± fxe4 1 6.fxe4 tt:Jes 1 7 .ltgs �d7 I 8 .�h6 and Black has nothing for his lost pawn) I S . . . a6 1 6.exfS ? ! ( 1 6 . .l::!. g l �xh4+ 1 7 .Wd l nf7 1 8 .exfS tt:Jf6 1 9 .tlJe4±) 1 6 . . . tlJeS 1 7 .ltxg7 'itixg 7 l 8 .tlJe4 ? ! ltxfS 1 9 .tlJg S ? �e7 + Hammer-Stokke, Nor­ way tt 2 0 1 2 / 1 3 . Black has the much better pawn structure with no obvious compensation and went on to win; B) 8 .ite3 e6 9 . �d2 �h4+ I O.itf2 �f6 I I .tlJa3 exdS 1 2 .cxdS a6 1 3 .tLlc4 bS 1 4.tlJb6 .l:!.a7 I S .a4 b4 1 6.tLldl and White is for choice due to the c4 out­ post square, but Black's better develop­ ment makes the game nearly even; C) 8 .ite2 ! This was not mentioned by Svetushkin. C I ) 8 . . . tt:Ja6 9 . .1e3 eS 1 0. 0-0 ( I O.g3 + 0 . 2 6) I O . . . tt:Jf4 l l .�d2 ( + 0 . I S ) I I . . . .1d7 1 2 .�dl tt:Jc7 1 3 .tt:Je2 gs 1 4.tt:Jbc3t

Analysis diagram

66

a powerful outposted bishop.

1 2 . . . a6 1 3 .tLlc2 exdS 1 4.tlJxd4! dxe4 I S . 0-0 cxd4 1 6 .�xd4 tt:Jc6 1 7 .�fl exf3 1 8 .itxf3 tt:JeS 1 9 .h3t

Analysis diagram

+ 0 . 7 6 . White has the bishop pair and pressure down the f-file, and Black is missing his important king-protecting bishop on g 7 . 8 itf6?! ...

Both bishops are very valuable, White's because his pawns are mostly on light squares, Black's because it defends the king. It's hard to say who benefits from the swap. However the text is an admis-

Chapter

sion that �gs was better than �e3 , be­ cause White can now retreat to e3 when the bishop is worse on f6 than on g 7 . A ) Anand suggested 8 . . . h6 9 . �e3 fS ! ? 1 O .exfS gxfS 1 1 .°tWd2 f4 1 2..�.f2 tll d 7 1 3 .�e2t; Black's edge knight and backward e-pawn are White's edge; B) 8 .. .fS ! Svetushkin recommends Anand's 8 . .tgs for White, but he does not mention this reply. 9 .°tWd2 f4 1 O .�e2 tll d 7 1 1 .tll a 3 tll e s 1 2 .�h4 h6 1 3 . 0-0 tll f6= - 0 . 0 7 . I agree with Ko­ modo's slight preference for Black here due to the powerful knight on e S . 9.�xf6 9 .�e3 t +0 . 3 3 . Komodo prefers this to the text while Houdini prefers the game move, probably because Komodo gives more weight to good and bad bishops. Both moves should keep some edge I think. This position looks better for White than the position after 8 . .te3 be­ cause the bishop on f6 blocks both the f-pawn and the knight's retreat to f6. 9 ... exf6 9 . . . tll xf6 1 O .°tWd2 eS 1 1 .tll d l t + 0.44. White will place knights on f2 and c3 . 1 0.'Wid2 1 0 .g4!? tll f4 ( 1 0 . . . tll g 7 + 0. 1 1 is not bad) 1 1 .°tWd2 gs 1 2 .h4 tll d 7 1 3 .'iYh2 tll e S 1 4.tll d 2 hS 1 S . gxhS + 0 . S 6 . White will castle queenside and should have a plus, but it's rather messy. 1 O .. .f5 1 1 .exf5

2:

T h i rd M o v e O ffs h o o t s

11 ...�xf5 A) 1 1 . . . .l::!. e 8+ 1 2 .@dl .txfS 1 3 .g4 transposes to the game; B) 1 l . . . 'Wfh4+ 1 2 . @d l ( 1 2 . Wif2 +0.44) 1 2 . . . �xfS ( 1 2 . . . tll g 3? 1 3 .'iYf2 tll xfS 1 4. 'iYxh4 tll x h4 1 S . ttJ b S ±) 1 3 .g4 �xb l 14 . .l:!.xb l tll g 7 ( 1 4 ... tll g 3 ? 1 S .'iVe l +-) 1 S .@c2 tll d 7 ( 1 S . . . fs 1 6.Wt'e l t) 1 6.�e l ! ?

Analysis diagram

+0. 2 9 . This is Shipov's suggestion. The resulting endgame after 1 6 . . . 't!Wxe 1 1 7 .ld.xe 1 =/;\; gives White a small, but riskless edge. 1 2.g4! .l::!. e 8+?! Inaccurate, as it chases the king into a safer place. A) 1 2 . . . .txb l 1 3 .l:f.xb l tll g 7

Analysis diagram

1 4.h4 ( 1 4.�d3 + 0 . l S ) 1 4 . . . h S ( 1 4 . . . tll d 7 l S .hS tt:J e s 1 6.�e2 gxhS 1 7 . gxhS tt:Jfs 1 8 .b4! + 0 . 2 9) 1 s .@d 1 67

S a b o t a g e t h e G riinfeld

tlld 7 l 6.Wc2 tll e S ( 1 6 . . . a6 + 0 . 3 7 ) l 7 .�e2;!; Anand. + 0 . 6 0 . White has an attack for free; B) 1 2 . . . 'f¥h4+ ! ? 1 3 .Wd l �xb l trans­ poses into the l l . . . 'f¥h4+ line. 1 3.'it>d1 �xb1 1 4.t!xb1 'f¥f6? A blunder, Black apparently overlooked or underestimated White's 1 7 th move. A) 1 4 . . . tll f6 I S .W e l t +0.68 ( 1 5 .h4-+) . Again, White has a ' free' at­ tack; B) 1 4 . . . tll g 7 1 5 .h4t + 0 . 5 2 . Although White i s better i n both cases, the game is still far from over. Gelfand wants to avoid this and secure the f4-square for his knight. 1 5.gxh5! With the text-move possible, there was no need to calculate and evaluate the consequences of 1 5 .Wc2 tll f4 1 6.tlle 4 .!:!.xe4 ! ? ( 1 6 . . . 'f¥e5 1 7 .lle l tll d 7 l 8 .�d3 ;!;) l 7 . fxe4 tlld 7g?. 15 ...'f¥xf3+ 1 6.@c2 'f¥xh1 1 7.'f¥f2 Black resigned - hardly premature: A) 1 7 . . . tll c 6 1 8 .dxc60 ( 1 8 .�d3 ?? tll d4+ l 9 . Wd2 'f¥f3 ) White's minor pieces dominate Black's rook : 1 8 . . . 'f¥xc6 1 9 .�g2 1 9 .�d3 i:!.eS 2 0 .l::rfl +- (+ 1 .45) 1 9 ... 'f¥d7 1 9 ... 'f¥c8 2 0 .1::r fl +- ( 2 0.hxg6 hxg6 2 1 ..!:!.fl +-) 2 0 .tll d s +- + 1 . 2 7 . Probably White will emerge with a bishop for two pawns, but it should be an easy win; B) 1 7 . . . tll a 6 and now : 1 8 .�d3 ( 1 8 .a3 !?+-; 1 8 .�h3 +-) 1 8 . . . tll b 4+ l 9 . Wd2 tllxd3 2 0 .Wxd3 +- . (A5 7) Game 2.7 Ortiz Suarez, Isam Reynaldo Perez Mitjans, Orelvis Badalona 20 1 1 (6) 1 .d4 tll f6 2.c4 c5 2 . . . g6 3 .f3 cS 4.dS bS 5 . cxbS a6 would be the move order that is relevant to this 68

book, which would transpose to this game. 3.d5 b5 4.cxb5 a6 5.f3

This f3 line is not so great against the Benko, when Black has not yet commit­ ted to . . . g7-g6. In our move order, it should favor White. 5 ...96?! 5 . . . axb S 6 . e4 'f¥a5 + 7 .�d2 b4= ; 5 . . .e 6 ! = . 6.e4 d 6 I f Black omits this, White has the stan­ dard Samisch maneuver tllh 3 -f2 . 7.tll a 3 7 . tll c 3 �g7 8 . a4 0-0 9 .�e3 e6 (9 . . . tll b d7 1 O.tllh 3;!;) 1 O . dxe6 �xe6 l 1 .tllh 3 ! ? �xh3 1 2 . gxh3 ;!; + 0 . 5 2 is a suggestion of Svetushkin. I prefer the game continuation to avoid the ugly kingside pawns, even though the com­ puter says White is doing well. 7... �g7 8.tll e 2 0-0 9.tll c3

Chapter

9 ...ttJeS 9 . . . e6 1 0 .�c4 ( 1 0 . �e2 ! + 0 . 4 8 ) 1 0 . . . axbS 1 1 .lUaxbS exdS ( 1 1 . . .tl:ibd7! + 0. 1 1 ) 1 2 .ltJxdS tl:ixdS 1 3 .�xdS J:i.a6 1 4.0-0 tl:ic6.

2:

T h i r d M o v e O ffsh o o ts

1 o ...tl:id7 1 1 .0-0 axb5 1 2.tl:iaxb5 An

outpost knight on bS can

© sometimes make the Benko Gambit look like a blundered pawn.

1 2 ...�a6 1 3.a4 �b6 14.�e3 tl:ic7 1 5.�d2 �fb8 1 6.tl:ixc7 �xc7 1 7.tl:ib5 'il'b 7 1 8. .l::!.a c1 ±

Analysis diagram

This was the game Kovalenko­ Zviagintsev, Magnitogorsk 2 0 1 1 , which was eventually drawn. 1 5 .�e3 ! (+ 0 . 7 2 ) 1 5 . . . �xb2 1 6 .J:i.b l �f6 1 7 .tl:ixd6 tl:ib4 1 8 .tl:ixf7 l:!.xf7 1 9 .�xf7 + Wxf7 2 0 .�xd8 �xd8 2 1 .J:i.fc l tl:ic6 2 2 .�xcS .l::!.x a2 2 3 .�b6 �xb6+ 24.J:i.xb6 tl:ie7 25 . .l:!.c7± .

By retaining the outpost piece on bS , White is well on his way to consolidat­ ing the extra pawn. 1 8...tl:ie5 1 9.�e2 c4? 1 9 . . . .l::!. c 8 2 0 .b3 tl:id7 2 1 . l:tc2 �es 2 2 . .l:Ifc l ±. 20.f4 �xb5 21 .axb5 ttJd3 22 . .l:ixc4 tl:ixb2 23.l:tc2 �a3 24. .l::!.fc1 +-

Analysis diagram

White is only up 1/4 pawn, but the ab­ sence of queens favors him, the pin on the 7th rank is very strong, and he has the potential for two connected passed pawns. 1 0.�d3 +0.48.

White has a strong extra passed pawn, the bishop pair, and the black knight is awkward. 24 ... �aa 25.e5 tl:ia4 26 . .l:i.c7 tl:ic5 27.�xc5 dxc5 69

S a b o t a g e t h e Gr ii n f e l d

28.l::r 1 xc5 2 8 . �c4! �f8 2 9 .d6+-. 28...g5 29.�h5? 2 9 .d6 exd6 3 0 .exd6+-. 29...l'ta4? After 2 9 . . . �a l + 3 0 . l'tc l .lhc l + 3 1 .l::rx c l gxf4 3 2 .�xf4 �xdS 3 3 .l:rb l l:rf8 34.�f3 ± the win is by no means certain. 30.�c2?

70

30 ...�xf4? Black should have played 3 0 . . . l:ra l + 3 1 .�d l �as 3 2 .�e2 �a4 3 3 . .l:!.c l l:rxc l 3 4 . .l::!.xc l �xf4 3 5 .l'tc4 �xeS 3 6.�xeS �xeS with equality. 31 .g3 l:!.b4 32.�xf7+ @h8 32 . . . @xf7 3 3 .�fs + @g8 34.�c8++-. 33Jk8+ �f8 34.�e6 l't4xb5 35.�f5 Black resigned.

Chapter 3

The Neo-Griinfeld This is the heart of the book. Black refuses to be forced into a King's Indian or a sideline and plays his intended Griinfeld even though there is no knight on c3 to exchange, believing that White's move 3 .f3 will prove silly or wasted. A heavy struggle is guaranteed. Let's look at the moves one by one. 1 .d4 tt:Jf6 2.c4 g6 3.f3 d5

4.cxd5 White really has no choice here. 4.lLic3 looks reasonable planning 5 .e4, but 4 . . . c 5 ! makes the f3 move look silly. 4...tt:Jxd5 Black can play a gambit here with 4 . . . c6 5 .dxc6 lLixc6 6.e3 e 5 , but after 7 .�b5 his compensation for the pawn is clearly inadequate. You might try this in blitz though. Also favoring White is 5 . e4 cxd5 6 .e5 . See notes to Game 3 . 1 . 5.e4 tt:Jb6 With White's knight on c3 instead of the pawn on f3 , this position is known to be clearly better for White. The black knight is not so well placed on b6 and White is ahead in space and time. But clearly 3 .f3 is not as useful a move as 3 .lLic 3 , mainly because it steals the best square from White's king 's knight. So the line makes sense for both sides. We can state that White is now clearly worse than clearly better ( ! ) but h e may still b e better. 6.tt:Jc3 White can also choose 6 . �e3 �g7 7 .lLic3 , transposing. The text order is usually chosen under the principle of always making first the move you are most sure you want played. 6 ...�g7 7.�e3 71

S a b o t a g e t h e Grii n fe l d

7 0-0 Here we come to the first real branch. Black can postpone castling and play 7 . . . tUc6 , which has the point that it attacks the d4-pawn before White can defend it by castling queenside. White can reply with the direct 8.dS, which is likely to lead to a more pleasant position for White as he can force the exchange of Black's 'pride and joy ' , the g 7 -bishop. Or he can go for more by the pin 8 .�bS , which should lead to a wild line where Black gives up a knight for two pawns and the bishop pair. White should be better, but I sure wouldn't play 8 . .tbs in a blitz game! See Game 3 . I for both options. 8.�d2 Here the alternative is 8.f4, favored by Evgeny Postny recently. Amazingly, no one plays the reply both engines swear is the best and fully satisfactory, namely 8 . . . cS ( ! ) . If Black doesn't play this move, he must thread a very narrow path or risk seri­ ous trouble, but it seems that if he plays perfectly he can force White to settle for perpetual check. This is a very wild line and likely to be very effective against an unprepared opponent, but not against someone who has read this book! See Game 3 . 2 . 8 tt:Jc6 A major alternative is the direct 8 . . . eS , as played by Gelfand in the 2 0 1 2 World Championship match. Up to Black's 1 4th move play is virtually forced, with White advancing and exchanging his h-pawn while Black isolates and attacks White's d-pawn. This is a very complex line where Black often cedes the exchange for a pawn and some positional compensation. The computers insist that White keeps a decent (t) advantage, and I don't disagree, but this is a playable line for Black and one in which full equality may yet be proven. Game 3 . 3 is the Anand-Gelfand game from the last World Championship match, as well as analysis of the probably infe­ rior defense 1 4 . . . �d6 instead of the recommended 1 4 . . . �d7 . Black missed total equality on move 23 but still held the draw. Game 3 .4 shows a somewhat more promising line for White against Gelfand's defense with 1 4 . . . �d7 . White missed multiple chances for at least a small edge on move 2 0 , although in most cases Black reaches an inferior but drawable endgame, though my novelty 2 0 .�e l may keep a real edge. ...

...

72

Chapter

3:

T h e N eo - G r ii n fe l d

9.0-0-0

9 . 'i¥d6 Here there are two major alternatives to the recently hot text move, which intends . . . .l::!. d8 to pressure the d4-pawn. The move 9 . . . es is logical and formerly popular, but now it is almost refuted. After the virtually forced moves 1 O .dS tl:Jd4 l l .f4 , Black can maintain his strong knight by 1 1 . . . cS (or l l . . . �g4 1 2 . .l:le l cS) , but if he allows it to be exchanged off by an eventual tl:Jf3 he will be clearly worse; see Game 3 . S . He can avoid this by artificial means as in Game 3 . 6 , but White can sacrifice the exchange for a dangerous attack combined with threats against the exposed black queen. Although 9 . . . es is the most frequent move in Powerbase 2 0 1 3 , White scores really well ( 64%) , so I would avoid this line as Black. More promising is 9 . . .fS , when 1 0 .h4 appears to lead only to perpetual check. So White normally plays 1 0.eS and answers 1 0 ... tl:Jb4 by l 1 .tl:Jh3 . White should generally not be in a hurry to attack the knight by a2-a3 because this can often be met by . . . a7 -aS , sacrificing the knight. Instead White should give priority to �b 1 , tl:Jf4 or tl:JgS , depending on circumstances, and sometimes h2-h4. Usually White plays a2-a3 only when there is some drawback to the reply . . . a7-aS . This line tends to end up in positions where Black is clearly somewhat worse but has decent draw­ ing chances due to the blocked nature of the position. See Game 3 . 7 . 1 0.�b1 Here the major alternative is 1 O .tl:JbS 'i¥d7 . White can now choose between the dangerous attacking moves 1 1 .f4, l 1 .�h6 , and 1 1 .h4 (or 1 0 .h4, which should transpose) , all played in recent high-level games, or just play 1 1 .�b l which may transpose to the text, though there are pros and cons to this transposition (dis­ cussed in Games 3 . 1 1 and 3 . 1 2) . l l .f4 leads to positions where White has an im­ pressive center for which Black has to demonstrate compensation. According to my analysis in Game 3 . 8 , Black must play very precisely to reach a position which can reasonably be called equal, so I would say that 1 1 .f4 is an excellent practical weapon; it was good enough to score against superstar Peter Svidler in that game, though it is rather double-edged. The alternative 1 l .�h6 is a pawn sacrifice which also appears to demand a very precise path for Black to equalize, one which was . .

73

S a b o t a g e t h e G rii n feld

demonstrated in Svidler-Caruana given in the notes to Game 3. 9. The key defensive idea is to play . . . f7-f6 and . . . e 7 -eS rather than the . . . c7 -cS move of the main game, giving back the extra pawn to equalize. In Game 3 . 1 0 we see another dangerous try with a quick h2-h4, which appears to offer only one never-before recommended way to equalize. So while from a theoretical standpoint these 1 1 th move offshoots are not dangerous, you better have a very good memory if you plan to defend the new 9 . . . '/Wd6 line. 1 0... ladS This was the move intended when Black played 9 . . . 'lWd6 , but there is a serious ar­ gument to insert 1 O . . . a6 now that tbbS is a more serious threat than on the previ­ ous move, though this is rarely played. It does however give White time to start his attack with l l .h4, when 1 1 . . . hS can lead to an ending where Black gets a pawn and a good position for the exchange, with decent holding chances. Houdini rather likes this 1 O . . . a6 move while Komodo does not approve; it comes down to j udging whether or not White has a meaningful plus in this endgame. For me it's a tough call. See the notes to Game 3 . 1 1 . 1 1 .tbb5 '!Wd7 1 2.d5 a6 This is forced, because 1 2 . . . tbeS is met by 1 3 . 'lWc2 ! , winning a pawn. 1 3.tbc3 Instead 1 3 .dxc6 wins a pawn and trades queens, but with precise play Black gets full compensation in terms of superior activity, and the result should be a draw. See the notes to Game 3 . 1 1 . After 1 3 . tbc3 Black has three choices, all discussed in Game 3 . 1 l . It seems to me that White keeps some advantage against all three. The key points to remember are that when Black threatens to fork the queen on d2 and the knight on e3 by . . . tbc4, the reply .id4 is best, that White should not take a knight on c4 with his bishop when it is defended by a knight on b6 (because doing so would permit . . . b7-bS) and that whenever the white queen is attacked, its best retreat is almost always to c 1 . On that square the queen defends all the important points, including the b2-pawn, the c3 -knight, and the e3 -square. Black's main trump should be his dark-squared bishop, but it seems White can always either exchange it or force Black to block the long diagonal by . . . e7-eS or . . . f7-f6. Then his space advantage and potential to attack by h2-h4 and h4-hS usually give him a plus.

£ 'Whenever the white queen is attacked, its best retreat is almost always to c l .'

How should Black play given that he has chosen the Neo-Grunfeld 3 . . . dS ? I recom­ mend following Game 3 . 1 1 note C to move 1 4 (i.e. 1 3 . . . '!We8 1 4.'!Wc l tbaS ) , al­ though White still gets a 'par' advantage. Sorry, chess is not a fair game; White starts out with an edge that should last into the endgame, though it's almost cer­ tainly not enough to win by force. If you don't like these positions, you can agree to transpose to the Samisch King's Indian, relying on the Dzindzi Gambit for equality. In my opinion this is objectively the better choice for Black, although since White keeps a small plus in both cases the choice (between 3 . . . dS and 3 . . . �g7) can be made by personal preference. 74

C h a p t er 3 : T h e Neo - Gr iin fe ] d

Game 3.1 (D7 0 ) Khismatullin, Denis Yandemirov, Valery Voronezh 2006 (8) 1 .d4 lZ:Jf6 2.c4 g6 3.f3 d5 4.cxd5 lZ:Jxd5 4 . . . c6?! is a dubious but not ridiculous gambit.

5.e4 lZ:Jb6 6.lZ:Jc3 .ig7 7...te3 lZ:Jc6

This is fairly rare before castling, but it leads to totally different play. 8.�b5

Analysis diagram

S .e4 (after S .dxc6 lLlxc6 6.e3 eS 7 .�bS exd4 8.�xd4 �xd4 9 . exd4t White's d-pawn is isolated and a bit weak, but it is an extra pawn) S . . . cxdS 6.eS and now: A) 6 . . . lZ:JhS 7 .�d3 lZ:lg7 8 .lZ:lc3 lZ:lc6 9 .lZ:lge2t and White has an advantage in space for free; B) 6 . . . lZ:Jfd7 7 .�d3 lZ:lc6 8 .lLle2t e6 9 .�e3 (Svetushkin says only 9 .h4 hS 1 O .a3 gives an edge. I agree that White is better here, but the text may be even stronger) 9 . . . �b6 1 0.�d2 lZ:lb4. Svetushkin stops here, saying that White loses his light-squared bishop and implying that Black is no worse then. But after 1 l . lZ:lc 1 t, recapturing with the knight will give White the edge despite the loss of his better bishop, as the knight is ideally posted on d3 and his space and development advantages are significant due to the open c-file.

Valery Yandemirov

8 .dS is the usual and safer move. The game move is more ambitious. 8 . . . lZ:JeS 9.f4 lLJg4 1 0 . .ibS + and now: A) 1 O . . . c6 1 l .dxc6 0-0 1 2 .�xd8 �xd8 1 3 .c7 .:af8 1 4 . .ixb6 .ixc3 + 1 S .bxc3 axb6 1 6.lLJf3 t . Black will need to make some positional concessions to regain the pawn; B) 1 O . . . �d7 1 l ...td4t ..txd4 1 2.�xd4 0-0 (see diagram next page)

75

S a b o t a g e t h e Gr iinfeld

( 1 1 .b3 (+0 . 2 2) is the way to wimp out with a tiny edge) l l . . . axb6 l 2 .b4 c6 1 3 ..2.d3 bS 1 4.bxaS b4 l S .tll b l cxdS l 6 .tll d 2 l:l'.xaS 1 7 .tll b 3 dxe4 l 8.fxe4 .l:i.a3 1 9 . 0 - 0 .2.e6 2 0 . tll e c l �h6 2 1 .@h l �b6 22 . .l:i.b l .l::!.c 8 2 3 .�f3t

Analysis diagram

1 3 . .2.xd7 Svetushkin's novelty, though it is the first move most amateurs would think of. 1 3 . . . �xd7 1 4.h3 tll f6 1 S .tll g e2 c6 l 6.a4t + 0 . 3 7 - Svetushkin. 8 ... 0-0 9.tll g e2 Analysis diagram

+0.3 7 . Black has substantial but proba­ bly not full compensation for the piece. Two pawns and the bishop pair for a knight works out to down a pawn, and although Black's pieces are more active, he does have a crippled queenside ma­ jority. 1 0.b3 a6 1 O ... eS 1 1 .dxeS �xeS 1 2 . 0-0 �e7 9 ...tll a 5 A) 9 . . . �d6 1 0 .0-0 �d8 l l ..2.xc6 bxc6 1 2 .�c2 aS 1 3 .l::rfd l (Komada also likes 1 3 .eS �e6 1 4 . .l:i.fe l (+0 . 2 9) , but the rook move looks better to the human eye) l 3 . . . tll c4 1 4 . .2.f2t +0.3 3 . White's much superior pawn structure and center control more than offset the bishop pair; B) 9 . . . �d7 1 0.b3 e6 1 1 . 0-0 �e7 1 2 .�c l t + 0 . 3 7 . With a strong center and ideas of .l:i.d l and .2.h6 or .2.gS White has a nice edge; C) 9 . . . eS ! ? This piece sacrifice is the critical line; otherwise White is just better. Surprisingly Svetushkin does not mention it. 1 O .dS tll a S l l .�xb6 76

Analysis diagram

1 3 . �e l t Svetushkin. 1 3 . f4 �g 7 1 4.�e l is at least equally good. Both lines are +0.48. Black's problem is that if he plays . . . c7-c6, his knight on aS is stranded.

C h a p t er

1 1 .�d3 0ic6 1 2.�c2 e6 1 3.0-0 'iYe7

1 4.�c1 1 4.'iYe l (+0 .48) is Komodo's prefer­ ence, with a nice plus due to center control, development, and ideas of t!.d 1 and 'iYf2. 1 4 ... es 1 5.dS f8 2 2 .tl:ixe8 \tlxe8 2 3 .'iito>d 2 tl:ixc6 24.tl:if3 ;\; . White is up % pawn, but Black's superior pawn struc­ ture was enough to hold the draw in Postny-Rodshtein, Acre ch-ISR 2 0 1 3 ; B) Svetushkin does not mention l 5 .�e2 ! , for example 1 5 . . . cS l 6 .'iVf2 tl:iac4 1 7 .�xc4 tl:ixc4 l 8 .tl:if3

Analysis diagram

White has full compensation for the pawn with his superior minor pieces, but Black should be able to equalize. 2 2 . . . l:Id8 is j ust + 0 . 04. 1 4.. .i::r d s 1 4 . . . �e8 1 5 .b4 tl:iac4 1 6 .�xc4 'iYxb4 1 7 . .ld.c l cS 1 8 .dxc6 'iVxc4? ( 1 8 . . . .ie6! 1 9 .tl:ie2 tl:ixc4 20 . .i::f.b l 'iVaS 2 1 .cxb7 l:i'.ad8 2 2 . .i::rb s .l:!.xd4 (22 . . . 'iVa6 ! 0 . 00) .

Shakhriyar Mamedyarov

Analysis diagram

+ 0 . 2 6 . White should be better after Black regains his pawn, due to his cen­ tral superiority. 1 5.b4 tt:Jac4 1 6 ..ixc4 ?! l 6.tl:if3 ! ? is apparently best though it seems only to draw: 1 6 ... .ig4 1 7 .Lc4 � 1 8 .0-0 �xb4 1 9.e6 fxe6 20.�xf3 'iVxc4 2 l .'iVf6 exdS . White obviously has perpetual check, but no more than that. 1 6 ...'iVxb4 1 7.tl:if3

1 7...'iVxc4 79

S a b o t a g e t h e G rii n fe]d

A good alternative was 1 7 ... cS! 1 8 .a3 �xc4 1 9 .�e3 �g4 2 0 .tll d 2 �d4 -. 1 8 . Black looks for choice here. 1 8.�e3 c6 1 8 . . . .l:!.e8 1 9 . .l::!.d l �g4 2 0.e6 �xf3 2 1 .exf7 + @xf7 2 2 .�xf3 + Wg7 and due to the white king being stuck in the center, the engines and I slightly prefer Black here. 1 9 . .ld.c1 .ig4 20.tll d 2 �b4 21 .0-0 cxd5? This is wrong because it allows an even­ tual tll d 2-e4-f6. 2 l . . . .ld.ac8= . 22.exd5 tt::Jxd5 23.tt::Jxd5 .l:l:xd5 24.tll e4;!; 'li'd4 25.tllf6+ @g7 26.tt::Jxd5 �xd5 27.'£c7 .ie6 28.a3

So White is up % pawn for j ust moder­ ate compensation (Black's king is a bit better protected) . 28 ... b5 29.h4 h5 30 ..ld.f2 .l:rd8 31 .l::!.x a7 �d 1 + 32 ..l:f.f1 �c2 33.'li'f2

80

33 ...�xf2+? In principle a very bad trade for Black, because White's weaker king no longer matters after this, and also because rooks are primarily endgame pieces. Probably Black thought he would win a pawn by force, but it turns out to be an illusion. 3 3 . . . �c3 should probably en­ able Black to draw, but even the move played should probably have drawn. 34..ld.xf2 .l:!.d3 35.Wh2 .l:!.e3 36.l::!. b 2 .ic4 37. .ld.a5 So that 3 7 . . . .l::!.xeS ? can be met by 3 8 .a4. 37... g5 38.hxg5 @g6 39 . .ld.f2 .laxes 40.a4 1:!.xg5 41 ..l::r a 6+ @g7 42.a5 b4 43.1:!.a7 �e6 44..ld.e2 @g6 45.�aa @g7 46 . .l:!.a7 @g6 47..l::!. e4 �b5 48.a6 b3 49 ..l:!.b7 �a5 50.a7 .id5 51 . .l::i. b 6+

51 ...@g7?? A losing blunder. Instead 5 1 . .. �e6 should hold the draw. 52.�e5 .ld.a2+ 53.Wg1 �e6 54.ri.'.b7 @g6 55 . .l:I.eb5 �c4 56.J::r c5 .ie6 57..ld.cc7 �d5 58.J:lb6+ @g7 59Jk5 �e6 60.J::f. b 7 @g6 61 ..l:re5 .l:!.a1 + 62.Wf2 J::l'. a 2+ 63.We3 ld.a4 64.Wd3 I;:!.a1 65.Wc3 l:!.c1 + 66.Wb2 .ld.c2+ 67.Wb1 .l:!.a2 68.Wc1 J::r c 2+ 69.Wb1 .l:!.a2 70..ld.e2 J::l'. a4 71 ..ld.d2 @g5 72 ..ld.dS 1 -0 .ld.a2 73 . .ld.b5+

Chapter

(D7 0) Game 3.3 Anand, Viswanathan Gelfand, Boris Moscow Wch-m 2 0 1 2 (3) 1 .d4 tt:Jf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tt:Jb6 6.tt:Jc3 �g7 7.�e3 0-0 8.'iYd2 e5 9.d5 c6 9 .. .f5 I 0 . 0-0-0t. 1 0.h4

3:

The Neo - G riinfeld

A) l 1 . . .h5 1 2 .a4!N (after 1 2 .g4 hxg4 1 3 .h5 �f5 l 4.hxg6 White is for choice, but this is not very convincing if Black makes the correct recapture l 4 . . . �xg6 instead of 1 4 .. .fxg6 ? as played in Moiseenko-Timofeev, St Vincent tt 2 0 0 5 ) 1 2 . . . �f5 1 3 .a5 tZ'lc8 1 4.tZ'lh3 tZ'ld7 1 5 .tZ'lf2 tZ'lf6 1 6 . .tg5 tZ'ld6 1 7 .g4±

Analysis diagram

1 0...cxd5 1 0 . . . h5 l l .d6 ( I 1 .g4 cxd5 1 2 .exd5 transposes to the next move note)

Boris Gelfand

l l . . .�e6 1 2 . 0-0-0 tZ'l8d7 l 3 .tZ'lh3 ± + 0 . 6 0 . White's strong passer and ideas of tZ'lg5 or g 2 -g4 give him the edge. 1 1 .exd5 tt:J8d7

+ 0 . 6 8 . White's attack is quite effective; B) l l . . . e4 1 2 .fxe4 f5 1 3 .h5 fxe4 l 4.hxg6± Svetushkin. 1 2.h5 tlJf6 1 3.hxg6 fxg6 1 4.0-0-0

1 4...�d7 ! This i s better than l 4 . . . 'iYd6 , when there can follow: A) 1 5 .@bl and now: A l ) 1 5 . . . l:!'.d8 1 6 .@a l tZ'lbxd5 1 7 .tZ'lxd5 'ifxd5 ( 1 7 . . . tt:Jxd5 1 8 .�g5±) l 8 . ..tc4 'ifxc4 l 9 .'ii'x d8+ �f8 2 0 .'ii'd 3 'ii'xd3 2 I ..l::i.xd3 ..tf5 22 . .l:!.dl h5 2 3 .a3 81

S a b o t a g e t h e G r iinfeld

2 4.d6 �xg2 2 5 .l::t h 2 �c6 26.tt:Je2 tt:Jc4 2 7 . .ld.h4 tt:Je3 28 . .l:!.g 1 tt:Jg4 2 9 .c4 b6 3 Q .J::[h 3 ;£.

Analysis diagram

White is up only a quarter pawn by my count, but there is no compensation for it with the queens off and only Black having a weak pawn; A2) 1 5 . . . fif5 + 1 6.'it>a l h5 and now: A2 1 ) 1 7 .tlJh3 .l:!.ac8 1 8 .tt:Jf2 tt:Jc4 1 9 .�xc4 1::!'.xc4 20 . .l:!.c1 (also reasonable is 2 0 .fi..h 6 �d4 2 l .Vi'g5 tt:Jxd5 2 2 .�xg7 'it>xg 7 23 . .ld.xd4 exd4 24.tt:Jxd5 Vi'xd5 25 . .l:!.xh5) 20 . . . .ld.fc8 2 1 .�h6 .ld.d4 2 2 .'ifg5 l:!.xc3 23 . .l::!.xc3 tt:Jxd5 24.tlJe4 �xh6 2 5 . Vi'xh6 Vi'f8 2 6 . Vi'xf8 + Wxf8 2 7 . .l:!.cc l We7 2 8 . .l:!.he l We6 2 9 .g3;!;.

Analysis diagram

White is up % pawn, but Black's con­ nected passers give him good drawing chances. I think White's winning chances are better with this line than with the previous one because this vari­ ation is more unbalanced. B) An alternative is 1 5 .tlJh3 fixh3 1 6 . .l::!.xh3 .l::!. ac8 1 7 .l:ih4 Vi'xd5 1 8 .Wb l 'i¥xd2 l 9 . .l:!.xd2 .l::!. fe8 ! 20.�b5 .l:!.e6 2 I .fi..d 3 fi..f8 2 2 .fi..g 5 'it>g7 2 3 .tt:Je4 fie? 2 4.�h6+ 'it>h8 2 5 . .l:!.h l ;t

Analysis diagram

White is up % pawn, but Black's well-centralized pieces offer reasonable compensation for this though White is still for choice; A 2 2 ) 1 7 .fi.. d 3 e4 1 8 . fxe4 tt:Jxe4 l 9.�xe4 �xc3 2 0 .Vi'xc3 �xe4 2 l .fi..c 5 Vi'f6 2 2 .�xf8 Vi'xc3 2 3 .bxc3 I:lxf8 82

Analysis diagram

+0. 2 6 . White is down half a pawn, but he has the better pawn structure and a better piece positioning. In human play I think Black would have a tough time. 1 5.Wb1 I;!.c8

Chapter

1 6.@a1 l 6 . d 6 , transposing to Grischuk­ Dominguez Perez, is probably best but a tad risky. An interesting but probably only draw­ ing try is 1 6.tll h 3 e4 1 7 .tll g S exf3 1 8 . gxf3 �fs + 1 9 .@a l tll fxd S 2 0 .tll x dS .&!.c2 2 1 .itWb4 tll x dS 2 2 .�c4 .l::i. x b2 2 3 . itWxb2 �xb 2 + 2 4.@xb2 itWf6 + 2 5 .�d4 itWxgS 2 6 .�xdS + .i::i.f7 2 7 . �xf7 + @xf7 2 8 . .&!.xh 7 + @e6 2 9 . .&!.xb7 �g2 + .

3:

T h e Neo - G rii n f e l d

Analysis diagram

l 8 .tll x a4! (this improves on l 8 .tll g e2 as in Feller-Salgado Lopez, France tt 2 0 1 2 , when 1 8 . . . itWaS fully equalizes) 1 8 . . . �xa4 1 9 . .&!.cl .!:!.xc l + 2 0 .'iYxc l t + 0 . 2 2 . White protects his d6-pawn by the threat of forking on c4. 1 7...tll a4 1 7 . . . 'i¥c7 1 8 .�e2 tll a4 1 9 .tll h 3 itWg3 2 0 .tll x a4 �xa4 2 1 ..l:!.c l .l::!. x c l + 2 2 . .&!.xc l �d7 2 3 .tll g S exf3 24.gxf3 �h6 2 5 .f4 �xgS 2 6.fxgS tll e 4 2 7 .°iYd3 'i¥xd3 2 8 .�xd3 �fs 2 9 .�xe4 �xe4 3 0.d6t + 0 . 1 S . White's advanced passer is clearly worth more than Black's back­ ward h-pawn, but the bishops of oppo­ site color make it drawish.

Analysis diagram

While White's two rooks should be su­ perior to Black's queen, it's hard to imagine any result other than a draw here with only one white pawn remain­ ing and bishops of opposite color. 1 6 ...e4 l 6 . . . tll a4 1 7 .tll e 4 tll xe4 l 8.fxe4 tll c S 1 9 .d6t +0.44. 1 7.�d4 l 7 .d6 tll a4

1 8.tll g e2 l 8 . tll x e4! tll x e4 l 9 . fxe4 �xd4 2 0.\lfxd4 itWgs 2 1 .@b l .!:!.xfl 2 2 .�xfl itWxg2 2 3 .tll e 2 itWxe2 24.@a l .!:!.c4 2 5 .Wff6 itWxb 2 + 2 6 .Wk'xb2 tll x b2 2 7 .@xb2 laxe4 2 8 . .&!.e l .!:!.xe l 2 9 . .l::!.xe l 83

S a b o t a g e t h e G rii n f e l d

@£7 30 . .l:!.cl @e7 3 1 ..l:!.c7 hS 32 . .ld.xb7 h4 3 3 .@c3 h3 3 4.@d4 @d6 3 5 .@e3 �fS 3 6 . .l:!.xa7 @xd5 3 7 .Wf3 .

Analysis diagram

White has the sunny side of a drawn ending. The computer's + 0 . 64 score has no real meaning here, as Black eas­ ily holds. 1 8...�aS 1 9.tt::J xe4 1 9 .a3 (+0 .0 7 ) is a try. 1 9...�xd2 20.tll xf6+ J::!.xf6 21 .J:Ixd2 J::!.fS 22.�xg7 @xg7 23.d6

23 ....l:!.fcS? 2 3 . . . tll b 6 24.tll c 3 .l:!.dS = ; Black exploits the weak back rank to regain his pawn. The evaluation is 0 . 0 0 . 24.J:Id1 a s 2S.J:Ih4 l::r c 2 26.b3 tll b 2 27..l:!.b1 tll d 3 28.tll d 4 .a:d2 29.]lxd3 .ld.xd3 30.l::r e 1 .a:d2 31 .@b1 ]lfS+ 32.tll xfS+ gxfS 33 ..l:!.e7+ @g6 84

34. .l::i.c 7 White had two better winning tries, though since this is not an endgame book I won't try to determine if either actually wins : 34.d7 .ld.cc2 3 5 . .ld.c4;l;; or 34.l::r e 6+;!;. 34....i::r e a 3S. .ld.h1 .l::te e2 36.d7 l::r b 2+ 37.@c1 .ld.xa2 112-112 (D7 0) Game 3.4 Grischuk, Alexander Dominguez Perez, Leinier Thessaloniki 20 1 3 ( ! 0.4) 1 .d4 tllf6 2.c4 g6 3.f3 dS 4.cxdS tt::J x dS S.e4 tt::l b 6 6.tll c3 ]lg7 7.]le3 0-0 8.'iWd2 es 9.dS c6

1 0.h4 Svetushkin recommends the rare 1 O . .l:!.dl because he feels that this game shows how Black can equalize against 1 O .h4. Although I don't agree about 1 O.h4 being neutralized, I do agree that 1 O . .l:!.dl is also a strong move. Much of

C h ap t e r

the following analysis is from Svetushkin, who deserves credit for revitalizing this line. As to whether 1 0 . .t!.d1 or 1 0 .h4 is better, I would say that they are of comparable objective merit, both sufficient for advantage, but obviously 1 O . .!'.;ldl being a rare move will have more surprise value. A) 1 O . . . .l::!. e 8 l 1 .�d3 cxd5 1 2 .exd5 tt:lc4 1 3 .�xc4 'iVh4+ 1 4.�f2 'iYxc4 1 5 .tt:lge2 tt:ld7 1 6.0-0 b6 1 7 .b3 'ifb4 l 8 .a3 (this is only possible because tt:lb5 -c7 will fork the rooks) 1 8 . . . 'iYxb3 1 9 .�b l 'iYxa3 2 0.tt:lb5 'iff8 2 1 .tt:lc7 +0.4 1 . Black will have two pawns and the bishop pair for the exchange, which is more than enough, but the open c-file for White's rooks together with the passed pawn give him the edge; B) 1 O . . . cxd5 1 1 .exd5 and now:

3:

T h e Neo - G rii n fe l d

3. tt:lh3 ! ! . This surprising pawn sacri­ fice is what makes 1 0 . .l::!. d 1 a good move. Congratulations to Svetushkin on a very important find here. 1 3 . . . tt:lc4 1 4.�xc4 'iYxh4+ 1 5 .tt:lf2 �xc4 1 6.g4 ( 1 6 . �h6 �xh6 1 7 .'iYxh6 'iVf4 1 8 .'iVxf4 exf4 l 9 . .t'!.h4 �d7 2 0 . .l::!. xf4t + 0 .48 - the passed pawn in the endgame should give White some plus) 1 6 . . . f5 ( 1 6 . . . e4 1 7 .fxe4 tt:lc5 1 8 . gxh5 +-) 1 7 . b3 'iYb4 1 8 . ti:ld3 'iYd6 1 9 . gxh5 f4 2 0 .�g l ! �f5 2 i .tt:if2 t 1

Analysis diagram

Analysis diagram

B l ) 1 1 . . .tt:la6 1 2 .h4 h5

Analysis diagram

+ 0 . 5 6 . The combination of edge attack with a knight outpost on e4 gives White a serious edge; B2) 1 1 . . . tt:lc4 1 2 .�xc4 'iVh4+ 1 3 .�f2 'iVxc4 1 4.tt:lge2 tt:ld7 1 5 .d6 ( 1 5 . 0-0 b6 1 6.tt:le4 �a6 1 7 .l:!.fe l t +0.3 7 . The passed center pawn is well supported here, and Black lacks enough counterplay to offset it) 1 5 . . . tt:lf6 1 6.b3 'iYb4 1 7 . 0-0t �d7 1 8 .'i!Ye3 .l::!. fe8 1 9 .tt:lg3 l:!.ac8 2 0.tt:lge4 tt:lxe4 2 1 .tt:lxe4 b6 2 2 .'i!Yd3 .l:!.f8 2 3 . �fe l �b5 24.°iVb l 'i:Va3 2 5 .�h4 f6 2 6.h3 t , White's ad­ vanced passed pawn and more active pieces give him the edge; B3) 1 l . . .ti:l8d7 1 2 .tt:lh3 e4 ( 1 2 . . . tt:lf6 1 3 .tt:lf2t +0.68) 1 3 .fxe4 tt:le5 1 4.tt:lf2 tt:lbc4 1 5 .�xc4 tt:lxc4 1 6. 'i!Ye2 tt:lxe3 1 7 .'i¥xe3 .l::!. e 8t 85

S a b o t a g e t h e G riinfeld

+ 0 . 64. Although White's king is not as snug as Black's, his powerful extra passed pawn is a big deal; BS) The move 1 l . . .l::l e 8 shows as rela­ tively best on my tree. The idea is to avoid blocking the c8-h3 diagonal for a move to deter tt:lh3 while still making a useful move. Svetushkin does not men­ tion it. 1 2 . .id3 tt:l8d7 (for 1 2 . . . tt:lc4 see 1 0 . . . l:te8) 1 3 .tt:lge2 fs 1 4.0-0 tt:lf6 1 s .�c2 @h8 1 6 . .ibs �d7 1 7 .�b3t .

Analysis diagram

+0.48. Black doesn't have much to show for the half-pawn deficit;

Analysis diagram

+ 0. 3 7 . Again the passed pawn is White's trump. 1 o ... cxd5 1 1 .exd5 tt:\8d7 1 2.h5 tt:lf6 1 3.hxgG fxg6 14.0-0-0 �d7 Alexander Grischuk

B4) 1 1 . . .�fs 1 2 .g4 .ic8 1 3 .h4 e4 1 4 . .id4t e3 1 S . �xe3 .l:!.e8 1 6.@f2t

Analysis diagram

86

Svetushkin claims this equalizes for Black, but based on my notes to move 2 0 I disagree. 1 5.dG 1 5 .@b 1 .!:tc8 1 6.d6 just transposes to the game.

C h a p t e r 3 : T h e Neo - G riinfeld

1 5 ... l:i.cS 1 6.@b1 e4! A) 1 6 ... ltfs + 1 7 .@a l e4 (Postny­ Greenfeld, Tel Aviv 2 0 0 1 ) and now:

A 1 ) 1 8. tLlh3 exf3 1 9 . gxf3 tt:Jhs 2 0 .tLlgS .l:f.xc3 2 1 .bxc3 'iYxd6 2 2 .ltd4t + 0 . 4 8 . Black shouldn't have quite enough for the exchange after this move; Al) A good alternative is 1 8 .lth6 �xh6 (Kortchnoi favors the exchange sacrifice 1 8 . . . ilh8 , but Komodo says it is totally unsound here) 1 9 .'i¥xh6 .l:f.f7 20.tLlh3 exf3 2 1 .gxf3 l::rd 7 2 2 .tLlgS .l:f.xc3 2 3 .bxc3 'iYc8 24.°iYh2t + 0 . 5 2 . Black will have less than full compensa­ tion for the exchange; A3) 1 8 . .l:!h4 tbe8 1 9.l:!.f4t +0.48 is Svetushkin's preference. B) 1 6 . . . tLlc4 1 7 .ltxc4+ .l:f.xc4 1 8 .tt:Jge2 ltfs + 1 9 .@al lte6 (P.H. Nielsen-Tazbir, Helsingor 2 0 0 9 ) 2 0.lth6±

+ 1 . 1 2 . Here White has both the safer king and a strong passer; C) 1 6 . . . lte6 1 7 .tLlh3 tLlbdS 1 8.tLlxdS ltxdS 1 9 .ilgs ile6 20.ltxf6 + 0 . 8 0 . 1 7.fxe4 tt:Jg4 1 8.ltgS 'l!Ve8 1 9.tt:Jf3

1 9 ... .l:f.f7 1 9 . . . .l:f.xc3 20.bxc3 'iYxe4+ 2 1 .ltd3 'iYc6 2 2 .ltcl ltxc3 2 3 . 'iYel !H7 24.tLld4 �cs 2 5 .tt:Je6 �b4+ 2 6 .ltb3 �bs 2 7 .d7 tt:Jxd7 2 8 .'iYcl ltc4 2 9 .�dl ltxb3 3 0 .axb3 �xd2 3 1 ..1::!'.x dl tt:Jgf6 3 2 .J.:ld4 �bs 33 . .1::!'.e l ;i; + 0 . 6 8 . While White i s actually a quarter pawn down, his superbly posted knight and very active major pieces give him the edge. 20.ile7 A) 2 0 .�d3 tt:Jes 2 1 .a'.c l ! .

Analysis diagram

Analysis diagram

Minimizing any danger from . . . .!::!'.x c3 later. I think this novelty, not mentioned 87

S a b o t a g e t h e G r iinfeld

by Svetushkin, is significant and casts some doubt on the 8 . . . es line. 2 1 . . .tll bc4 (2 1 . . .'it'e6! 2 2 .tll x eS .itxeS +0. 1 8 is a better line for Black) 2 2 .�xc4 tll xc4 2 3 .'it'd3 bS 24.�h6 �f6 2 5 . I:the l l'.:tc6 2 6 . 'it'e2 .ixc3 2 7 ..l:hc3 tll x d6 2 8 .'tWd3 t.

most the same pawns remaining but fewer pieces; C) 2 0.'it'e l ! is a computer novelty: 2 0 ... .ixc3 2 1 .bxc3 tll a4 2 2 . .l:!.c l tll e S (22 ... �e6 2 3 .Wa l tll c S 2 4 . .l:!.c2;!;

Analysis diagram

Analysis diagram

White has the safer king, the more effective bishop, and a passed central pawn; B) 2 0 . .ld. e l tll e S 2 1 . �h 6 tll b c4 2 2 .�xc4 tll xc4 2 3 .'it'c l .iteS 24 . .l:l.hfl tll x d6 2 5 .tll x eS 'it'xeS 2 6.'it'e3 .l:!.xfl 2 7 . .l:!.xfl .itc6 2 8 .Wa l tll f7 2 9 . �f4 'it'f6 3 0 . .ld.e l .l:l.e8 3 1 .eS 'it'e6 3 2 .tll e 4 �xe4 3 3 .'it'xe4 b6 3 4.'i¥a4 ld.e7 3 5 .'it'c2t .

+ 0 . 7 2 . Svetushkin doubts that White 'has anything substantial here ' , but Komada gives White quite a serious plus. With both sides having weak kings, I'd certainly prefer to play White with his 1 . 5 pawn material plus) 2 3 .tll d 4 tll c S 2 4. 'it'h4 tll a 4 2 5 .Wa l 1hc3 2 6 . .l:!.xc3 tll x c3 2 7 .�d2 tll a4 2 8 .�e2;!;

Analysis diagram

Analysis diagram

White is for choice due to the passed pawn, although Black should be able to hold. White's chances look better here than in the previous line, which had al88

+ 0 . 5 6 . Although White's king is less well guarded than Black's, Black is miss­ ing the bishop that would guard the weak squares around his king. White's bishop pair and passed pawns give him the advantage.

Chapter

20 ..Jhc3! 21 .bxc3 �ca 22 ..l:l'.c1 tt:Ja4 23.tlJgS �cs

24.tt:Jxf7 A good alternative is 24 . .l:th3 ! .l:tf2. 2 5 .�d5 + �xd5 2 6 . exd5 tlJxc 3 + 2 7 . .l:l'.cxc3 .l:txfl + 2 8 . �c2 �xc3 29 . .l:l'.xc3 h6 3 0 . tlJe4 +0. 1 1 . White's ad­ vanced passers, though doubled, are more dangerous than Black's split extra pawns, but the bishops of opposite color make a draw quite likely. 24 ...tt:Jxc3+ 25 ..l:txc3 �xc3 26.tlJhG+ �g7 27.�dS �b4+ 28.�b3 �xe4+ 29.�c2 �e1 + 30.�c1 �e4+ 31 .�c2 �e1 + 32.�c1 �e4+ 33.�c2

3:

T h e Neo - G r ii n fe l d

(D7 0) Game 3.S Leitao, Rafael Mekhitarian, Krikor Sao Paulo 2008 (2) 1 .d4 tlJf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tlJb6 6.tlJc3 .tg7 7.ile3 0-0 8.�d2 tt:Jc6 9.0-0-0 9 . .l:td1 ? ! is an unsuccessful attempt to bring back this old, discredited line : 9 . . . e5 1 0 .d5 tlJd4 l 1 .tLlb5 f5 1 2 .tLlxd4 exd4 1 3 .ilxd4 .txd4 1 4.�xd4 fxe4 l 5 .fxe4

Analysis diagram

l 5 . . . kg4 (after l 5 . . . c5 ! = ( 0 . 0 0) l 6.dxc6 �h4+ 1 7 .g3 �e7 Black will have tremendous compensation for the material due to White's problems with his king) l 6 . ..te2 This was ultimately drawn in Gunina-Lagno, Beijing 2 0 1 3 . Instead l 6.lld2 shows as slightly better at + 0. 1 1 , but it's rather moot in view of l 5 . . . c5. 9 ...es 1 O.ds tt:Jd4 1 1 .f4

Neither side can profitably avoid the perpetual check. Personally, I think perpetual check should be forbidden, as it is in the Asian versions of chess. There would be a lot fewer draws! 89

S a b o ta g e t h e G r ii n fe l d

1 1 ...cS 1 2.fxeS �g4 1 2 . . . �xeS 1 3 .tll f3 tll xf3 1 4.gxf3 c4 1 5 .h4 h5 1 6.�d4t. 1 3.�e1 �xe5 1 4.h3 �d7 For the alternative 1 4 . . . 'iih 4, see next game. 1 5.tllf3 tllxf3 1 6.gxf3

1 6 .. .fS A) 1 6 . . . 'iih4 (just to stop h4-h5) 1 7 .'itib l .l::r fe8 1 8 .�gs 'iig 3 1 9 .�e2± + 0 . 7 2 . Black can only extricate his queen without losing material by giv­ ing up his valuable dark-squared bishop for the knight on c3 ; B) 1 6 . . . !'.!.e8 1 7 .�d3 ( 1 7 .h4t) 1 7 . . . .l::!. c 8 1 8 .'itib l 'iih4 1 9 .f4:t. Here too Black must give up the vital bishop for a knight to avoid even worse; C) 1 6 . . . 'iie 7 1 7 .h4 and now: C l ) 1 7 . . .fs 1 8 .�gS °iYd6 1 9 .hS fxe4 2 0 .fxe4 �g3 2 l .hxg6 hxg6 2 2 .�e2 ! ±

+ 1 . 0 7. White gets a strong attack and the bishop pair and has connected passed pawns in return for the sacri­ ficed exchange; C2) 1 7 . . . c4 1 8 .�d4 tll a4 1 9 .�xeS 'ii x eS 2 0 . f4 'ii c 7 (I. Sokolov­ Krasenkow, Wijk aan Zee 2 0 0 2 ) 2 1 .tll xa4 �xa4 2 2 .hs+-; C3) 1 7 ... hS 1 8 .'itib l fS 1 9 .!'.!.g l t. 1 7.�xcS A) Perhaps better, preferring attack to material, is 1 7 .h4 fxe4 1 8 .fxe4 �g4 1 9.�e2 �f3 2 0 .�xf3 .l::rxf3 2 1 .'iie 2 °iYf6 2 2 .�gS �fl 2 3 .hS �f8 (23 ... gxhS 24.�d2:t + 0.44. White's edge is two connected passers vs. doubled isolated h-pawns) 24.hxg6 hxg6 2 5 .tll d l ±

Analysis diagram

+ 0 . 9 3 . White has two connected pass­ ers and a safer king; B) 1 7 .�gs 'iic 7 1 8 .h4t is yet an­ other route to an advantage. 1 7.. .fxe4 1 8.�xfS 'iixf8 1 9 ..llxe4 �g7 20.�d3 �cs 21 . ..tib1 �ts

.1 •• ,• ,•

[l] j. ,

. . • 8 •.t •• :g: . • . �·• 8 ·• 8 1 --::- �. """"" -

Analysis diagram

90

:g:

Chapter

3:

T h e Neo - G r iin fe l d

The game was drawn here. The following moves are a plausible 2 2 . l:the l continuation : Wh8 (22 . . . �xe4 2 3 .tt:lxe4 'i'xf3 24.d6±; the dangerous passed pawn gives White the advantage) 2 3 . .ld.h4 (with 2 3 . d 6 ! ? �xc3 2 4 . bxc3 .ixe4 2 5 .�xe4 'i'f6 2 6 .'i'd4 'i'xd4 2 7 . cxd4 J::!. d 8 28 . .ixb 7 .ld.xd6 2 9 . :i'.e4 White stays a pawn ahead, but with all his pawns isolated the advantage is small) 2 3 . . . tt:lc4 2 4 . .ld.xc4 ld.xc4 2 5 .'i'c2 �xd3 2 6 .'i'xd3 ld.d4 2 7 V&' b s ( 2 7 . 'i' c 2 'i'xf3 2 8 . l:! e 8 + �f8 2 9 .'i'e2

1 S ... l:tac8 A) 1 5 . . . �d7 1 6.tt:lf3 tt:lxf3 1 7 .gxf3 fS 1 8 .�gS �g3 1 9 .f4 �xc3 2 0 .bxc3 fxe4 2 l . .l::!. e 3 'i'xe3 2 2 .'i'xe3 exd3 2 3 .'i'xd3 .l::!. ae8 24.d6 .ld.e6.

Analysis diagram

Analysis diagram

.

White retains a small plus due to the passed pawn) 2 7 . . . 'i'fs + 2 8 .Wa l hS 2 9 .'i'xb7 +0.44. Maybe White can emerge a pawn ahead.

Game 3.6 (D70) Postny, Evgeny Greenfeld, Alon Givataim 2000 (3) 1 .d4 tt:lf6 2.c4 g6 3.f3 dS 4.cxdS tt:JxdS S.e4 tt:lb6 6.tt:lc3 �g7 7.�e3 0-0 8.'i'd2 tt:Jc6 9.0-0-0 es 1 0.ds tt:Jd4 1 1 .f4 �g4 1 2 . .ld.e1 cs 1 3.fxeS ilxeS 14.h3 'i'h4 1 S.�d3 1 S . Wb 1 ( + 0 . 3 3 ) is an alternative.

Evgeny Postny

91

Sabotage the

Griinfeld

So far Taleb-T.L. Petrosian, Dubai 2 0 0 6 , which was eventually drawn. Now 2 5 .h4 (+0. 80) would have left Black with very little to show for the queen vs rook and knight material disadvantage; B) 1 5 .. J:He8 1 6.�gS 'lWhS So far Dzindzichashvili-Kudrin, Denver ch­ USA 1 9 9 8 , which was eventually drawn. There can follow: 1 7 . 'iit> b 1 (ideas include llJbS and l::ic l ) 1 7 . . . �d7 1 8 .J::!. f l fS 1 9 .llJf3 �g7 2 0 .d6±.

1 6.'iit> b 1 Another option is 1 6 .�gS 'lWhS 1 7 . 'iit>b l �d7 1 8 .llJge2 ( 1 8 . l:rfl t) 1 8 . . . f6 1 9 .�e3 fS 2 0 .l::rhfl llJxe2 2 l .�xe2 'lWh4 2 2 .�f4 .l:!.ce8 (in case of 22 ... �xc3 2 3 .'lWxc3 fxe4 24.l::i d l ± White has a solid bishop pair advantage plus the more dangerous passed pawn. His bishop is defended by the g2-g3 fork) 2 3 . �xeS J:!.xeS 2 4.g4 (24 . .l:!.f4t + 0 . 2 6) 24 .. .f4 2 5 . l:rd l t

Analysis diagram

Analysis diagram

White has a dangerous passer, kingside threats, and Black's queen is a bit un­ comfortable; C) 1 5 . . . �g3 1 6 . J:!.fl fS 1 7 . hxg4 'lWxh l 1 8 .llJh3 t; D) 1 5 . . .fS 1 6.�gS 'lWhS 1 7 .exfS �g3 1 8 .hxg4 'lWxh l 1 9 .J::!. d l J:!.ae8 2 0.llJbS± (Svetushkin)

+0. 1 8 . The connected passed center pawns, even though blockaded, are an asset. 1 6 fS A) 1 6 . . . �d7 1 7 .tt:Jf3 tt:Jxf3 1 8 . gxf3 t + 0 . 64; B) 1 6 .. .f6 1 7 .J:!.fl �hS 1 8 .tt:Jf3 �xf3 1 9 .gxf3 c4 20.�xd4 �xd4 2 1 .�c2 �cs 2 2 .f4 I:rcd8 2 3 .'lWg2± .. .

Analysis diagram

+ 1 .2 2 . 92

Analysis diagram

Chapter

+ 0 . 64. The mobile center pawns and kingside attack give White a clear edge; C) l 6 .. JHe8 1 7 .�f2 'iVh5 l 8 .tll b 5 tll xb5 l 9.�xb5 �d6 2 0.�xe8 ld.xe8 2 1 .a l +- .

Analysis diagram

Black has little compensation for the ex­ change. 1 7.�g5 �h5 1 8.hxg4 'iVxh1 1 9.exf5 'iVh2 20.tll h 3 2 0 .fxg6 ! hxg6 2 1 .tll h 3 'iVg3 2 2 .d6± +0.80.

20..J:keS?? 2 0 . . . c4 2 l .�e4 gxf5 2 2 .gxf5 tll xf5 2 3 .'iVc2 'iVg3 24 . .l::i.h l �xc3 2 5 . �xf5 ld.xf5 2 6.'iVxf5 'iVe5 2 7 .'iVc2 �d4 2 8 .�c l ld.f8 ( 2 8 . . . 'iVh5 +0 .48) 29 .lld l + 0 . 6 8 . White supports the passer while aiming for tll g 5 . 21 .fxg6 hxg6 22.�xg6 ( + 1 . 5 5 ) 2 2 ...'ii' g 3 23.tll e4+'ii'xg4 24.�xeS llf1 ? 24 . . . ld.xe8 2 5 .'ii'd 3 +- .

3:

T h e Neo - G rii n f e l d

25.tll hf2 'ii'x g2 26 ..l:!.xf1 'ii'xf1 + 27.'ii' d 1 'ii'c4?

28.'ii' d 3 2 8 .�f7 + ! xf7 2 9 . tll d 6+ �xd6 3 0 . 'ii'h 5 + leads to mate in 1 5 . 28...'ii' x d5? 29.tll c3 2 9 .tll g 4! wins quickly. 29...'ii'e 6 30.'ii' g 6+ 3 0 .�g6 should lead to a quicker win. 30...'ii'x g6+ 31 .�xg6 tll f3 32.�c1 tt:Jc4 33.�d3 tll d 6 34.tt:Jce4 tll e 1 35.�f1 tt:Jxe4 36.tt:Jxe4 b6 37.tll g 5 g7 38.�d2 @96 39.'.tc1

1 -0

(D70)

Game 3.7

Zhao Xue Laylo, Darwin Moscow Aeroflot B 2 0 I I ( S) 1 .d4 tllf6 2.c4 g6 3.f3 d5 4.cxd5 tt:Jxd5 5.e4 tll b 6 6.tll c3 �g7 7.�e3 0-0 8.'iVd2 tt:Jc6 9.0-0-0 f5 93

S a b o t a g e t h e G r ii n fe l d

@f8 25 . .ld.fl + �f6 2 6.�h8+ @f7 2 7 .�h7+ @f8 with a draw by perpet­ ual check; B) 1 2 .d S ? ! tl'ieS 1 3 . �h 6 tl'iec4 1 4.�gS .ld.f7

1 0.e5! 1 O.h4 is a very interesting line, but it appears to burn out to perpetual check: 1 O . . . fxe4 1 1 .hS gxhS Analysis diagram

Analysis diagram

A) 1 2 . .l:!.xhS �fs 1 3 . .l::! g S �g6 1 4.�e2 e5 1 5 . dS exf3 ( 1 5 . . . tl'id4 1 6 .fxe4;t) 1 6 .tl'ixf3 tl'id4 1 7 .tl'ixeS tl'ixe2+ 1 8 .�xe2 .l::! e 8 1 9 .�g4 �xe5

-0.4 1 . White scored only 2 out of 6 from here in grandmaster level games and the computers say White's initiative isn't worth the material given up. Black aims to play . . . �d6. 1 0...tl'ib4 1 O . . . aS 1 l .h4 tl'ib4 1 2 .hS �e6 1 3 .hxg6 hxg6 1 4.�h6 f4 1 5 .tl'ige2 (+0 . 7 2) 1 5 . . . tl'i6d5 1 6.a3 �d7 1 7 .�xf4 �c6 l 8 .�h6 .

Analysis diagram

Analysis diagram

2 0 . .l::!. xg 6 + hxg6 2 1 . �xg 6 + � g 7 2 2 .�d4 �e7 2 3 . .l::!.h l tl'i d 7 24.�h7+ 94

Black has nothing to show for the lost pawn. 1 1 .tl'ih3 �e6 A) 1 1 . . .aS 1 2 .@b l c6 1 3 .ltc l �e6 1 4.tl'if4 �c4 ( 1 4 . . . �d7 1 5 .tl'ixe6 �xe6

Chapter

3:

T h e N e o - G r ii n fe l d

l 6.f4 + 0 . 3 3 . White has the bishops and attacking potential by g2-g4 or h2-h4) l 5 .h4 'i¥d7

Analysis diagram

1 6.a3;\; ( 1 6.d5 jlxfl 1 7 .jlxb6 �c4 l 8.'i¥d4 cxd5 1 9 .:i:thd l is equally strong) 1 6 . . . .txfl 1 7 . l:thxfl 4.Jc4 l 8 .�e2 b5 1 9 .b3 4.Jxe3 2 0 .�xe3 4.Ja6 2 1 .d5

Zhao Xue

1 2.@b1 '!Wd7 A) 1 2 . . .4.J4d5 1 3 . 4.Jg5 4.Jxc 3 + 1 4.°iYxc3 .td5 1 5 .h4 °iYd7 1 6 .i.d3 .l::!. ac8 ( 1 6 . . . a5 1 7 .'i¥c5 ;\;) 1 7.�c2 4.Jc4 (Ivanisevic-Gabrielian, Plovdiv Ech 2 0 1 2) 1 8.4.Jh3 ! ±

Analysis diagram

+0. 5 2 . White's advantage is becoming clear. He has a large space advantage and kingside attack potential; B) l l . . .4.J4d5 1 2 .4.Jf4 ( 1 2 .@b l (+ 0 . 3 7 ) may be even stronger) 1 2 . . . 4.Jxf4 1 3 .�xf4 c6 1 4.�h6;\; .txh6 1 5 .°iYxh6 @h8 1 6 .h4 l:tg8 1 7 .@b l (+0 . 2 2) 1 7 . . . °iYf8 1 8 .°iYg5 and White's kingside initiative gives him an obvious plus; C) l l . . .c6 1 2 .4.Jf4 4.J4d5 l 3 . 4.Jcxd5 4.Jxd5 1 4.�c4 e6 1 5 .@b l t.

Analysis diagram

+ 0 . 7 6 . White aims to play 4.Jf4 next. If both knights are traded for bishops, White's light-squared bishop will be much more active than Black's dark­ squared bishop; B) 1 2 . . . c6 1 3 .4.Jf4 'i¥d7 1 4.h4 4.Jc4 1 5 . .txc4 �xc4 1 6.b3 i.f7 1 7 .4.Ja4! 4.Ja6. This was my game as White vs. World under- 1 4 Champion Kayden Troff, Concord 2 0 1 3 . 95

S a b o t a g e t h e G r iin fe l d

1 3 �f7 A) 1 3 . . . �c4 1 4.h4 ( 1 4.�xc4+ tl:lxc4 1 5 .�e2 b5 1 6 .�c l ± (+0. 7 6) looks even better than the game cited) 1 4 . . . �xfl 1 5 . .l::!.hxfl .l::!.a d8 1 6.d5 �xe5 1 7 .�xb6 axb6 1 8 .tl:le6 �xe6 1 9.dxe6 .l::!.xd2 20 . .l::!.x d2 @g 7 ...

Analysis diagram

My l 8.h5 made things unclear. Better is 1 8.�e l !± +0.93 , avoiding tactics in­ volving . . .�xe5 in some lines and leaving White with space and a kingside attack; C) 1 2 . . . tl:lc4 l 3 .�xc4 �xc4 1 4.b3 �fl 1 5 .�h6 tt:Jd5 1 6.�xg7 @xg 7 1 7 .tl:lxd5 �xd5 l 8 .tl:lf4 c6 1 9 .h4±.

Analysis diagram

2 1 ..ld.fd l (2 1 ..l::!.e l t+0.72) 2 l . . .@f6 (Nikolov-Heimann, Basel Open 20 1 2) 22 . .l::!.e 2!t +0.44. As is often the case when Black is down the exchange for a pawn, he retains good chances for a draw; B) 1 3 . . . .l::!.fd8 1 4.a3

,.n Deferring this until White has \d..l played tl:lf4 makes sense. Analysis diagram

White has an attack, space, and knight vs. slightly bad bishop; D) 1 2 . . . tl:l6d5 1 3 .tl:lg5 �d7 1 4.tl:lxe6 �xe6 l 5 .tl:lxd5 tl:lxd5 l 6.h4;!; + 0 . 6 8 . 1 3.tl:lf4

1 4 . . . a5 ( 1 4 . . . tl:l4d5 1 5 .tl:lcxd5 �xd5 1 6 .tl:lxd5 'ifxd5 1 7 .�c l ±) 1 5 .d5 ! tlJ4xd5 l 6 .tl:lcxd5 tl:lxd5 1 7 .�c4 c6 1 8 .�b6 .ld.dc8 l 9.tl:lxe6 �xe6 2 0.f4 a4 2 1 .h4;!;.

Analysis diagram

96

Chapter

This was Zhou Jianchao-Mu Ke, Beijing Zonal 2 0 1 2 . White has tremendous compensation for the pawn - bishop pair, attack, pin, imprisoned black bishop. He can regain the pawn at will, for example 2 1 . . . 'W'fl 2 2 .hS ( 2 2 .�xdS cxd5 2 3 .'W'xdS .ld.c6 2 4.�e3t) 2 2 . . . e6 2 3 .hxg6 hxg6 24.'W'fl + 0 . 6 0 . White has more than enough for half a pawn with the open h-file against the black king, the g2-g4 lever, and a space advantage. 14.a3 tt:J4d5 1 4 . . . aS

3:

T h e N e o - G r ii n fe l d

B) 1 6 . . . cxbS 1 7 .�xb6 �xe5 1 8 . .l:i.he l �f6 1 9.�d4 �xd4 2 0 .'W'xd4 0ia6 2 l . d6 exd6 2 2 . ti:JfdS �xd5 2 3 .ti:JxdS 'W'g7 2 4.'W'b6± + 0 . 6 8 . White will regain his two pawns (or the ex­ change) and retain a much superior position. 1 5.tt:JcxdS tt:Jxd5 1 6.�c4

1 6 ....ld.adS l 6 . . . tt:Jxf4! 1 7 .�xfl + .i::!.xfl l 8 .jf_xf4 e6 1 9 .h4 h5 2 0 .�gS 'W'ds 2 l .�e2 .l:i.af8 2 2 . .l:!.d3 c6 2 3 . .ld.c l t Analysis diagram

1 5 .�bS ! c6 1 6.d5 ! and now: A) 1 6 . . . �xeS 1 7 .jf_xb6 cxb5 1 8 . .l::!.he l �f6 1 9 .jf_d4 �xd4 2 0 .�xd4 0ia6 2 l .d6±.

Analysis diagram

Analysis diagram

White will bring a knight to d5 next with a very strong attack;

+0.3 7 . White has the initiative with queenside play and the latent possibility of g2-g4, though Black should probably hold with perfect defense. 1 7.h4 1 7 . .ld.he l ! ;!:; 0ixf4 ( 1 7 . . . l::rfe8 1 8 .h4 +0 .48) 1 8 .�xfl+ .ld.xf7 1 9.�xf4 'W'b5 2 0. d5 . Without the rook on e 1 this could be met by . . . e7-e6, but not here. 97

S a b o t a g e t h e G rii nfeld

1 7...tl:Jxt4 1 a.�xt7+ l:!.xt7 1 9.�xt4 �b5 20.h5 2 0 .�c2 e6 2 1 .h5 gxh5 2 2 .l::!.xh5 t +0.4 1 . Black's weakened kingside gives White the edge. 20 ... 95 21 .�x95 �xe5 22.�c3t

22 ...�t6? 2 2 . . . �a4 2 3 .l::!.de l �xd4 24.�xc7 l:!'.e8 2 5 .'iHc l t + 0 .4 1 . White has the safer king and fewer pawn islands. 23.�xc7± ( + 1 . 1 7) 23 ...e5?! 24.'ti6c1 +- exd4 25 ..2.xt6 l:!.xt6 26.�95+ wt7 27.h6 l::!. 9 a 2a.�h5+ wta 29.�h4 �b6

40.�e 5 + +- . 40 ...@t6? 40 . . . Wd6=. 41 .�t4+ We7

42.�h4+? 42 .�e5++- �e6 43 .�xe6+ Wxe6 44.l:!'.xb6+ axb6 45 . Wb3 Wf7 46.Wc4 Wg6 4 7 . Wd5 Wxh6 48.f4 Wg7 49.We6 h5 5 0 . WfS Wh6 5 1 .a4+- . 42 ... @ea The game went on for another 2 3 moves, eventually ending in a draw. (D70) Wang Hao Svidler, Peter

Game 3.8

Stavanger 20 1 3 (2.5)

1 .d4 tl:Jt6 2.c4 96 3.f3 d5 4.cxd5 tl:Jxd5 5.e4 tl:Jb6 6.tl:Jc3 �97 7.�e3 0-0 a.�d2 tl:Jc6 8 . . . �d6 9.f4 l::!.d 8 1 0.tl:Jf3 �b4 l l .a3 �b3 1 2 .�d3 t. 9.0-0-0 'ti6d6 1 O.tl:Jb5 �d7 1 1 .f4 30.l::!.x d4? 3 0. g4+- . 30 ... l::!.x 92 31 . .l:!.b4 l:!'.91 + 32.l::rx 91 �x91 + 33.Wa2 �9a+= 34.l::!. b 3 l:lb6? 34 . . . l:!.d6=. 35.�da+ wt1 36.WVd5+ wta 37.�c5+ @ea 3a.'ti6ca+ wt7 39.�xt5+ We7 40.�e4+? 98

C h ap t e r

This rare move is dangerous and prom­ ising but somewhat unclear. 1 1 ... 'iYeG A) After l 1 . . .l::rd 8 1 2 .tbf3 a6 l 3 . tbc3 'iYe8 1 4.'iVf2;!; White's powerful center is not easy to challenge; B) l l . . . a6 1 2 .tbc3 tbas 1 3 .b3 'iYd6 1 4. 'itib 1 ::!; . 1 2.tbc3 t2Jc4 1 2 . . . l:td8 1 3 .tbf3 tbc4 1 4.'iYe2 tbxe3 ( 1 4 . . . tb6a5 1 5 .�£'2 bS 1 6.b3 tba3 1 7 .'iYe3::!;

3:

T h e Neo - G riinfeld

sition of the black knight offsets his bishop pair, and White's extra center pawn is more valuable than Black's crip­ pled queenside majority. Here too Ko­ modo favors White while Houdini says it's dead equal; B) 1 5 . . . tbb4? 1 6.'itib l and White is better; C) 1 5 . . . 'iYd6 1 6.eS 'i¥b4 1 7 .dS tbb8 1 8 .�bS a6 1 9 .�e2 1 9 .a3 ! + 0 . 6 8 . 1 9 . . . tb d 7 2 0 .h4::!; according to Svetushkin.

Analysis diagram

+ 0 . 2 6 . White's center remains strong. Komodo and I prefer White, while Houdini thinks it's dead even) 1 S .'iYxe3 and now: A) 1 5 . . . tbas 1 6.dS 'iYb6 1 7 .'iYxb6::!;

Analysis diagram

+ 0 . 5 2 . Black's bishop pair cannot com­ pensate for White's huge space advan­ tage and pawn center. 1 3.'iYe2 t2Jxe3 After 1 3 . . . tb6a5 1 4.tbf3 we have a crossroads: A) 1 4 . . . tbxe3 1 5 .'iYxe3 'iYb6 1 6 .h3 �e6 1 7 .Wb l =

Analysis diagram

1 7 . . . cxb6 ( 1 7 . . . axb6 1 8 .eS cS 1 9 .h3 ±; Black has little hope of being able to get his knight off the edge without the loss of a pawn) l 8.h3 ;!; + 0 . 1 8 . The bad po-

Analysis diagram

99

S a b o t a g e t h e G rii n fe l d

Analysis diagram

Wang Hao

+ 0 . 1 5 . Black's bishop pair offsets White's space advantage, although White's game looks easier to play; B) For 1 4 . . . .l:!.d8 see the previous note on move 1 4; C) l 4 . . . c5 ! may be the only fully sat­ isfactory line for Black against 1 l .f4, al­ though it involves a rather unclear piece sacrifice: C l ) 1 5 .�f2 .

+ 0. 1 1 . Black obviously has tremen­ dous compensation for the piece. Maybe White is okay in computer-as­ sisted correspondence play, but I wouldn't play this for White over the board; C 2 ) 1 5 .e 5 cxd4 1 6 . lll x d4 'iWb6 ( 1 6 . . . 'iYa6 1 7 . �f2 b5 1 8 .b3 lll b 3 1 9.ab3 'iWa3� (Svetushkin) transposes to the 1 5 . �f2 line) 1 7 .lll f5 lll e 3 1 8.lllx e3 lll c 6 1 9 .lll e d5 'iWd8.

Analysis diagram

Analysis diagram

Svetushkin recommends this l 4 . . . c5 defense for Black, but does not mention this move, perhaps because it may transpose to the 1 5 .e5 line. For exam­ ple, 1 5 . . . cxd4 1 6.lllxd4 'iWa6 1 7 .e5 b5 1 8 .b3 lllxb3 + 1 9 .axb3 'iWa3 + 2 0.'itic2 'iWb2 + 2 1 .'itid3

100

This was played in Karjakin-Giri, Wijk aan Zee 2 0 1 3 . C2 1 ) Now Tibor Fogarasi in Year­ book 1 0 7 suggests 2 0. h4 h5 2 l .'iWb5 'itih8 2 2.�e2 with attack, but 2 1 . . .�d7 is better and gives Black equal play; C 2 2 ) 2 0 .g3 �e6 2 1 .�g 2 ! + 0 . 1 8 . White's powerful knight outpost and space advantage (compare the king's bishops! ) mean more than Black's

C ha p t e r

3:

T h e Neo - G rii n fe l d

bishop pair. Also possible is 2 l .'lWb5 t +0.29. 1 4.VWxe3 tt:Jb4 For 1 4 . . . tlJa5 1 5 . tlJf3 see the previous note. 1 5.@b1

1 8 .�c4 ! ± . Probably Black overlooked this on move 1 5 ; B) 1 6 . . .a 5 1 7 .tt:Jg5 �b6 1 8 .�c4 e6 1 9 .f5 t ; although Black is suffering, this was probably the lesser evil. 1 7.a3

1 5 ..JidS? A) 1 5 . . . c6 1 6.l:!.e l ! (to avoid . . . tlJd5) 1 6 . . . l:!'.d8 1 7 .tlJf3 a 5 1 8 .h4 'tWd6 1 9 .a3;!; + 0 . 3 3 . White's center and better attacking prospects seem more important than Black's bishop pair; B) 1 5 . . . c5 1 6.d5 'tWb6 1 7 .a3 .td4 1 8 .'ii'd 2 tlJa6 1 9 .h3 ( 1 9 .tlJf3 ! .tg4 2 0 .h4 e6 2 1 .i::!.c l ± + 0 . 60 ) 1 9 . . . i::!.d 8 ( 1 9 . . . 'tWa5 ! ;!; + 0 . 3 7) 2 0 .tlJf3 .tg7 2 l .g4 (2 1 ..tc4! :t-0. 98) 2 1 . . .tlJc7 . Now White lost his advantage with 2 2 .e5 ( 2 2 .'lWe3 ! + 0 . 6 0) , but went on to win in Matlakov-Ftacnik, Czech tt 2 0 1 3 / 1 4. 1 6.tt:Jf3± (+0. 5 2 ) 1 6 ... b5?1 A) 1 6 . . . c5? 1 7 .d5 �xc3

1 7...tt:Ja6?1 After 1 7 . . . tLld5 ! 1 8 . tt:Jxd5 1:!.xd5 1 9 .tlJg5 1::!.x g5 Black has some compen­ sation for the exchange. 1 8.�xb5 VWb6 1 9.tt:Je5 iLxe5 20.fxe5 l:!'.b8 21 .1::!.d 2

Analysis diagram

Black has no compensation at all for the lost pawn, and eventually loses the end­ game. 21 ...'lWa5 22.�xa6 �xa6 23.tt:Jd5 l:!.b3 24.tt:Jxe7+ @fa 25.'ii'x b3 VWxd2 26.tt:Jc6 VWxg2 27.1::!.e 1 1::!. e 8 28.�c2 �xc2+ 29.@xc2 �b5 30.tt:Jxa7 �d7 31 .1::!.f 1 f5 32.exf5 gxf5 101

S a b o t a g e t h e G rii n feld

Game 3.9 (D7 0) Jones, Gawain Aronian, Levon London 20 1 2 (7 .4) 1 .d4 t2Jf6 2.c4 g6 3.f3 d5 4.cxd5 t2Jxd5 5.e4 t2Jb6 6.tt:lc3 �g7 7...te3 0-0 S.Wid2 tt:lc6 9.0-0-0 �d6 1 0.t2Jb5 �d7 33.@c3 3 3 .l:rf3 ! nas 34.e6 �a4+ (34 . . . �xe6 3 5 .t2Jb5 .l::r c 8 3 6 . nc3+-) 3 5 .b3 nxa7 3 6 . nxf5 + We7 3 7 .bxa4 Wxe6 3 8 .ne5+ @d7 3 9 .a5 +- . The second extra white pawn, although not directly useful, has enough nuisance value to ensure the win. 33 ... .i:ras 34.e6 �xe6 35.t2Jc6 @g7 36.d5 �xd5 37.t2Jd4 �e4 3S.t2Jxf5+ @g6 39.t2Jd4 nas 40.tbe6 �f5 41 .t2Jd4 �d7 42 ..l::[f2 h5?! 43.@d3 .l::[ g 5 44.t2Jf3 nas 45.@d4 na4+ 46.@e3 �f5 47.t2Jd4 �d7 4S.l::r c2 na7 49.h4 ..t>f6 so.ncs �es s1 .nts+ we7 52 . .b!:e5+ @d7 53.t2Jb3 �f7 54.tt:lc5+ @d6 55.@d4 .l:!.aS 56.l:i.f5 ..t>e7 57.a4 l::r g S ss.nt4 .l::[g 2 59.b4 .l:!.d2+ 60.@c3 l:ra2 61 .a5 .b!:a1 62.a6 �es 63 ..l::!. e4+ 1 -0

6 3 . . . Wf7 64 . .b!:e5 wins a second pawn. 1 02

1 1 .�h6 This looks really good at first, but it gambits the d-pawn so it's not that con­ vincing. It seems Black can equalize af­ ter grabbing the pawn by playing . . . f7-f6 and . . . e7-e5 , as Caruana did against Svidler after I had already in­ cluded the line up to move 2 3 in my notes here! 1 1 ...�xh6 1 2.�xh6 a6 1 3.t2Jc3 t2Jxd4 14.f4

14 ... c5? 1 4 . . .f6 ! 1 5 .tt:lf3 ( 1 5 .h4 e5 1 6 .tbf3 �e7 1 7 .tbxd4 exd4 1 8 . .l::!.xd4 �e6=) 1 5 . . . e5

Chapter

3:

T h e Neo - G r ii n fe ] d

is the surest equalizer ( l S . . . cs trans­ poses to the game while avoiding the next note) : 1 6.fxeS fxeS 1 7 .tlJxeS �d6 l 8.tlJf3 cS and now:

Analysis diagram

Analysis diagram

A) l 9 .�d3 �f4 + ? ! ( 1 9 . . . �c7 ! = 0 . 0 0 ) 2 0 .�xf4 .l:!.xf4 2 l . tlJxd4 (2 1 ..ld.hfl ! +0.4 1 ) 2 1 . . .cxd4 2 2 .tlJe2 + 0 . 2 2 ended in a draw in Sadler­ Svidler, London 2 0 1 3 ; B) 1 9 . tlJ g S �e7 2 0 .tt:Jds tt:JxdS 2 1 .�c4 'iilh 8 ( 2 1 . . .�g 7 2 2 .�h4 'iilh 8 23 ..l:!.hfl !;!;) 2 2 .exdS �g 7= 2 3 .�xg 7 + 'iitx g7 24.d6 �fs ( 2 4 . . . bS ! 2 s .�ds .l:!.a7 (0.00) 26 . .ld.de l .l::i.fS =) 2 S .b4 b6 2 6 .l:ihe 1 ld.fe8 and now:

2 7 . .l:!.e7+ (2 7 .'iitb 22;; 2 7 .a4! (+0 . S 2 ) is recommended Svetushkin) by 2 7 . . . .l:!.xe7 2 8 .dxe7 ld.e8 2 9.bxcS bxcS 3 0 . .l::!. e l tlJc2 3 l .ld.e2 tlJd4 32 . .l:!.e3 tlJc2 3 3 . l:re 2 . After more repetitions Svidler-Caruana was drawn in Thessaloniki 2 0 1 3 . 1 5.tt:Jf3? l S .h4! f6 ( 1 S . . . 'i¥c6 ! l 6 .tt:Jf3 'i¥f6 1 7 .hS;l; +0. 1 S ) 1 6.hS �e8 ( 1 6 . . . �g4 1 7 .tt:Jf3 - see next note) 1 7 . .l::!. d3 gS 1 8.fxgS and White is clearly better. 1 5 .. .f6 1 6.h4

1 6 �ea A) 1 6 . . . �g4? 1 7 .hS gS 1 8 .fxgS tt:Jxf3 l 9 . gxf6 i&' g S + 2 0 .i&'xg S + tlJxgS 2 l .fxe7 .l::!.e 8 2 2 .l::rd 6 tt:Jd7 2 3 . tt:Jds± tt:Jxe4? (23 ... tt:Jf7 24.l::r f6 ! 2;) 24.l:re6 tt:Jgs 2 S .�e3 bS 2 6 .h6 �b7 27 . .l:!.hs tlJf7 2 8 .tlJc7 c4? ( 2 8 . . . tlJde S ) 2 9 . �e 2 +- .l:!.ac8 3 0 . tlJxe8 .l:!.xe8 3 1 .Jlfs tt:Jxh6 3 2 . l:lf4 @g7 3 3 .�hs ...

Gawain Jones

1 03

S a b o t a g e t h e G r iinfeld

tl:Jf6 34 . .!:!'.g3+ 1 -0 Hammer-Erdos, Greece tt 2 0 1 2 ; B) Black's best option is 1 6 . . J:ld8 ! 1 7 .h5 �e8 1 8 .hxg6 �xg6 1 9 .�h4 �g4 20.�f2 (unpinning with 2 0.lld3 may be marginally better for White) 2 0 . . . '.tih8 2 I .tl:Jxd4 �xd l 2 2 .tLle6 �a4 2 3 .tl:Jxd8 .l:!.xd8 24_jLe2 iLc6 2 5 .�xc5 tLlc8 2 6 .iLf3 tl:Jd6 2 7 .�e3 tl:Jc4 2 8 .�e2=

Analysis diagram

+ 0 . 0 7 . White's edge is tiny. Houdini has a slight preference for White, but Komodo and I think it's really bal­ anced. 1 7.h5 1 7 . tl:Jxd4 cxd4 1 8 . .ld.xd4 jLg4 1 9 .f5 and again White has an edge.

Analysis diagram

+0.26. 1 7...iLg4 1 8.tl:Jxd4 cxd4 1 9.hxg6 WVxg6 20. .l:!.xd4 1 04

20....l:i.acS? 20 ... �xh6 2 1 ..!:!'.xh6 .l:!.ac8 2 2 . .l:!.b4 .l::!.c 6 2 3 .\tid2;!; + 0. 1 5 - a slightly better pawn structure and endgame king position. 21 .f5 2 1 .�h4 ! ± + 0 . 6 8 . With the safer king, White shouldn't trade queens. I'm sure Jones knew this, but perhaps in view of Aronian's astronomical rating he headed for a draw. 21 ...�xh6+ 22 ..ld.xh6 .ld.fd8 23 . .l:i.xdS+ l::!. x d8= 24..ld.h4 h5 25.iLe2 �xe2 26.tl:Jxe2 lacs+ 27.'.tib1 tl:Jc4 28.\tic2 tl:Je3+ 29.'.tid2 tl:Jc4+ 30.\tic2 tl:Je3+ 31 .'.tid2 tl:Jc4+ 1/2-1/2

Game 3.10 (D70) Postny, Evgeny Mikhalevski, Viktor Belgium tt 2 0 1 2 / 1 3 ( 1 1 .4) 1 .d4 tl:Jf6 2.c4 g6 3.f3 d5 4.cxd5 tl:Jxd5 5.e4 tl:Jb6 6.tl:Jc3 �g7 7.�e3 0-0 8.�d2 tl:Jc6 9.0-0-0 �d6

C h ap t e r

1 0.h4 This rare line looks like it allows Black only one narrow path to equality, which is recommended nowhere but here. 1 O.tl:ib5 'i¥d7 l 1 .h4 .ld.d8 transposes ( 1 1 . . .a6 1 2 .tl:ic3 'i¥d8 1 3 .d5;l; ) . 1 0....a'.dS 1 1 .tl:ib5 'i¥d7 1 2.h5 a6 1 3.tl:ic3

3:

T h e N eo - G r ii n fe l d

B) 1 5 .tl:id5 tl:ixd5 1 6 ..tc4

Analysis diagram

1 3 ...tl:ixd4 ?I Better is 1 3 ... .txd4! 1 4.hxg6 fxg6 and now: A) 1 5 .�xd4! 'i¥xd4 1 6 .'i¥f4.

1 6...e6 ( 1 6 . . . �g 7 1 7 .'iYe l e6 ( 1 7 ... 'i¥e8 ! - 0 . 4 1 ) 1 8 .exd5 tl:i e 5 1 9 .VWh4 was Piorun-Svidler, Germany Bundesliga 2 0 1 3 / 1 4; after 1 9 . . . h5 the evaluation is -0.29) 1 7 .�xd4 tl:ib6 1 8 . .txb6 (Gonda-Gledura, Zalakaros 2 0 1 3)

Analysis diagram

Analysis diagram

Here Fogarasi gives a clear advantage to White, but . . . 1 6 . . . 'i¥f6 1 7 Jhd 8 + tl:ixd8 1 8 .'i¥xc7 (first 1 8 .'i¥h2 h5 and then 1 9 .'i¥xc7 didn't help White much in Santiago-Ragger, Graz Open 2 0 1 4) 1 8 . . . �d7 1 9 .tl:ige2 .ld.c8 2 0 .'i¥f4 1l?Vxf4+ 2 1 .tl:ixf4 tl:ie6= -0. 04. White's slightly better pawn structure offsets Black's better bishop;

1 8 . . . 'i¥e7 ! 1 9.'i¥e3 .ld.xd l + 2 0.xd l cxb6+ -0.29, Black is slightly for choice due to White's poor king position. 14.hxg6 fxg6 Not l 4 . . . hxg 6 ?? 1 5 .�xd4 'i¥xd4 ( 1 5 ... �xd4 1 6.'i¥h6 'i¥d6 1 7 .'i¥h7 + f8 1 8 . .ld.xd4 �xd4 1 9 .�h8+ wins a piece) 1 6.'lWe l ( 1 6.'i¥f4 also wins) , winning the queen, as happened in the top-level game Nakamura-Caruana, Paris 2 0 1 3 . 1 5.g4 1 05

S a b o t ag e t h e G r iinfeld

1 6.Wh2 @f7? 1 6 . . . h6 1 7 .�xh6 �xh6+ 1 8 .Wxh6 �g7 (Postny-Givon, Acre ch-ISR 2 0 1 3) 1 9 .�e3 .l:l'.f8 2 0 .@b l :t

1 5 ... e5 1 5 . . . �c6 1 6 . �f2 ( 1 6 .'Wh2 ! hS 1 7 . gxhS tll a 4 1 8 . .l:l'.d3 ± + 0 . 44) 1 6 . . . tll e 6 1 7 ..ld'.xd8+ tllxd8 l 8 .tll e 2 tll f7 ( 1 8 ... �f6 !) 1 9 .iVh4 h6 2 0 .�xe7 �d7 2 1 .Wxd7:t �xd7

Analysis diagram

+ 0 . 2 6 . Black's weakened king position gives White a small plus. 1 7.f4± We7 1 8.f5

Analysis diagram

2 2 .�g2 (after 2 2 .b3:t White's passer, better king position, and pressure on h6 give him the better endgame) 2 2 . . . tll c4 2 3 .�fl �bS 24.tll d4 tll c eS 2 5 .�e3 �d7 2 6 . .l:!.d l .l:!.f8 2 7 . tll d S @h7 2 8 .@b l tll c4 2 9.�c l c6 3 0 .tll c 3 �c8 3 1 .�fl tll ce5 3 2 .�e2 cS 3 3 .tll c 2 bS 34.tll d S c4 3 5 .tll c e3 tll d 3 3 6.�xd3 cxd3 3 7 .f4 tll d 6 3 8 .fS gxfS 3 9 . exfS tll e 4 40 .tll f4 tll f6 4 1 .tll x d3 .l:!.e8 42 . .l:l'.g l �f8 43.gS hxgS 44 . .l:IxgS �h6 45 . .l:l'.g6 �xe3 46 . .l:l'.xf6 @g7 47 . ld'.d6 �xfS 48.�xe3 .l:l'.xe3 49.@c2 1/2- 1/2 Grischuk- Mamedyarov, Moscow 2 0 1 O ; B ) 1 5 . . . c s 1 6.�h2 @f7 1 7 .Wfl Wg8 1 8 .Wh4 @f7 1 9 .�d3:t +0.48. Black's king is unsafe. 1 06

+ 1 .0 2 . 1 8....l:l'.hS 1 9.�f2 @es 20.tll f3 c5 21 .�g5 Wf7 22.Wg3 h6 23.tll xe5 �c7 24.f6 hxg5 1 -0 25 . .l:l'.xhS+ �xh8 26.f7+

Game 3.1 1 (D7 0 ) Aronian, Levon Caruana, Fabiano Moscow 2 0 1 2 (9) 1 .d4 tllf6 2.c4 g6 3.f3 d5 4.cxd5 tll x d5 5.e4 tll b 6 6.tll c3 �g7 7.�e3 0-0 8.�d2 tll c6 9.0-0-0

Chapter

9 ...'iVd6 This is recently popular as the tradi­ tional moves have not done well. 1 O.@b1 ! This clears c 1 for the queen, which is usually the best square for her when . . . tLlc4 is a threat. 1 0 .tLlbS ! . Whether this move order or the game move order is better is hard to say; they may transpose, but in the lines where they don't transpose it seems that White has the choice of great compen­ sation for a pawn with this move order, or winning the exchange for a pawn with Black holding some compensation in the game move order. I guess Shirov would like the 1 O .tLlbS order while Kortchnoi would prefer the 1 O .@b1 order. 1 O . . . 'i¥d7 1 1 .@b l and now: A) l l . . . a6 l 2 . tLla3 fs ( 1 2 . . . I:!.d8 1 3 . tbe 2 - this shows the point of tLla3 , as now the bishop need not guard c4) 1 3 .dS tLld8 1 4.tLle2 (+0. 2 6) 1 4 . . . fe4 l S .fe4� ; B) l l . . . I:!.d8 transposes to the game; C) 1 1 . . .fS 1 2 .tLlh3 ( 1 2 .h4 fxe4 1 3 .hS exf3 (for 1 3 ... eS see the next game) l 4.tLlxf3 ( + 0 . 1 S) is a promising gam­ bit) 1 2 . . . a6 1 3 . tLla3 (in the line 1 3 . tbc3 fxe4 1 4.fxe4 eS l S .dS tLld4 1 6 .tLlgS 'i¥d6 1 7 .�e2 'iVb4 1 8 .I:!.c l White is slightly for choice. Black has a great knight outpost, but everything else favors White here) 1 3 . . .fxe4 1 4.fxe4

3:

T h e N eo - G r ii n fe l d

Levon Aronian

'iVg4 l S .tLlgS �xd4 1 6.�e2 �xe3 1 7 .'iVxe3 'YWf4 1 8 .�b3 + e6 1 9 .h4 'iVeS 2 0.tLlc4 tLlxc4 2 1 .�xc4 h6 2 2 .tLlf3 �g7 2 3 .eS tLlxeS 24.tLlxeS �xeS 2 S .hS bS 2 6 .'YWc2� + 0 . 1 S . White has huge compensation for the pawns. 1 0 l:ldS 1 O . . . a6 is rare though not clearly worse than the usual 1 0 . . . l:rd8 : 1 1 .h4 and now: A) 1 1 . . .hS 1 2 .�h6 ( 1 2 .g4! (+0. S 6) may be better) and now: A l ) 1 2 . . . 'iYxd4 1 3 .'iYxd4 �xd4 1 4.�xf8 'it1xf8 l S .tLldS tLla4 1 6 .b3 ( 1 6.I:!.d2 �es 1 7 .tLle2 + 0 . 3 3 . White is up 1/4 pawn by my count, for which Black has just slight compensation) 1 6 . . . tLlc3 + 1 7 .tLlxc3 �xc3 1 8 .g3 as l 9 .�h3 �xh3 2 0 .tLlxh3 a4 2 l .I:!.d3 ...

Analysis diagram

1 07

S a b o t a g e t h e G r iinfeld

+ 0 . 2 6 . Now White is up % pawn, but the strong Black bishop and isolation of White's pawn give him significant compensation; 1 3 .�xh6 �f6 A 2 ) 1 2 . . . �xh6 1 4.�e3 ( 1 4.tbge2 .!:!.d8 1 S .'/We3 trans­ poses) 1 4 ... l:td8 1 S .tbge2 tbc4 1 6.�c l tlJb6 1 7 .dS tt:Jes 1 8 .tbg3 �d7 1 9 .f4;t

Analysis diagram

+ 0 . 3 3 . White's mobile center and Black's slightly weak king should give White a plus. B) 1 1 . . . 1d.d8 1 2 .ds tt:Jas ( 1 2 . . . �b4 1 3 .a3 '1Wb3 1 4.�c2 �xc2 + 1 S .'it'xc2 tbeS 1 6 .hS + 0 . 2 6 ) 1 3 .�c l tbac4 1 4.�d4 es 1 s .�f2 �b4 1 6.g4 �d7 1 7 .'it'al c6 1 8 .�xc4 tbxc4 1 9.tbge2 1d.ac8 2 0.hS �h6 2 1 .�c2;1;

1 1 .tt:Jb5 �d7 1 2.d5 a6 1 2 . . . tlJeS ? ! 1 3 .�c2 ! c6 1 4.tbxa7 ± Anand-Mamedyarov, Corsica rapid 201 1 .

1 3 .tt:Jc3 Now there are many options for Black, but as White you just need to remem­ ber one thing: in every line given be­ low, when the queen must move, it should move to c 1 ! A) 1 3 .dxc 6 ! ? is the greedy move, winning a pawn for limited compensa­ tion, but Black can regain it with equal­ ity: 1 3 . . . '/Wxd2 1 4. 1d.xd2 1d.xd2 1 S .�xd2 axbS 1 6.cxb7 �xb7 1 7 .�xbS and now:

Analysis diagram Analysis diagram

+ 0 . S 2 . Black has the bishop pair and some activity, but White's h-file pres­ sure and kingside attack look more im­ portant here. 1 08

A l ) 1 7 . . . l:rd8 1 8 .�c l l:rd4 ( 1 8 . . .fs 1 9 .exfS �c8 2 0 .tlJh3 �xfS + 2 1 .'it'a l .i::i.d S 2 2 .�e2 l:ras 2 3 . .l:f.d l ) 1 9 .�fl tba4 2 0 .tbh3 l::lb 4 2 1 .b3 tbc3 + 2 2 . @c2 tbxa2 2 3 .�a3 .!:l'.b6 2 4.�xe7 �es

Chapter

1 S .�c4 �a6 1 6 . .l:Ib l and Black doesn't have enough for the pawn; Al) 1 7 . . .fS ! 1 8 .exfS �dS (So-Lou Yiping, Ho Chi Minh City Ach 1 0 1 2) l 9 .b3 ( 1 9 . tbel �xa l + 1 0 . 'iil c l + 0 . 04) 1 9 . . . gxfS 1 0 .h4 ji e 6 1 1 ..l:Ihl �xb3 1 1 . axb3 .ld. a l + 2 3 .'iil c l .ld.xg l = 0.00. B ) 1 3 . tba3 'iY e 8 1 4. 'iYe l i s an obscure alternative: 1 4 . . . tba7 l S .h4 ( 1 S .'iYh4 tbbS=) l S . . . tbbS = 0 . 0 0 .

3:

T h e N e o - G r ii n fe l d

A l l l) 1 S . . . �xd4 1 6 . l:lxd4 tbes ( 1 6 ... tbd6 1 7 .h4±) 1 7 .h4± + 0 . S l . White has an attack at no cost; A1 1 3 ) l S . . . eS 1 6 . .tcs ( 1 6 . dxe6 �xe6 1 7 .�xc4 �xc4 l 8 .b3 �c6 1 9.tbgelt + 0.4 1 ) 1 6 . . . tbd6

Analysis diagram

1 3 �ea Alternatively Black has the choice be­ tween two knight moves: A) After 1 3 ... tbeS White can play: A l ) 1 4.�c l 'iYd6 1 S .h4 hS 1 6 .�h6 (in case of 1 6 .tbgel �d7 1 7 .'iil a l per­ haps White is a tad better) l 6 . . . �xh6 1 7 .�xh6 �f6. This was Wojtaszek­ Areschenko, Wroclaw 1 0 1 3 . Now best was 1 8.tbge l , which the engines rate as equal, though I would rather play White myself; Al) 1 4 . .td4! Al l ) 1 4 . . . tbec4 l S .�c l ! ( 1 S .�xc4 tbxc4 transposes to the next note; if 1 S .�f2 .txd4 ( l S . . . eS 1 6.�cS �f8 1 7 .h4;\;) l 6 . .ld.xd4 tbd6 1 7 .tbgel es l 8.dxe6 �xe6 l 9 .tbf4 �es 1 0.tbd3 �e7 l l .tbcS t) and now: A l l l ) 1 S . . . �d6 1 6 . .txg 7 'iit x g 7 1 7 .h4 hS 1 8 .tbgel �d7 1 9.tbd4± +0.72; ...

1 7 . b3 ( 1 7 .�d3 transposes to the next note) 1 7 . . . VWe7 1 8 .h4 tbd7 ( 1 8 . . . hS ! + 0 . 2 2) 1 9 . .te3 fS 1 0 .hS f4 1 1 .�fl (+0.60) 1 1 . . . gS 2 2 .h6 �f8 1 3 .g3± and White eventually won in Berczes­ Chirila, Dallas l 0 1 3 . A l l ) 1 4 . . . tbbc4? was my game as Black vs. near-GM Daniel Ludwig (At­ lantic Open, Washington D. C. l 0 1 2) , when l S .�c l (ct-O.S 6) followed by f3-f4 would have been very strong; A 2 3 ) 1 4 . . . cS l S .�xcS tbec4 was Elianov-Areschenko (Kiev ch-UKR 1 0 1 2) .

Analysis diagram

1 09

S a b o t a g e t h e G r ii nfeld

A 2 3 1 ) Against World under- 1 4 Champion Kayden Troff I played l 6 .'iVf2 in the 2 0 1 3 World Open in Arlington, Virginia. After 1 6 . . . .¥lxc3 1 7 .bxc3 he should have played 1 7 . . . 'iVc7 when 1 8 .@al is just a tad better for White. He actually played 1 7 . . . ttJa4?, when after 1 8 . .¥lxc4 'iVc7 1 9 . .¥ld4 'iVxc4 White would be a safe pawn up if he had played 2 0 .@a l . Un­ fortunately I blundered by 2 0 .'iVc2 ?? laxd5 ! and Black wins; A 2 3 2) 1 6 . .¥lxc4 is the move to play if you don't want to trade your queen for three minor pieces in line A2 3 3 below, but I see no need to avoid that line. 1 6 . . . ttJxc4 1 7 .'iVc l 'iVc7 1 8 . .¥ld4 e5 1 9 . .¥lf2 b5 2 0 .ttJge2 ttJd6 2 1 .@a l (Svetushkin recommends 2 1 . 'iVe3 , but I think it's better to get the knight out of the way of . . . b5 -b4. There can follow 2 1 . . . .¥ld7 2 2 .1::1'.c l 'iVb7=) 2 1 . . .�d7 2 2 .tLlb l t. The endgame would favor White, and otherwise the attack with h2-h4 is planned; A2 3 3) 1 6 .'iVc l ! was best, leaving Black little compensation for the pawn: 1 6 . . . ttJxb2 ( 1 6 . . . 'iVc7 1 7 .i.d4 i.xd4 l 8 . l::i'. x d4 'iVc5 l 9 . ttJge2t + 0 . 5 6) 1 7 .'iVxb2 tLla4 l 8 . tLlxa4 i.xb2 l 9.tLlxb2 'iVc7

just to maintain the balance' . To me this is a surprising comment from such a strong grandmaster. The superiority of three minor pieces to the queen is uni­ versally accepted, especially when the minors include the bishop pair and when all rooks are on the board, as here. White's king is a bit bare but it is not evident how Black can exploit this. Komodo rates this as + 0 . 7 6 , which feels about right to me. Houdini may rate it close to even, which may have in­ fluenced Svetushkin here, but Houdini overvalues the queen. 2 0 . . . .¥ld7 2 l .tLle2 ± ; B ) On 1 3 . . . ttJa5 there can follow: B l ) 1 4.i.d4 and now: B l 1 ) 1 4... e5 1 5 .hb6N cxb6 1 6.'iVe3 b5 1 7 .'iVb6 tLlc4 l 8.itxc4 bxc4 1 9 .ttJge2 'iVe8 ( 1 9 ... 'iVd6 20.tLla4t) 2 0 .d6 i.e6 2 l .'iVxb7 i.f8 2 2 .'iVb6 Itac8 23 .ttJd5 .¥1xd5 24 . .l::i.xd5 �c6 2 5 .'iVa5 l::i'.dxd6 26 . .l:!.xd6 hd6 2 7 .l::i'.c l @g7 2 8 .'iVa4t

• •

Analysis diagram

Analysis diagram

2 0 .i.e3 Svetushkin says here 2 0 . .l::i.c 1 ' gives White enough compensation, 110

+ 0 . 64. White's better pawn structure and the plan to bring his knight to d5 give him a distinct advantage; B l 2 ) 14 ... ttJbc4 (so far Gustafsson­ Mekhitarian, Reykjavik Open 2 0 1 1 ) 1 5 .i.c4 ( 1 5 . 'iVc 1 transposes to the pre­ vious note) 1 5 . . . tLlc4 1 6.'iVc l .¥ld4 1 7 . .l::!. d4 b5 1 8 .tLlge2 i.b7 1 9 .�hd l 'iVd6 2 0 .f4t

Chapter

3:

T h e N eo - G rii nfeld

1 4.�e1 1 4.'ifc l ! is better: A) 1 4 . . . tt:Jes 1 S .�e2 tt:Jec4 1 6.�d4 eS 1 7 . .tf2;t +0.4 1 . White still has the usual h2-h4-hS attack, and the knight on c4 is unstable. 1 7 .�cs ( + 0 . 3 7) is also favorable for White; B) 1 4 . . . tlJa7 1 S .�d3 e6 1 6.�gS .l:!.d6 1 7 . .tf4 .l:!.d7 1 8 .dxe6 �xe6 Analysis diagram

+ 0 . 3 3 . The mobile center favors White; B2) 1 4.'ifc 1 , as preferred by Svetushkin, limits Black's options by avoiding 1 4 . . . eS, otherwise it should transpose to a l 4.�d4 option: l 4 . . . tlJbc4 ( l 4 . . . tlJac4 1 S .�d4 transposes to 1 4.�d4 tlJac4 l S.'ifc l ; 1 4 . . . 'iie 8 trans­ poses to the 1 3 . . . 'iie 8 line) 1 S ...td4 �d4 l 6 .lld4 bS 1 7 .h4 cs ( 1 7 . . . es l 8 .lld3 +0. 1 8) 1 8.Iid 1 f6 1 9 .g4.

Analysis diagram

Analysis diagram

1 9 .�h6 ( 1 9 . tt:Jge 2 tlJc6 2 0 .�e3 + 0 . 48) 1 9 . . . tlJc6 ( 1 9 . . . �xh6 2 0 .'iix h6 tlJc6 2 1 .tlJge2 + 0 . 3 3 ) 2 0 .�xg 7 Wxg7 2 l . tlJ g e 2 t + 0 . 44 . Analysis by Svetushkin. White's edge is obvious, the black king has no defenders and dS may be a good home for a knight. C) If 1 4 . . . tt:Ja S White has two ways to start the attack: C l ) 1 S . .th6 �xh6 1 6 .'iix h6 e6 1 7 .tlJh3 �f8

White has an attack while Black's pieces are clumsy.

Analysis diagram

111

S a b o t a g e t h e G rii n fe l d

l 8 .�f4 ( 1 8 .�xf8+ �xf8 l 9 .tlJf4 exd5 l O .tZ:lfxdS tZ:lxdS l 1 ..l::!.x d5 + 0. 1 1 . The endgame is slightly more pleasant for White though undoubtedly drawn with correct play) and now: C l 1 ) 1 8 . . . �e7 1 9 .tZ:lfl exd5 l0.exd5 il..fs + l l .�a l .tel 2 2 . .l::!. c l �b3 2 3 . axb3 tZ:lxb 3 + 2 4 . �b l tZ:lxc l 1 5 .�xc l .l::!.a c8 l 6 .�d3 tZ:lxdS 2 7 .�c4 c6 1 8 .1::!'. e l ;l;.

+ 0 . 1 6 . Black has reasonable but not full compensation for the pawn. Cl) 1 5 .h4 e6 and now: Cl I ) l 6 .�f4 �e7 1 7 .�g s f6 l 8 .�e3 exd5 1 9 .�xb6 cxb6 l O.tZ:lxdS �f7 l l .h5 �e6 2 2 . hxg6 hxg6=.

.1 •

Analysis diagram

Analysis diagram

White is down a quarter of a pawn, but the presence of all major pieces favors him, and the weak squares f6 and h6 can potentially be exploited by White's knights. White's pieces are all well placed; C l l ) 1 8 . . . exdS 1 9 .�xc7 �d6 l 0.�xd6 .l::!.x d6 l l .exd5 il..fs + 2 2 .�a l �xh3 l 3 . gxh3 l:lad8 2 4.tZ:le4 .l::!. c 6 l 5 .�xa6 .l::!.x d5 l6 . .l::!.x d5 tZ:lxdS 27 .�b5 .l:!.b6 l 8 .il..a4 .l:!.b4 l 9 . n d 1 tt:Je3 3 0 . .l::!.d 3 .l:!.xa4 3 l . .l::!.x e3 f5 3 l .b3 .l:rd4 3 3 .tZ:lcS ;l;

This was Rodshtein-Bok, Biel 1 0 1 l , a game which was eventually drawn. Black's bishop pair offsets White's out­ posted knight and better pawn struc­ ture. Vigorito feels that it looks a little better for White, but Komodo and Houdini both say dead even. Probably Black still needs to play one or two accurate moves to prove equality; C l l ) 1 6 .�g S ! f6 1 7 . .ildl tZ:lac4 1 8 .�f4

Analysis diagram

Analysis diagram

1 12

C22 1 ) 1 8 . . . eS ! 1 9 . .ile3 (Svetushkin prefers 1 9 . .ilh6 , but it trades a good bishop for a poor one: 1 9 . . . .ilxh6

Chapter

2 0 .�xh6 �e7 2 1 .g4 aS 2 2 .�d3 a4 2 3 . g S '1Wg7 + 0 . 1 1 ) 1 9 . . . ti:Jxe3 2 0 .�xe3 �f8 2 1 .�c l �d7 2 2 .ti:Jh3 �d6 2 3 .hS gs 24.ti:Jf2 �e7 2 S .ti:Jcd l fS 2 6 . exfS �xfS + 2 7 .ti:Je4 @h8 (27 . . .tlJxdS 2 8 .�a3 �e6 2 9 .�c4;l;) 2 8 .�d3 h6 2 9 .ti:Je3 �h7 3 0 .�c2;!;

Analysis diagram

+ 0 . 3 7 . If White can force Black to exchange his good light-squared bishop for either bishop or knight, which looks fairly likely, he will have a clear advantage; C 2 2 2) 1 8 ... �e7 1 9 .hS ( 1 9 .ti:Jh3 ;!;) 1 9 . . . gS 2 0.h6 �f8 2 1 .�g3 c6 2 2 .dxc6 bxc6 2 3 .eS �xd l 24.tlJxd l tlJxeS 2 S .�xeS fxeS 2 6.ti:Jf2;!;. All six black pawns are weaknesses, so White has more than enough for the pawn, while Black's bishop pair looks pretty useless; C 2 2 3 ) 1 8 ... �f7 1 9.hS exdS 2 0.hxg6 hxg6 2 1 .exdSt or 2 1 .�h6t; the open h-file is White's edge.

3:

T h e N eo - G r ii n fe l d

14... tt:Je5 1 4 . . . ti:Ja7 ! 1 S .h4 ti:Jbs 1 6 .ti:Jge2 ( 1 6.ti:JxbS axbS 1 7 .hS ? ti:Jc4 is given by Fogarasi in Yearbook 1 0 7 as promising for White (correctly so, I think) , but unfortunately for White 1 7 ... tlJxdS ! ! turns out to pretty much win on the spot. I doubt any human would find this over the board) 1 6 . . . tlJc4 0 . 0 0 . This fully equalized i n Gelfand­ Caruana, Zurich 2 0 1 3 . 1 5.�e2 1 S .f4 is the safer move: 1 S ... ti:Jec4 1 6.�d4 e6 1 7 .�xc4 tlJxc4 1 8 .�xg 7 Wxg7 1 9.ti:Jf3 t +0.26. White is usually better in this line once he exchanges Black's powerful dark­ squared bishop. 1 5 ... eG 1 S . . . ti:Jec4! 1 6.�d4 eS 1 7 .i.cS �f8 1 8 .�f2 �xcS 1 9.�xcS �f8= + 0 . 0 7 . White's superior pawn (dS vs. c7) is offset by his inferior bishop. 1 6.�xbG cxb6 1 7.f4 tt:Jd7 1 7 . . . exdS 1 8 . fxeS dxe4 1 9 . �xd8 �xd8 2 0 .ti:Jxe4 �fS 2 1 .�f3 �xeS 2 2 .Wa l ;t

.1 •

Analysis diagram

+ 0 . S 2 . Black's two pawns, bishop pair, and better development offer substan­ tial but not quite full compensation for a knight. 1 8.dxeG �xe6 1 13

S a b o t a g e t h e G rii n f e l d

1 9.tllf3 1 9 .�f3 ! +0. 7 6 . There was no need for the speculative two pawn sac in the game, when this move keeps a clear advantage with a nice pin, better pawn structure, and the powerful move tll d S planned, all for the bishop pair. 1 9 ... 'lWeS Black could try 1 9 . . . .txc3 ! 2 0 .'lWxc3 'lWxe4+ 2 1 .�d3 'lWxf4 2 2 . .l:i.hfl ( 2 2 . .tc2 is similar) 22 . . . �f8 . White has great development and attacking pros­ pects, but at the price of two pawns, so this would have been unclear. 20.'lWh4t .tf6 21 .tll g 5 tllf8 22.�c4 Wg7?! 23.'lWg3 �xg5 24.fxg5 .te6 25.tll d 5 �xd5 26.�xd5± .i::!. d 7?! 27.h4+- .l:i.c8 28.a3 'lWd8

29.'lWf2? 2 9 .hS +- . 29 ...tll e 6 30.@a2 'lWe7 31 . .l:i.hf1 ? 3 1 .'lWxb6+- . 1 14

31 ... b5 32.lad3 a'.cc7 33.Wb1 ? 3 3 . .l:!.h3 ± . 33 ...a5 33 . . . 'lWcs ± . 34.g4 a4 3 5 . .&!.f3 'lWd6? 36 ..l:!.f6+- 'lWc5 37.'lWg3 b4 38.axb4 'lWc2+ 39.Wa1 a3 40.bxa3 I:!.xd5 41 .exd5 tll d 4 42 ..l:i.xf7+ .!axf7 43.'lWe5+ Wf8 44.'lWb8+ Wg7 45 . .l:i.xf7+ Wxf7 46.'lWxb7+ We8 47.'lWb8+ Wd7 48.'lWa7+ Wd6 49.'lWxd4 'lWc1 + 50.@a2 'lWc2+ 51 .'lWb2 'lWc4+ 52.@a1 'lWxg4 53.'lWf2 @xd5?? 53 . . . 'lWdl + 54.@b2+-. 54.'lWc5+ 1 -0

Game 3.1 2 (D7 0) Laznicka, Viktor Zhigalko, Andrey Warsaw Ech-tt 2 0 1 3 (8.6) 1 .d4 tll f 6 2.c4 g6 3 .f3 d5 4.cxd5 tll x d5 5 .e4 tll b 6 6.tll c3 �g7 7.�e3 0-0 8.'lWd2 tll c 6 9.0-0-0 'lWd6 1 O.tll b5 'lWd7 1 1 .@b1 f5

A new try. 1 2.h4 l 2 .tll h 3 + 0. 1 5 was analyzed in the pre­ vious game. The game continuation is a rather good gambit. 1 2 ...fxe4 1 3.h5

Chapter

3:

T h e N eo - G r ii n fe l d

1 6 ..."iYxdS? 1 6 . . . �f5 ! 1 7 .'it> a l �xd5 1 8 . hxg 6 �xd4 l 9 . gxh7 + 'it>h8 2 0 .�xd4 .2.xd4 2 l .�xd4 �f5 2 2 .�e2 c5 2 3 . l:rd6 �ae8 2 4.b3 exf3 2 5 .�xf3 l:re l + 2 6 .l:rd l .!lxd I + 2 7 .�xd 1

1 3 ... es 1 3 . . . exf3 1 4.tLlxf3 a6 1 5 .tLlc3 °iYd6 1 6 .hxg6 �f5 + 1 7 .'it>a l �xg6 1 8.�d3 �xd3 l 9.°iVxd3 �g6 2 0.tLle4 tLlb4 2 1 .�b3+ e6 2 2 .�xb4 'i!Yxe4 2 3 .�c l = Analysis diagram

Analysis diagram

+0 . 1 1 . White cannot keep his extra pawn, but he retains some endgame ad­ vantages. 1 7.hb6 �xd2 1 8.J:i'.xd2 axb6 1 9.�c4+ 'it>h8 20.hxg6 h6 2U xe4 + 0 . 8 0 . White is half a pawn up with a good position. 21 ... j)_g4 22.ttJf3

0 . 0 0 . White has full compensation for the pawn, but not more. 1 4.dS 1 4.hxg6 (+0. 1 8) was a good alterna­ tive. 1 4...ttJd4 1 5.ttJxd4 exd4 1 6.�xd4

22 ....l::!.a eS? 2 2 . . . �fe8 + 0 . 7 6 . 23.tLlgS �f6? 24.tLlh7+- l:!.xe4 25.�d3 .l:td4 26.tLixfa �gs 27.�dd1 'it>g7 28.�c2 l:!.xd1 + 29.�xd1 j)_fs+ 30.'it>a1 j)_e7 31 .�f1 1 -0 115

Chapter 4

The Samisch with . . . c 7-c5 In this section we consider the transposition to the Samisch Variation of the King's Indian with the move 6 ... cS. Our main move order is: 1 .d4 tt:Jf6 2.c4 g6 3.f3 �g7 4.e4 d6 5.tt:Jc3 0-0

Now White can choose between 6.ile3 and 6.tLlge2 (I don't consider 6.�gS due to the reply 6 ... a6, intending . . . c7-cS and ... b7-bS with a fairly good version of the Benko Gambit) . These moves usually transpose, but when Black replies 6 ... cS each move has an independent follow-up. It was the discovery that 6.�e3 cS is a sound gambit that turned the Samisch into a secondary line against the King's Indian. Before this discovery in the 1 9 70s by my good friend grandmaster Roman Dzindzichashvili, Black tried various moves like 6 ... b6 and 6 ... tLlbd7 to prepare ... c7-cS, but it is obviously better to play it right away if the gambit is sound. Since the gambit lacks a name I suggest 'Dzindzi Gambit' . But if 6 .ile3 fails to prevent ... c7 -cS , perhaps it is not the best move. This is the thinking behind 6.tLlge2, intending to meet 6 ... cS by 7 .dS (7 .�e3 is of course okay but that's just a transposition to 6.�e3 cS 7 .tLlge2 ) . Then the dark-squared bishop will usually be happier on gS than on e 3 . If Black kicks it with . . .h7-h6 the bishop then goes to ile3 , in which case the eventual �d2 will gain a tempo by hitting h6. In Game 4. 1 we examine the gambit accepted ( 6.�e3 cS 7 .dxcS dxcS 8.�xd8 .ld.xd8 9 .�xcS tLlc6) . Notice that Black has three more pieces developed than White (counting the castled king as a developing move) , which is traditionally considered adequate compensation for a pawn. White has several serious tries here; I believe that Black can always achieve full equality, but he has to react differently to each line so there is a fair amount Black should learn if he intends to play the gambit. De117

S a b o t a g e t h e G r iinfeld

pending on White's tenth move, Black may reply . . . b 7 -b6, . . . �e6 , or . . . tbd7 . I give all the white tries together with what I consider the best black replies to them, plus an unknown miniature with the gambit by Dzindzi himself. Since I am inclined to favor switching to the King's Indian as Black against 3 .f3 , I felt it was necessary to include this analysis for the benefit of those readers who need a defense to 3 .f3 . Next we begin the analysis of the main line 6 .�e3 c S 7 .tbge2 tbc6 8 .dS . In Game 4. 2 we explore the rather rare 8 . . . tbaS . In the game Black played Benko­ Gambit style and didn't get enough for the pawn, but frankly I can't understand why he didn't prepare . . . b7-bS by . . . i::rb 8 and . . . �d7. The computers rate this as fully equal, although I suppose the two elite players must have assessed this as at least slightly better for White due to Black's pieces being in each other's way to some degree. In any case I show that 9 .b3 instead of the played 9 .tbg3 may save a tempo against the . . . l:!.b8 , . . . .itd7 plan and should keep a plus. In Game 4.3 we look at the usual 8 . . . tbeS , which led to a Kasparov-recom­ mended piece sacrifice. It led to a three minor pieces vs queen and pawn battle, which Svidler (White in the game) , Komodo and I all rate as much in White's favor had he played differently on move 24, although Houdini disagrees, probably due to a tendency to overvalue the queen. Black has several ways to deviate, but none seem to equalize, although the positions are messy and hard to evaluate. Svetushkin thinks that one of these is okay for Black, but he does not mention the line given here. So at this writing, the line looks good for White by a three to two vote! The next two games feature 9 ... e6 rather than the 9 . . . hS of Game 4.3 . After trading pawns and inserting ... a7-a6 and a2-a4, Black can choose between 1 2 .. hS of Game 4.4 or l 2 . . . �d7 of Game 4. S . It seems to me that 1 2 . . . hS is not as logical here as in Game 4.3 because White can reroute the knight via h l to the good square f2 , and in­ deed White keeps a plus in more than one way here. Instead 1 2 . . . �d7 achieves the advance . . . b7-b S , which is usually thought to mean equality for Black in the Benoni, but here White achieves a quick attack with f3-f4 and e4-e S , which is likely to in­ duce Black to play a reasonable but perhaps not quite fully sound piece sacrifice. In Game 4. 6 we consider the alternative defense 7 . . . 'ifaS , which is probably the safest move and perhaps the best. As the notes show, a likely result (if both oppo­ nents are well-prepared) is a transposition to a Maroczy Bind vs. the Accelerated Dragon, but with Black having the option of reaching a normal position of the bind with the queens off, based on new original analysis here. I think that this increases Black's chances to equalize, so although I would still prefer the white side I don't think it would be very difficult for Black to hold. Finally we come to the critical line 6.tbge2 cS 7 .dS in Game 4. 7. This is the way for White to play if he really wants to win the game, as it avoids the rather drawish line mentioned above. I think White keeps a plus in all lines, but the rare l l . . . tbe8 ! line given there (planning to prepare . . . b7-bS by a later . . . tbc7 while avoiding the pinning �gS) leaves some doubt about it. Usually Black instead plays . . . h7-hS at some point, but Komada and I like White in those lines. If Black avoids 6 . . . cS , then 6 . . . eS 7 . .ite3 or 6 . . . tbc6 7 . .ite3 will transpose to lines given in this book via the 6 . .ite3 move order in most cases. 118

C h ap t e r

4:

The Somisch wi th . . . c 7 -c5

Game 4. 1 (E8 I ) Kovalenko, Igor Chojnacki, Krzysztof Trzcianka 2 0 1 3 (7 .4) 1 .d4 tLlf6 2.c4 g6 3.f3 ii.g7 4.e4 0-0 5.tLlc3 d6 6.ii.e3 c5 7.dxc5 dxc5 8.�xd8 .l:!.xd8 9.ii.xc5 tLlc6

Roman Dzindzichashvili

1 0..ie3 This move is rare but it requires accu­ racy by Black. White postpones the committal I O.tLlge 2 . The alternatives are: A) I O .tLldS A l ) 1 0 ... tLlxdS ?! 1 1 .cxdS b6 1 2.ii.a3 tLld4 1 3 .0-0-0 e6 1 4.tLle2 tLlbS 1 5 .�e7 .l:!.d7 1 6.d6?! ( 1 6.�gS!t) 1 6 ...ii.es 1 7.f4 hd6 1 8.�f6? �b7 1 9.eS nc8+ 20.@b l

given to me from memory by 'Dzindzi' , the game is not in any data­ base. I included it both because the game is pretty and shows the power of the gambit, and because 'Dzindzi' was the inventor of the gambit; A2) I O . . . tLld7 1 1 .�xe7 ( 1 1 .tLlxe7+ tLlxe7 1 2 . .txe7 ilxb2 1 3 .�b l (in case of 1 3 .ilxd8 ilxa l 1 4.tLle2 + 0 . 0 7 , Black's better development and better structure compensate for the pawn mi­ nus) 1 3 . . . .tc3 + 1 4.Wdl ld.e8 1 5 .tLle2 .tes 1 6 ..tb4 tLlb6 1 7 .tLlc3 aS 1 8 .ila3 ii.xc3 I 9 .l:rxb6 ilb4 20.�c l nds+ 2 l .@c2 .td7 2 2 . ile2 ilc6 2 3 .�e3 fS= ; White will have to surrender his extra pawn for tactical reasons) I 1 . . . tLlxe7 1 2.tLlxe7+ Wf8 1 3 .lLldS �xb2 1 4 . .l:!.b l ii.g7 I 5 .tLle2 tLlcS I 6.tLlec3 .te6=.

Analysis diagram

2 0 . . . tLla3 + ! ( 2 0 . . . ilxeS ! ! was even better) 2 1 .bxa3 ? .ixe S ! 2 2 . .ixeS .l:!.xd l + 2 3 . Wb2 .l:!.d2 + 24.Wal ii.e4 2 5 . tLlc3 .l:!.xc3 ! and White resigned, Dlugy-Dzindzichashvili, rapid, New York, circa 1 99 0 . The moves have been

Analysis diagram

1 19

S a b o ta g e

the

G rii n fe l d

Black's better pawn structure together with White's 'bad' bishop offset the half-pawn material deficit. Both sides have good knight outposts, but White's second knight is redundant. B) 1 O.tll g e2 b6 l l . .ia3 .ia6 1 2 .tll b S .ib 7 1 3 . .l::!. d l tll d 7 1 4. tll e c3 tll c 5 1 5 .�xcS l:lxd l + 1 6 .@xd l bxc5 1 7 .tll a4 a6 1 8 .tll bc3 �d4 1 9 .@c2