Aveskulov - Attack with Black 2012.pdf

Attack with Black Valery Aveskulov

First published in the UK by Gambit Publications Ltd 2012 Copyright© Valery Aveskulov 2012 The right of Valery Aveskulov to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form or by any means (electronic, mechanical, photocopying, recording or otherwise), without prior permission of the publisher. In particular, no part of this publication may be scanned, transmitted via the Internet or uploaded to a website without the pub­ lisher's permission. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damage. ISBN-13: 978-1-906454-39-5 ISBN- 10: 1-906454-39-6 DISTRIBUTION:

Worldwide (except USA): Central Books Ltd, 99 Wallis Rd, London E9 5LN, England. Tel +44 (0)20 8986 4854 Fax +44 (0)20 8533 5821. E-mail: [email protected] Gambit Publications Ltd, 99 Wallis Rd, London E9 5LN, England. E-mail: [email protected] Website (regularly updated): www.gambitbooks.com Edited by Graham Burgess Typeset by Petra Nunn Cover image by Wolff Morrow Printed in Great Britain by the MPG Books Group, Bodmin and King's Lynn 10 9 8 7 6 5 4 3 2 1 Gambit Publications Ltd Managing Director: Chess Director: Editorial Director: German Editor: Webmaster:

Murray Chandler GM Dr John Nunn GM Graham Burgess FM Petra Nunn WPM Dr Helen Milligan WPM

Contents Introduction

4

Symbols

6

White Avoids the Benko 1

Diemer, Veresov and Trompowsky

7

2

Colle, Zukertort, London and Torre Systems

22

3

Anti-Benoni 4lt:Jc3

37

4

Blumenfeld Gambit

42

5

Blumenfeld Gambit: 5 .tg5

55

6

1 d4 lZ'lf6 2 c4 c5: 3 e3 and 3 dxc5

65

7

Vaganian Gambit

80

8

Vaganian Gambit: 7 e3

92

The Benko Gambit 9

Benko Gambit Declined

104

10

Benko: Zaitsev, Dlugy and Modern Lines

120

11

Benko Gambit: 5 b6

137

12

Benko Accepted: Introduction and Rare Lines

148

13

Benko Accepted: King Walk

158

14

Benko Accepted: Fianchetto Lines

177

15

Benko Main Line with 10 J:.b1

192

Understanding the Benko

16

Dream Positions for Black

200

17

Positions to Avoid

207

18

Tactical Exercises

212

Solutions

217

Index of Variations

223

I ntroduction

The aim of this book is to present a complete repertoire for Black against 1 d4. The cornerstones of the repertoire are two gambits that are aggressive but positionally motivated: the Benko Gambit and the Blumenfeld Gambit. While seeking out lines that give Black attacking chances, I have also striven to recommend only those variations which are theoretically sound and reliable. Throughout the book, I have attempted to explain the most relevant strategic concepts for both players, cater for tricky move-orders, and to analyse new moves and ideas for both sides. I hope that by covering logical but untested ideas for White, I have to some degree 'future­ proofed' the book. But why did I decide to write this book on this topic, and at this point in time? Allow me a brief digression. Our generation has the honour to live in the computer era. Besides the obvious advantages, it also puts certain duties on us; this applies to practically all aspects of life, and chess is no exception. In the 1960s and 1970s, top chess-players spent months studying an opening to find new ideas, and just to gather relevant material they had to subscribe to many chess magazines and manually assimilate all the data. If somebody played a strong novelty, even in a major event, it could take months before it became well-known to the chess community. Nowadays we can learn the main line of any opening in an hour using a database (though this does not mean we can play it well!). If a novelty is played in a top tournament, a good chess-player watching it live online can figure out even before the game has ended whether it is a really strong idea or if it's just a one-time bomb to surprise an oppo­ nent. The speed at which information travels is amazing and we need to take this into account when choosing our openings. I feel it makes sense to have some lines that are special to us, that we have prepared in ways that cannot be replicated by our opponents working with their computer in the few hours before they face us. But which sort of openings should we study: sharp and fashionable or rare and half-correct? Poor opening choices can ruin your efforts in chess. If you like to play quiet positions with pawn-chains, then the Sicilian Dragon shouldn't be in your repertoire. And if you are a good tactical player, forget about defending the Queen's Gambit Declined with Black; learn the King's Indian or Benko Gambit! It will immediately help you to get better results. From my experiences as a coach of players of a wide variety of ages, I have discovered that many players have little notion of how to study a new opening

INTRODUCTION

5

variation. They have a limited grasp of ChessBase, don't know how to use an en­ gine effectively and as a result their repertoire barely gets them to move 10 with­ out a mishap. I assisted them first with advice about which opening lines to choose, and then I offered some brief lines of that or another variation. Later I wrote some opening articles that started to appear in chess magazines. Even­ tually, the idea formed: "Maybe I could write a book on the opening to assist even more players." The question "Which opening should it be about?" was solved very quickly. When I was 7 or 8, a coach presented me with a handwritten 1 70-page mono­ graph that unfortunately has never been published. It was dedicated to the Benko Gambit! If I had been asked at that point to explain why I should sacrifice my b-pawn, I could hardly have done so, but I started to sacrifice it and I loved the games I got. Later when I began to face stronger opponents, my Benko games became more difficult. I took up openings like the Nimzo-Indian and Queen's Gambit Declined in order to resist against solid players, but against weaker play­ ers I still preferred to play the opening from my childhood. As I became stronger, I began to understand more and more about the Benko, but there was never time to devote serious study to my favourite opening. But my students often asked about Benko lines, and I found myself increasingly drawn into the subject. Grad­ ually, more and more Benko analysis appeared on my computer. When I asked Graham Burgess (Editorial Director of Gambit Publications Ltd) whether a Benko guide might be of interest, he asked me if I could extend it into a complete open­ ing repertoire against 1 d4. I readily agreed with his idea since most of the other variations were also worked out on my computer and I already had many inter­ esting ideas for how to complete the repertoire. That is how the idea of this book was born. So, if you feel at home with aggressive gambit chess, this book is for you! The last three chapters of the book feature instructive material on the Benko Gambit: positions you should aim for, ones to avoid, and positions for solving (both stra­ tegic and tactical). If you are new to the Benko, you may wish to study these chapters before anything else - the more detailed discussion should make more sense then! I sought to make this book interesting for both professional players and ama­ teurs. There are a great many new ideas that should be viable in high-level games, while there are also plenty of verbal explanations. I very much hope you will like the results of my work. Also I shall be very thankful if you send me your feedback, ideas, questions or even complaints to me by email: [email protected]. Finally, I would like to thank Graham Burgess, who professionally and coop­ eratively assisted me throughout the whole process of writing this book; and Al­ exander Moiseenko, who helped me to believe that I can analyse openings at a high level.

6

A ITACK WITH BLACK

And many thanks to my darling wife Irina who gave us a boy, Ivan, on 9th March 2012 and made me unbelievably happy! Valery Aveskulov July 2012

Symbols X

+ ++ # !! !? ?! ? ?? +± ;!;

=

+ -+ Ch tt 1-0 lf2- lh 0-1

+

(n) (D)

capture check double check checkmate brilliant move good move interesting move dubious move bad move blunder White is winning White is much better White is slightly better the game is equal Black is slightly better Black is much better Black is winning championship team event the game ends in a win for White the game ends in draw the game ends in a win for Black nth match game see next diagram

1 Diemer, Ve resov a nd Trom powsky 1 d4

w

liJf6 (D)

A)

2 tt:Jc3 dS 3 e4?

This is a form of the Blackmar­ Diemer Gambit, a very rare opening at higher levels but it has many adher­ ents among club-level players. An objective evaluation of this gambit is 'not correct' . Its general idea is to give up a pawn for rapid development - a laudable enough aim, but right here this idea can hardly be recom­ mended. Black takes the pawn practi­ cally for free. 3 ...tt:Jxe4 4 tt:Jxe4 dxe4 (D)

Our main lines in this book start af­ ter 2 liJf3 and especially 2 c4. How­ ever, White can direct the game in a w very different direction on his second move, and this chapter is devoted to these lines. In increasing order of im­ portance, we have: 7 A: 2 tt:Jc3 dS 3 e4? 8 B : 2 liJc3 dS 3 ..tgS C: 2 ..tgS 12 Given our repertoire preferences, the move-order 2 c3 gives us no prob­ lems as we simply meet it with 2 ... d5. None of White's continuations give After the rare 2 g3, Black can reply him sufficient compensation. 2 ... c5 3 d5 b5 4 ..tg2 d6, with active 5 ..tc4 play. Or:

AITA CK WITH BLACK

8

a) 5 ..lte3 i.f5 6 g4 ..ltg6 7 lt:\e2 (Ferreira-Chauca, Rio de Janeiro 2008) 7 ...lt:\c6 8 lDf4 'ied6 + intending ...0-0-0 and ... e5. b) 5 i.f4 lt:\c6 6 c3 e6 7 'i¥c2 f5 ! 8 0-0-0 i.d6 + M.Pfeifer-Neckar, Czech Team Ch 2001/2. c) With 5 f3 e5! Black gives the pawn back in order to get a develop­ ment advantage. 6 dxe5 (6 fxe4? 'ii'h4+ -+; 6 ..lte3 exd4 7 'iiixd4 'i!Vxd4 8 ..ltxd4 lt:\c6 9 i.b5 i.d7 10 i.c3 exf3 I l lt:\xf3 0-0-0 12 0-0-0 f6 + Helin-Yurenok, Caleta 2012) 6 ...'iiix dl + 7 �xdl i.f5 8 f4 lt:\c6 9 c3 0-0-0+ 10 �e1 i.c5 +. 5 lt:\c6 6 d5 lt:\a5! This new move improves over Abbasifar-Tanaka, Istanbul Olympiad 2000, and forces the exchange of the light-squared bishop. 7 'i!ie2 (7 i.e2? e6 +; 7 i.b5+? c6 8 dxc6 'i!ixd 1 + 9 �xdl bxc6 +) 7 ... lt:\xc4 8 iixc4 e6 9 dxe6 ..ltxe6 10 'ii'xe4 'ii'd5 ! I I 'ii'xd5 i.xd5 + gives Black a very pleasant endgame with the bishop-pair. .••

B)

2 lt:\c3 d5 3 i.g5 (D)

This is the most common move­ order to reach the Richter-Veresov At­ tack (we shall call it the 'Veresov' for short). Other versions are 1 d4 d5 2 lt:\c3 lt:\f6 3 i.g5, I d4 lt:\f6 2 Ji.g5 d5 3 lt:\c3 and I lt:\c3 lDf6 2 d4 d5 3 i.g5. The opening was named after the Ger­ man International Master Kurt Richter (who played it from the 1920s to the 1940s) and later the Soviet master Gavriil Veresov ( 1950s-1970s ). The Veresov has never been popular at the highest levels, though some top players have employed it on occasion. It has appealed mostly to maverick players such as Spassky, Tal, Larsen, Bronstein, Miles, Alburt and Moroze­ vich, who have never objected to irreg­ ular play on the board. Nowadays you can find this variation in games of such GMs as Hector, Chemyshov, Khachian and Miladinovic. White has two main ideas: the first is to damage Black's pawn-structure with i.xf6 (not considered so danger­ ous nowadays) and the second is to make a pawn advance in the centre with f3 and e4 (often after 0-0-0). 3 lt:\bd7 (D) This is the most popular response to the Veresov, as Black prepares a strong response to the e4 plan, while also avoiding doubled f-pawns. Now: 9 B l : 4 iid3 10 B2: 4 e3 B 3 : 4 f3 10 12 B4: 4lt:\f3 The third of these, preparing the e4 advance, is the most consistent with White's aims in the Veresov, but ...

DIEMER, VERESO V AND TROMPOWSKY

9

bl) 5 iLf4 c6 6 f3 b5 7 a3 trans­ poses to note 'b' to White's 5th move in Line B3. w b2) 5 iL.h4 c6 6 f3 e6! (this quiet move prevents White's e4 advance) 7 e3 (7 e4? tLlxe4 + Misanovic-D.Ivan­ isevic, Belgrade {women} 1990) 7 ... b5 (Black starts an attack on the queenside that is effective since White hasn't created any tension in the centre) 8 ..if2 a6 (Black adopts a French De­ fence plan that is especially good with white pawns on e3 and f3) 9 tt:Jge2 c5 doesn't work very well if Black plays 10 g4 iL.b7 I I iL.g2 �c8 + Moroze­ precisely. The fourth is the most popu­ vich-Ehl vest, Podolsk 1993. lar move. 81) There are a few alternatives: 4 'iWd3 a) White can sacrifice a pawn with White prepares the e4 advance, but 4 e4?! tLlxe4 5 tLlxe4 dxe4 6 f3 but this Black has a nice way to hinder this clearly can't work out well. 6 ... h6 and thrust: now: 4 ...h6 5 iLh4 c6! (D) a1) 7 iLh4 loses control of the e3square: 7 ... c5 ! 8 dxc5 (8 d5? 'iWb6 9 l:tb1 g5 10 iL.g3 iL.g7 1 1 c3 'iWg6 + Gomes-Epishin, Las Palmas 1997; w Black has kept the extra pawn) 8 ... e3 9 'iWd4 e5 ! +. a2) 7 iL.f4 (Lodi-G.Kovacs, Hun­ garian Team Ch 1999/00) 7 ... c6 (as is often the case, the simplest reply to a gambit is to give back the pawn for rapid development) 8 fxe4 e5 ! (and Black can even sacrifice one of his own) 9 dxe5 'i¥a5+ 10 'iVd2 'iWxd2+ 1 1 Wxd2 tLlc5 1 2 ..id3 ..ie6 1 3 tt:Jf3 0-0-0 with ideas of ...tt:Jxe4+ and ...tt:Ja4. 6 tLlf3 Or: b) White can also play 4 'iWd2 in­ tending 0-0-0, f3 and e4 - a delayed a) 6 e4? is not good now because of form of the idea we see in Line B3. a geometric idea: 6 ...tt:Jxe4 7 tt:Jxe4 4 ...h6 and now: dxe4 8 'iWxe4? (it's better not to take

AITA CK WITH BLACK

10

this pawn, even though 8 'ili'd2 tt'lb6 leaves Black with a clear extra pawn) 8 ... g5 9 i.g3 'i!Va5+ 10 c3 f5 1 1 'i!Vf3 f4 + and Black wins a bishop. b) 6 0-0-0?! invites a quick attack on the white king: 6 ... b5 ! 7 tt'lf3 (Ge­ lashvili-Halkias, Erevan Zonal 2000) 7 ... b4! 8 tt'lbi (8 tt'la4? 'ti'a5 9 b3 i.a6 10 'ife3 i.b5 +) 8 ...'ili'a5 9 a3 e6 +. 6 'i!Vas 7 tt'l d2 7 0-0-0? tt'le4! +. The c3-knight is tied to the defence of a2. 7 a3 e6 8 tt'ld2 (Sengupta-N.Mame­ dov, Hastings 2007/8) 8 . .'it'b6!? (a new move) 9 0-0-0 e5 ! 10 e3 ii.d6 +. 7 'ili'b6 8 0-0-0 eS ! 9 dxeS tt'l xeS 10 'i!Vg3 Now: a) Giannakoulopoulos-Dvoirys, Ano Liosia 2000 featured 10 ... tt'lg6?! I I ii.xf6 gxf6, but now 12 tt'lb3!? (in­ tending e4) 12 ...f5 13 e3 ;;\; would have led to a position with a weird pawn­ structure for Black. b) It is safer to avoid the doubling of the pawns: 10 ...tt'lfd7!? Il f4 tt'lg6 12 e4 d4! (the natural 12 ...tt'lxh4? leads to problems due to 13 exd5!; e.g., 13 ... tt'lf5 14 'i!Ve1 + 'iti>d8 15 tt'lc4 'ili'c7 16 dxc6! with a very strong attack) 13 tt'le2 c5. Black has the ... tt'lxh4 idea in hand and has avoided problems with the king in the centre. .••

that there is no guard on the queenside - the bishop is on the other side of the wall. s 'ifb6 6 .litbt White can't defend the b2-pawn in­ directly by 6 a3? because of 6 ... h6 7 i.h4 tt'lg4! 8 'iVc1 g5 ! 9 i.g3 (9 fxg5 hxg5 I 0 i.xg5 tt'lxh2 {intending to play ... tt'lf3+} 1 1 'ifi>d1 e5 + with a strong initiative) 9 ... tt'ldf6 10 tt'lf3 tt'lh5 1 1 i.e2 i.g7 + intending to take on f4 and then on d4. ...

6 e6 7 tt'l f3 (D) •••

•••

Now I recommend 7 ... i.b4! (a new move, improving over Napoli-Mrsevic, Nis 2008) 8 ii.d3 c5 9 0-0 c4! 10 i.e2 i.xc3 1 1 bxc3 'ii'a5 12 tt'le5 tt'le4 +. 83) 4 f3

This aggressive interpretation of the Veresov turns out not to work well. 5 tt'lf3 h6 6 i.h4 transposes to Line 4 c6! B4. This move is important, as it pre­ With the text-move, White seeks pares a strong reply to White's e4 ad­ slow manoeuvring play in the Stone­ vance. wall style. However, the difference is 5 e4?! 82) 4 e3 c6! 5 f4? !

..•

DIEMER, VERESOV AND TROMPOWSKY

White pushes on regardless. This leads to the most interesting play, but it is objectively better to delay this ad­ vance with 5 'i¥d2. Then 5 ... b5 ! is a strong and typical idea. Black starts queenside play thanks to the fact that White has created no tension in the centre. Besides the obvious idea of playing ... b4, Black prepares an out­ post on c4. Then: a) 6 e4 involves a pawn sacrifice: 6 ... b4 7 lt:Jd1 dxe4 8 lt:Jf2 (8 'iVxb4 e5 9 �a4 exd4 10 'iWxc6 l:tb8 1 1 .i.xf6 'iYxf6 12 'ifxe4+ �d8! + Summer­ scale-Miezis, Cappelle la Grande 1999; Black intends to play ....1i.b4+ and ...l:.e8) 8 ... e3 ! (disrupting White's de­ velopment) 9 i.. xe3 e6 10 .i.d3 .1i.e7 1 1 lZJe4 lZJd5 12 i.. g5 0-0 13 ll'le2 c5 Heyken-Kengis, Hamburg 1995. Black has comfortable play; the f3-pawn looks ugly. b) 6 a3 h6 7 .1i.f4 e6 8 e4 lt:Jb6 9 lt:Jh3 .te7 10 lt:Jf2 a5 is equal, Ochoa­ Rivas, Medina del Campo 1980. Black has a good version of the French De­ fence. =

5 dxe4 6 fxe4 (D) •..

11

6 e5! •.•

Black grabs dark squares in the cen­ tre. 7 dxe5 White can't hold the dark squares by 7 lt:Jf3 because of 7 ... 1!i'b6 8 dxe5 'i¥xb2 9 i..d2 (9 :b1 'ti'xc3+ 10 i..d2 'iWc5 1 1 exf6 lt:Jxf6 + Berges-Prie, Fouesnant 1 997; White has no com­ pensation for the pawn) 9 ... lt:Jg4 10 Ilb1 'iVa3 II e6 fxe6 12 lZJg5 lZJde5 +, when Black is a pawn up. 7 'iVa5 ! This is the main point of Black's play. 8 i..xf6 Or: a) White can win a pawn by 8 exf6 'i!Vxg5 9 fxg7 i..xg7, but Black's total control of the dark squares promises him the better chances. For example, 10 'iVd2 �xd2+ 1 1 �xd2 lt:Jc5 12 .1i.d3 .1i.e6 13 lt:Jf3 0-0-0 14 �e2 b5 15 a3 a5 + Alburt-Tal, USSR Ch, Baku 1972. b) 8 'iYd2lZJxe5 9 0-0-0 .i.e6 10 lt:Jf3 (Vooremaa-Bronstein, Tallinn 1981) 10....i.b4 +. White's pawn-structure will be terrible after ... i..xc3. 8 gxf6 White is fighting to avoid a very se­ rious disadvantage: a) 9 e6 fxe6 10 i..c4 i.. a3 ! (a typi­ cal tactic) 1 1 �cl (Vallieres-Lesiege, Quebec City 2004) 1 1 ...'i\Yxc3+! 12 bxc3 i..xc 1 13 l::txc 1 lt:Jc5 ! 14 i.d3 .l:.g8 15 'ifi>f2 e5 +intending ... i..e6. b) 9 exf6 .i.a3! (the strongest move, again emphasizing the weakness of White's dark squares) 10 'i!Vcl lt:Jxf6 1 1 i..d3 lt:Jg4! (intending a double attack by ... 'i\Yb6) 12 h3 (12 bxa3? allows .•.

...

A ITACK WITH BLACK

12

Black a very strong attack: 12 ...1'ixc3+ 13 �e2 'i¥d4 14 tt:lh3 l:tg8 + Brauer­ Crosa Coil, Mendoza 2004) 12 ...1'ib6! 13 tt:ld1 �b4+ 14 c3 tt:le5 15 �e2 �e7 leaves White's position badly dam­ aged. 84)

4 tt:lf3 Thi� is the number one choice for White although it does not fit with the word 'Attack' in the name of the opening: White plays solid but pas­ sive chess. Naturally, this type of de­ velopment doesn't put much pressure on Black. 4 b6 Black immediately determines the position of White's bishop. 5 .i.h4 5 .i.f4 e6 (5 ... c6 allows White to support the e4 advance with 6 1'id3) 6 e3 (now 6 'i¥d3 can be met by 6 ... c5 ! =) and now 6 ... a6! is the most aggres­ sive. Black prevents tt:lb5 and prepares to seize space on the queenside. 7 �d3 c5 8 0-0 b5 9 .l:.e1 �e7 10 tt:le5 �b7 1 1 tt:lxd7 'ii'xd7 12 dxc5 �xeS 13 �e5 1'ie7 = Mestrovic-A.Kovacevic, Nova Gorica 2004. Black has solved all his problems. 5 c6 5 ... e6 can lead to a strange position: 6 e4 g5 7 .i.g3 tt:lxe4 8 tt:lxe4 dxe4 9 tt:le5 �g7 10 h4 with complicated play. 6 e3 6 'i¥d3 transposes to Line B 1. 6 ... e6 7 �d3 �e7 8 0-0 0-0 (D) Black has fully equalized. For example: ...

••.

a) 9 �xf6?! tt:lxf6 10 e4 c5 ! gave Black a slight advantage in Deriabin­ Moiseenko, Ukrainian Team Ch, Al­ ushta 2005. Black opens lines for his bishops. b) 9 .l:r.e 1 c5 1 0 tt:le5 tt:lxe5 1 1 dxe5 tt:ld7 12 ii.g3 (Emodi-Groszpeter, Hun­ garian Team Ch 2002/3) and here I propose 12 ... a6!? intending ... b5. The position after 13 a4 'ii'c7 14 f4 c4 15 �e2 �b4 is rather unclear but I prefer Black. c) 9 tt:le2 b6 10 c3 .i.b7 1 1 .l:te1 c5 = Gukasian-Zarnicki, Internet blitz 2004. C)

2 �g5 (D)

This move defines the Trompowsky Attack. It is named after the Brazil­ ian player Octavio Trompowsky, who played it in the 1930s and 1940s. In modern times, the opening received more widespread popularity on ac­ count of the efforts of players such as Lev Alburt, Vlastimil Hort, Konstan­ tin Chernyshov, Igor Miladinovic and especially Julian Hodgson.

DIEMER, VERESOV AND TROMPOWSKY

13

is the easiest way to get a safe posi­ tion, rather than opening the h-file for White. Black will methodically com­ plete his development while the h­ pawn's advance looks irrelevant. 4 liJd2 lLlxd2 5 'iWxd2 .tf5 6 e3 h6 7 .tf4 e6 8 lL\f3 .i.d6 9 .td3 .txd3 10 'iWxd3 lL\c6 II c3 'iWe7 = Galego-Ziiger, Eu­ ropean Team Ch, Haifa 1989. Cl) 3 .i.h4 c5

The primary idea of 2 .i.g5 is to take on f6, doubling Black's pawns. White will then seek a slow manoeuv­ ring game: he will restrict Black's dark-squared bishop with pawns on c3, d4 and e3, and put his bishop on d3, and his knights on d2 and e2. The plan is that the knights will prove more nimble than the black bishops in this scenario. Although Black has reli­ able ways to handle this set-up, it is more in keeping with the aim of our repertoire to cut across White's de­ signs with an active move: 2 lL\e4 This leads to concrete positions. White will seek to gain time by attack­ ing this knight, but Black aims to make use of its influence in the centre. Now: C l : 3 .th4 13 C2: 3 .i.f4 15 The bishop retreat to f4 is the main move, but in addition to the alterna­ tive of dropping it back to h4, White can also support it with his h-pawn, odd though 3 h4 appears. Then 3...d5 •••

Naturally, Black attacks the centre. 4 f3 Or: a) 4 lLld2 'ifa5 5 c3 lL\xd2 6 'i!Vxd2 cxd4 7 cxd4 'iVxd2+ 8 �xd2 d5 offers White absolutely nothing: 9 e3 lLlc6 1 0 lLle2 e6 1 1 lLlc3 .i.d7 12 .i.e2 i..d6 13 .l:.hcl 0-0 14 lLlb5 .i.b8 = Rausis­ Inkiov, Gausdal 1989. b) 4 dxc5 lLlxc5 5 lL\c3 lL\c6 6 lLlf3 (6 e4?! reaches a kind of Sicilian, but the bishop on h4 appears awkward: 6 ...g6 7 .tb5 .tg7 8 lL\ge2 a6 9 .i.xc6 bxc6, and Black was slightly better in the game Sazhinov-Bocharov, Tomsk 2006) 6...g6 7 lLld5 .i.g7 8 c3 d6 9 e3 0-0 10 .i.b5 .l:le8 =. Black has no prob­ lems. c) 4 d5?! is dubious due to the striking manoeuvre 4 ...'iVb6 5 'ifc1 g5 ! 6 .i.g3 .tg7 7 c3 'i!Vh6! (a key move: Black exchanges off the g3bishop in excellent circumstances) 8 .i.xb8 (8 lLlh3 d6 9 lLld2 lLlxg3 10 fxg3 'ii'g6 + Rossetto-Sanguinetti, Buenos Aires 1975) 8 ... l:.xb8 9 liJd2 lLlf6 I 0 e4 0-0 1 1 .i.d3 d6 + Glek-Kharitonov, Vilnius 1984. 4 g5 5 fxe4 gxh4 (D) •••

A ITA CK WITH BLA CK

14

w

6 e3?!

I am giving this as the main line only due to its great popularity. In fact, I think it's a poor move that lets Black seize the advantage. Other moves: a) 6 lt:Jc3?! allows Black to isolate White's e2- and e4-pawns. 6 ... cxd4 7 'iVxd4 l:[g8 8 e5 lt:Jc6 9 'ifxh4 lt:Jxe5 10 0-0-0 (Teske-Kocwin, Guben 2008) 10 .. .'i!i'b6! + intending to exchange queens by .. .'i!Vh6+. White's isolated e-pawn provides Black with a long­ lasting advantage. b) 6 c3!? is not the most popular move, but it looks clearly the most logical. White avoids isolated pawns on the e-file and does not weaken any new squares. There is limited practical experience here; the main line so far is 6 . .'i¥b6, but I prefer to offer some­ thing new and interesting. One idea is 6 ... i..h6!?, but I'll focus on 6 ... d5 !?, blowing open the centre to provide more lines for his bishops. Then: b1) 7 e5? is an attempt to keep the game closed, but it doesn't work well. 7 . ..ti::Jc6 8 lt:Jf3 �b6 9 'ifb3 (9 b3?! i..h6 + gives Black a strong initiative

thanks to the absence of a pawn on f2) 9 ... cxd4 10 'ii'xb6 axb6 1 1 cxd4 i..h6 12 lt:Jc3 lt:Jb4 13 xf2 dxe3+ 15 'it>fl , Mc­ Donald-Wells, Southend 2008, and now 15 .. .'iM4! wins right away: 16 lita2 i..e6 17 litc2 .litfd8 1 8 bxa5 .Yi.b3 -+) 12 ... 'iVg5 13 0-0 li.g4 14 Ae2 i..h 3 15 i.. f3 .Yi.c7! +. 7 ltJxc3 Or: a) 7 ... ..1tf5?! is unconvincing due to an immediate attack on the b7-square: 8 �b3 ltJxc3 (d5 was also hanging) 9 bxc3 �d7 10 ltJe5 ltJxe5 1 1 Axe5 (in­ tending e4 and il.b5) l l...a6 12 e3 and ..•

DIEMER, VERESOV AND TROMPOWSKY

17

Black should play ... f6 at some mo­ ment to develop the f8-bishop. Later White will play c4, exposing Black's vulnerability on the a2-g8 diagonal. b) But Black can start with 7 .. .f6!?. This covers the e5-square and amongst other things prevents l2Je5. The c6knight is useful in these positions - it avoids unpleasant checks on the a4-e8 diagonal, attacks d4 and is ready to move to b4 or a5 to attack on the queenside; on the other hand, the f3knight only defends d4. Black's next few moves could be ... i.f5, ... e6, ... i.d6, ... 0-0, etc., or the same but leaving the bishop on c8. 8 bxc3 g6 This is a simple solution to Black's problems, leaving him a tempo up compared to a position that can arise from the Exchange Slav or a quiet form of Griinfeld (e.g., Portisch-Kram­ nik, Biel Interzonal 1993). After 9 e3 i.g7 10 i.e2 0-0 1 1 0-0 i.f5 Black is comfortably equal.

partial compensation for the pawn). 6 ...l2Jd5 ! is a key move for Black. avoiding any complications related to accepting the gambit pawn. Now: b1) 7 %Vd2 l2Jxf4 8 %Vxf4 %Vxd4 9 l:.d 1 e5 ! (forcing a favourable liquida­ tion) 10 l:txd4 (10 %Vg5? h6 is much better for Black) 10 ...exf4 1 1 l:txe4+ 'iit>d8 12 l:txf4 (12 i.c4, Dunworth­ D.Buckley, Monmouth 2000, 12 . .f5 ! 13 l:txf4 l2Jc6 + intending ... g5 or ... i.d6) 12 ... ..ie6 +. Black's bishop­ pair will be powerful in the endgame. b2) 7 l2Jxd5 %Vxd5 8 i.xc7 (8 %Vd2? yields nothing good: 8 ... l2Jc6 9 i.e3 e5 ! 10 fxe4 %Vxe4 1 1 ..id3 %Vh4+ 12 g3 %Vh5 13 d5, Kanep-Nureev, Internet blitz 2003, 13 ...l2Jb4 +) 8 ... l2Jc6 9 c3 .i.f5 10 .i.e2 (Djurhuus-Elsness, Gaus­ dal 1995) IO ... e6 1 1 fxe4 i.xe4 12 l2Jf3 l:tc8 13 .i.g3 .i.d6 =. Black is fine here. c) 5 c4 e6 6 l2Jc3 c5 ! (D) .

C22) 4 f3 l2Jf6 5 l2Jc3

w

Or: a) 5 e3?! doesn't fit in with White's previous play (i.e. f3), and 5 ... c5 ! ex­ ploits this. Black has a development advantage after 6 c3 l2Jc6 7 l2Jd2 e6 8 i.g5 e5 9 dxe5 l2Jxe5 10 e4 i.e7 + M.Gurevich-Cvitan, Vrsac 1985. b) 5 e4 is an interesting pawn sacri­ fice. After 5 ...dxe4 6 l2Jc3 (following 6 fxe4 l2Jxe4 7 %Vd3, as in R.Leyva­ Again Black attacks his opponent's Vazquez Igarza, Cuba 1997, Black can awkward pawn-structure in the centre simply continue 7 ...l2Jf6 8 l2Jc3 l2Jc6 9 - the pawn on f3 looks disharmonious. 0-0-0 l2Jd5 +, when White has only Now:

A ITACK WITH BLACK

18

c1) With the continuation 7 tlJb5 tt:Ja6 8 e4 (Devereaux-Gallagher, Brit­ ish League (4NCL) 2004/5) White tries to get some initiative but his pieces are unprepared for such play. Here I rec­ ommend the new move 8 ...cxd4! - as is often the case in complicated positions, the most important thing is to castle; af­ ter that everything becomes clear. 9 cxd5 i..b4+ 10 'iii>f2 0-0 + gives Black an obvious advantage. c2) 7 e3 cxd4 8 exd4ltJc6 9 c5 i.e? was equal in the game Santacruz­ Pineda, Thessaloniki Olympiad 1988. Everything would be quite typical if White's pawn weren't on f3. It just gets in the way ofWhite's plans, while Black has no such problems. 5 e6 6 e4 After 6 'ifd2 c5 ! (with a pawn on c5 it's very hard for White to arrange his planned e4 advance) 7 e3 Black can advance his queenside pawns in much the same way that we saw in a couple of Veresov lines: 7 ... a6 8ltJge2 lDc6 9 g4 b5 10 i..g2 i.e? 1 1 0-0 b4 1 2 lDd1 a5 13 c3 i.. a6 + Rusanov-Emelin, St Petersburg 1998. 6 c5 ! 7 tt:Jb5 tt:Ja6 8 e5 8 c4 transposes to note 'c1 ' to White's 5th move above. 8 4J d7 9 c3 tt:Jab8! Improving the worst piece. Now: a) 10 c4? a6 1 1 ltJd6+ i..xd6 12 exd6 (Treppner-Assmann, 2nd Bun­ desliga 1998/9) 12 ... cxd4! 13 'i!i'xd4 e5! 14 'ii'e3 ( 14 i..xe5?ltJc6 costs White his bishop) 14 ... d4 15 'i!i'e4 0-0 +. b) 10 dxc5 i.. xc5 1 1 b4 i.e? 12 i..g3 0-0 13 f4 tbc6 14 tt:Jf3 (Vatter­ Blauert, Germany 1989) 14 ... a6 15

lDbd4 tt:Jxd4 16 lDxd4 a5 17 a3 tlJb6 + followed by ... i..d7, ...'ii'c7 and .. J1fc8. White's play on the kingside can be stopped with ... g6. C23) 4 e3

This is the most popular move. Black immediately puts pressure on White's centre. 4 c5 (D) .••

••.

•.•

.•.

5 i.. d 3

There plenty of alternatives at this point: a) 5 4Jf3?! transposes to the note to White's 5th move in Line C21 . b) 5 c3ltJc6 (Black creates the po­ sitional threat of gaining the bishop­ pair by ... g5; the immediate 5 ... g5 is met by 6 i..e5 !) 6 lDf3 ( 6 ltJd2 i..f5 7 tbgf3 e6 leads to the same position; 6 i..d3?! is dubious in view of 6 ... g5 ! 7 i..g3 tbxg3 8 hxg3, as in Budrewicz­ Shishkin, Wroclaw 2009, when V.Mi­ khalevski recommends 8 .. .'i!i'b6 9 'i!i'b3 h6 ; Black has the bishop-pair and faces no problems) 6 ... i..f5 7 tlJbd2 e6 are

=

DIEMER, VERESOV AND TROMPOWSKY

8 tt:lxe4 il.xe4 9 il.d3 Axd3 10 Wt'xd3 and now: b1) It's important for Black to avoid taking on d4 for the time being, as this would give White nice attacking pros­ pects on the kingside since the e-file and 3rd rank prove very useful. Just take a look at how easily White won the following game: 10 ... cxd4?! 1 1 exd4 ..td6 12 il.g3 a6 13 0-0 h6 (al­ though 13 ... 0-0!? looks scary for Black because of 14 i.xd6 �xd6 15 tt:lg5, in fact it's not fatal for him yet: 15 ... g6 16 'i¥h3 h5 keeps a fighting position) 14 liae1 0-0 15 tt:ld2! (preparing a pawn-storm) 15 ... ..txg3 16 'i¥xg3 tt:le7 17 lle2 tt:lf5 18 'iVh3 'iVa5?! 19 tt:lf3 'ii'xa2? (Black wins a pawn but gives White time to pursue his attack) 20 g4 tt:le7 21 tt:le5 �a5 22 f4 �d8? (the black queen does not get back in time; 22 .. .'ifb5 ! ?) 23 f5 tt:lc6 24 f6 tt:lxe5 25 dxe5 'ii'b6+ 26 .:tef2 'i¥c7 27 'ili'h5 with a winning position for White, Miladinovic-Ftacnik, Croatian Team Ch, S ibenik 2007. Black could have defended better in this game, but there is nothing to be gained from inviting White to attack in this way. b2) 10 ... .te7 1 1 0-0 0-0 leaves Black fine; e.g., 12 dxc5 (12 tt:le5 .l:.c8 13 a3 i.f6 14 tt:lf3 {White has just wasted two tempi} 14 ... a6 Schu­ bert-Navara, 2nd Bundesliga 2001/2) 12 ... i.xc5 13 e4 dxe4 14 'iVxe4 'ii'e7 15 l:lfe1 h6 Shmirin-Uhlmann, Leu­ tersdorf 200 1. c) 5 tt:ld2 and now: c 1) The popular 5 ... 'i!Vb6 does not equalize because of 6 tt:lxe4 dxe4 7 dxc5 ! 'i¥xb2 8 'i¥d4 'ii'a3 (White is =

=

19

better in the endgame after 8 ... 'ii'xd4 9 exd4 tt:lc6 10 ..tb5 ;l;;) 9 ..txb8 'i¥a5+ (forced) 10 c3 .l:.xb8 (Mesias Rojas­ Lopez Silva, Chilean Ch, Fenach 2007) 1 1 .:tdl ! (a new move) 1 l...i.f5 12 f3 ;l;; with a strong initiative - Black's king is in big trouble. c2) 5 ... tt:lxd2 6 1!Vxd2 tt:lc6 7 tt:lf3 and here: c21) Black can't develop his queen's bishop because 7 ... .tf5? is answered by 8 dxc5 !. Then 8 ... e6 9 tt:ld4! i.g6 (9 ... il.xc5 10 tt:lxf5 exf5 1 1 0-0-0 ± and the d5-pawn drops as 1 1 ...tt:le7? fails to 12 'ii'c3 +-) 10 'i¥c3 ± gives White a clear extra pawn. 8 ...f6 looks a better try but still does not help to win the pawn back: 9 tt:ld4 il.d7 10 tt:lb3 e5 1 1 il.g3 i.e6 12 l:i.dl ± leaves White a pawn up. c22) 7 ... e6 8 c3 i.d6 9 il.b5 0-0 10 0-0 'iVe7 (a new move; 10 ... i.xf4?! strengthens White's control of the e5square: 1 1 exf4 cxd4 12 cxd4 i..d7 13 lirfc1 ;l;; Pelletier-Vogt, Swiss Team Ch 1998) 1 1 .l:lacl i.d7 with an equal po­ sition. 5 cxd4 (D) ...

A TTACK WITH BLACK

20

6 il.xe4

This is the most critical reply White creates a target on e4 that he will attack in the near future; fortu­ nately, Black can defend it or in some cases successfully give it up. If White wants to avoid risk he can simply take on d4. 6 exd4 tDc6 and then: a) Black feels OK in the case of 7 c3 il.f5 (D). Now:

b) 7 i..xe4 dxe4 8 tDe2 (the aggres­ sive 8 d5 is met by a typical opening counterbreak: 8 ... e5 ! 9 il.g3 tDe7 10 tDc3 tDg6 and in Stefanova-Kurajica, Solin 2007 Black was even slightly better considering that the e4-pawn is indirectly defended: 1 1 tDxe4? f5 12 tDc3 f4 +) 8 ... il.g4 9 tDbc3 'tleb6 trans­ poses to line 'c' of the next note. c) 7 tDe2 il.f5 8 0-0 (8 f3 tDd6 9 i..xd6 i..xd3 10 'ii'xd3 'i!Vxd6 ) 8 ... e6 9 c3 il.g6 10 tbc 1 i..d6 1 1 ..ltxd6 tDxd6 A.Smimov-lordachescu, Paleochora 2010. 6 dxe4 7 'ilxd4 White can also take on d4 with the pawn. 7 exd4 Vi'b6 8 tDc3 tDc6 (D) (8 ...�xb2?! 9 tDge2 gives White the initiative) and now: =

=

•••

al) 8 tDe2 e6 9 f3 tLld6 10 i..xd6? (10 0-0 is more solid) 10 ... il.xd6! (unexpectedly for White, his opponent is not scared by the doubling of his pawns, since in return Black gets an open file and weak squares to target on e3 and e4) 1 1 il.xf5 exf5 12 f4 0-0 13 0-0 .l::te 8 14 tLld2 'ilc7 15 g3 tDa5 16 .l::tf2 tDc4 + Vitiugov-Kariakin, Rus­ sian Ch, Moscow 2010. Black's ad­ vantage is obvious. a2) 8 �e2 e6 9 tLld2 tLld6 10 tDgf3 il.xd3 1 1 �xd3 ttJf5 12 g4 tDfe7 13 h4 tDg6 14 il.g5 �c7 15 h5 tDf4 16 il.xf4 'ii'xf4 17 :g1 h6 18 0-0-0 i..e7 Timman-Van Wely, Breda rapid 1998. =

=

a) After 9 't!Vd2? Black can take the pawn: 9 .. .'ii'xd4! 10 tLlb5 'i!Vxd2+ 1 1 i..xd2 .l::tb 8! + Stefanova-Mamedya­ rov, Wijk aan Zee 2005. b) 9 d5 e5 ! (again we see this typi­ cal counterblow - it is worth memo­ rizing it!) 10 i..e3 'iVxb2 (this is forced, but there doesn't appear to be any way

DIEMER, VERESOV AND TROMPOWSKY

to punish Black for his greed; instead, IO .. .lbd4? loses a pawn with no com­ pensation: 1 1 lZJge2 i.. g4 12 �xd4 exd4 13 �xd4 �xd4 14 LDxd4 ± R.Janssen-Visser, Dutch Ch, Rotter­ dam 1999) 1 1 lZJge2 (the only move) l l ...lZJb4! 12 0-0 �xc2 (the white queen cannot avoid exchange) 13 lZJb5 ..ig4 14 �xc2 lZJxc2 15 lZJg3 �b4! +. c) 9 lZJge2!? �g4 10 0-0 (a new move, improving over 10 h3? �xe2 1 1 lZJxe2 .l:l.d8 1 2 c3 e5 ! + Laurie-Tse­ sarsky, Tel Aviv 1997) 10 ....l:.d8 1 1 d5 e6!? is unclear. Black's king is still uncastled but his bishop-pair and well-coordinated pieces don't allow White to exploit this fact. 7 ...lZJc6! 7 ...�xd4?! is in my opinion not advisable: 8 exd4lZJc6 9 c3 f5 (9 ... g5? 10 ..ixg5 .l:.g8 1 1 i.f4 ;!; and 1 1 .. Jhg2? 12 ..ig3 traps the rook) 10 lZJd2 ;!;, Once White plays f3, Black will have problems on the e-file. 8 �xd8+ 8 �xe4? is not good because of 8 ... �a5+ 9 c3 i.f5 10 �f3 (1 0 �c4? e5 1 1 ..ig3 lZJb4! 12 cxb4 i.xb4+ and ...�c8 wins) 10 ... e5 1 1 ..ig3 �b5 ! 12 b3 ..ic2 + followed by ....l:!.d8, when White is in trouble. 8 lZJxd8! ? (D) With this new move (varying from the game Stefanova-Mongontuul, Rus­ sian Women's Team Ch, Sochi 2006), Black preserves his right to castle and doesn't experience any problems with the king on e8. •••

21

w

9 lZJc3 f6!

That's the point behind Black's 8th move. He temporarily sacrifices a pawn in order to get a pleasant pawn­ structure. 10 lZJdS Attacking c7 and delaying ... e5 for a while. If White decides to take the pawn by 10 lZJxe4 he should be pre­ pared for 10 ... e5 1 1 ..ig3 ..if5 12 f3 lZJf7 with ... ..ixe4 and ...lZJg5xe4 com­ ing next; Black is preferable due to his better pawn-structure. 10 0-0-0 also does not change the character of the play; after 10 ... e5 1 1 i.g3 ..if5 the pawn is defended and the g3-bishop is out of play. Black is OK. 10 ...lZJe6 1 1 lZJe2 ..id7 12 lZJec3 .l:!.c8 13 0-0-0 lZJcS

Next Black will play ... e5 (proba­ bly after ... e6 to push the d5-knight back). Considering that Black has the bishop-pair and the white bishop will be bad on g3, Black's chances are even preferable.

2 Co l le , Zu ke rtort , London a nd Torre Systems These four opening systems have no main concrete order of moves; that's why they are called 'systems'. The common thread is that White plays d4 and an early tt:'lf3, and meets our ... c5 by supporting his pawn with c3 and/or e3. The defining features are as fol­ lows: Colle: e3 and c3, with Si.d3 Zukertort: e3 intending a later b3 London: Si.f4 Torre: Si.g5 Let's note that against the Zuker­ tort, we shall be adopting a set-up with ... g6, which may discourage White from playing b3. In other lines we need to be careful that White can't play an early dxc5 to good effect, so in most cases we shall be occupying the centre with ... e6 and ... d5. Each sys­ tem has its own unique themes, so let's move on to a discussion of each of them: 22 A: Colle 27 B: Zukertort 30 C: London 34 D: Torre

Capablanca in 1929) was the Belgian player Edgar Colle. This variation is characterized by the following moves for White, in one order or another: d4, e3, tt:'lf3, i.d3, 0-0, c3 and tt:'lbd2. This is similar to Black's set-up in the Semi-Slav, and therefore it's no sur­ prise that one of White's main ideas in the Colle is to make the e4 advance. The main advantage of the Colle is its solidity, and there is the practical ad­ vantage that White can save a lot of time in the opening making these au­ tomatic moves. The main disadvan­ tage is that White practically ignores his opponent's replies, and Black has a wide choice and can direct the game towards the kind of position he wants. Also, while the Semi-Slav works well as a counterpunching opening, the re­ versed form proves less effective for taking the initiative. That's why you rarely see games with this system at GM level. So, let's see our way to re­ spond to the Colle system! 1 d4 tt:'lf6 2 tt:'lf3 cS 3 c3 We shall discuss the consequences of 3 e3, which we shall meet with 3... g6, in Section B of this chapter. A) Colle 3 e6 The first notable master to adopt this Here 3... g6 can be met by 4 dxc5 !?, opening (including a game against after which Black has no simple way ...

COLLE, Z UKERTORT, LONDON AND TORRE SYSTEMS

to take his pawn back (. . .'�i'a5 is not check any more). Usually Black sacri­ fices a pawn by ... b6, but it does not look very convincing to me. 4 e3 dS Black captures space in the centre, and we now see the contours of a Semi-Slav with reversed colours. Another popular continuation for Black is 4 ... b6, planning ... ii.b7, ... d6, ... tZ:lbd7, ... i..e7, ... 0-0, etc. The disadvantage of this arrangement is that White can make the e4 advance: 5 tLlbd2 i..b7 6 ii.d3 ii..e7 (Black also can switch back to the idea we recom­ mend in the main line by 6 ... d5, but if Black wanted to play ... d5 it was better to do so on move 4 as now it lets White proceed with another typical Colle idea: 7 tbe5 tLlbd7 8 f4 i..e7 9 0-0 0-0 10 'i¥f3 with attacking chances on the kingside) 7 e4 cxd4 8 cxd4 tbc6 9 a3 0-0 10 0-0 d6 1 1 b4 with a space ad­ vantage for White, Bricard-Bischoff, Bissen 1995. 5 i.. d3 Of course, White can make other moves in this position, but this move is logical, and is a basic part of the Colle set-up. Having played c3, there is little point in playing b3 and i..b2, while dxc5 is premature before Black has moved his king's bishop. 5 tLlbd2 tbc6 6 i..d3 ii..e7 7 0-0 would lead to the same position. 5 tbc6 6 0-0 White can also try to get a space ad­ vantage with 6 tbe5 tZ:lxe5 7 dxe5 tLld7 8 f4. However, Black easily obtains in­ teresting active play: 8 ... c4 9 i..c2 (this position was reached in the game ...

23

Kemeny-P.Richardson, New York 1 894) 9 ...'iVh4+! 10 g3 'iVh3 1 1 tLld2. Black has prevented White from cas­ tling kingside and now it's time to start play on the queenside: 1 1 ... b5 (a new move, deviating from Atanas­ kovic-Lajthajm, Serbian Ch, Lesko­ vac 2002) 12 �f3 :b8 with an unclear position. 6 i..e7 7 tLlbd2 0-0 (D) ...

w

Black continues developing and is ready for play in the centre. 8 dxc5 Besides this exchanging operation, which is also a standard theme in the Semi-Slav, White can continue play­ ing slowly: a) After 8 'i¥e2 b6 White has plenty of possibilities but none of them should scare Black too much, as he is ready to meet all White's active ideas such as tbe5 or e4: a1) 9 dxc5?! (giving up control of the centre does not look good) 9 ... bxc5 10 e4 'i¥c7 1 1 e5 (1 1 l;lel does not change too much - Black is still pref­ erable: l l ...Jtb7 12 e5 tbd7 13 tZ:lfl,

24

A ITA CK WITH BLA CK

Sydor-Smejkal, Sandomierz 1976, 13 ... c4! 14 �c2 �a6 + intending to occupy the d3-square with ... lt:'lc5-d3) 1 l ...lt:'lg4 12 :e1 c4 13 .i.c2 �c5 ! (Black simply ignores the scare tac­ tics) 14 �xh7+ 'ii?h 8! (now White needs to defend f2; 14 ...'ii? xh7? leads to an unclear position after 15 lt:'lg5+ �g8 16 �xg4 f6) 15 lt:'ld4 �xeS + Guimard-Granda, Porto Velho 1988. Black is better since he has exchanged a rook's pawn for a central pawn. a2) 9 e4?! is another self-destruc­ tive idea. 9 ... cxd4 10 e5 (10 cxd4 can lead to the same position after 10 ...lt:'lb4 1 1 e5 lt:'ld7) 10 ... lt:'ld7 1 1 cxd4 lt:'lb4 12 �b5 a6 13 �a4 (D) and now:

a21) 13 ... a5?! allows the white bishop to return to the a6-fl diagonal: 14 �b5 (14 a3 happened in F.Cruz­ C.Cruz, Sabadell 2009) 14 ... .i.a6 15 .i.xa6 lt:'lxa6 16 lt:'lbl ! and Black cer­ tainly isn't better. a22) I prefer 13...b5 !. With this new move, Black temporarily blocks the a6-fl diagonal in order to open it with greater effect later. 14 �d1 (14 �b3

�6 15 lt:'lbl a5 followed by ... a4, ... �a6, etc., gives Black an advan­ tage) 14 ... �6 15 lt:'lb1 a5 with the same plan: ... �a6, ... lt:'lc6, ...b4, etc. a3) White again gets nothing with 9 lt:'le5 lt:'lxe5 10 dxe5 lt:'ld7 1 1 f4 c4 12 .i.c2 lt:'lc5 13 .l:tf3 (Muse-Kritz, Ger­ man Ch, Hoeckendorf 2004) 13 ....i.a6! intending ... lt:'ld3 (to block the c2bishop) and then ... f6. a4) 9 b3 is the most solid move for White since, as we have seen, all the active options rebound on him. After 9 ... �b7 10 �b2 the natural 10 ...�c7 followed by ....l:tad8 leads to simple and equal play, while in Bruno-Gyi­ mesi, European Team Ch, Gothenburg 2005 Black played a somewhat fussy but still viable plan: 10 .. Jk8 1 1 .l::tac1 :c7 12 .l:r.fd1 'ifa8, after which 13 dxc5 bxc5 14 c4! d4 15 exd4 cxd4 16 a3 a5 led to a very complex and un­ clear position. b) White can again try to attack with 8 lt:'le5. Then: bl) Black gets a strategically risky pawn-structure after 8 ... lt:'lxe5 9 dxe5 lt:'ld7 10 f4 f5 1 1 exf6 lt:'lxf6 12 e4!, al­ though he is not worse at the moment, Bareev-Tukmakov, Tilburg 1994. But anyway I prefer to recommend some­ thing clearer. b2) Black has a typical way to stop White's lt:'le5-based attack: 8 ... lt:'ld7!. After 9 f4 Black plays 9 ...f5! = Araya­ Vazquez Igarza, Asuncion 1991. Now Black is ready to take on e5 since fxe5 does not create any attacking potential for White any more. If White takes a risk with 10 c4?!, Black gets the better game thanks to his more harmonious

COLLE, Z UKERTORT, LONDON AND TORRE SYSTEMS

25

pieces: 10 ... cxd4 1 1 tt'lxc6 bxc6 12 giving perpetual check or continuing the game with three pawns for a piece. exd4 c5 ! 13 cxd5 tt'lf6! +. c 1) In his book A Rock-Solid Chess 8 i.xc5 Now: Opening Repertoire for Black, Ein­ 25 gom suggests 12 ...tt'le4 13tt'lbd4 tt'lxd4 Al : 9 e4 26 14 tt'lxd4 'ild6 intending ... i.c7, which A2: 9 b4 "maintains the tension and keeps the Al) initiative". c2) 12 ... i.xh3!? 13 gxh3 �g3+ 14 9 e4 This is a logical continuation, but xh7 12 �h5+ 'it>g8 13 t:bg6 e6 3 ii.g5, when 3 ... c5 4 c3 transposes 'iVe8 14 t:bf3 t:be7! + Blatny-lotov, Eu­ to our position). It's named after the famous Mexican master Carlos Torre ropean Ch, Kusadasi 2006. Repetto. Compared to the London Sys­ l l ... bxc6 12 0-0 (D) tem, the key difference is that White places his bishop on g5, exerting a lit­ tle more pressure on Black, but the general set-up is quite similar: c3, e3, lL'lbd2, .i.d3, 0-0, etc. Against accurate play, such modest development is un­ likely to produce any advantage in the opening. 4 ... d5 I prefer this move-order over 4 ... h6, since then besides 5 i..h4 White can take on f6 followed by e4. 5 e3 h6 6 i.. h4 Now there is no reason for White to exchange on f6. 12 ... e5! ? 6 ... t:bbd7 7 lL'lbd2 i.d6 8 i.. d 3 0-0 Black can also close the centre by 12 ... f5 !? with an absolutely safe po­ 9 0-0 (D) sition. 9 ... b6 This quiet-looking but actually 13 fxe5 fxe5 14 'iVh5 lL'lf6 15 �h4 h6! ? rather ambitious move has recently This new idea removes the pawn been tried out in high-level chess. from the firing line. Not 15 ... i..d7?? Black wants to play ... i.. b 7, when he 16 dxe5 i..xe5 17 .txh7+ +- Liogky­ will have a stable and pleasant posi­ Fingerov, Odessa rapid 2005. tion, since if White then plays e4, it =.

•••

=.

=,

.15

COLLE, Z UKERTORT, LONDON AND TORRE S YSn'MS 10 e4

Otherwise Black plays ... .ib7, when White's e4 advance will evidently be less dangerous. 10 cxd4 1 1 cxd4 White has nothing special after I I .ixf6 �xf6 (l l ...tt:lxf6? 12 e5 ±) 12 cxd4 dxe4 1 3 tt:lxe4 'fie7 l l... dxe4 1 2 .ltxe4 Or 12 tt:lxe4 i.e7 12 J:tb8 Black can be happy from a posi­ tional viewpoint, so from now on he can focus on simplification. White in turn tries to complicate matters. 13 .itc6 Another attractive-looking continu­ ation is 13 'ii'e2 but here Black also copes with his problems: 13 ... .ib7 14 .ltxb7 llxb7 15 tt:le4 (15 tt:lc4 .if4! puts the bishop on a very good square, where it controls c 1 ; then 16 tt:lfe5 "file? is equal) 15 ... i.e7 16 .l:.acl tt:ld5 (Black exchanges more and more pieces) and now 17 .ltg3 tt:l7f6 18 "ila6 .l:!.d7 19 tt:lxf6+ .ltxf6 17 .ltxe7 'fllxe7 18 tt:le5 tt:lxe5 19 dxe5 .Uc7 and 17 tt:ld6 l:.c7 18 .ltg3 �hc1 19 l:.xc1 .ixd6 20 i.xd6 .l:!.e8 21 'ii'a6 tt:l7f6 22 i.e5 'i!Vd7 all lead to equal play. .•.

=.

=.

..

will merely lead to exchanges that are by no means unfavourable for Black. So if White is going to create active play, or prove an opening advantage, he needs to take forcing measures right now. In case of the natural 9 ... e5, Black must be prepared to meet the aggres­ sive 10 dxe5 tt:lxe5 1 1 tt:lxe5 .ltxe5 12 f4!. Then: a) 12 ... .itd6?! 13 e4 c4 14 .lte2 �b6+ 15 .ltf2 .ltc5 16 e5 tt:le4 (the al­ ternative 16 ....1txf2+ is similar: 17 l:txf2 tt:le4 1 8 tt:lxe4 dxe4 19 �d4 �xd4 20 cxd4 :l.d8 21 l:!.d 1 .lte6 22 f5 .itd5 23 .ifl ! t intending l:.c2 and �f2-e3) 17 tt:lxe4 dxe4 1 8 .txc5 'i!Vxc5+ 19 'ii'd4 'ii'xd4+ 20 cxd4 with a slightly better endgame for White, Morozevich-Alekseev, Moscow 2008. b) The computer offers the cold­ blooded 12 ... .1tc7 13 e4 g5 14 fxg5 tt:lg4 with an 'equal' evaluation. You may wish to investigate this further, but of course you would need to pre­ pare it very carefully, since in such a position even the most natural move can be a fatal mistake.

= ,

13 ... .ib7 14 l::.cl (D) 14 ... .if4!?

Again the bishop is good on f4 since it pins the d2-knight and pre­ pares ... i.xc6 and ... .l:tc8, which had been impossible due to l::txd6. I am offering this new move as an improvement over 14 ... b5?!, which allows White to achieve his desired complications: 15 d5 ! exd5 16 .ltxd7 'ii'xd7 17 i.xf6 gxf6 18 tt:\d4 with

36

A ITA CK WITH BLACK

quite an unpleasant position for Black, Andreikin-Kovchan, Moscow 2012. 1S �g3 This is the safest move for White. Other continuations tend to rebound on him: a) 15 g3?! forces an exchange of Black's annoying bishop but leaves White's h4-bishop in a precarious sit­ uation. 15 ... ..txc6 16 llxc6 ..txd2 17 1!Vxd2 l:r.c8 1 8 !!xc8 (18 llfcl ?! llxc6 19 .Uxc6 1Wa8 ! 20 'iVc 1 lt::lb 8 21 llc3 lt::ld5 +) 1 8 .. .'i!i'xc8 19 ..txf6 lt::lxf6 leaves Black more comfortable. b) 15 d5?! does not work so well now because Black's pieces are more active than they were in Andreikin­ Kovchan. 15 ...exd5 16 ..txd7 'i!i'xd7 17 �xf6 gxf6 18 g3 �e5! (not allowing White to obtain the d4-square for free) 19 lt::lxe5 fxe5 20 'ii'h5 'i!i'd6 21 lt::lf3 �fe8 22 l:tfel e4 23 lt::ld4 'ii'g6 +. White has taken one blockade square but lost ground in all other parts of the board.

c) 15 'i!Va4?! is strongly met by 15 ...b5 !, gaining an important tempo for Black's development. Then: cl) 16 'fixa7 ..txc6 17 l:txc6 .l:.a8 and here: el l) 1 8 'fib7? loses to 18 ... 'ii'a5 ! (intending ... l:ta7) 19 lt::lb 3 (19 �xf6 .l:ta7 20 lt::lb 3 'fia4 -+ is the same) 19 ... 1!Va4! (defending b5) 20 ..txf6 .l:la7, trapping the queen. It can only be saved at the cost of a piece: 21 lt::lc 5 lt::lx c5 22 b3 lt::l xb7 23 bxa4 gxf6 24 axb5 l:txa2 -+. c12) 18 ..txf6! lt::lxf6 (18 ... gxf6? 19 'fib7 ;!; leaves the d7-knight hang­ ing; e.g., 19 .. .'i¥a5 20 .l::!.c 3) 19 1\Vc5 .l:txa2 +. c2) 16 1Wxb5 ..txc6 17 'ifxc6 ..txd2 1 8 lt::lxd2 .l:r.xb2 19 lt::lc4 (unexpectedly Black wins a pawn since after 19 l:lc2, the d4-pawn drops: 19 ... lt::lb 8! 20 'ii'c 3 .l:txc2 21 'fixc2 'fixd4 +) 19 .. Jha2 20 lt::le5 aS ! + and White needs to fight for a draw. c3) 16 'ii'b4 'ifb6 17 �xb7 .l::!.xb7 + followed by ... lt::ld5. 1S �xg3 16 hxg3 �xc6 17 .l::!.xc6 .l:!.c8 •••

Black has solved his opening prob­ lems; for example, 1 8 'i!i'a4 .l:lxc6 19 'ili'xc6 lt::lb 8 20 'ii'b7 'i!Vd7 21 'iWxd7 lt::lbxd7 22 .:!.c1 lt::ld5 White has captured the only open file but Black has safely blockaded the d4-pawn and taken control of the only dangerous square on the c-file - c7. The position is equal. =.

3 Anti - Benon i 4 ltJc3

1 d4 lDf6 2 lDf3 c5 3 d 5 (D)

B

White gains ground in the centre, and shows that he is in an uncompro­ mising mood. 3 e6 This move is my recommendation for Black. After White's main reply, 4 c4, our repertoire choice is to reach the Blumenfeld Gambit by playing 4 ... b5. In this chapter we look at White's al­ ternative options, most notably 4 lbc3. Many Benko players choose 3 ... g6 or 3 ... b5, hoping to transpose to a Benko. In practice, White rarely coop­ erates, and some of his alternatives are by no means easy to face. Also, some of our repertoire choices in the Benko are based on an early ... e6 rather than ... g6, so opting for 3 ... g6 would require additional preparation. A few •••

specific thoughts on these possibili­ ties: a) 3 ... g6 and now: al) 4 c4 b5 5 cxb5 takes us outside our Benko repertoire if White meets 5 ... a6 with 6 b6 or 6 e3. a2) I don't like 4 lbc3 .ii. g7 5 e4 d6 (a form of Schmid Benoni) for Black, since the positions that arise are of a completely different character from the Benko. Black's position is cramped and if there is a good technical player on the white side, it is usually very hard to get a comfortable game. There is little scope for activity for Black un­ less White is particularly careless. Of course, it's possible for Black to play this way, but I think when you exam­ ine the lines after 3 ... e6 you will agree that they offer Black far better win­ ning chances. b) 3 ...b5 transposes to a Benko De­ clined if White replies 4 c4. But 4 i.g5 is a popular alternative that leads to play of a different nature. Then you may wish to investigate 4 . .'ti'b6, which leads to interesting irregular positions. 4 lb c3 This is White's last chance to avoid serious complications. White seeks to control the d5-square with his pieces and hopes to establish this as an out­ post that gives him a long-lasting edge. Other moves:

ATTA CK WITH BLACK

38

a) 4 c4 is the main move, when our choice is 4 ... b5, the Blumenfeld Gam­ bit. See the next two chapters. b) 4 dxe6 fxe6 5 c4 is a rare se­ quence in practice, as it allows Black easy play in a variety of ways. If you are really keen on the Blumenfeld, then you can choose 5 ...b5, transpos­ ing to Line C of Chapter 4, but there are safe and very satisfactory alternatives such as 5 ... d5, while 5 ... i..e7 and 5 ... lLlc6, with ... d5 often coming soon, have achieved solid plus scores for Black. c) Black has several good answers to 4 i..g 5, but I recommend 4 .. .'ifb6!?, which immediately attacks two pawns - b2 and d5. Black is OK; e.g., 5 dxe6 dxe6 6 'i'c1 lLlc6 7 e3 i..e7 8 lLlbd2 0-0 9 lLlc4 'i!Vd8 4 ...exd5 It doesn't make any real difference if Black takes with pawn or knight, as they both lead to the same position: 4 ... lLlxd5 5 lLlxd5 exd5 6 'itxd5. If you were hoping for coverage of the Pseudo-Blumenfeld, 4 ...b5?!, then I'm sorry, but I consider the text-move to be objectively better. 5 lLlxd5 lLlxd5 6 'i!Vxd5 (D) White's queen has taken up an in­ fluential but exposed central post. If White can establish firm control of e4 and d5, then Black will come under strong pressure. His main mission is therefore to push the queen back in the quickest possible way, which is ...lLlc6, ... d6 and ... i..e6. 6 lLlc6! This move appears to invite an im­ mediate attack on the f7 -square, but =.

.••

careful analysis shows that there is no real problem. In case of 6 ... i..e7 Black needs to be ready to face 7 i..f4 lLlc6 8 0-0-0 0-0 9 i..d6!? (a new move), which leaves Black rather cramped. I don't enjoy such positions, and won't recommend them to my readers. Now: A : 7 i.. g5 39 B : 7 e4 39 Or: a) Now White does not have time to blockade the d7-pawn with 7 i.. f4 because of7 ... d6 (Rakhmangulov-Efi­ menko, Ukrainian Team Ch, Alushta 2002) 8 e3 i..e6 9 'i!Vd2 i..e7 10 .!:i.d1 0-0 with an equal position because the pawn is indirectly defended: 1 1 i..xd6? i..xd6 12 'i!Vxd6 'i!Va5+ 13 'iWd2 �xa2 +. b) 7 i..d2 d6! (again the priority is to push the white queen back; the next move is ... i..e6 with a comfortable po­ sition; Black can also play 7 ...i..e7 8 e4 0-0 9 i..c4 d6, which proved unclear in P.H.Nielsen-Jobava, European Team Ch, Khersonissos 2007, but I don't see

ANTI-BENONI 4 tiJc3

any reason to allow White to arrange his pieces in this way) 8 liJg5 'i/e7 9 e4 h6 10 .i.b5! (this is the only way for White to keep the balance; 10 liJf3? -te6 1 1 'ilid3 .i.f5 ! + costs him the e4-pawn) 10 ... -td7 1 1 liJf3 (Fodor­ Paschall, Paks 2005) l l ...a6 12 .i.e2 _te6 13 'ilid3 g5 ! (intending ... g4 or ... .i.g7) 14 .i.c3 .litg8 15 0-0-0 0-0-0 = followed by ... .i.g7. c) 7 c3 is a prophylactic move White stops ... liJb4. Then 7 ... d6 8 liJg5 'ilie7 9 .i.f4 (D) was played in Rombal­ doni-Vezzosi, Arvier 2009.

39

8 ltJxe5 9 'ilixe5+ 'ilie6 10 'ilic7! ...

Here I would like to offer a new move, 9 .. .f5!. This cuts off the white queen from defending the g5-knight and prevents ltJe4. The idea is simple: ... h6 and ... .i.e6 (the immediate 9 ... h6? allows 10 ltJe4 ±). After 10 0-0-0 h6 l l liJf3 i.e6 12 'i¥xd6 'iVxd6 13 -txd6 i.xd6 14 .l:txd6 i.xa2 = Black is OK.

White threatens mate in one. The endgame after 10 'i!Vxe6+ dxe6 1 1 g3 f6 12 i.e3 i.d7 13 i.g2 0-0-0 14 0-0-0 i.e7 is approximately equal. 10 'i¥b6!? This is a new move (varying from Berkes-Wojtaszek, Balatonlelle 2002) . Black's safest path is to move towards an endgame. 1 1 'i¥xb6 1 1 'i¥e5+ 'i¥e6 leads to a repetition. ll axb6 12 e4 i.e7 13 i.f4 It's useful for White to keep the dark-squared bishops on the board since his bishop is going to be more active than Black's. After 13 i.xe7?! 'i;; xe7 14 a4 d6 15 b3 i.e6 intending ... c4 Black is comfortably equal. 13 d6 White retains some pressure in the endgame, but Black has sufficient re­ sources to hold the balance, and White can easily get in trouble if he is care­ less. Here are some sample lines: 14 a3 i.d7 15 0-0-0 0-0-0, and now: a) 16 i.xd6? i.g4 17 f3 i.xd6 18 e5 (18 fxg4?? ..tf4+ 19 'i;;b l .Uxdl+ -+) 18 ... i.xe5 19 .l:!.xd8+ %hd8 20 fxg4 iLf4+ 21 'i;;b 1 l:td 1 + 22 'it>a2 'i;;c7 +. b) 16 iLc4 i.e6 17 i.d5 Wc7 18 l:l.d3 l:the8 19 .l:.hdl iLf6 is rather un­ clear. White may enjoy slightly better chances, but Black's position is very solid.

A)

B)

B

...

...

...

7 e4 d6 i.gS 'ilib6 8 liJeS As we know, it's important to pre­ After 8 0-0-0 d6 = followed by pare a quick ... i.e6 to prevent White ... i.e6 Black is fine. 7

40

A TTA CK WITH BLACK

from establishing control over the a2g8 diagonal. 8 ltJ g5 After 8 .tc4 .te6 9 Vi'd3 ltJb4 1 0 �e2 .txc4 1 1 �xc4 d5 ! Black ex­ changes his backward pawn. 12 exd5 'itxd5 and here: a) 1 3 �xd5 ltJxd5 14 .td2 (Mar­ tinez Rodriguez-Veingold, Alicante 2000) 14 ... f6! (an important restrict­ ing move) 15 0-0-0 0-0-0 is equal. Of course, it would be much better for Black if the pawn were back on c7, but thanks to the good location of the black knight, his pawn-structure should not be a big problem. b) Black should not be scared of the check 13 'ife2+ (D). Then:

l:td6 18 a3 ltJd5 19 'it>cl ! (19 c3?! ltJf4 ) 19 .. J:U8 20 c3 g5 21 h3 h5 22 .l:te4 l:tf5 23 'it>c2 followed by .l:.ae1, with a slight advantage for White. b2) I prefer the untried 13 ... .te7!: b21) After 14 0-0, 14 .. .'ii'e6 ! now works well because White has prob­ lems with the c2-pawn. 15 'ii'xe6 fxe6 16 c3 ltJd3 is equal. b22) 14 ..tg5 f6 15 a3 (retreating the bishop is worse: 15 i.f4? 'i¥f5 ! + or 15 i.h4?! 0-0-0 16 0-0 { 16 a3? llhe8 17 0-0 ..td6 + } 16 ... ..td6 + and ...lir.he8) 15 ... fxg5 16 axb4 g4 17 ltJe5 0-0 18 0-0 cxb4 19 ltJxg4 i.c5 Black has solved his problems. =

=.

8 'iVe7 (D) •••

w

bl) 13 . .'ir'e6?! 14 'ii'xe6+ fxe6 cre­ ates a weakness on e6 that White will exploit later. 15 'it>dl ! and now: bl l) 15 ...lir.d8+ 16 .td2 .te7 17 .l:te 1 l:!.d6 18 a3 ltJc6 19 .l:.e4 i.f6 20 c3 �f7 21 �c2 lir.hd8 22 .li1.ae1 ;!; Epi­ shin-Lubbe, Deizisau 201 1 . bl2) 15 ... 0-0-0+ doesn't change matters much. 16 i.d2 .te7 17 .l:tel

9 i.c4 Another interesting idea for White is to start with 9 .tb5, forcing 9 ... i.d7, and then drop the bishop back: a) After 10 .tf4?! h6 l l ltJf3 a6 12 i.e2, as in McDonald-Kaufman, Bu­ dapest 2006, Black can win material by 12 ... ltJb4! 13 'ii'd2 'ii'xe4, when I don't see full compensation for White; e.g., 14 0-0-0 (14 0-0 Vfxc2 15 'ii'e 3+

ANTI-BENONI 4 li:Jc3

iL.e6 16 l:tadl .l:.d8 +) 14 ...lt::lxa2+ 15 �b1 iL.e6 16 l:thel 0-0-0! + followed by ...�a4. b) 10 JL.c4!? (this untried move looks more critical to me) 10 ... lt::ld 8 (Black retreats for the time being, but plans to return soon; the point of White inserting i..b 5 is revealed by 1 o ... f6? 1 1 ti:Jf7 lt::lb4 12 'i!fxb7 { with the bishop on c8, this capture was im­ possible! } 12 ... iL.c6 13 �xe7+ iL.xe7 14 iL.f4! lt::lxc2+ 15 'iti>d2 lt::lx al 16 ti:Jxh8 ±) 1 1 0-0 h6 12 lt::lf3 lt::lc 6 13 iL.f4 iL.e6 14 'i!Vd3 iL.xc4 15 'ii'xc4 'ife6 16 'i¥b5 0-0-0 with an unclear position. 9 f6! (D) •••

41

iL.xb5 13 lt::lxd6+ 'ii'xd6 14 ifxb5+ 'iti>f7 +. 10 'iti> d7! Black gives up his castling rights. If White could keep the queens on, this would be a real problem, but since the queens are going to be exchanged, the black king can feel quite secure. 1 1 'ifxe7+ 1 1 iL.e6+?! 'iti>c7 12 iL.xc8 fxg5 13 'i!fxe7+ JL.xe7 14 iL.e6 lt::ld4 15 iL.b3 .i.f6 16 'iti>d 1 lt::lxb3 17 axb3 l:the8 1 8 liie l a6 + Hebden-Plaskett, British League (4NCL) 1998/9. Next Black will double on the e-file. ll JL.xe7 12 lt::l e6 Black has no problems after 12 lt::lf3 (S.Muhammad-Plaskett, Internet blitz 2000) 12 ... f5! 13 exf5 r:J.c7 12 lt::lb4! (D) •••

...

=.

•••

w

Black pushes White's pieces back. 10 'iff7+ The over-ambitious 10 ti:Je6?? can be punished by 10 ...lt::lb4 1 1 iL.b5+ rli;f7 and Black simply takes the knight after 12 lt::lg5++ 'iti>g6 13 h4 h5 -+ or 12 ti:Jd8++ 'iti>g6 13 h4! h5 -+ (not 13 ... lt::lxd5?? 14 h5#, with a spectacu­ lar checkmate). White must also avoid 10 lt::lf7? lt::lb4 1 1 iL.b5+ iL.d7 12 'ii'xb7

Black's ideas include ...lt::lxc2+ and ... b5, and White can't stop them both. 13 0-0 b5 14 iL.d5 lt::l xd5 15 exd5 (Glud-Skovgaard, Skanderborg 2005) 15 ... i..b7 16 .l:l.d 1 iL.xd5 17 lt::lxc5+ dxc5 18 .l:.xd5+ 'iti>c6 with an equal endgame.

4 B l u m e nfeld Ga m bit

1

tt:'lf3 c5 3 d5 e6 4 c4 b5 and other strong grandmasters. It's notable that scrutiny with modem com­ puter engines has, it seems, suggested that this speculative-looking gambit has a good deal more substance than tended to be thought for much of its earlier history. In common with the Benko Gam­ bit, Black sacrifices a pawn by ... b5. But in the Blumenfeld it is more com­ mon for White to refuse this gift and focus on securing his central foothold. Also, the aims of Black's pawn sacri­ fice are different. In the Benko, Black seeks play on the half-open a- and b­ files using the power of the g7 -bishop; We have now reached the basic po­ in the Blumenfeld Gambit, if White sition of the Blumenfeld Gambit. It's accepts the gambit pawn, Black plays named after the Russian master Benia­ in the centre, setting up a powerful min Blumenfeld. The greatest adher­ pawn-centre. Thus, the inclusion of ent of the gambit in its early years was the moves tt:'lf3 and ... e6 (compared Rudolf Spielmann (author of the fa­ with the Benko Gambit) changes a lot; mous book The A rt of Sacrifice in most significantly, the fact that Black Chess) , who used it successfully in nu­ has played ... e6 means that there is far merous games. Later on, Lev Alburt more tension in the centre, and this (who has played almost all the varia­ can work for or against either side, de­ tions featured in this book) scored pending on the specific features of some high-profile victories with the each position. gambit that considerably increased its The next chapter deals with the popularity. Nowadays you can find the main line, 5 i.g5, which aims to exert Blumenfeld Gambit in games by positional pressure on Black. In this Ehsan Ghaem Maghami, Yuri Krivor­ chapter we shall analyse all the rare uchko, Liviu Dieter Nisipeanu, Fran­ moves White can play on move 5, to­ cisco Vallejo Pons, Andrei Volokitin gether with lines where he accepts the (D)

d4 tt:'lf6 2

BLUMENFELD GAMBIT

gambit pawn. This chapter therefore divides into the following sections: A : 5 a4 43 44 B: 5 lt::lc3 47 C: 5 dxe6 Other moves can be dealt with more briefly: a) The odd-looking 5 cxb5 can be met by the natural 5 ...lt::lxd5 or 5 ... exd5 6 .ltg5, transposing to the note to White's 6th move in Chapter 5. b) 5 'i!Vc2?! does not look like a correct pawn sacrifice. 5 ... bxc4 6 e4 exd5 (D) and then:

43

plans) 9 �dl (or 9 i.e2 lt::lb4 10 'i!Vdl i.a6 1 1 i.e3 'i!Vd6 +) 9 ...lt::lc7! 10 l:.el lt::le6 +. The knight on e6 looks like a great defensive piece; Black is still a pawn up and White needs to prove he has sufficient compensation. c) 5 e4 lt::lxe4 and now: c 1) 6 i.d3 lt::lf6 7 dxe6 fxe6 trans­ poses to line 'c3'. c2) In the case of 6 cxb5 i.b7 7 dxe6 fxe6 8 i.d3 Black needs to be precise. 8 ... i.e7?! allows White an ad­ vantage after 9 'ifc2 lt::lf6 10 lt::lg5 ! ;l;, so 8 ... a6! is best. Now 9 ll¥c2 lt::lf6 10 lt::lg5 can be met by 10 ... axb5 1 1 i.xh7 (1 1 lt::lxh7? 'i!fc7! +) l l...l!Vc7 with a complicated and unclear position. c3) 6 dxe6 fxe6 7 i.d3 lLlf6 and then: c31) The aggressive 8 lt::lg5 bxc4 9 i.xh7 (D.Paulsen-Krause, 2nd Bun­ desliga 200112; 9 i.xc4? lt::lc6 +) can be met with 9 . .'�'c7! (a new move) 10 i.g6+ hl (15 'it>xf2? ltJg4+ 16 'ifi>fl c3+ -+) 15 ... exd5 +. c2) 1 3 b3? c3 14 lLlbl ltJe4 1 5 ..te3 iVa5 is much better for Black, Cemou­ sek-Navara, Czech Team Ch 2008/9. c3) The attempt to undermine the black centre by 13 ltJg5 .l:.e8 14 e4!? is the most critical reply. Black can safely answer 14...h6 15 exd5 exd5 16 l:.xe8+ 'ii'xe8 17 ltJgf3 'ii'f7, when he is still a pawn down but his centre and better piece coordination provide him with pleasant prospects. l l ltJc6 (D) ...

a) 12 .l:r.e l .l:r.b8!? (this untried move looks the most useful, especially con­ sidering that e4 does not give White any advantage; 12 ...'ii'c 7 was played in Babula-Krejci, Czech Ch, Pardu­ bice 201 1 ) 13 e4 d4 and then: al) 14 e5 lLlxe5 ! 15 lLlxe5 ..txe5 1 6 lLlb5 ! (this surprising move aims to avoid ... dxc3 followed by a further ad­ vance of the pawn towards promotion; not 16 .l:.xe5?? dxc3, when Black is winning after both 17 'ii'c2 ltJg4 and 17 'ii'xd8 .l:tfxd8, threatening ... cxb2 and ...l1dl+) 16 ... l:.xb5 17 llxe5 'ii'd6 +. The material balance is restored and now Black just has a powerful centre for free. a2) 14 'ii'a4 l:tb6 15 e5 (after 15 lLlbl ? ltJg4 + Black dominates the board) 1 5 ... ltJxe5 16 ltJxe5 dxc3 17 bxc3 ..txe5 18 .l:lxe5 ..tb5 19 'ii'c2 ltJg4 20 l:.el llxf2 21 'iVe4 h5 is un­ clear but probably roughly balanced. Just a few variations to show this: a21 ) 22 ..te3 l:hg2+! 23 'iVxg2 ..tc6 24 'iVd2 'iVf6 (intending ... 'iVf3) 25 .:!.fl 'iVe5 26 'iVd8+ 'it>h7 27 'iVd3+ ..te4 28 'ii'e2 (28 'ii'd2? 'ii'b 8! + intend­ ing ....li!.b2) 28 ... ..td5 29 'ii'd 3+ ..te4 a22) 22 h3 .i.d3 23 'ii'a8 l:tb8 24 ikc6 l:.b6 25 'i!ka8 (25 'ii'a4 'ii'd6! with an attack) 25 ... .l:.b8 b) 12 a4 has a similar strategic idea to those we see in the Benko Gambit White prepares an outpost on b5. But in the Blumenfeld Gambit it's often con­ nected with a pawn sacrifice in order to exchange the a6-bishop and get the bishop-pair. 12 ...lLlb4 (a perfect out­ post for the black knight - it defends the d5-pawn and prepares ... e5) 1 3 .l:ta3 =.

=.

12 ..tg5 For some reason, this move is the most popular in my database. But the idea of exchanging one of the bishops in so open a position looks weird to me. Other moves:

54

AITA CK WITH BLACK

'ike8 14 lt:Jb5 (completing the idea) 14 ... i.xb5 15 axb5 .Ub8! (15 ... 'iixb5?! lets White grab the initiative by 16 i.h3 l:!ae8 17 lt:Jg5 l::te7 18 i.xe6+ �h8 19 e4!) 16 i.h3!? (White gets no­ where with 16 lt:Jg5? h6 17 lt:Je4 lt:Jxe4 18 i.xe4, Nikolic-Volokitin, Bundes­ liga 2004/5, 18 ...'i¥xb5 ! 19 i.g2 llf7 +) 16 ... e5 17 'ika4 'ikxb5 18 'i!i'xb5 .:txb5 19 i.e6+ �h8 20 lt:Jg5 .l:tb7 c) 12 lt:Jg5 (D) is an attempt to trap Black with some unexpected tactics. =.

c21) The point is that 14 lt:Jxd5? does not work so well now due to 14 ... lt:Jxd5 15 i.xe6+ (after 15 �xd5?! lt:Jd4! + followed by ... i.b7, ...h6, etc., Black seizes quite a dangerous initia­ tive) 15 ....:txe6 16 'ikxd5 .Uae8 17 lt:Jxe6 .l:.xe6 18 i.e3 lt:Jb4 +. capturing the long light-square diagonal. c22) 14 .l:r.e 1 (intending e4) 14 ... h6 15 lt:Jxe6! (with a rook on e1, this line is more attractive for White) 15 ....:txe6 16 lt:Jxd5 lt:Jxd5 17 'i!ixd5 l:1ae8 18 i.xe6+ .l:txe6 is somewhat unclear but I think Black should be fine. 12 h6! 13 i.xf6 'ikxf6 14 :tel The alluring but untried 14 lt:Jxd5 leads to equality after 14 ... exd5 15 'i!i'xd5+ �h8 16 'itxc6 i.xe2 17 lt:Jh4! i.xfl 18 .l:!.xfl g5 (the easiest way for Black) 19 lt:Jf5 ! (19 lt:Jf3? l:ha2 +) 19 ...'fixf5 20 'fixd6 .l:tf6 14 �h8 Avoiding lt:Jxd5 tactics. Black has the bishop-pair and a powerful centre; his compensation is obvious. 15 b3 Now: a) Not 15 ... c4? 16 bxc4 i.xc4 17 lt:Jd2! (a crushing move) 17 ... i.b4 (or 17 ... i.a6 1 8 lt:Jce4! and White is much better) 18 lt:Jxc4 i.xc3 19 lt:Jb6 �a7 20 lt:Ja4! +- Lputian-Babujian, Armenian Ch, Erevan 2008. b) 15 ... .l:.ab8!? is a new move that keeps an eye on the b5-square and evacuates the rook from the a8-h1 di­ agonal. After 16 lt:Ja4 'W/e7 17 lie1 , 17 ... c4! works well. Black has a strong initiative for the pawn. ...

B

=.

...

Now: c1) After 12 ... 'ike7? (Akhmadeev­ A.Zhigalko, Saratov 2006) White wins too many pawns with a combination you should memorize: 13 lt:Jxd5 ! exd5 14 i.xd5+ lt:Jxd5 15 'i¥xd5+ �h8 16 'ifxc6 i.b7 17 �b5 ±. Of course, Black has some positional compensa­ tion for the material damage, but three pawns is too much. c2) 12 ...'i!Vd7!? (this new move de­ fends c6 in order to avoid the tactic we have just seen) 13 i.h3 .l:fe8 and now:

5 B lumenfel d Ga m bit: 5 i.. g 5

1 d4 liJf6 2 5 .ig5 (D)

lDf3 c5 3 d 5 e6 4 c4 b5

a) 7 ... d6 8 lDc3 i..e7 (this new move covers d5 indirectly; in Khurtsidze­ Schuurman, European Clubs Cup (women), Saint Vincent 2005 Black played 8 ...liJbd7) 9 i..e2 (9 i..xf6 i..xf6 10 ltJxd5 i.. xb2 1 1 .l:tbl 'ii'a5+ 12 'ii'd2 'ii'xd2+ 13 ltJxd2 i.. xd5 14 lhb2 liJd7 leads to an equal endgame) 9 .. 0-0 10 0-0 a6 with unclear play. Now it's clear why the knight stayed on b8; from here it has much better prospects now. b) 7 ... a6 8 ltJc3 i..e7 (D) and now: .

This is the most popular continua­ tion and the most dangerous for Black. White strengthens his centre (by pinning a piece that attacks d5) and claims that the ... b5 advance has weakened Black's queenside. Often White gets pressure thanks to his con­ trol of the c4-square. Note that having played ... e6, Black cannot revert to Benko-style play: the central tension ensures that there will be a hand-to­ hand fight for the central squares in the near future. 5 exd5 6 cxd5 For some reason, 6 cxb5 is very rarely played, though it is by no means bad, and leads to interesting positions after 6 ... i..b7 7 e3. Then: •..

bl) 9 i..xf6 i..xf6 10 ltJxd5 is natu­ ral, taking the d5-pawn without further ado. But 10...i..xb2 I I .l:bl 'iia5+ 12 'ii'd2 axb5 ! works out fine for Black: bl l ) After 13 'ii'x a5?! .l:!.xa5 14 ltJc7+ 'iii>d 8 15 ltJxb5 i..f6 Black has preserved the bishop-pair and feels

AITA CK WITH BLA CK

56

good; e.g., 16 i.c4 (after 16 l"Lld6 i.xf3 17 gxf3 'iitc 7 1 8 l"Llxf7 l:.f8 19 i.c4 .l:.a4! 20 i.d5 l"Llc6 White is a little worse due to his badly placed knight) 16 ... 'iite7 + Breier-Feigin, Bundesliga 201 1/12. Black's bishops are too good. a2) 13 lhb2 i.xd5 14 i.xb5 'ifxd2+ 1 5 'iii>xd2 i.xf3 !? 16 gxf3 cJ;e7 with an equal endgame. b2) 9 a4!? is untried but logical. White does not have to take on d5 right away and plays this useful stan­ dard move instead. 9 ... 0-0 10 i.e2 axb5 1 1 axb5 (the b5-pawn plays an important role by restricting the b8knight but here Black can break through in the centre) l l .. . .l:txal 12 'iix a1 d4! (Black has to play aggres­ sively while White has not completed his development; otherwise, he is lia­ ble to end up cramped) 13 exd4 i.xf3 14 ..txf3 cxd4 15 ..txf6! (in the case of 15 l"Lle2?! ..i.b4+ 16 �fl h6 17 i.xf6 'ti'xf6 18 'i¥d1 .l:te8 + Black solves his problems and now it's White's tum to solve his own) 15 ... i.xf6 16 l"Lle4 'i!i'b6 17 l"Llxf6+ gxf6! (Black's main target is the b5-pawn) 18 i.e2 l:.e8 19 h7 13 lt:le4 (the knight moves to c5 to at­ tack e6 and b7 but thanks to his harmo­ nious development, Black can ignore this) 13 ... l::tc 8!? 14 lt:lc5 i..g4 15 lt:lxb7 'ifb6 16 lt:lc5 l:!.fd8 17 lt:la4 'i!Vb4 18 lt:lc3 lt:lb6 with active play to com­ pensate for the small material deficit, Plaskett-Bellon, Bahrain 1990. b) Now it is too late to push the knight back: 10 'i¥b3?! i..e6 1 1 'ifxb7 li:lxd4 12 li:lxd4 i.. xd4 13 i..h6 (13 litd1 i..xc3 14 bxc3 'i¥c8! + Barle­ Lalic, Yugoslav Team Ch, Brezovica 1988) 13 ...ltb8 14 'i!Va6 l:r.xb2! 15 li:lxd5 'iVxd5 16 i..xf8 'ifi>xf8 + Kliuner-Sie­ brecht, Duisburg 1999. Black's domi­ nation of the board is worth more than the sacrificed material. c) 10 h3?! i..e6 (D) is instructive because, without making any blatantly tempo-losing moves, White has ended up playing the black side of a main­ line Tarrasch QGD, without being a move up. How has this happened? In the reg­ ular Tarrasch line, 1 d4 d5 2 c4 e6 3 lt:lc3 c5 4 cxd5 exd5 5 li:lf3 lt:lc6 6 g3 li:lf6 7 i..g2 i..e7 8 0-0 0-0 9 i..g5 cxd4

74

w

AITA CK WITH BLACK

+) 14 ...'f6d6 15 lt::lc5 lt::lxd4 16 tbxd4 �xd4 17 lt::lxe6 'i!Vxe6 18 i.f3 'f6d6 +. c42) 13 ... 'ii'd8 14 lt::lc 5 lbxd4 15 lt::lxd4 ..txd4 16 lt::lxe6 fxe6 17 ..tf3 Wlb6 18 lie2 ( 1 8 W/e2? I!c2 19 .:!.ad 1 lDf4 -+) 18 .. Jhf3 ! 19 gxf3 'ifd6 + followed by ... ..txb2 provides nice play for Black. 10 i.f5 10 ... i.e6!? also deserves attention. ll �g5 1 1 'iVb3 (D.Allan-Fishbein, Chicago 1989) l l ...i.e6! 12 'iYxb7 tbxd4 13 lt::lxd4 �xd4 + is similar to note 'b' to White's lOth move. The fact that the rook is on e 1 instead of f1 does not amount to much. ll h6 Now: a) 12 �h4 (Akatova-Gunina, Rus­ sian Women's Team Ch, Dagomys 2010) allows Black to put pressure on d4. There is nothing to defend it, so White needs to take some action, ready or not: 12 ... lt::ldb4!? 13 d5 �xc3 14 bxc3 lt::lxd5 + with an extra pawn. b) 12 i.e3 lt::lxe3 13 fxe3 (Vidarte Morales-Peralta, La Bordeta 2010) 13 ... e5 ! (a new move) 14 d5 e4 15 dxc6 exf3 16 �xf3 bxc6 17 i.xc6 l:lb8 +. The two bishops have great prospects. ••.

10 lt::lxd4 h6 1 1 �e3, the white bishop has reached e3 via g5 (where it moved to provoke Black to resolve the central tension). Here, the black bishop has moved directly to e6 after White ex­ changed voluntarily on d5. It's un­ likely that even the most ardent fan of the Tarrasch will choose to reach this position as White, but you may find some opponents stumbling into it by accident. Rather than repeating heavy­ duty Tarrasch theory, I shall just men­ tion a few interesting lines that keep pressure on White: c 1) 1 1 lt::le4 'iYb6 12 lt::lc5 �f5 13 tba4 W/c7 +. c2) 1 1 �e3 Ii!.c8 12 'ifd2 'ii'b6 13 .l:.fd 1 .l:.fd8 +. c3) 1 1 i.g5 'i!Va5 ! 12 lt::la4 (12 'i!Vd2 lbxd4! 13 lt::lxd4 �xd4 + gave Black an extra pawn in Masic-Pap, European 822) Ch, Budva 2009) 12 ... l:r.fd8 13 lt::lc5 8 c5 This advance gives Black new pos­ �c8 14 lbb3 'iYb6 +. c4) 1 1 .l:!.e 1 'iYa5 !? 12 i.d2 .l:!.ac8!? sibilities thanks to the lack of pressure 13 tba4 ( 13 �c4 tbe3! +) and either on the d5-pawn. 8 lt::l c6 9 0-0 lt::le4 (D) queen retreat leads to interesting play: Now Black has two main ideas. The c41) 13 ... 'ii'c 7 14 .l:.c1 (14 tbc5 lt::l xd4 15 lt::lxd4 i.xd4 16 lt::lxe6 fxe6 first is to attack the c5-pawn by ... b6 ...

.••

1 d4 tbj6 2 c4 c5: 3 e3 AND 3 dxc5

75

ltJxd4! 15 ltJxd4 bxc5 16 Ji.d3 cxd4 1 7 cxd4 e5 + with a useful extra pawn) 12 ...bxc5 13 "fixd5 cxd4 14 ltJxd4 ltJxc3 15 bxc3 l:!.c8. This looks better for Black thanks to the weak c3-pawn. b3) 1 1 i.b5 i.b7 12 l:.c1 bxc5 13 dxc5 ltJxc3 14 bxc3 e5 is unclear, but Black has a comfortable game. 10 .tg4 (D) •••

and play on the queenside; the second is to play in the centre and on the kingside, with a possible advance of the e-, f- and g-pawns. 10 .tf4 Or: a) After 10 h3 Black can initiate play against White's pawn-centre by 10 ... b6!?, seeking to target the d4pawn. 1 1 cxb6 (in case of 1 1 ii.b5, as in Bachmayr-Wendt, 2nd Bundesliga 199112, Black can defend the knight from b7 with an advantage: 1 l .. . .tb7! 12 �a4 liteS 13 cxb6 axb6 +) 1 l...axb6 12 ii.e3 Ji.b7 with a comfortably equal position for Black. b) 10 ii.e3 b6!? (more aggressive than 10 ... ltJxc5, which simply leads to equal play: 1 1 dxc5 d4 12 ltJxd4 ltJxd4 13 l:lc1 ii.e6 Petrov-Guidarelli, Pula 2003) and now: b1) 1 1 cxb6 ltJxc3 12 bxc3 axb6 looks preferable for Black, since the a2- and c3-pawns are not good at all. b2) White does not equalize by 1 1 'ii'a4 ii.d7 12 'ii'b 3 (Black gets an ad­ vantage in case of the inaccurate 12 Ji.b5?! ltJxc3 1 3 bxc3 "fic7 14 "fia3 =

l l .te3

Now this move does not allow tac­ tics with ...ltJxc5. Instead: a) 1 1 ltJe5? simply loses a pawn to 1 1 ...ltJxe5 ! (l l ...iLxe2?, as in Hruska­ Z.Ilic, Prague 1980, misses a zwischen­ zug: 12 ltJxc6! 1Lxd1 13 ltJxd8 ltJxc3 14 bxc3 i.e2 15 ltJxb7 i.xfl 16 'it>xfl with an unclear endgame) 12 1Lxe5 1Lxe2 13 "fixe2 1Lxe5 14 dxe5 ltJxc5 +. b) 1 1 l:.c1 e6!? (intending ... "fif6) 12 ltJe1 (Black need not fear the more solid 12 ltJd2 1Lxe2 13 ltJxe2 'i¥a5 ) and now Black has the interesting plan of advancing his kingside pawns: 12 ... 1Lxe2 (this new move varies from Nei-Jansa, Tallinn 1983) 13 ltJxe2 g5 !? 14 ii.e3 f5 with the initiative. =

76

A ITA CK WITH BlACK

ll ... b6! ?

Here l l...tt'lxc5?! does not work since the g4-bishop is loose: 12 dxc5 .i.xf3 (the problem is 12 ... d4? 13 tt'lxd4 +-) 13 .txf3 d4 14 'ii'b 3 ;!; with pres­ sure. 12 'ii'a4 Or: a) Black has the better position in case of 12 cxb6?! tt'lxc3 13 bxc3 axb6 + thanks to the weaknesses on the a­ and c-files. b) 12 h3?! .i.xf3 13 .txf3 bxc5 14 dxc5 ( 14 lbxe4? loses a pawn after 14 ... cxd4! +) 14 ... lbxc3 15 bxc3 e6. Although White has the bishop-pair, Black's pawn-structure is far superior and Black even has slightly the better chances; e.g.: bl) The exchange sacrifice 16 c4? can be answered by the cold-blooded 16 ... tt'le5 ! 17 cxd5 lbxf3+ 18 gxf3 (18 �xf3? doesn't give full compensation without the queen on the al-h8 diago­ nal: 18 ... .txal 19 :Xxal 'ii'xd5 20 'iVf6 'ii'f5 +) 18 ... exd5 +. b2) 16 'ifa4 l:tc8 17 l:.acl �a5 ! 18 'ifxa5 lbxa5 +. 12 .. .'iVd7 13 lladl 13 .i.b5?! lets Black destroy White's pawn-structure: 13 ... tt'lxc3 14 bxc3 .i.xf3 15 gxf3 (15 .txc6?? 'ii'g4 16 g3 �h3 -+) 15 ...l:tfc8 +. 13 ... bxc5 Now: a) Not 14 lbxe4? lbxd4! 15 �xd7 lbxe2+ 16 Whl .i.xd7 17 lbxc5 .i.b5 1 8 a4 (18 .l:r.xd5?? lbg3+) 18 ... i.c4 19 b3 lbc3 ! 20 bxc4 lbxdl 21 l:!.xdl dxc4 + and ... c3. b) 14 .i.b5 lbxc3 15 bxc3 l:.fc8 +.

c) 14 lbxd5 ! (thanks to this com­ bination, White holds the balance) 14 ... 'ii'xd5 15 dxc5 �e6 16 .tc4 'iff5 (16 ... �f6 17 i.d5 i.xf3 1 8 gxf3 lbxc5 ! ) 17 �xc6 Jtxf3 1 8 gxf3 �xf3 19 i.e2 ii'xe2 20 �xe4 .l:.ad8 =

=.

823) 8 0-0 lbc6 (D)

White now has a huge range of pos­ sibilities. The points to bear in mind are that Black's principal plans are 9 ... Jtg4, putting pressure on d4 and encouraging White to release the ten­ sion by 10 cxd5 or 10 c5, and 9 ... dxc4, opening up a direct attack on d4. Against the latter plan, it is possible for White to make a pawn sacrifice of it by answering with 10 d5. With his next move, White can try to be ready for either plan, but most moves are committal in some way, and give Black a hint as to which plan works best. 9 h3 White prevents ... .tg4, but this move is less useful against the ... dxc4 plan. Other moves:

1 d4 ti:Jj6 2 c4 c5: 3 e 3 AND 3 dxc5

a) 9 c5 transposes to Line B22. b) 9 cxd5 ti:Jxd5 is covered in Line B21 . c) 9 ti:Je5?! is poor: 9 ... dxc4 10 ti:Jxc6 bxc6 1 1 .i.xc4 'ii'd6 + Ruiz Vinals-Lalic, Dos Hermanas 1998. The c6-pawn is less weak than the d4pawn. d) 9 a3 prepares to meet 9 ... dxc4 with 10 .i.xc4 ti:Ja5 1 1 .i.a2, but is much less relevant if Black replies 9 ... �g4. e) 9 .i.g5 dxc4 10 .i.xc4 .i.g4 (forc­ ing White to advance the pawn) 1 1 d5 .i.xf3 12 'i!Vxf3 ti:Je5 13 'ii'e2 ti:Jxc4 14 'i!Vxc4 ltc8 15 'ii'b3 and now 15 ...'ii'd7 is more precise than 15 ...'ii'c 7?! (Jelen-Dizdarevic, European Clubs Cup, Portoroz 1993), which can be met by 16 .li!.adl ! intending d6, .i.xf6 and ti:Jd5. f) 9 .i.f4 dxc4 and now: fl) 10 d5 ti:Ja5 1 1 �e5 (1 1 d6?! exd6 12 .i.xd6 lte8 1 3 ti:Jb5? .i.f5 14 ti:Jc7 l:1xe2! 15 'i!Vxe2 .l:!.c8 -+ Adly­ lvanchuk, World Team Ch, Ningbo 201 1) l l ....i.g4 12 �d4 .i.xf3 13 .i.xf3 ti:Je8 14 .li!.fel f6 15 .i.g3 e5 ! 16 �d2 ti:Jd6 is unclear, Raicevic-Velimirovic, Yugoslav Ch, Belgrade 1978. f2) 10 .i.xc4 ti:Ja5 !? (again this idea; 10 ... ..tg4 1 1 d5 ti:Ja5 12 i..e2 .l:r.c8 is also possible) 1 1 ..te2 ..te6 12 .l:r.cl .tlc8 13 �a4 a6 Nei-Velimirovic, Tallinn 1977. g) 9 l:tel .i.g4 (this seems to work out well enough, despite arguably be­ ing a little obliging; there may be a strong case for 9 ... .i.f5 ! ? , challenging White to find a way to make use of the move .:tel ; as Giddins points out in =

=

77

How to Build Your Chess Opening Repertoire,

there are even lines where the rook is vulnerable after ... ti:Jb4) and then: gl) 10 c5 ti:Je4 1 1 i..e3 gives us a position from Line B22 except for White's extra move .l:!.e l. However, even this does not give him an advantage. l l ...b6 ( l l ...f5!? is also interesting) 12 �a4 ti:Jxc3 13 bxc3 (S.Emst-L' Ami, Dutch Team Ch 2005/6) 13 ...'i!Vd7!? (intending ...ti:Jxd4) 14 'ii'b 5 a6! 15 �b2 (15 'iWxb6?? .l:r.tb8 -+) 15 ...bxc5 16 dxc5 e6 with a comfortably equal position for Black. g2) After 10 cxd5 ti:Jxd5 1 1 h3 i..e6 Black has fixed the isolated pawn and has serious ambitions. This is a reversed Tarrasch where White has the extra move .li!.el , which ought to be of some use, but in fact White has scored rather badly from this posi­ tion, so it may not actually change the nature of the play in any marked fash­ ion. Now: g21) 12 i..f l .l:tc8 13 .i.g5 h6 14 i..e 3 and here I like 14 .. .'ii' d6!? in­ tending ... .li!.fd8. Black should not rush to help White reconnect his pawns: 14 ...ti:Jxc3 15 bxc3 ti:Ja5 16 'ifd2 'ith7 17 ..ltf4 ..ltc4 Elianov-Strelnikov, Ukrainian Junior Ch, Kharkov 2000. g22) 12 i.. g5 h6 13 ..te3 (Kindl­ Milos, Groningen 1994) and again it's good to place the queen on d6: 13 ... �d6!? 14 �d2 ti:Jxe3 15 fxe3 f5 ! intending ... f4 with the initiative. h) 9 .i.e3 dxc4 and here: hl) 10 d5 needs study: 10 ... ti:Ja5 1 1 b4 (1 1 'ii'd2?! is too passive and lets Black complete his development and =

ATTA CK WITH BLA CK

78

keep the extra material: 1 l ...b6 12 l:.ad1, Darga-Hort, Bundesliga 1981/2, 12 ... i..b7 +) 1 1 ...cxb3 12 axb3 and then: h1 1) 12 ... lDxd5!? leads to compli­ cations. Let me show you the main line I found: 13 lDxd5 i..xal 14 'i¥xa1 (14 b4 i..g7 15 bxa5 e6 + and ... Vi'xa5 with an advantage) 14 .. .'ii' xd5 15 i..h6 f6 16 .l:.d 1 'i!Vc5 17 b4! (opening the diagonal to give check) 17 ... 'i'i'xb4 18 'iia2+ lDb3 19 i..xf8 �xf8 20 l:.d8+ �g7 21 i..d l ! (not 21 .l:te8? i..e6! 22 .l:txa8 lDd4 and Black wins, Inkiov-Gharamian, French Team Ch 2008/9) 2l...'i¥b6 22 l!e8 �f7 23 .l:!.h8 �g7 24 .l:!.e8 h12) 12 ...b6!? 13 b4 lDb7, followed by ... lDd6 and ... i..b 7, leads to un­ clear position. White's pieces are more harmonious, but Black has an extra pawn. h2) 10 i..xc4 lDa5 ! (a typical ma­ noeuvre one should remember: Black pushes the c4-bishop back so he can place his own bishop on e6) 1 1 i..e2 i..e6 12 iia4 (12 lDe5 lDd5 13 lDxd5 i.. xd5 14 iia4 lDc6 15 .l:Ifd1 e6 Kavalek-Matulovic, Havana Olympiad 1966) 12 ... lDd5 1 3 lDxd5 i.. xd5 14 .l:.ac 1 a6 15 lDe5 e6 intending ... lDc6 is fine for Black. White should then avoid 16 lDd7? i.c6 17 lhc6 iixd7! (better than 17 ... lDxc6? 1 8 lDxf8, when ac­ cording to my database Black resigned in Bregadze-Pavlidis, World Under-16 Ch, Kerner 2007, but it can't be true since the position is equal after, e.g., 18 ... lDxd4 19 i.xd4 'ifxd4 20 'i!Vxd4 i..xd4 21 .l:!.d 1 .l:.d8) 18 1'!c4 b5 19 'i!Vxa5 bxc4 + . 9 dxc4 10 i.xc4 lDaS! (D) =.

an

=

...

We have already seen this idea. The c4-bishop is pushed back so that Black can place his own bishop on e6. ll i..e2 Usually the bishop is worse placed on d3 but in this situation it makes sense in order to have the option of a later l:!.xe6 sacrifice. 1 1 i.d3 i..e6 12 .l:!.e 1 .l:tc8 (if Black is really scared of llxe6, he can continue 12 .. .'iVd7 or 12 ... 'ifd6) and now: a) 13 i..g 5 i.. c4 (an exchange of light-squared bishops will leave c4 and d5 less well defended) 14 i.c2 i.. a6 15 l:.c l (15 lDe4 lDc4 16 libl lDxe4 17 i.. xe4 lDd6 1 8 i.. d3 i..x d3 19 'i!Vxd3 .l:.e8 gave Black a pleasant position in Morchiashvi1i-Mi1anovic, Kavala 2007) 15 ... lDc4 16 b3 lDb6 17 lDe4 lDxe4 18 i.. xe4 'i¥d7 19 :f.xc8 .l:.xc8 and again Black has no problems, Epishin-Zviagintsev, Russia Cup, St Petersburg 2009. b) 13 .l:!.xe6!? is an absolutely cor­ rect exchange sacrifice, but White gets merely sufficient compensation. 13 ... fxe6 14 iie2 (Black's e6-pawn is doomed, and the white knight heads

1 d4 tbf6 2 c4 c5: 3 e3 AND 3 dxc5

for a nice outpost on e6; however, Black has counterplans) l4 ... 'it>h8 15 tbg5 tt:'lc6! 16 tDb5!? (16 tbxe6? tbxd4 +) l6 ... e5 ! (this is the most convinc­ ing move, and improves over the l6 ... a6 chosen in Lputian-Magerra­ mov, Daugavpils 1978) 17 tbe6 �d5 18 tbxf8 i.xf8 ! (it's important to take with the bishop in order to control c2 and c3; not 1 8 ... l:!.xf8? 19 dxe5 tbxe5 20 i.c2 ;!;) and then: b1) 19 dxe5? tt:'lxe5 (the d3-bishop has nowhere good to move) 20 tbxa7 (20 tbc3 .l:txc3 21 bxc3 tbxd3 and 20 i.c2 .l:txc2 21 'ii'xc2 'i!Vxb5 22 �c8 �g8 23 i.h6 tbed7 are both much better for Black) 20 . Jixc1+ 21 l:.xc1 tbxd3 22 .l:!.c8 'it>g7 gives Black a slight advantage. b2) After 19 tbc3 tbxd4 20 'ite3 'ita5 Black is fine. l l i.e6 12 l:te1 12 i.g5 l:tc8 13 l:!.e1 leads to the main line. 12 tbe5?! l:l.c8 1 3 1i'a4 (this doesn't make much sense since it just helps Black to strengthen his control of the light squares; 1 3 .l:tel transposes to note 'b' to White's 13th move) 13 ... a6! 14 .l:td1 b5 15 'i1Va3 'ifb6 gives Black a comfortable position; e.g., 16 �xe7 .l:.fe8 17 �b4 (17 �a3?? b4 1 8 tba4 bxa3 19 tbxb6 axb2 20 i.xb2 l:tc2 -+ Polak-Ftacnik, Pribram rapid 1997) 17 ...i.f8 1 8 tba4 (forced) 1 8 ...�c7 19 tbc5 i.f5 ! (White's pieces are in dan­ ger but there is a way to save them) 20 i.f3 ! .i.xc5 21 dxc5 .l:txe5 22 i.f4 tbc4 23 b3 'ili'xc5 24 �xc5 liexc5 25 bxc4 i.e4 12 l:tc8 (D) =

...

=.

•.•

79

13 i.gS

Or: a) 13 tbg5?! (there is nothing for the knight to do on g5) 13 ... i.c4 14 i.f3 (Rausis-Fominykh, Cairo 2001) 14...e6 15 i.e3 tbd5 +. Black makes desirable exchanges and gains total control of the light squares. b) Also there is no reason to acti­ vate the knight to e5: 1 3 tbe5?! tbd5 14 i.d2 (or 14 tbxd5 i.xd5 15 'ita4 tbc6 + Beutel-Mahdi, Aschach 1995) 14 ... !Dc6! (as everyone knows, the side facing an isolated pawn welcomes ex­ changes since they reduce counterplay and make the weakness more vulnera­ ble) 15 tbxc6 l:txc6 16 tbxd5 (16 i.f3 .l:c4 17 �e2 lhd4 + Zhukova-Beshu­ kov, Berlin 1995) 16 ... i.xd5 17 i.b4 i.f6 + Slogar-Sermek, Pula 1993. 13 tbc4 14 i.xc4 i.xc4 15 'itd2 l:te8 16 i.h6 i.dS The game is equal, Cabrilo-Leskur, Serbian Team Ch, Kragujevac 2009. Although White still has an isolated pawn, it's hard for Black to attack it ­ the bishop being on d5 disturbs rather than helps in this respect. .••

7 Vaga n ia n Ga m bit

1 d4 liJf6 2 c4 c5 3 liJf3 This is a popular way to avoid both the Benko Gambit and the Modem Benoni. In the theory books, you will find it classified as a Symmetrical English, with masses of theory on lines where Black develops with ... e6, etc. But there is no need for Black to play so slowly, as he can proceed in­ stead in gambit style. Moreover, it is a gambit with an excellent theoretical reputation and an unusually good re­ cord in practice. 3 ... cxd4 4 liJxd4 For 4 'ii'xd4 liJc6 5 'ii'd l e5 see note 'c' to White's 5th move. 4 ... e5 (D)

This is the most natural reply, eye­ ing the weakness on d6. White seem­ ingly stops 5 ... d5 (due to a double capture on d5 and liJc7+ ). Before we examine the main line, let's see the other possibilities for the knight: a) 5 liJb3?! d5 ! already affords Black an advantage in the centre: al) 6 e3 liJc6 7 i..e2 (7 cxd5 'i¥xd5 8 'ti'xd5 liJxd5 9 a3 i..e6 10 liJld2 :c8 + Oll-Westerinen, Oviedo rapid 1992) 7 ... i..e6 (attacking c4 and obliging White to take on d5) 8 cxd5 'ti'xd5 9 0-0 :d8 10 liJc3 �xdl 1 1 :xdl .Uxdl + 12 i..xdl i..b4 + followed by ... 0-0 and ... l:td8 gives Black the better position. a2) 6 cxd5 'r�Vxd5 ! (6 ... liJxd5 is also possible but then White has fewer problems activating his pieces: 7 e4! liJb4 8 i..b 5+ liJ8c6 9 'i¥e2 a6 10 i.. xc6+ liJxc6 1 1 liJc3 ) 7 'r�Vxd5 liJxd5 (now the c2-square is weak­ ened and the black knight heads there) 8 e4 (8 a3 prevents ... tt:Jb4 but leaves the b3-knight vulnerable; after 8 ... liJc6 9 e4 liJf6 10 liJc3 i..e6 1 1 liJd2 i.. c 5 + Black's pieces are more active) 8 ... tt:Jb4 9 liJa3 i..e6 10 i..b 5+ liJ8c6 1 1 0-0 (Popelka-Pacher, Teplice 201 1) 1 1...a6! (this new move is better than 1 l...liJxa2?! 12 liJa5 ! with an un­ Black weakens the d5-square but clear position) 12 i.. c4 (12 i..e 2? liJxa2! + costs White a pawn, while af­ it's only temporary. 5 liJb5 ter 12 i..xc6+ liJxc6 Black is better =

HI

VAGANIAN GAMBIT

due to the bishop-pair) 12 ... lDc2! and then: a21 ) 13 i.xe6? simply loses to 1 3 ... ttJxal -+. a22) Black needs to be precise in the case of 1 3 i.xa6!? i.xa3 ! (the only move that gives an advantage) 14 i.xb7 'it>d7! (a critical move!) 15 i.xa8 �xa8 16 bxa3 (16 �bl ? i.d6 -+) 16 ... i.xb3 17 litbl i.e6 + (17 ... i.xa2? leads to a pawn-down endgame after 18 .l::tb2! i.c4 19 .l:.xc2 i.xfl 20 'itxfl ;!;). a23) 1 3 tDxc2 i.xc4 14 l:td 1 l:td8 +. Black has the bishop-pair. b) 5 lDc2 d5 and now: b 1 ) After 6 e3 lDc6 7 i.e2 (7 ltJc3? d4! 8 exd4 exd4 9 lDb5 i.c5 + fol­ lowed by ... a6 gives Black an advan­ tage, Vasina-Melnichuk, Kiev 2006) 7 ... i.e6 8 cxd5 lDxd5 9 0-0 i.c5 + Kopnicky-V.Sergeev, Tatranske Zruby 2002. Black is better because of his more active pieces. b2) 6 cxd5 'ilixd5 7 'ilixd5 lDxd5 and here: b21 ) 8 g3?! is too modest; White should fight for the centre more radi­ cally. 8 ... lDc6 9 i.g2 i.e6 10 0-0 0-0-0!? (an interesting way to develop the rook faster) 1 1 i.d2 'itb8 12 ltJc3?! (relatively best is 12 .l:!.cl !? i..e7 +) 12 ... liJdb4! 13 lDxb4 i..xb4 14 i.. g5 f6 15 i..xc6 (to avoid ... ltJd4) 15 ... fxg5 16 i.g2 �d2 + Andersson-Gelfand, Tilburg 1990. b22) 8 e4 lDb4 9 lDxb4 i.xb4+ 10 i..d2 lDc6 is totally equal; e.g., 1 1 i.b5 i..d7 12 i.xc6 i.xd2+ 13 'iitxd2 i.xc6 14 ltJc3 0-0-0+ 15 'ite3 .:td6 16 .l:.ad1 .l:thd8 17 .l:!.xd6 �xd6 18 .l:td 1 .l:.xd 1 19

ltJxdl 1h- lf2 G.Mohr-A.Grosar, Por­ toroz 1993. c) 5 liJf3 attacks the e5-pawn, but after 5 ... ltJc6 Black's plan remains the same: ... d5. 6 lDc3 (6 g3?! d5 7 cxd5 'ilixd5 +) 6 ... i.b4 (renewing the ... d5 threat) 7 i.d2 (after 7 a3?! i.xc3+ 8 bxc3, as in Krasnopeev-Kron, Siberian Ch, Irkutsk 2009, 8 ... d6! + prepares to besiege White's doubled c-pawns by ... 0-0, ... i..e6, ... l:.c8, ... lDa5, etc.) 7 ... 0-0 and now: cl) 8 e3 e4 9 lDg5 .l:!.e8 10 a3 (T.Balogh-Rachela, Slovakian Team Ch 2000/1 ) 10 ... i.xc3 1 1 i.xc3 d6 (intending ... h6, meeting lDh3 with ... i..x h3) 12 h4 (in order to take on h3 with the rook) 12 ... ltJe5 (Black is in no hurry to chase the g5-knight back, in­ stead developing his own pieces) 13 i.e2 i.f5 14 Vi'd4 .l:tc8 15 .i:!.dl b6 +. White can't take on d6: 16 'ii'xd6? lDd3+ +. c2) 8 a3 i.xc3 9 i.. x c3 ltJe4! 10 �c2 ltJxc3 1 1 'ilixc3 d6 12 l:.d1 'ilie7 Munkhgal-Brandenburg, Shenzhen 201 1 . 5 ... d5! Black will not be left with a back­ ward pawn, and opens lines so he can develop his pieces rapidly and aggres­ sively. 6 cxd5 i.. c5 (D) Not 6 ... ltJxd5?? 7 �xd5 ! �xd5 8 lDc7+ �d8 9 lDxd5 +-. This move defines the Vaganian Gambit. Though the Armenian grand­ master did not invent the 4... e5 5 lDb5 d5 gambit as a whole, he did introduce this vital follow-up which was essen­ tial to make it respectable. =

=

82

A ITA CK WITH BLA CK

Black sacrifices a pawn for long­ lasting compensation in the form of free piece development while White suffers from a lack of space and slow development, and can face a strong at­ tack on the kingside (after ... e4 is played). The extra pawn on d5 is hard to defend, though it can sometimes be­ come a strength or be given back for positional dividends. Before delving into specific varia­ tions, we should talk about Black's plan. Considering that White can't play a quick e4 to support the d5-pawn be­ cause of ... ltJg4, Black will shortly play ... e4 himself. This keeps the d5pawn cut off and provides the black pieces with a natural deployment: e5 is a good square for a knight, f5 for the bishop, e7 for the queen, e8 and d8 for his rooks, while the c5-bishop often moves back to d6. Once this arrange­ ment is achieved, Black starts to create serious threats: attacking d5, entering on d3 or - the most dangerous for White - attacking on the kingside. For a while, Black will simply ignore the d5-pawn since it cannot go anywhere.

White has a choice between two main schemes: he can play e3 and �e2, or fianchetto this bishop on g2. In this chapter we shall look at the g3 plan and a variety of other set-ups for White. The next chapter focuses on 7 e3. Our coverage divides into two main moves here: 83 A: 7 ltJlc3 85 B : 7 ltJ5c3 In both cases, we only deal with lines that don't transpose to the next chapter; that is, White doesn't follow up with 8 e3. A few other moves are worth men­ tioning too: a) After 7 'ifc2 ltJa6 it's important to point out that White can't defend the d5-pawn with the e-pawn. White should certainly avoid 8 e4? ltJg4 +, but he also has trouble after 8 lDlc3 ltJb4 9 �a4 0-0 1 0 �e3 (this looks odd but it's an absolutely required step; 10 e3? .1f5 1 1 ltJa3 ltJfxd5 +; 10 �d2? a6 1 1 ltJa3 ltJbxd5 +) 10 ... .txe3 1 1 'ifxb4 �b6 12 .li!.d1 .i.a5 1 3 �b3 �f5. Black has a firm grip on the posi­ tion thanks to his better development, while White's extra pawn means noth­ ing in such circumstances. b) 7 d6?! is a tricky move. White supports ltJc7+, which at first sight appears quite dangerous for Black. However, the natural 7 . . . 0-0 demon­ strates that this threat can be ignored. Then 8 ltJc7 (after 8 lD1c3 ltJc6 +. White no longer threatens ltJc7 and Black will gradually surround the d6pawn, after which he will have an

VAGANIAN GAMBIT

advantage due to his better develop­ ment) 8 .. .lbe4! (threatening instant mate) 9 e3 (9 .i.e3? .itxe3 l 0 fxe3 '1if6 1 1 'ii'd5 'ii'f2+ 12 'it>dl 'ii'xe3 followed by .. :�:Jf2+ is hopeless for White) 9 ... .i.b4+ (D) leaves White no good defence against Black's activity. w

83

c) If White wishes to reach Line B2, then he can also start with 7 g3 because Black can't land any tactics based on f2 and b5. 7 ... 0-0 (the tactic 7 ... .i.xf2+?? does not work because after 8 'it>xf2 'ii'b6+ White can defend king and knight by 9 e3 +-) and now: cl) 8 .i.g2? does not work well because the b5-knight and f2-pawn can't be defended simultaneously af­ ter 8 ... 'ii'b6 +. c2) 8 tblc3 transposes to Line A. c3) For 8 tb5c3 see Line B2. A)

tLl lc3 This is not a good idea because the b5-knight will be pushed back to a3, after which the knights will be awk­ ward on a3 and c3. They are better placed on c3 and d2. 7 0-0 8 g3 Not 8 e4? l2Jg4 +. For 8 d6?! tbc6, see note 'b' to White's 7th move above. 8 e3 transposes to note 'b' to White's 8th move in Chapter 8. 8 a6 9 tb a3 (D) 7

bl ) After 10 tbd2? �xd6 1 1 tbxa8 .l:i.d8 -+, Black will first win the d2knight, and then capture the other white knight, securing a material ad­ vantage. b2) 10 .i.d2 tbxd2 1 1 tbxd2 �xd6 12 tLlb5 (after 12 tbxa8? I!d8 -+ Black wins both white knights) 12 ... 'ii'g6!? (keeping an eye on the g2-pawn) 1 3 a3 (or else Black plays ....l:!.d8) 13 ....i.xd2+ 14 'ii'xd2 tbc6 + followed by .. Jld8 and ... .i.e6. b3) 10 tbc3 tbxc3 1 1 bxc3 .i.xc3+ 12 .i.d2 .i.xal 1 3 'ii'x al 'ii'xd6 14 tbxa8 .i.e6 (intending ...tbd7 or ...tbc6, when Black keeps an extra pawn) 15 'ii'b2 tbc6 16 �xb7? (it is better to give up the knight) 16 ... .l:!.b8 1 7 'ii'a6 �b1+ 1 8 We2 l2Jd4+ and the white queen is lost.

...

...

B

84

AITA CK WITH BLACK

b) After 16 lt'lxc8 'fia5+ 17 �e2 It's extremely important for Black (D) Black gets a decisive advantage by to provoke e3 to gain access to f3 and unexpected means: d3. 10 e3 Now we examine two moves for Black: A l : 10 e4! ? 84 A2: 10 i.xa3! ? 85 9 JWb6! ••

•.•

•••

Al} 10 e4! ? •••

This typical thrust is appropriate here since it is in keeping with all the main ideas of the gambit: the e4-pawn stakes out Black's space advantage while the pawns on e3 and g3 leave weaknesses on f3 and d3. l l lt'lc4 'fic7 12 a4? ! We shall take this as the main line here, even though it is not White's ob­ jectively best option. He should first look after his king and only then think about the queenside. After 12 i.g2 l:r.e8 13 0-0 b5 14 lt'ld2 i.f5 followed by ... lt'lbd7, ... :ad8, ... i.d6, ...lt'lc5, etc., Black has sufficient compensation. 12 ...11d8 13 'fib3 lt'lxd5! This is better than 1 3 ... i.g4?!, as played in Inneman-Prymula, Czech Team Ch 1997/8. With the text-move, Black recap­ tures the gambit pawn, but White can immediately take another pawn: 14 lt'lxe4 White can't win the exchange by 14 lt'lxd5? .l:.xd5 15 lt'lb6 because of 15 ... l:id8 !. Then: a) 16 lt'lxa8? 'fia5+ 17 'it>e2 (17 'fic3 i.b4 -+) 17 ... i.g4+ 18 f3 exf3+ 19 'it>f2 .l:td2+ leads to checkmate.

17 ... i.b4! ! (the idea is ... 'iYh5+; the simple 17 . Jhc8? leads to White's ad­ vantage after 18 i.g2 ;\;). Then: bl) 1 8 h3 'iYc5 (intending ... l:r.d3) 19 lt'lb6 lt'lc6! (Black brings his pieces into the battle as quickly as possible; 19 ... l:.d3? 20 'fic4 'fixb6 21 i.g2 is unclear, while 19 ... 'fixb6?! 20 i.g2 lt'ld7 + is advantageous but less con­ vincing) 20 lt'lc4 :d3 21 'iYc2 l:r.ad8 -+. b2) 18 i.g2 'iYh5+ 19 f3 lt'lc6 20 lt'lb6 (otherwise Black takes the knight with a winning position) 20 ... lt'le5 (20 ...l:ld3 is also good) 21 g4 'iih4 gives Black a decisive attack; e.g., 22 lt'ld5 exf3+ 23 i.xf3 lt'lxg4 -+. 14 i.e7! Leaving the b4-square for the knight. White's main problem is that he can't easily castle: 15 i.g2?! lt'lb4 16 0-0? l:td3 17 lt'lc3 i.e6 -+. After 15 i.e2 i.h3 ! +. preventing 0-0, Black's com­ pensation is beyond doubt. .•.

VAGANIAN GAMBIT

85

( 14 g5 ltJe4 15 iLd3 ltJxg5 16 0-0 lbd7 17 l:tb 1 'ii'c5 + followed by ... e4 and This new move is a concrete and ...ltJe5) 14 ....l:l.xe3+!? (14 ...ltJxg4 leads to a draw by repetition after 15 1'ixg4 straightforward continuation. .l:.xc 1 + 16 Iixcl 'ii'xe3+ 17 d2 J:ld8 ! 1 3 tt'lxc5 tt'le3+ 14 tt'ld3 tt'lxd 1 15 'it>xd 1 tt'la6 + ; 1 1 g5?! is less solid, and after 1 l ...tt'lxc3 12 WUxd8 Iixd8 13 tt'lxc3 i.. f5 Black is better) 1 l ...tt'lxc3 12 'V.!Vxd8 l::r xd8 13 tt'lxc3 i.. c 6 with a slight advantage for Black. b) After 1 0 tt'lxe4?! tt'lxe4 1 1 e3 Black gets a strong attack by 1 l...'V.!Vh4 12 �c2 f5 ! (the only move, but good) 13 ii.g2 tt'la6 14 gxf5 (not 14 tt'lc3?? tt'lb4 15 'ife2 tt'ld3+ 16 'ifxd3 'i¥xf2+ 17 c;t>d 1 'ifxg2 -+ Anton-Salem, World Under-18 Ch, Kerner 2009) 14 ...tt'ld6 15 0-0 .l:i.xf5 16 tt'lc3 .l:g5. c) 10 .:l.h2 is definitely White's best response. 10 ... i..g6 1 1 g5 !? (White should continue with his plan of ad­ vancing his pawns; after 1 1 i..g5?! tt'lbd7 12 tt'ld2 .:l.c8 Black had the better chances due to the awkwardly placed white pieces in Raetsky-Kae­ nel, Lenk 1995) 1 1 ...tt'lh5 (l l ...tt'le8?! is less convincing due to 12 h4 tt'ld6 13 tt'ld2 intending h5, e4 and tt'lc4; White is preferable since he is still a pawn up and it's not clear how Black is going to gain compensation) 12 tt'ld2 (Black can meet 12 e4 with 12 ... tt'ld7 intend­ ing . Jk8, ... i..b4, ... tt'lc5 or . .f5, open­ ing the f-file for more counterplay; with the knight on h5, Black clearly has more options) 12 ... i..b4 is a rather un­ clear position where the white pieces look awkwardly placed for the coming battle. Black's next moves could be ...tt'ld7, ... .l:tc8, . .f6 and ... e4. 9 tt'lbd7 10 g4 10 �f3 i.. g6 1 1 e4 allows Black to initiate play against the e4-pawn: 1 1...i..d4! (intending ... tt'lc5) 12 i..d3 ...

VAGANIAN GAMBIT

tbc5 13 .tc2 (13 0-0? ttJxd3 14 'ii'xd3 ttJxd5 + Palus-Vaulin, Polanica Zdroj 1998) 13 ... b5 and now: a) 14 tiJd2?! .txc3! (a draw was agreed in an unclear position after 14 ... b4?! 15 tbe2 tbxd5 in Papaioan­ nou-Stefanopoulos, Greek Ch, Athens 1998) 15 bxc3 tDxd5 + followed by .. lieS, ...ttJf4, etc. b) 14 a3 has not so far been tried. Then 14 ... a5 (renewing the ...b4 threat) 15 0-0! (this tricky idea helps White to maintain the balance) 15 ...b4 16 axb4 axb4 17 .l:txa8 'ifxa8 18 tiJd2! (a key move: the d4-bishop is going to be trapped) 18 ...bxc3 19 bxc3 .txf2+ 20 'ii'xf2 (20 .l:.xf2? tbxd5 +) 20 ...ttJfxe4 21 .txe4 tbxe4 22 ttJxe4 .txe4 23 d6 'i!Vc6 reaches an equal endgame. .

10 .tg6 (D) ...

Now: a) 1 1 g5?! ttJe4 12 h4 looks like a dubious continuation of the pawn ad­ vance: a1) 12 ... f6?! leads to problems: al l) In the only practical example, 13 .tg2?! tiJd6 14 h5 gave Black a

87

chance to maintain the balance. After 14 ... .te8!? he is a pawn down but his pieces are better prepared for the com­ ing battle. Instead, after 14 ... .txbl ?! (the light-squared bishop is too impor­ tant to exchange, especially after ...f6) 15 .l:.xbl fxg5 16 �g4 a draw was agreed in Polak-A.Rotstein, Austrian Team Ch 199617, despite White obvi­ ously having the better chances. al2) 1 3 h5 ! .tf5 14 g6! (White cracks open the light squares around Black's king; at first glance 14 'iff3?! should win but Black has a strong de­ fensive resource that leads to crazy positions: 14 ... tiJd6 15 e4 .txe4! 16 tbxe4 fxg5 17 'ii'g4 tbxe4 18 'ifxe4 .txf2+ 19 �dl i.d4 intending ... ttJf6; the position is very complicated and hard to evaluate without any practical tests) 14 ... hxg6 (the natural 14 ... h6? simply loses a piece to 15 'iff3 +­ since there is no play connected with ... fxg5 any more) 15 tbxe4 .txe4 16 hxg6! i..xg6 (16...i.xhl?? 17 �h5 .l:.e8 18 d6! +- intending i..c4+) 17 'ii'g4 gives White a strong attack. a2) I think 12 ... ttJd6!? is the right approach, keeping the pawns together. White has weakened his kingside but achieved nothing concrete, so it's time for Black to fight back. For example, 13 i..g2 (the instinctive 1 3 h5?! i..f5 + increases White's problems - now he needs to take care of the g5-pawn) 1 3 .. Jic8 14 tiJd2 i..b4 15 0-0 tbc5 + followed by ...ttJd3. b) 1 1 i.g2!? (this untried move is more solid than 1 1 g5) l l ..J::t c8 12 0-0 h6 (Black prevents g5 and makes some luft for his king) 1 3 e4 a6. Black

A TTACK WITH BLA CK

88

prepares ... b5 and frees the a7-square for his bishop. Although White has a protected passed pawn on d5, Black still has good compensation thanks to the d6-square for the knight ( ... CLJe8d6), control of the a7-g1 diagonal and the c-file, weak dark squares on White's kingside (thanks to the g4 advance) and simply better develop­ ment. 82) 8 g3 (D)

B

it should give Black sufficient com­ pensation. The most popular continuation is 9 ... .ltg4?! but after 10 .lte2 .ltxe2 1 1 'iUxe2 e4 (otherwise White plays e4 with advantage) 12 a3 ! White gets very good results. 12 ...CLJbd7 1 3 CLJd2 llfe8 14 b4 .ltf8 15 0-0 ;\; Aronian­ Sutovsky, European Team Ch, Kher­ sonissos 2007. Black's pieces are too far from White's weakened kingside. Lines like this forced me to seek an­ other way to get counterplay. Actually, 9 ... e4 also leads to the main line after 10 .ig2 .if5. 10 .i g2 (D) B

White's plan is to fianchetto his king's bishop, castle, and then gradu­ ally develop his queenside. And in­ deed things would look grim for Black if he could not weaken White's pawn­ structure by the following move: 8 .. Wb 6 ! 9 e3 Now the light squares in White's camp are vulnerable and Black has po­ sitional advantages to compensate for the gambit pawn. 9 ... .ltf5! Another critical move. It's actually very rarely chosen, but in my opinion J

10 ... e4

Black gives White no chance to support the d5-pawn with his e-pawn. 1 1 0-0 CLJbd7 Black slowly completes his devel­ opment while White suffers from an obvious space problem. 12 CLJd2 Sooner or later White will make this move anyway. 12 .. Ue8 (D) J

VAGANIAN GAMBIT

89

shows that he can ignore this idea. 14 lbdxe4 lbxe4 15 lbxe4 'i!Vg6! 16 li'a4 (the only move, eyeing the d7-knight) 16 ... lbb6 17 �xe8+ .l:i.xe8 18 lbxc5 lbxd5 ! (Black is in time to take an im­ portant pawn) 19 .i.xd5 �d6 20 .i.xb7 �xc5 d) 13 h3 intends slower play, but Black can create threats right away: 13 ... .i.b4! (threatening to take on c3 and d5) 14 .l::!.b 1 (defending the pawn but now Black has new ideas) 14 ...�d6 15 g4 .i.g6 16 lbc4 �a6 17 �b3 .i.xc3 18 bxc3 .:.ac8 19 lba3 lbc5 20 �c4 �xc4 21 lbxc4 l::ted8 (2l ...lbxd5? 22 tbd6 lbxc3 23 l:!.b2 ±) 22 lba5 (White will win a pawn to compensate for the one on d5, but Black's initiative grows) 22 ....l:i.xd5 23 lbxb7 h5! 24 gxh5 .i.xh5 intending ....l:i.g5 and ... .i.e2-d3, etc. Black is OK. e) 1 3 a3 intends b4, and 13 .. J::tac8 prevents it. Now: el) 14 b4?! would be met with 14 ... .i.xe3, discovering an attack on the c3-knight. e2) After 14 �a4, Black can't play ... .i.b4 to free the c5-square for the knight, but there is another square­ vacating move: with 14 ...�c7! Black prepares ... lbb6 while playing b4 still does not give White any advantage: e21) 15 lbcxe4? lbxe4 16 lbxe4 b5 ! 17 �c2 �e5 ! -+. e22) 15 b4?! .i.xe3 ! 16 lbdxe4 lbxe4 17 .i.xe4 (17 lbxe4? .i.xc 1 18 l::taxcl �XC 1 19 ltJd6 �g5 f) 17...ltJb6 18 �d1 (after 18 �xa7 .l:!.a8 19 lbb5 �e5 20 �xb7 .i.xe4 21 .i.xe3 lbxd5 the white king is in danger) 18 ....i.xf2+ 19 .l::!.xf2 .i.xe4 20 lbxe4 .l::!.xe4 +. =.

White has prevented a frontal as­ sault on his king, but this has not been cost-free: his king's bishop is out of play on g2, and his queenside is some­ what vulnerable. Also, White's d5pawn is still weak. I have found just three games played from this position, so most of what fol­ lows is the result of pure analysis. Now we examine: 90 B21 : 13 lbc4 B22: 13 b3 91 Or: a) After 1 3 �a4?! (as played in the game Francsics-G.Feher, Budapest rapid 1996) 1 3 ... .i.b4! followed by ...tbc5 or ... .i.xc3 and ... lbxd5, Black has sufficient pressure for the pawn. b) 1 3 lba4?! doesn't make much sense. Allowing the black knight into d3 is too large a price to pay for exchanging off the c5-bishop. After 1 3 ...'i!Va6 14 lbxc5 lbxc5 the d5-pawn is going to fall, and Black has pleasant prospects. c) 1 3 �c2 threatens to take on e4, but with the smart 1 3 ....l::!.ac8!, Black

90

A TTACK WITH BLA CK

e23) 15 'i¥c2 i.f8 gives Black many possibilities, such as ...lt:le5-d3, ...lt:lc5d3 and ... lt:lb6xd5; White's position is cramped and Black can be happy with the results of the opening. e3) 14 ltbl prepares b4 but Black again easily stops it: 14 ... i.d6 15 h3 (preventing ...i.g4) 15 ... i.g6 gives Black a very comfortable position. 82 1 ) 13 lt:lc4 'ifa6 1 4 'ifb3 .l:.ad8! ? iLf8

14 .. (as played in Khetsuriani­ Pountzas, Greek Team Ch, Ermioni Argolidas 2005) leaves the c5-square vacant for the d7-knight, and also de­ serves attention. The text-move brings a piece into play, which is why I prefer it. 15 .Ud1 15 a4?! weakens the b3-square, which Black can exploit immediately by 15 ... i.f8! (much more effective than it was a move ago) with more than enough compensation.

Black unexpectedly starts the sort of kingside attack that we shall see in the next chapter, even though the bishop is on g2 here. 17 lt:lxe5 :!.xe5 18 i.fi This is the most natural. After 1 8 a3 b6! Black prepares ...�c8 followed by ... i.f3 or ... i.h3, with a dangerous initiative; e.g., 19 l:.bl 'ii'c 8 20 'ii'c4 i.h3 (D) and now:

.

15 ...i.g4 16 l:.d2 (D)

B

16 ...lt:le5!

a) 21 b4 .l:!.h5 ! throws all Black's pieces into the attack: al) 22 lt:le2 i.xg2 23 lt:lf4 l:.h6 24 bxc5 (24 'iti>xg2 g5 +) 24 ... i.f3 intend­ ing ... g5 and ... 'ii'h 3. White is in trou­ ble. a2) 22 i.xe4 lt:lxe4 and then: a21 ) 23 lt:lxe4?! 'iff5 24 lt:lc3 l1e8! (preventing 25 'it'e4 while intending ... 'ii'f3 with mate threats) 25 'ii'f4 l:r.xe3 ! ! 26 l::td l (26 fxe3 i.xe3+ 27 l:tf2 i.xf4 28 i.xf4 'ii'd3 +) 26 ... 'ii'xf4 27 gxf4 l:!.xc3 28 bxc5 l::.X c 5 29 d6 l:thd5 30 i.e3 .l:r.a5 + followed by ... 'iii>f8-e8-d7. a22) 23 'iVxe4 i.f5 24 'iVf3 i.xe3 ! 25 fxe3 (25 VWxe3? :es 26 'ifxe8+

VAGANIAN GAMBIT

'it'xe8 27 .l:te2 'ifxe2 28 lt:Jxe2 .i.xb1 29 lt:Jc3 .l:te5 +) 25 ... 'ifxc3 26 'i!Vxh5 ..ixb1 27 1'kd1 'i!Vxe3+ 28 l:tf2 �d3! + gives Black good chances in the end­ game. b) 21 .i.h 1 keeps a key defensive piece on the board but fails to repel the attack. 21 ...'iff5 22 b4 .i.xe3! 23 fxe3 l:tc8 24 'ife2 .l:.xc3 25 .i.b2 .l:b3 26 'it'c4 :xd5 ! 27 .li!.f2 .l:!.xb2! 28 .l:.fxb2 h6 gives Black powerful compensa­ tion for the exchange.

9/

822) 13 b3 .i.b4 14 .i.b2 (D)

18 'it'b6 19 'iixb6 ..ixb6 (D) ••.

14 .i.xc3! ? •..

In Beliavsky-B.Vuckovic, Euro­ pean Ch, Budva 2009 Black continued 14 ....l:.ac8?! 15 lt:Jc4 'ili'a6. Then Beli­ avsky chose 16 lt:Je2?! b5 17 a3 .i.f8 18 lt:Jd2, when Black could have obtained a safe position with 18 ...'ifb7! and taking on d5 (instead of 18 ...lt:Je5?!, which he actually played). But the problem is that White can secure a clear edge by 16 a4!? lt:Jb6 (16 ... lt:Jc5 ?? 17 lt:Ja2 .i.a5 18 lt:Jxa5 'ili'xa5 19 .i.xf6 gxf6 20 b4 +-) 17 lt:Ja2 lt:Jxc4 18 .i.xf6 'i!kxf6 19 bxc4 ;!; and �b3. That's why I prefer the text-move, which is un­ tested so far. 15 .i.xc3 lt:Jxd5 16 ..id4 li'e6 Black has solved all his opening problems, and we can assess the game as equal. The black rooks have enough open lines, the f5-bishop and queen support the e4-pawn and control the situation on the kingside, while the knights have a variety of squares at their disposal. =

The d5-pawn is hanging and White must defend it by 20 ..ic4 if he hopes to get any advantage. But then 20...:h5! simply continues the plan of attacking the white king; e.g., 21 b3 .i.a5 22 .i.b2 ..if3 23 l:tc1 lt:Jg4 24 h4 g5 25 .i.e2! gxh4 26 .i.xf3 exf3 27 lid4 f5! 28 gxh4 .i.c7 29 d6 .i.xd6 30 lt:Jb5 (forcing simplifying exchanges) 30 ... .i.h2+ 3 1 �fl .l:txd4 3 2 lt:Jxd4 .i.e5 3 3 �g1 .i.h2+ 34 �fl ..ie5 with a draw by repetition. Of course, this line is not forced - it's just a demonstration of the power of Black's attack.

8 Vaga n ia n Ga m bit: 7 e3

l2'lf6 6 This is the most common choice in practice, even though the lines with g3 in the previous chapter may be a more critical test of Black's gambit. The main drawback of playing e3 (rather than fianchettoing) is that the b8-h2 diagonal remains open for future play against the white king. 7 0-0 8 tLl5c3 White can also reach this position using the move-order 7 tLl5c3 0-0 8 e3. Or: a) 8 i.c4?! does not make any sense since it does not disturb Black's plans but rather provokes them - when the knight moves to e5, it will win a tempo by attacking this bishop. Also, the white bishop is much better on e2 in these positions, since it covers f3 and g4. 8 ... a6 9 tLl5c3 e4 10 l2'ld2 �e7 fol­ lowed by ... i.f5 and ... tLlbd7-e5 leads to similar play to the main line, but substantially improved for Black. b) 8 l2'l1c3 suffers from the same defects as the analogous line (7 l2'l1c3) that we saw in the previous chapter. After 8 ... a6 9 tLla3 b5 10 tLlc2 b4 1 1 tLla4 i.e7 12 i.c4 i.b7 Black regained the pawn with a comfortable position in Benderac-T. Vasilevich, Belgrade (women) 1997. 1 d4 2 c 4 c 5 3 l2'lf3 cxd4 4 tLlxd4 e5 5 t2'lb5 d5 cxd5 i.c5 7 e3

As always, this thrust is a major part of Black's plans. He stakes out a space advantage, cuts off support for the d5-pawn, frees the e5-square for a black knight, and opens the b8-h2 di­ agonal.

...

8 e4 (D) ...

Now we consider: 92 94 l2'l 97

A: 9 a3 B: 9 d2 C: 9 i.e2 A) 9 a3

Although this move does not look so logical (White advances a rook's pawn while his king is still uncastled), it is quite dangerous. White intends b4 followed by i.b2, which is an excel­ lent place for the bishop. Control of the long diagonal will be major a

VAGANIAN GAMBIT: 7 e3

factor if White manages to simplify the position. 9 .. :fle1 My general advice is not to be too scared by White's b4 idea. Often Black can allow this advance and exploit it later by playing ... a5. 10 b 4 �d6 1 1 lLld2 �e5 1 1 ...it.f5?! is not good because of 12 ltJc4 ;!; followed by lLlb5. Black should almost never allow the exchange of his dark-squared bishop since this leaves the b2-bishop unopposed and sharply reduces Black's attacking chances on the kingside. 12 i.. b 2 l:!.d8 13 "ii'b 3 (D)

13 ... a5!

With this new move, Black gets ei­ ther an outpost on c5 for his queen's knight or creates a weakness on b4 to attack. 13 ... i..f5?! allows the exchange of the e5-bishop and gets nothing in re­ turn: 14 ltJc4 i.. xc3+ 15 'i!Vxc3 .l:txd5 16 ltJa5 ;!; followed by i..c4 with ad­ vantage, A.Onishchuk-Summerscale, Hastings 1995.

9J

14 ltJc4

In case of 14 bxa5 l:.xa5, we have the following lines: a) 15 ltJc4?! lt.xc3+ 16 i..xc3 .l:r.axd5 17 liJb6 (17 i..e2?! .:g5 18 g3 i..h 3 + shuts the hl-rook in) 17 ....Ug5 stops the fl-bishop's development, and Black is OK. b) 15 i..c4 lLla6 gives Black suffi­ cient compensation since 0-0 is im­ possible: b1) 16 0-0? i..xh2+! 17 �xh2? (better is 17 �h 1 lLlc5 18 "flc2 i..e5 +) 17 ...ltJg4+ 18 �gl (18 �g3 'iee5+ 19 f4 exf3+ 20 �h4 'iVh2+ 21 �g5 'ieh6#) 18 ... 'ii'h4 19 Il.fd1 'ii'xf2+ 20 �h1 !ld6 leads to checkmate. b2) 16 .:I.e 1 lLlc5 17 "fib4 l:.a8 gives Black very good compensation. His pieces have taken up ideal squares and now the question is what White can do. Castling is still impossible because of ... i..xh2+, while there are no objects for White to attack or pieces for him to im­ prove. White's game is more difficult, which is why I prefer Black here. 14 ... axb4 15 axb4 Or: a) 15 "fixb4 'i!Vxb4 16 axb4 .l:txa1+ 17 i..x al i..x c3+ 18 i.. xc3 lLlxd5 19 i..d2 i..e6 +. With his rook playing no active role, White hardly can hope to maintain the balance. b) 15 lLlxe5 (here the exchange of the dark-squared bishop is no big deal for Black since he has other advan­ tages, most notably the uncastled king on e1) 15 ... bxc3 16 St.xc3 lLlxd5 and then: bl) After the slow 17 i..b2?!, Black gets strong play by 17 ... lLla6! 18 i.e2

94

ATTACK WITH BLACK

( 1 8 .i.xa6 .l:txa6 19 0-0 .i.e6! + intend­ ing ...f6 and ... lbb6) 18 ...f6 19 lbc4 lbc5 20 �c2 lbd3+ 21 .i.xd3 exd3 -+ with a crushing attack. b2) 17 ii.c4 (hurrying to castle) 17 ... ..te6 1 8 0-0 lba6 +. The e5-knight feels especially uncomfortable with ... f6 threatened. 15 ... l:!.xa1+ 16 .i.xa1 .i.c7 (D)

his light squares are left undefended) 1 8 ... .i.e5 19 .i.c4 .i.f5 and now: a) 20 0-0? is still impossible, due to 20 ... ii.xh2+! 21 �xh2 (21 �h1 lbg4 -+) 2l ...lbg4+ 22 �g3 (22 �g1 'ii'h4 23 l:!.d1 'ii'xf2+ 24 �h1 .l:!.d6 -+) 22 ...'ii'd6+ 23 f4 'ii'h6 -+. b) 20 h3 prevents tactics involving ... ..txh2+ and also rules out ... lbg4. After 20 ... lbc7 21 0-0, Black must choose carefully: b1) Regaining the sacrificed pawn by 2l ....i.xc3? 22 "ii'xc3 lbcxd5 leads to rather an unpleasant endgame of a type Black should usually avoid in this opening: 23 ii.xd5 .l:!.xd5 24 'iVc4 .l:!.d8 25 .i.xf6! 'ifxf6 26 lbd4 ;!;;. The knight dominates the bishop. b2) 2l.....td6! intends ...'i!i'e5, at­ tacking h2. White's move h3 was a good way to prepare castling, but it also weakened the b8-h2 diagonal, and now Black makes use of this. Black's attack almost plays itself. White can try 22 h4, planning to meet ...'iVe5 with g3. But naturally this gives Black new pos­ sibilities, such as 22 ...lbh5 23 g3 ii.h3! followed by ...lbxg3 with a practically decisive attack - all White's pieces are on the other side of the board. 17 lbxd5 18 lbxc7 Now 1 8 ...1\Vxc7? 19 ..te5 +- costs Black his b8-knight, but 1 8 ... lbxc7 19 ii.e2 lbba6 gives him easy play on the queenside.

Black's main idea now is to attack the b4-pawn with ... lba6. If White plays b5, the knight uses another route: via d7 to c5. 17 lbb5 This is the safest for White, ex­ changing off Black's dark-squared bishop as it is likely to become his main attacking piece later. White can stubbornly defend the extra pawn by a somewhat awkward knight manoeuvre from c4 to c2, but in this case he can expect serious prob­ lems: 17 lba3?! (freeing the c4-square B) for the fl-bishop) 17 ... lba6 18 lbc2 9 lDd2 'iVe7 10 a3 (exchanging the light-squared bishop Compared with Line A (9 a3), the by 18 ii.xa6?! bxa6 + is very risky for extra moves lbd2 and ... 'ii'e7 have White - as it often is in this gambit- as been made. This difference means that •••

=

VAGANIAN GAMBIT: 7 e3

White can prevent the ... i.e5 idea that Black used in that line. 10 ....l:.d8 1 1 b4 1 1 i.e2 transposes to Line C 1 1 . l l ... i.d6 12 lLlc4 This is the point of White's move­ order. For 12 i.b2 i.e5, see Line A. 12 i.c7 (D) .••

As we already know, Black should avoid the exchange of his dark-squared bishop unless he gets something really substantial in return. 13 i.b2 Or: a) 13 .l:!.a2?! (an awkward and un­ natural idea) 13 ... i.g4! 14 i.e2 i.xe2 15 .l:r.xe2 lLlc6! and the knight moves to e5 with obvious compensation, Kra­ senkov-Smirin, Vilnius 1988 (note that Black could not develop this knight by 15 ... lLlbd7?? because of 16 d6 +-). b) 13 'iii'c 2?! is not good because it allows Black to regain the pawn un­ der good circumstances right away: 13 ... lLlxd5 ! (this new move improves over the 13 ... i.f5?! played in Bou­ aziz-Othman, Arab Ch, Dubai 1995)

95

and unfortunately for White he can ' 1 take the e4-pawn either way: b1) 14 'i!i'xe4?? surprisingly leads to a quick mate: 14 ... lLlxc3 15 'iixe7? :d1#. b2) 14 lLlxe4? is bad because i t opens attacking lines for the black pieces. After 14 ... i.f5 15 i.d3 lLlc6, intending to take on b4, Black is prob­ ably winning already; e.g., 16 i.d2 lLlb6! (planning to take on d3) 17 f3 'iih4+ 18 �e2 :txd3 ! 19 'i!Vxd3 l:.d8 20 'iic2 i.xe4 21 'ifxe4 (21 fxe4 lLlxc4 22 'iixc4 'iVg4+ -+) 2l...'i!Vxe4 22 fxe4 lLlxc4 -+. b3) 14 i.b2 appears best, but of course White's unintentional generos­ ity can't bring him any dividends. 14 ... lLlxc3 15 'ifxc3 f6 followed by ... lbc6, ... i.e6, etc., gives Black com­ fortable equality. c) 13 'ii'b3 a6! (intending ... b5 to push the c4-knight to a less comfort­ able square; the idea behind White putting his queen on b3 is shown by 13 ...lLlxd5? 14 lLlxd5 :txd5 15 lLlb6! axb6 16 'iixd5 ;!;;) 14 a4 lLlxd5 ! (this is a new move; with pawns on a4 and a6, Black can already sacrifice an ex­ change; 14 ... i.f5?! 15 i.a3 lLlbd7 16 .l:td1 gave White strong pressure in A.Donchenko-Kotainy, German Un­ der-18 Ch, Oberhof 201 1) 15 lLlxd5 :xd5 16 lLlb6 i.xb6 17 'i!Vxd5 lbc6 gives Black enough for the exchange: the uncastled white king, Black's better development and the weaknesses on b4 and d3. Here I shall show just a few lines that demonstrate how Black de­ velops his initiative. 18 i..d2 i..g4!? (D) (preparing .. J:td8) and now:

96

A TTA CK WITH BLACK

13 i.e6 Black is ready to exchange the d5pawn for the e4-pawn to get more open lines. 14 �c2 (D) ••.

w

c 1) 19 i..c4 ?? loses a piece after 19 ... l:td8 20 �xf7+ 'i!Vxf7 21 Ji.xf7+ 'it>xf7 -+. c2) 19 'iVc4 l:tc8 20 'iib 3 l:r.d8 trans­ poses to line 'c3'. c3) 19 'i!Vb3 .l:.d8 20 i.c3 a5 ! 21 bxa5 i.xa5 ! (Black creates mating threats) 22 i.. xa5 lLlxa5 23 �c3 lLlb3 ! (distracting the white queen away from the a1-e5 and el-a5 diagonals) and then: c31) 24 'ii'xb3?? 'i!Ve5 25 .l:.c 1 'iia5+ -+. c32) 24 llb1 'iid7! 25 f3 exf3 and now 26 .:txb3?? loses to 26 .. ."Vi'd1+ 27 'it>f2 .:td2+ 28 'it>g3 fxg2 29 i.xg2 l:!.xg2+ 30 �xg2 'ti'e2+ 3 1 'it>gl (31 'it>g3 �f3+ 32 'i2th4 i.e6 -+) 31...i.h3 -+, while 26 gxf3 lLld2 affords Black a strong attack. c33) 24 i.e2 i.xe2 25 'iixb3 i.d3 followed by ... 'i!Vg5 gives Black suffi­ cient compensation. c4) 19 i..e2!? lLlxb4 20 �b3 i.c5 ! 21 llcl (21 i..xg4?! tLld3+ 22 'it>e2 'i!Vh4 23 'ii'xb7 .l:!.d8! 24 i.el 'i!Vxg4+ 25 'it>fl �f5 +) 21...b6 and again Black has enough compensation.

14 tLlxd5! •.•

This new move improves over the previously played 14 ... i.xd5?! 15 �dl lLlc6 16 lLlxd5 l:.xd5 17 l:txd5 lLlxd5 18 i.e2 ;l; Bartels-Mainka, Luxem­ bourg 1989. Now White can take the e4-pawn in two ways but neither leads to an ad­ vantage: 15 lLlxe4 15 'iixe4 lLlc6 leaves Black far ahead in development, and his com­ pensation is rather convincing; e.g., 16 i.e2 llac8 17 0-0 lLlxc3 18 i.. x c3 i.xc4! 19 'i!Vxc4 (19 'ii'xe7? loses one of the bishops: 19 ... lLlxe7 20 i.xc4 Ji.d6 -+) 19 ... lLld4! 20 'i!Vxd4 (20 i..xd4? i.. xh2+ 21 'it>xh2 l:lxc4 22 i..xc4 'ifc7+ 23 'it>g1 'ii'x c4 is some­ what better for Black) 20 ... l:lxd4 21 i..xd4 b6 =.

15 lLld7 •.•

VAGANIAN GAMBIT: 7 e3

White hasn't castled yet, and this provides Black enough time to secure good compensation. The most natural plan is ....l:tac8, threatening ... b5 fol­ lowed by a move by the c7-bishop. Unfortunately for White, there is no quick way to get castled. Let's see a few example lines: a) 16 i..d3?! .lir.ac8 (intending ... b5) and then: a1) 17 'i!ke2? offers Black a choice of good replies: 17 ... b5 ! ( 17 ...lbxb4 18 axb4 i..xc4 19 i.. xc4 'i!Vxe4 + is also good) 1 8 lbcd2 (or 18 lba5 lbe5 19 0-0 il.g4 20 f3 lbxd3 21 'i!ixd3 lbxe3 ! -+) 18 ... ..te5 ! 19 lbf3 lbxb4! -+. a2) 17 .l:icl b5 18 lbcd2 i..e5 19 'ili'b1 (19 'i!Vxc8 i..xb2 20 'i!Vc2 i..xa3! 21 .:tal li:Jxb4 22 'i!Vb1 li:Jxd3+ 23 'ii'xd3 i..b4 24 0-0 f5 ! 25 'i!Vxb5 fxe4 26 li:Jxe4 a5 +) 19 ... ..txb2 and here: a21) 20 'itxb2 l:txcl + 21 'ii'xcl f5 ! (a critical idea for Black: the pawn ad­ vances to f4 with tempo) 22 li:Jg3 (22 lbc5 lbxc5 23 bxc5 li:Jxe3 +) 22 ... f4 with an attack. a22) 20 .lir.xc8 .l:.xc8 21 'ii'xb2 f5! 22 li:Jg3 f4 gives Black an obvious ini­ tiative. b) 16 .l:.dl ?! l:tac8 17 'i!Vbl b5 18 li:Jcd2 f5 19 li:Jc3 li:Jxc3 20 i..xc3 i..b6 + followed by ...f4 again brings Black a strong initiative. c) 16 l::tc l l::tac8 17 'i!Vbl b5 18 lbcd2 (18 li:Ja5? il.xa5 19 bxa5 i..f5 20 .lir.xc8 .lir.xc8 21 il.d3 l:te8 + followed by ...li:Jxe3) 18 . .f5 19 lbc5 li:Jxc5 20 l:txc5 (the only move, since 20 bxc5? il.a5 -+ leaves the d2-knight in big trouble) 20 ...f4! + with strong pressure.

97

C) 9 ..te2 'ii/e7 (D)

Here White has a choice of three main continuations: 10 a3, 10 li:Jd2 and 10 0-0. There are mostly not very serious differences between them and often they lead to the same positions. We take two of them as main lines: C 1 : 10 li:Jd2 98 C2: 10 a3 101 Or 10 0-0 .l:r.d8: a) 1 1 li:Jd2 transposes to note 'a' to White's l ith move in Line Cl. b) 11 'ili'c2 il.f5 12 l:td1 (for 12 li:Jd2, see note 'b' to White's 12th move in Line C12) 12 ... li:Jbd7 (12 ... li:Ja6!? also deserves attention since from here the knight can head to either b4 or c7 to take the d5-pawn; the text-move is more centralizing, which is why I recommend it) 13 li:Jd2 and now we have reached note 'b2' to White's 12th move in Line C 12. c) 1 1 a3 is the only independent continuation. Since there is no threat to the e4-pawn, Black can postpone

IJH

A TTA CK WITH BlA CK

. . . lLf5 e5

in order to bring the knight to more quickly. 1 1 ...lt::lbd7 12 b4 1Ld6 1 3 lLb2 and then: cl) 13 ...lt::lb6?! is unconvincing be­ cause it gives White an opportunity to remove Black's dark-squared bishop by 14 lt::lb5 ! (rather than 14 lt::ld2, Oatlhotse-Ezat, African Ch, Windhoek 2007), securing good prospects for White. c2) After 1 3 ... .ie5 ! 14 'il:Vb3 lt::lb6 15 l:!d 1 White has defended the pawn three times but because the white queen is overloaded, Black can never­ theless play 1 5 ... lt::lbxd5 ! 16 lt::l xd5 lt::lxd5, when the position is equal; e.g., 17 lt::ld2 .ixb2 1 8 'i!Vxb2 .i.d7! (an important move: if the white knight comes to d4, White can hope for some advantage; that's why Black brings his bishop to a4 to eliminate the knight) 19 lt::lb 3 .i.a4 =.

Cl} 1 0 lt::ld2

White first attacks the enemy e4pawn. to :ds Now we have two main sections: C11: 1 1 a3 99 100 C12: 1 1 'ili'c2 Or: a) I I 0-0 lt::la6 (Black will capture on d5 without any complications; 1 l ....if5?! 12 lt::lb 3 ! {Altini-Matamo­ ros, Forni di Sopra 201 1 } followed by lt::ld4 looks too good for White; at least, I could not find any smart ideas there) 12 lt::lb 3 .ib6 followed by ... lt::lc7/b4xd5; Black is OK. ...

=

b) With 1 1 lt::lb3?!, the knight moves towards the d4-square, but takes the pressure off the e4-pawn. Black can get the advantage by simply taking on d5. Now I prefer l l.. . .ib6 over the two other options (l l ...lLd6 and 1 1 ....ib4) because the d5-pawn remains attacked by the d8-rook, and from b6, the bishop controls d4 and e3 and can switch to the b8-h2 diagonal. After 12 0-0 lt::lc6! the knight is ready to move to either b4 or e5. Then: b1) 1 3 'it>hl ?! (this strange move does not help White at all) 13 ... lt::lb4 14 .i.c4 lt::lbxd5 15 .ixd5 lt::lxd5 ! (the alternative 1 5 ... .ie6?! + gave Black less in Luczak-Tomaszewski, Porabka 1986) 16 lt::l xd5 'ilig5 +. b2) 13 a4!? plans to push the b6bishop back, but after 1 3 ... lt::lb4 14 aS .i.e? 15 lt::ld4 lt::lbxd5 + Black is simply better. c) 1 1 'ili'b3 lt::la6 and here: c1) 12 a3?! lt::lc7 1 3 'ili'c2?! .i.f5 14 b4 .id6 1 5 lt::lc4 lt::lcxd5 16 lt::lxd6 l:txd6 17 .ib2?! l:tc8 1 8 lt::lxd5 lt::lxd5 19 'i!Va4 occurred in Stepak-Birnboim, Tel Aviv 1 988. Black has played very well so far, but now gave his opponent a last-minute reprieve by 19 ... lt::lb 6?. Instead, 1 9 . �g5 ! 20 0-0 (20 �fl lt::lxe3+! 21 fxe3 .:ld2 22 l:.c 1 l:.cd8 -+) 20 ... .i.h3 -+ is strong. c2) Black can meet the untried 12 0-0 with 1 2 ...lt::lc7 1 3 .l:!.d1 b6!?, in­ tending ... .i.b7 and ... lt::lcxd5 (and not 1 3 ... lt::lcxd5? 14 lt::ldxe4! ±). Usually, Black does not develop his pieces like this, but since the white queen is block­ ing its own pawns, it works well here; e.g., 14 'ifa4 (attacking e4) 14 ... .i.d7 15 . .

VAGANIAN GAMBIT: 7 e3

99

'i!Vc4 b5! 16 'iVb3 .tf5 17 ltJxb5 lLlcxd5 the balance, though you'd probably and Black obtains good compensation. need to be a computer to find them over the board: 14 fxe3 'ifxe3 15 .l:!.fl Cl l) lLlc6 16 'i!Vb3 (16 Vi'a4? ltJe5 ! 17 �d I 1 1 a3 (D) .te6 + Postny-Smirin, Israeli Team Ch 1999) 16 ... ..te6 17 lLlc4 'ti'd4 18 ltJb5 li'c5 (18 ...�h4+ 19 Vi'g3 'ti'xg3+ 20 hxg3 .txc4 21 .txc4 �d7 ! 22 l:txf7 ! l:!.xf7 23 ltJd6 lLle5 24 .ta2 ) 19 .te3 ! B .txc4 20 .txc4 'i!Vxe3+ 21 li'xe3 .txe3 22 .txf7 + 'iit>h8 b) The less dramatic 13 ... ltJxc3 ! 14 bxc3 i.f5 gives Black very strong play for the pawn. 12 Jtxd5 13 b4 1 3 'i!Va4?! is most energetically met with 13 ... ltJd7! followed by ... lLlf6, ... .tf5, etc. Then White can't take on e4: 14 lLlxe4? (also not 14 'i!Vxe4? This move provokes an opening of :xd2 15 Wixe7 Itxe2+ 16 'iit>xe2 i.xe7 the position, which in principle fa­ +) 14 ... lLlb6 15 'i!Vc2 i.f5 16 f3 .l:.c8 +. vours the better-developed player. Instead, 13 ... ..tf5?! proved less effec­ ll ... ttJxdS 12 ltJxdS tive in Vakhidov-Nesterov, Bishkek 12 lLlcxe4 attempts to justify White's Zonal 1993. play by exchanging the d5-pawn for 13 .tb6 14 .tb2 ltJc6 15 li'b3 the one on e4. However, it does not Another queen move, 15 'ifc2, trans­ lead to desirable simplifications. After poses after 15 ... .tf5 to a well-known 12 ... .tb6! Black develops his pieces game by the young Kasparov. After 16 naturally and strongly - the knight 0-0, he chose 16 ...'ti'g5?! (16 ... l:!.c8 17 goes to c6, the queen's bishop can be ltJc4 .tc7 1 8 l:!.ad 1 llcd8 is better, with brought to e6 or f5 and g6, while the equality, while 16 ... ..tc7 and 16 ... lLle5 queen's rook will move to c8. Then can also be tried), which from a hu­ White's pieces will be even more un­ man perspective looks very natural, as comfortable. For instance, after 13 0-0 the black pieces move closer to the lLlc6 it's not clear what White should white king. But in the cold light of day, do, and Black's game feels good. And and under computer scrutiny, this plan 1 3 ltJc3?! is extremely risky: does not look so scary: a) 17 'iit>h 1?! .l:!.d6 (17 ... .l:tc8!?) 18 a) 13 ... ltJxe3 !? has been success­ fully tried in practice, but may not be lLlxe4 i.xe4? (Black should prefer objectively best because White has 1 8 ... 'i!kg6 19 f3 .txe3 with an unclear enough defensive resources to hold position) 19 li'xe4 l:.d2 and i n =

=.

••

.•.

/ ()(}

A 1TACK WITH BLACK

Mikhalchishin-Kasparov, USSR Ch, Frunze 198 1 White played 20 b5? and Black later won. Both players evi­ dently missed 20 �a6! bxa6 2 1 �xc6 .l:.ad8 22 �c3 ;!;, when White has an extra pawn with good chances to real­ ize it. b) 17 �c4!? even helps Black com­ plete his plan. After 17 ...l:ld6 18 l:lfd1 (continuing to ignore Black!) 18 ... .l:tg6 19 g3 we see the point. Black spent three tempi to create a mating threat, but White just played g3, and it's not easy to see how Black should proceed. 19 ...l:!.e8 20 �b3 'i!fe7 21 �e2 ;!; leaves White's pieces more harmonious. We now return to 15 'ifb3 (D) :

15 ... �e6!

This is a new move. 15 ... l:.g5 (as played in Panno-Glavina Rossi, Ar­ gentine Ch 1 989) is also interesting, but there is a strong case for develop­ ing the queenside pieces without de­ lay. 16 lDxe4 :tc8! Now the white queen has serious problems:

a) 17 �c4 lDxb4! and then: a1) 18 axb4? l:.xc4 19 �xc4 litd1+ 20 .l:.xd 1 i..xc4 21 ltJd6 �a6 +. a2) 1 8 J.. xd5? ltJxd5 (intending ...ltJxe3) 19 1i'd1 'i'h4! 20 ltJg3 'ii'c4! (planning ... J..a5+) 21 lDe2 .l:.d8 -+ followed by ...lDf4. a3) 1 8 0-0 .l::i.d3 ! 19 'ii'xb4 'ii'xb4 20 axb4 l:txc4 21 lDc5 �xc5 22 bxc5 a6 23 J..d4 f6 with a slightly better end­ game for Black. b) 17 l:tcl :d2! 1 8 lDxd2 (18 J..c4 ltJa5 ! 19 bxa5 .l:.xb2! 20 'ii'xb2 J..xa5+ + and ... J.. xc4) 1 8 ... �xb3 19 lDxb3 litd8 +. C 1 2) 1 1 'ii'c2

�f5 12 a3 White does not rush to castle king­ side since the king cannot expect any real safety on that wing. Other possi­ bilities: a) 12 g4 leads to another opening of the position that Black definitely likes. 12 ... lDxg4 13 �xg4 �xg4 14 'i!Vxe4 (avoiding the queen exchange would favour Black since the white king is still in the centre: 14 ltJdxe4? �b4 15 J..d2 ltJd7 + followed by ... lDe5, etc.) 14 ...�xe4 15 lDdxe4 (af­ ter 15 ltJcxe4 J..b4 Black will soon capture on d5 with a better endgame) 15 ... J..b4! (a new move; 15 ...lDa6?! is less aggressive, though it's enough for equal chances, Morovic-Liew Chee Meng, Dubai Olympiad 1986) 16 l:tg1 �f3 17 �d2 ltJd7. Black is a pawn down but his pieces are much better organized and his chances are better in this endgame. b) 12 0-0 ltJbd7 (D) and then:

VAGANIAN GAMBIT: 7 e3

/ (J /

is more important. After 13 li::l xd5 llxd5 14 b4 �b6 15 g4 �g6 16 h4 h6 17 �c4!? (this new move improves over the 17 h5?! played in Karasev­ Montero Martinez, World Seniors Ch, Rijeka 201 1) 17 ... .l:.d8 18 �b2 White has escaped from the pressure and the game is unclear.

13 b4 �d6 14 �b2 .l:.ac8 15 'ii:Vb3 li::le 5 ! ?

Black should generally preserve his dark-squared bishop since it's a key piece in his kingside attack. 15 ... li::lb6 16 li::lb5!? (a new move improving over 16 0-0?!, Goldsztejn-Degraeve, La Fere 2008) 16 ... a6 (16 ... �b8?! 17 d6 �xd6 18 li::lxa7 .:!.a8 19 li::lb5 ;!;) 17 li::lxd6 'ifxd6 18 0-0 li::lbxd5 19 .l:.fd 1 is unclear. 16 0-0 li::ld3! Now: a) 17 �xd3? �xh2+! is a typical idea - the bishop has been distracted from e2 by ... li::ld 3, and ... �xh2+ wins. 18 'it>xh2?? loses at once to 18 ... li::lg4+ 19 'it>g1 (19 'it>g3 'ti'e5+ -+) 19 . .'ir'h4 20 l:.fc1 'i¥xf2+ 21 'it>h1 .l:r.d6 -+, while 18 'it>h1 �e5 ! (18 ... exd3?? 19 'it>xh2 li::lg4+ 20 'it>g 1 'ti'h4 21 li::lf3 +-) 19 �c4 (19 �e2? li::lxd5! -+ threatens both ...li::lxc3 and . .'ii' h4+) 19 ... li::lg4 20 g3 'ir'g5 gives Black a decisive at­ tack. b) 17 li::lb5 li::lxb2 18 li::lxd6 'i!Vxd6 19 'i!Vxb2 'i!Vxd5

b1) 13 li::lc4 liac8 14 lld1 li::le5 ! (an instructive moment: Black gets the d6and e5-squares for his bishop and queen) 15 li::lxe5 'ti'xe5 16 �d2 li::lxd5 ! + (16 ... �g4?!, as in Servat-A.Hoff­ man, Argentine Ch 1995, is a very 'human' move but there is no need for creative measures when it's possible simply to win the pawn back while keeping all Black's positional advan­ tages). The trick is that White can't take on d5: 17 li::lxd5?? �d6 -+. b2) 13 l:.d1 l:!.ac8 14 li::lf l �d6!? (better than 14 ... li::le5?!, as in Cher­ nin-Hebden, London 1989, because Black needs to keep the b8-h2 diago­ nal open so he can line up his bishop and queen on it) 15 li::lg 3 �g6 fol­ lowed by ... h5-h4, etc. White's pieces are cramped and can hardly expect any harmony in the near future. 12 li::lbd7 ! ? The rook will move to c8 and the knight heads for e5. The d5-pawn will C2) most likely to be taken later anyway. 10 a3 12 ... li::lxd5?! is less effective. In my This is the most popular move. opinion, there is no need to take the White aims for a quick b4 followed pawn at once, and rapid development by developing his queenside pieces. •••

=.

AITA CK WITH BLACK

1 02

10 :1.d8 •••

My feeling is that Black still should not worry about the b4 advance be­ cause it gives Black possibilities of ... a5 with play on the queenside. 1 1 b4 For 1 1 lt:Jd2, see Line C 1 1 . 1 1 0-0 transposes to note 'c' to White's lOth move in Line C. l l ...i.. d6 (D)

From here, the bishop keeps an eye on two pawns: b4 and h2. 12 i.. b2 White has two main ideas, both connected with exchanges: 1) To swap Black's dark-squared bishop and remain with an unopposed bishop on b2; 2) To exchange all the minor pieces except for a white knight that will come to d4, and Black's light-squared bishop. Many of the following variations will make more sense if you keep these two ideas in mind. White can also defend the d5-pawn with the rook, but it takes a lot of time

and Black obtains good play without real problems. 12 l:!.a2 lt:Jbd7 13 l:td2 lt:Je5 (Black's main idea is ... i..g4) and now: a) After the manoeuvre .l:r.a2-d2, White gains little from exchanging the d6-bishop: 14 ti:Jb5 i..g4! and then: a1) 15 lt:Jxd6? is not good because of the zwischenzug 15 ...i..xe2 16 'ili'xe2 (16 ti:Jf5 'i¥d7 17 lt:Jh6+ gxh6 18 'i¥xe2 'ii'f5 ! 19 0-0 ti:Jf3+! +) 16 ...:1.xd6 17 0-0 lt:Jxd5 + followed by ... lt:Jd3. a2) 15 0-0 i.. xe2 16 'iNxe2 ti:Jd3 17 ltJ1c3 (17 lt:Jxd6 'i¥xd6 18 i..b2 lt:Jxd5 19 f3 f5 leaves Black OK) 17 ... .litac8 18 l:tc2 (18 lt:Jxd6? lixc3 19 i..b2 lt:Jxb2 20 ti:Jb5 lt:Ja4 +) 18 ... i.b8 with obvious compensation for Black. b) The cold-blooded 14 0-0 is also possible, but then White must defend against an attack on his king after 14 ... i.g4!. Incidentally, 14 ...lt:Jf3+!? is also quite interesting; e.g., 15 i.xf3 (15 gxf3?? loses immediately to 15 ... exf3 16 i.xf3 'ili'e5 17 lie 1 i.h3 ! -+) 15 ... exf3 16 gxf3 (16 'ili'xf3?? i.. g4 -+ traps the queen) 16 ... i..h3 17 .l:!.el . Black has no direct ways to break through to the white king at the mo­ ment but its cover is so weak that I'm pretty sure a new wave of attack will come at some point. Right away even 17 ... a5 can be played, with reasonable compensation. c) 14 h3 (in order to prevent .. . ..ltg4) 14 ... i.f5 15 .litd4 (15 0-0 .l:!.ac8 16 i.b2 i.b8 gives Black compensation) and here Black must choose between tak­ ing an exchange for two pawns or keeping his positional advantages and developing his initiative:

VAGANIAN GAMBIT: 7 e3

cl) Playing to win an exchange by 15 ... liJd3+ 16 .i.xd3 exd3 is not so good in my view. Then 17 0-0 .i.e5 1 8 l:txd3 .i.xd3 19 'i!Vxd3 was at best unclear for Black in Peng Zhaoqin­ V.Mikhalevski, Dieren 1998. c2) 15 .. Jbc8!? 16 0-0 i..b 8 gives Black pleasant compensation . ... liJd3 can be played at practically any mo­ ment, and in addition Black has the idea of ... a6 (preventing lLlb5) and ... 'iVd6 with threats to the white king. 12 .i.e5 12 ... a5?! appears to lead to prob­ lems after 1 3 bxa5 .i.e5 (Timman­ Topalov, Wijk aan Zee 1996) 14 lLld2! (a new move), when White's main aim is to exchange off the e5-bishop: a) 14 ... .i.xc3 15 .i.xc3 lLlxd5 16 .i.b2 l:txa5 17 0-0 lLlc6 18 l:.cl t. b) 14 .. Jlxa5 15 �b3 .i.xc3 (or 15 ...liJxd5 16 lLlcxe4 t) 16 .i.xc3 .l:taxd5 17 l:dl t and 0-0. c) 14 ....!bxd5 15 lLlxd5 lhd5 16 .i.xe5 'iVxe5 17 :I.e 1 i..e6 1 8 'ii'c2 lLld7 19 0-0 l:tdxa5 20 'itxe4 'ii'xe4 2l lLlxe4 l:.xa3 22 l:[fd1 gives White an obvious edge in the endgame. 13 �b3 lLlbd7 14 lLld2 14 0-0 transposes to note 'c2' to White's lOth move in Line C.

](}3

•••

14 ...liJb6 (D) 15 .l:!.d1

Black has strong play after 15 lLlc4 lLlxc4 1 6 .i.xc4 a6!? (a new move; 16 ... .i.f5 lets White execute his plan of exchanging dark-squared bishops by 1 7 lLlb5, as in Khodos-Gofshtein, Daugavpils 1978) 17 a4 (17 lLle2?? b5 -+) 17 ... .i.f5 1 8 lLle2 ( 1 8 l:tcl l:!.ac8 19 h3 .i.d6 20 .i.a3 liJd7! + with

... lLle5 to follow) 1 8 ... i.. xb2 19 �xb2 l:tac8 ! and then: a) 20 l:tcl ?! lLlxd5 21 .i.xd5 l:hc1 + 22 'ii'xc1 (after 22 lLlxc1 ?! lixd5 23 0-0 'ili'd6! + White cannot bring his knight to d4) 22 ...�xb4+ 23 'ii'd2 'ii'xd2+ 24 'it>xd2 .l:.xd5+ 25 liJd4 'iii>f8 +. b) 20 i.a2 .i.g4! 21 liJd4 lLlxd5 22 .i.xd5 :xd5 23 0-0 'iVh4 +. Although White has achieved his aim (d4-knight vs light-squared bishop) Black has a powerful attack on the white king. 15 ... .i.f5 16 lLlc4 lLlxc4 17 �xc4 .l:l.ac8 18 'ii'b3 i.. g4! ? Black prepares for the endgame by exchanging his worst piece. 1 8 ... a6?! 19 lLla4 .i.xb2 20 'iVxb2 .l:txd5 21 l:!.xd5 lLlxd5 22 0-0 t Lau­ tier-Illescas, Linares 1995. Black has some problems with his light-squared bishop, which tends to be bad in this type of endgame. 19 h3 i.. xe2 20 lLlxe2 After 20 'it>xe2? 'ii'd7! + Black plans to take on c3 and check from b5.

20 ... i.. xb2 21 'ii'xb2 lLlxdS 22 0-0 lLlb6

The game is totally equal.

9 Ben ko Ga m bit Decl i ned

1 d4 tt:\f6 2 c4 c5 3 d5 b5 (D)

w

Finally, we have reached the main topic of this book - the Benko Gambit. Before we start analysing the varia­ tions, let me make some brief historical notes. The first article about 3 ... b5 was published by Argunov, a chessmaster from Kuibyshev (which is nowadays called Samara) in the magazine Shakh­ maty SSSR in 1946. Since this city is on the Volga River, the opening was named the Volga Gambit in Russian­ language literature. Some time later, the Hungarian/American GM Pal Benko made a huge contribution to the development of this gambit line, in particular by showing that Black can rely on long-term positional compen­ sation rather than quickly opening the centre with ... e6. He also published an v

influential book, The Benko Gambit, in 1974. The opening then quickly found followers, both at club level and among grandmasters. White's at­ tempts to refute the gambit kept run­ ning into brick walls, while the ease and elegance with which Black's ideas worked in practice, without White ap­ pearing to do anything wrong, made a strong impression on the chess world. As White started to lower his ambi­ tions, and merely seek a workable edge, the Benko began to acquire a more normal body of opening theory, but still one with more thm. ;ts fair share of spectacular and dynamic vari­ ations. Although the Benko Gambit is not so common at present in games be­ tween the world elite, many of the top players have this opening in their ar­ senal, including Carlsen, Ponomariov, Ivanchuk, Vachier-Lagrave and Caru­ ana. The main idea of the gambit is to create an initiative on the queenside. By sacrificing a pawn, Black opens the a- and b-files for his rooks and queen. At the same time, the g7bishop takes control of the al-h8 diag­ onal. There are also pleasant prospects for the black knights: there are so many good squares for them on the queenside that it is hard for White to

BENKO GAMBIT DECUNED

cover them all: a4, c4, d3 (after White plays e4 or e3), b5 and d4. Besides piece-play, very often Black blows apart the centre with ... e6 or ... f5, in­ tending to advance his centre pawns or create a weakness on d5. It's not easy for White to counter all Black's manoeuvres, which is why many players prefer not to accept this sacrifice. But acceptance is certainly critical, and if White adopts a logical plan, Black needs to be very accurate if he is to retain a strong enough initia­ tive to compensate for the pawn. On move 4 White has a large choice of continuations if he does not want to take on b5. We shall analyse the fol­ lowing moves: 105 A : 4 b3 106 B : 4 f3 107 C: 4 a4 109 D: 4 � g5 111 E: 4 'ii'c2 F : 4 lt:Jd2 1 14 1 17 G : 4 lt:Jf3 A) 4 b3

This is one of the less ambitious moves against the Benko Gambit. White spends time defending the c4pawn but does very little for his own development, since b2 will not be a safe or effective post for the bishop with an open b-file. There are better and more natural ways to defend the pawn, especially considering that the structure after an exchange on c4 isn't ideal for White, as the c4-pawn will be somewhat vulnerable in the long term. 4 ... e6! ?

1 05

Black uses the respite granted by the slow move b3 to seize the centre. This seems very logical to me. 5 dxe6 fxe6 (D)

w

6 lt:Jf3

In the case of 6 cxb5 a6! White is unable to defend b5 normally. 7 bxa6 �xa6 gives Black a pleasant merged version of the Benko and Blumenfeld Gambits; e.g., 8 lt:Jf3 lt:Jc6 and now: a) 9 �b2 �e7 l 0 g3 'ilia5+ and now 1 1 'ii'd2 1i'xd2+! 12 lt:Jfxd2 ( 1 2 lt:Jbxd2 lt:Jb4 +) 1 2... lt:Jd4! 13 �g2 d5 + or 1 1 lt:Jbd2 0-0 12 �g2 c4! 13 bxc4 .:ab8 14 'ifc2 �xc4 15 1i'xc4 l:txb2 16 0-0 l:.fb8 with at least enough com­ pensation. b) 9 g3 c4! exposes White's weak dark squares on the queenside. After 10 bxc4 �b4+ 1 1 �d2 �xc4 12 �g2 lt:Jd5, intending 13 ... 'iVf6 1 4 jhb4 lt:Jcxb4, Black has a strong initiative. 6 bxc4 This new move improves over 6 ...d5, which was played in M.Jovanovich­ R.Garcia, Argentine Ch, Santa Fe 1973. .••

1 06

A ITACK WITH BLACK

i.e7 Now: a) White can stop ... d5 by playing 9 e4 but it weakens his dark squares in the centre. After 9 ...l:tb8 10 i.e2 0-0 1 1 0-0 lbg4 + intending ..."fic7 or . . i.f6, Black is better. b) 9 e3 0-0 10 i.e2 d5 ! (if White does not prevent ... d5, just play it!) 1 1 cxd5 exd5 12 0-0 llb8 +. c) 9 g3 0-0 10 i.g2 .l:.b8 1 1 0-0 l:tb4 (the c4-pawn is undefended and Black exploits this fact to sow disorder in White's camp) 12 lDd2 .l:!.b6!? (free­ ing the b4-square for the c6-knight and avoiding the tempo-gaining i.a3; the idea is to play ... d5 - the immedi­ ate 12 ... d5?! is poor due to 13 i.a3 l:!.b8 14 cxd5 exd5 15 lbb3 c4 16 lbd4! ;!;) 13 i.a3 1li'e8!? leaves Black with no problems. The queen is headed for h5, when ... lbg4 becomes an idea. 7 bxc4 lbc6 8 lbc3

B)

4 f3

This move looks odd because the pawn takes a natural square away from the king's knight and gets little benefit in return - Black is not made to pay any real 'price' for playing the aggres­ sive ... b5. There are more harmonious ways to support the e4 advance, such as 4 'ii'c2 or 4 lbd2. If White wants to play f3 and e4, then a better option is 4 cxb5 a6 5 f3 (see Line B of Chapter 10), when the move e4 will come with greater force. 4 ... bxc4 Black can also play 4 ...'ii'a5 + 5 i.d2 b4; e.g., 6 e4 d6 7 lba3 g6 8 lbc2 'ii'b6 9 i.d3 lbbd7 10 f4 i.g7 1 1 lbf3 0-0 12

0-0 i.b7 intending ... e6 with unclear play. 5 e4 d6 Black simply develops his pieces to natural squares. 5 ... e6!? is a worthy al­ ternative . 6 i.xc4 g6 (D)

White needs to decide where his king's knight will be developed. 7 lbe2 This knight heads for c3, while the queen's knight will move to a3 to sup­ port c4. Another option is to play f4 and lDf3, but this costs more time. 7 lbc3 i.g7 8 f4 0-0 9 lDf3 i.a6!? 10 i.xa6 lbxa6 1 1 'ii'e2 1li'a5 12 0-0 lbc7 13 .l::td 1 .l:!.fd8! (preventing the e5 advance) 14 i.d2 'ii'a6! 15 'ii'e 1 .l::tab8 gives Black somewhat the better chances, L.Da­ vis-D.Gurevich, Palo Alto 198 1 . 7 ...i. g7 8 0-0 0-0 9 lbec3 9 lbbc3?! is wrong for two reasons: it fails to cover the c4-square, and it leaves the e2-knight with no pros­ pects. After 9 ...lbbd7 it's already not so clear what White should do. 10 l:[b1

BENKO GAMBIT DECLINED

lt'le5 1 1 .i.b3 .i.a6 + was Dalmau Comas-Lanka, Badalona 1994, while 10 lt'lg3 (Sieciechowicz-Tokarski, Pol­ anica Zdroj 2008) fails to improve the knight's prospects of finding a useful role. Black can reply 10 .. J::tb 8 + fol­ lowed by . .'ifa5, ... .i.a6, ... lt'le5, etc. 9 lt'lfd7 Black can also play the more regu­ lar 9 ... lt'lbd7 (Liang Chong-Zhou Jian­ chao, Beijing Zonal 2005) but the text-move gives him more options: the queen's knight can be developed via a6 to b4 (if White plays a4) or to c7; also ... .i.a6 is still an idea. 10 a4 10 lt'la3 lt'la6 1 1 .i.e2 .l::!.b 8 12 lt'lc4 lt'lc7 followed by ... .i.a6 is fine for Black. 10 lt'la6! Now 1 1 lL'la3 lt'lb4 12 .i.g5 lt'le5 13 .i.e2 f5 !? gave Black an active posi­ tion in Segura Ariza-Minzer, Palma de Mallorca 2002.

107

while White will defend his queenside and try to make progress in the centre. w

•.•

=

•.•

C) 4 a4

This move was popularized by So­ sonko. White spends a move to force a resolution of the queenside tension. 4 ...b4 (D) The most solid move. Black's play is simple and guarantees equality. An alternative is 4 ... bxc4; e.g., 5 lt'lc3 d6 6 e4 g6 (if Black defends c4 with 6 ....i.a6?!, then White grabs the centre by 7 f4 lt'lbd7 8 lt'lf3, with pleas­ ant prospects) 7 .i.xc4 !ILg7 8 lt'lf3 (Wojtaszek-Jianu, World Junior Ch, Kochin 2004) with a typical Benko po­ sition. Black will play on the b-file,

5 lt'ld2 Intending a rapid e4 advance. In­ stead, 5 b3 g6 6 JL.b2 JL.g7 7 lL'ld2 0-0 8 e4 d6 transposes to the next note. White can also start with a double fianchetto: 5 g3 d6 6 .i.g2 g6 7 b3 (if White does not develop his bishop to b2, then it can become a problem piece; after 7 lt'lf3 .i.g7 8 0-0 0-0 it is too late to play b3 because of the ...lt'Je4-c3 manoeuvre, while !1Lf4 or !ILg5 will be met with ...lt'lh5 or ...lt'Je4) 7 ... !1Lg7 8 JL.b2 0-0 9 lt'ld2 JL.b7 (it's important to attack the centre since otherwise White will gradually de­ velop his pieces and increase his cen­ tral preponderance) 1 0 lt'lgf3 e6! 1 1 dxe6 (forced since after 1 1 e4? exd5 the white king is trapped in the cen­ tre: 12 cxd5 JL.a6! + or 12 exd5 .l:.e8+ +) 1 l ...fxe6 12 0-0 file7 Zsinka­ K.Rovid, Hungarian Team Ch 2005/6. The b8-knight can be developed to ei­ ther d7 or c6, and Black has solved all his opening problems. =

1 08

A TTA CK WITH BLA CK

tiJgf3 Or 7 b3 .i.g7 8 .i.b2 0-0, and now: a) 9 g3?! is an unfortunate moveorder since Black immediately blows apart the centre by 9 ... e6!, when it's al­ ready not easy to give good advice to White: al) 10 i.g2? is a tactical miscalcu­ lation that almost finishes the game on the spot. 10 ...exd5 1 1 cxd5 .l:.e8 12 tDe2 tDxe4! 13 i.xg7 tDxd2 14 .i.h6 i.g4! 15 i.e3 tiJf3+ 16 .i.xf3 i.xf3 -+ Mileika-Tal, Latvian Ch, Riga 1953 and Surjadnji-Ponomariov, Swidnica rapid 1998. a2) 10 dxe6 i.b7 ! (10 ... fxe6? 1 1 e5 !) 1 1 i.g2 (after I I exf7+? l:txf7 12 i.g2 'fie7! + the threats against e4 are mounting) 1 l...fxe6 12 tDe2 'fie? 13 0-0 e5 + followed by ...tDc6-d4. b) After 9 i.d3, I feel 9 ... e5 ! is the most comfortable way to equalize. Then: b1) 10 dxe6? is a typical mistake due to an idea that everyone should remember: after 10 ... fxe6! + Black's plan is ... tDc6, ... e5 ! and ... lDd4 with the better game. b2) 10 lDgf3?! is a poor choice of square because of 10 ... tDh5. White can prevent ...tDf4 by playing g3, but this invites ... i.h3; both are equally unpleasant. b3) 10 tDe2 prevents ...tDf4 ideas and supports the f4 advance. After 10 ... tDh5 1 1 0-0, 1 l ...a5 was played in Kanakaris-Mastrovasilis, Thessaloniki 2001, but Black should take the initia­ tive on the kingside by 1 l ...i..g4! (a new move, renewing ideas of ... tDf4). Then: 5 g6 6 e4 d6 7 •••

b3 1) 12 h3?! lDf4! (this is the point of 1 l ...i..g4) 13 hxg4 tDxd3 14 .l:.b1 lDxb2 15 .l:.xb2 �h4 16 tDc1 tiJd7 + followed by ... tDf6 and ... h5 gives Black a strong position on the king­ side. b32) 12 f3 pushes the bishop back, but the calm 12 ... i.. c 8 + leaves the f3-pawn looking ugly. Black's main idea is play on the kingside with moves like ... a5, ... f5, ... l:ta7-f7, etc. Meanwhile, White can't do anything active. b33) 12 'fic2 'figS gives Black some pressure on the kingside. 7 i. g7 8 i.d3 0-0 9 0-0 e5 ! (D) •••

w

Again I recommend the ... e5 idea. This position has not been played very often, so there is no detailed theory just general thoughts. Black has closed the queenside and centre and now has a free hand on the kingside. The main plan is naturally ... tDh5 and ... f5. The queen's rook can be transported to f7 via a7 (after ... a5). The knight can remain on b8 for some time, or it can move to b6 to attack the

BENKO GAMBIT DECLINED

a4-pawn (if the white knight is on b3, this can constitute unpleasant pres­ sure). ... h6 and ... g5-g4 is also a possi­ ble idea. White's plan tends to be .l:.el, i.fl (anticipating ... tt::lf4), tt::lb 3, i.e3, 'ili'd2, etc. Objectively the position is probably equal. Let's see some lines. 10 l:te1 In my opinion, 10 dxe6?! is again a strategic mistake. Then: a) 10 ... fxe6?! (intending ... e5) does not work here because of the typical central break 1 1 e5 ! dxe5 12 tt::le4!, when White's initiative causes Black problems. b) I like Black's position after 10 ... i.xe6 since he controls the dark squares in the centre and has free play. 1 1 h3 tt::lc6 12 tt::lb 3 tt::ld7 13 tt::lg5 tt::lde5!? (this new move varies from 13 ... h6 14 tt::lxe6 fxe6 15 f4, which was unclear in Hubner-Hodgson, Bundes­ liga 1995/6) 14 f4 tt::lxd3 15 tt::lxe6! (af­ ter 15 'ifxd3?! Black has the very strong 15 ... i.c8! +. leaving the knight on g5 and bringing the bishop to a better place on b7 or a6) 15 ... fxe6 16 'iexd3 �h8!? (Black wants his oppo­ nent to commit to how he is going to develop his queenside; the immediate 16 ... g5?! achieves little: 17 fxg5 tt::le5 18 lixf8+ 'ili'xf8 19 'ili'e2 ;!;) 17 'ili'e2 (another way to defend b2 is 17 l:.f2 but then 17 ... g5 ! works well: 18 fxg5 .l:.xf2 19 �xf2 tt::le5 20 'ili'e2 d5 ! + with a strong position) 17 ... 'iee7 18 i.e3 e5 ! +. Black seizes the e5- and d4squares. 10 tt::lh5 ! ? I suggest this untried move. The following line is instructive: 1 o ... a5 1 1 .••

109

b3 tt::lh5 12 l:.a2 f5? (Black is fine after the quiet 12 ... tt::lf4 13 iH1 l:.a7; the point is that Black is not ready for an immediate opening of the position) 13 exf5 ! (this gives White a large advan­ tage; 13 tt::lb 1? was played in Mam­ brini-J.Horvath, Montecatini Terme 1999) 13 ... gxf5 (13 ... tt::lf4 1 4 i.b1 gxf5 15 tt::lfl tt::lg6 16 tt::lg5 ± intending tt::le6) 14 tt::lfl ! e4 15 tt::lg5 tt::lf6 16 i.b1 ±. Black's queenside pieces are not ready for the coming battle in the cen­ tre. u i.n tt::lf4 Intending ... tt::lh 3+ if White plays g3; after 1 l...a5 White can improve his position by 1 2 g3 and i.g2. 12 b3 a5 I like Black's game thanks to his f4-knight. The plan is a general ad­ vance on the kingside - all three pawns can be advanced with suitable preparation. D)

4 i.g5

This is a rare but interesting reply to the Benko, with some ideas in com­ mon with the Trompowsky Attack: White invites the knight into e4, hop­ ing it will proved misplaced. Even though I have played the Benko since 1993, I had not come across this move until my work on this book in 2012! But the move looks healthy enough and could easily be­ come more popular as players seek new paths. 4 tt::le4 5 i.f4 'ifa5+ 6 tt::l d2 g5 ! White gets a pleasant position in case of the quiet 6 ... bxc4 ?! 7 'ikc2 tt::lf6 ...

AITA CK WITH BLACK

110

8 e4 d6 9 .ixc4 g6 10 tLlf3 .ig7 1 1 0-0 ;l;. The queen is then doing nothing on a5; if Black is to justify his aggressive posturing, he needs to make use of his active queen and knight immediately. 7 .ie5 A few games continued 7 'i!Vc2?! gxf4 8 1!Vxe4 but this position can be awkward only for White; Black rapidly develops his pieces and gets the advan­ tage. 8 ... bxc4 9 'ti'xc4 d6 10 'ti'c3 (10 'i!Vxf4 lLld7 + followed by ....l:l.b8 and ... .ig7; an exchange of queens is rela­ tively safer for White) 10 ...'ti'xc3 1 1 bxc3 .ig7 12 l:.cl (M.Maksimovic­ Vucinic, Belgrade tt 2012) 12 ... lLld7 1 3 lLlh3 .ie5 + followed by ...lLlf6 and ... l:.b8. 7 I!.g8! (D) Mostly Black prefers 7 ... f6 but af­ ter 8 .ic3 lLlxc3 9 bxc3 (I vanisevic­ B.Vuckovic, Valjevo 201 1) his pawn­ structure looks miserable - Black's f8-bishop is closed in, the e8-h5 diag­ onal is weakened, while there is also a problem with the b5-pawn. Mean­ while, White's position looks more healthy; he only needs to make a few more moves to bring new pieces into the game. Of course, this position is playable for Black, but the text-move looks stronger since it avoids most of these problems. 8 lLlf3 Besides this natural (but untried!) move, White can also play in gambit style with 8 b4 but Black is ready for this as well. 8 .. .'ii'xb4 9 lLlf3 and then: a) 9 ... g4? was chosen in Murshed­ Hodgson, London 1992, the only game so far in which 7 ... l:!.g8 was played. ...

After 10 l::tb 1 lLlc3 1 1 l:!.xb4 lLlxd 1 12 li!.xb5 gxf3 13 exf3 ! (the d1-knight can be taken later) 13 ... d6 14 .ial lLlxf2 15 �xf2 .ih6 16 lLle4 ;l; intending .id3, !:tel or .t:!.hbl, White's pieces were better coordinated and Black had a tough task ahead of him. b) I propose 9 .. .'i!i'a5 !, sidestepping the rook's attack, so that 10 l:!.bl can be met by 1 O ... a6, defending the pawn. Now White needs to think up some­ thing special to prove he has a good position. However, I couldn't find any­ thing interesting for him. 1 1 'i!Vc2 (or 1 1 .ic7 'i/xc7 12 lLlxe4 'i!i'a5+ 13 lLlfd2 d6 with an extra pawn) l l...f5 12 e3 d6 13 .ial g4 14 lLlh4 'i!Vxd2+ (this simple move clarifies the posi­ tion) 15 11i'xd2 lLlxd2 16 �xd2 e5! 17 dxe6 (or 17 cxb5?! axb5 18 .ixb5+ �d8 +; the only chance for White is to break up the enemy pawn-centre) 17 ... .ixe6 18 .id3 l:tg5 ! (the only way to defend the pawn; conveniently, the rook also heads for h5) 19 cxb5 l:1h5 20 g3 axb5 21 .ixb5+ �d8 22 a4 .id5 + and 23 ... .ie7 leaves White fighting for a draw.

BENKO GAMBIT DECLINED

8 lt::lxd2! .•.

This is the easiest way to get a safe position. Black can expect some trouble after 8 ... g4?! 9 b4!, which now works better since Black can't move the queen back to a5 (as in the 8 b4 line). We already saw the consequences of 9 ... 'iVxb4? in Murshed-Hodgson above, while after 9 ... cxb4 10 lt::lxe4 gxf3 1 1 exf3 d6 12 .tf4 Black's position has a lot of long-term strategic disadvantages: his uncastled king, the lack of prospects for the g8-rook, and the fact that the c­ and e-files are going to be opened and used by the white rooks and queen; generally the white pieces are more harmonious. 9 'ili'xd2 9 lt::lxd2? makes no sense. 9 ... d6 and now: a) 10 .tc3? b4 1 1 lt::lb 3 'iia6 12 .td2 �xc4 + and White needs to de­ fend d5, delaying his development further. b) After 10 b4 the simplest path is 10 ...'i¥a3!, avoiding complications. Af­ ter 1 1 i..g3 Ji.g7 12 l:tc1 (or 12 .:r.b1 ?! cxb4 13 cxb5 lt::ld7 + ) 12 ...lt::la6 13 bxc5 lt::lxc5 14 cxb5 i..d7 15 e3 .tc3 (intending ...lt::le4) 16 f3 a6!, planning to meet bxa6 with ... .ta4, Black seizes a strong initiative. c) 10 .tg3 .tg7 1 1 cxb5 .txb2 12 .l:.b1 .tc3 gives Black a slight advan­ tage.

E)

Ill

4 �c2

This move was strongly recom­ mended by the influential writer John Watson in his book A Strategic Chess Opening Repertoire for White, so you may find that quite a number of your opponents will be playing it. 4 ... bxc4 5 e4 Of course, White was not going to take on c4 with his queen: 5 �xc4? e6! 6 e4 exd5 7 exd5 lt::la6 + intending ... lt::lc7 or ... lt::lb4 to attack the d5pawn; then ... .ta6 and .. ."fie7+. 5 ...e6! (D) 5 ... d6 6 ..txc4 g6 7 lt::lf3 .tg7 8 0-0 0-0 is also playable, but the text-move is more ambitious.

6 i.. xc4

Or: a) 6 dxe6?! gives up the centre for free: 6 ... fxe6 7 Ji.xc4 (Walk-Micha­ lek, Millheim 1993) 7 ... d5! 8 .tb5+ 9 'ifxd2+ 10 lt::l xd2 d6 11 .tg3 .td7 9 .txd7+ lt::lbxd7 10 exd5 exd5 bxc4 12 e4 .tg7 13 lt::lxc4 .ta6 14 1 1 lt::lf3 ..td6 12 0-0 0-0 +. l:tb1 i.. xc4 15 .txc4 lt::ld 7 b) 6 lt::lc3 (a logical move) 6 ...exd5 With an equal position where Black and then: has no particular problems. ...

1 12

A TTA CK WITH BLACK

bl) 7 exd5 i..e7 8 i.. xc4 transposes to the note to White's 8th move below. b2) 7 t'Dxd5 (E.Pogorelow-Sieglen, Deizisau 2009) 7 ... t'Dc6! (a new move; before capturing on d5, Black will drop his knight into d4) 8 i..xc4 t'Dxd5 9 exd5 (in case of 9 i..xd5 Black can reply 9 ... .ia6) 9 ... t'Dd4 10 'ii'd3 i..d6 1 1 t'De2 fi/e7 with a comfortable posi­ tion for Black. b3) 7 e5 t'Dg4 8 t'Dxd5 sharpens the game and demands an accurate re­ sponse. In Mamedyarov-Bareev, Mos­ cow blitz 2009, 8 ... .ib7? 9 i.. xc4 t'Dc6 10 t'Df3 ± was clearly unsatisfactory for Black. With a little more time to think, maybe Bareev would have tried 8 ...t'Dxe5 !, when Black is going to get a lot of central pawns for the knight: b31) 9 f4? t'Dd3+! (winning time for development and taking control of the light squares) 1 0 .ixd3 cxd3 1 1 �xd3 .ie7 1 2 t'Df3 .ia6 1 3 'ii'e4 t'Dc6 + and ... 0-0. b32) 9 �e2 d6 10 f4 .ie6 1 1 fxe5 i.. xd5 12 exd6+ and now 12 .. .'�d7! is the fastest way to finish development: the knight goes to c6, the bishop takes on d6, and the king's rook moves across to e8. The game is rather un­ clear, but it is White who needs to be accurate because his king is in even greater trouble. b33) 9 file4 d6 10 f4 (10 li:Jf6+?? loses to lO .. .'iVxf6 1 1 'i'xa8 lt:Jec6 -+ and ... 'i'e7+ and ... i..b7, trapping the queen) 10 ... f5 1 1 'i'e3 i..e6 12 t'Dc3 t'Dbc6 13 fxe5 d5 gives Black a nice position. I would not like to be White here. 6 exd5 7 exd5 i.. e7 8 t'Df3 ...

Or 8 t'Dc3 d6 (D), and then:

a) 9 t'Df3 0-0 10 0-0 transposes to the main line. b) 9 h3 0-0 10 li:Jf3 t'Dbd7 1 1 0-0 can be found in the note to White's lOth move below. c) White can also bring his king's knight to g3 but it does not change Black's plans much: 9 t'Dge2 0-0 10 0-0 t'Dbd7 1 1 t'Dg3 t'Db6 12 b3 and now: cl) 12 ... t'Dg4?! 1 3 .id3 g6 14 h3 t'De5 is somewhat risky because, as Watson points out in A Strategic Chess Opening Repertoire for White, White can seize the initiative by 15 i..h6! (15 i..e2?! f5 was fine for Black in Erdos­ Bologan, Caleta 201 1) 15 ... lie8 16 i..b5 i.. d7; e.g., 17 f4 i.. xb5 18 t'Dxb5 t'Ded7 19 t'De4 t'Dxd5 20 t'Dbxd6 .ixd6 21 t'Dxd6 :e7 (2l....:te6? 22 t'Dxf7! Wxf7 23 f5 with an attack) 22 l:tad1 t. c2) 12 ...t'Dxc4 13 bxc4 .l:b8 is qui­ eter, and satisfactory; e.g., 14 lte1 :es 15 .if4 .l:.b4! 16 t'Dce4 (in case of 16 t'Db5 Black readily sacrifices an exchange for a pawn and a pair of

connected passed pawns in the cen­ tre: 16 .. Jhb5 ! 17 cxb5 liJxd5 with a fine position) 16 ... ltJxe4 17 ltJxe4 ( 17 'ir'xe4 Ji.d7 = and ... Ji.f8; 17 .l:.xe4 Ji.f8 18 lbe 1 l:!.xe4 19 ltJxe4 Ji.f5 =) 17 ... Ji.f5 =. 8...0-0 9 0-0 d6 (D)

w

10 ltJc3

White sees no necessity to prevent ... Ji.g4, and although 10 h3 has been played, it appears unnecessary. Then after 10 ... tiJbd7 1 1 ltJc3 (this position can be achieved from a variety of move-orders) 1 1 ...tiJb6 12 b3 l:tb8 (Schiendorfer-Ztiger, Swiss Team Ch 2010), Black has solved his opening problems; e.g., 13 Ji.f4 ltJxc4 14 bxc4 l:r.e8 15 lilfe1 h6 = followed by ... Ji.f8 and ... Ji.d7. Black is a bit cramped but his position is solid enough and he has no real problems. 10 ...tiJbd7 After 1 0 ... Ji.g4?! 1 1 tiJd2! the black bishop appears misplaced: 1 1 ... tiJbd7 12 h3 Ji.h5 13 f4! with an obvious advantage, Ju.Horvath-K.Rovid, Hun­ garian Team Ch 2003/4.

11 Ji.f4 tiJb6 12 tiJd2 ! ?

This move (suggested by John Wat­ son) looks more critical than 12 b3, when the black bishop reaches a secure post on g6: 12 ... Ji.g4 13 tiJd2 Ji.h5 = Drozdovsky-S.Kasparov, Internet blitz 2006. 12 ... Ji.b7 13 'ii'd3 tiJh5 ! ? 14 Ji.e3 Black gains a comfortable position if the bishop retreats to g3: 14 Ji.g3?! ltJxg3 15 hxg3 Ji.f6. 14... f5 Black has a pleasant game: a) 15 Ji.xc5? does not work since after the forced 15 ... dxc5 16 d6+ fl i.g7 + Rozum-Levin, Peterhof 2009. b) 9 i.e3 i.g7 and then: bl) 10 i.d3?! is strongly met by 10 ...lbg4!, attacking e3 and b2. 1 1 i.cl (passive, but what else is there? After As mentioned above, White's prin­ 1 1 i.xg5 i.xb2 12 .l::.b 1 i.g7 + the cipal tries here are based on the attempt b5-knight is in real trouble) l l ...'iWb6 ...

...

122

AITA CK WITH BLA CK

(putting pressure on the b5-knight; White's position is critical) 12 i.e2 (12 tLlh3 tLle5 13 .te2, Chetverik­ Stokke, Oslo 2006, 13 ... .td7 ! 14 a4 bxa3 15 l:.xa3 .l:.xa3 16 tLlxa3 .txh3 17 gxh3 �4+ 1 8 i.d2 'i!Vxe4 +) 12 ... tLle5 and now: bl l) 13 .txg5?! c4! gives White only unpleasant options: b1 1 1) 14 tLld4 tLld3+ 15 i.xd3 Vi'xd4 16 li!.cl (forced in view of 16 i.e2? 'ii'xb2 -+) 16 ... 'i!Vxd3 17 'it'xd3 cxd3 18 .l:.xc8+ �d7 +, when the a2and b2-pawns are helpless. bl l2) 14 a4 bxa3 15 .:txa3 (15 tLlxa3? 'i!Vxb2 -+) 15 ...l:.xa3 16 tLlxa3 1Wxb2 17 tLlxc4? (but after 17 'i!Vcl c3 + White will definitely have problems with Black's passed pawn) 17 ... tLlxc4 18 ..Wa4+ (18 .txc4 'ii'c 3+ and Black wins) 18 ... .td7 19 'ii'xc4 i.c3+ 20 Wdl (20 Wfl iVai+ with checkmate) 20 ... 0-0 -+. The white king is in a hopeless situation. b12) 13 a4 bxa3 14 .l:.xa3 lha3 15 bxa3 c4!, intending ... tLla6/d7-c5, is slightly better for Black. b2) 10 f3 h6 leaves Black fine; e.g., 1 1 a4 bxa3 12 .lir.xa3 .lir.xa3 13 tLlxa3 tLlbd7 14 tLle2 tLlh5 15 Vi'd2 i.a6 + Potapov-Degraeve, Cappelle la Grande 2002. Black has the better pawn-struc­ ture (look at the b2-pawn) and his pieces are more harmonious. 9 tLlxe4 The knight gains a tempo by attack­ ing the bishop that he lured to g5. 10 .tf4 After 10 tLlf3 Black even doesn't take on g5, but instead plays 1 O .. .tg7. Then: •.•

.

a) 1 1 .i.d3 tLlxg5 12 tLlxg5 h6 13 tLle4 was tried in Hernando Rodrigo­ Van Riemsdijk, Barbera del Valles 1999 (where 13 ... tLld7?! was played). However, this looks like a bluff, since Black can take the pawn: 13 ... .txb2! 14 llbl i.g7 + followed by ... 0-0, ... tLld7 and either ...tLle5 or ... tLlf6. I don't see any problems for Black. b) 1 1 .tel tLld7 12 i.d3 tLlef6 13 0-0 i.b7 +. The d5-pawn is too weak and Black will take it at some point. lO tLlf6! (D) •..

The most precise move - anticipat­ ing White's 1We2 idea, the knight drops back to a secure square and puts pres­ sure on White's isolated d5-pawn. With White's e-pawn gone, he will find it hard to open lines. l l .tc4 Instead, 1 1 tLlf3 i.g7 12 i.c4 trans­ poses to line 'b' below, but 1 1 'i!Ve2 is an independent idea, creating the threat of tLlxd6+. Then 1 l...l:a6! is forced, but good. Now White needs to regroup and defend the d5-pawn. After 12 'ii'd2, 12...tLlbd7! is a critical move

BENKO: ZAITSEV, DLUGY AND MODERN LINES

that secures Black's advantage - he prepares to meet ..tc4 with ... lt::lb6 and so White can't defend the d5-pawn (that's why the immediate 12 ... ii.b7?! is not so good: 13 ..tc4 lt::lbd7 14 'ife2 was unclear in Abrashkin-Aveskulov, Saratov 2004). Then 13 'iYe2 (13 ii.c4? lt::lb6 14 ..tb3 'it'd?! 15 'iYd3 lt::lbxd5! -+; 13 g3 ..tb7 14 ..tg2 lt::lb6 + and the pawn falls again, K.luss-Koch, Schon­ eck 1988; 13 lt::lf3 ii.g7 14 ii.e2 lt::lb6 15 0-0 lt::lbxd5 16 ..tg3 0-0 + gives Black a clear extra pawn) 13 . .'�a5! 14 lt::lxd6+ �xd6 15 ..txd6 b3+ 16 'ii?d 1 bxa2 17 ..tg3 lt::lxd5 gives Black strong com­ pensation for the exchange. l l ... i.g7 Now: a) 12 lt::le2 (moving the knight to g3, where it poses some danger to Black) 12 ...lt::lbd7 13 lt::lg 3 lt::lb6 14 b3 h5! (this new move seizes space on the kingside; after the natural 14 ... 0-0 15 0-0 i.d7 16 'iYd2 lt::lfxd5 17 ..th6 White gets counterplay, Arencibia-Ivanchuk, Cap d'Agde rapid 1998) 15 h4 (White cannot allow the pawn to run to h3) 15 ... ..tg4 16 f3 ..td7 17 0-0 'iiif8! (in­ tending ... ..txb5 and ...lt::lbxd5, which was impossible with the king on e8 be­ cause of ..txb5+) 18 ii.g5 ( 18 a4 does not help: 18 ... ..txb5 19 axb5 lha1 20 'ikxa1 lt::lfxd5 21 'i!i'c 1 lt::lxf4 22 �xf4 ii.d4+ 23 'iiih2 lt::lxc4 24 bxc4 b3 -+) 18 ...lt::lh7! +. b) After 12 lt::lf3 the knight can't create real threats but it is more solid than putting it on e2. 12 ... 0-0 13 0-0 lt::lbd7 gives Black a clear edge due to the standard problems: the b5-knight and the d5-pawn. White's only hope is

/ 23

to create some threats to the black k i n g but it's a tall order. A2) 8 lt::lf3

This move dates back to the semi­ nal encounter I.Zaitsev-Benko, Szol­ nok 1975. Again White supports the e5 advance but here he does not ex­ pose his bishop. On the other hand, the knight blocks the f-pawn. 8 lt::lxe4! ? The most popular line here is 8 ... g6, when one possibility is 9 e5 dxe5 10 lt::lxe5 ii.g7 11 ii.c4 0-0, but it seems that Black is able to take the e4-pawn without White getting enough com­ pensation. The fact that the move f3 is unavailable to White makes it harder for him to force open the e-file. 9 ..t c4 g6 10 'i!i'e2 (D) .••

10 ... f5 !

Only this move gives Black pros­ pects of an advantage. After 10 ... lt::lf6? a typical tactic works: 1 1 i.f4 l:!.a6 12 lt::lxd6+! �xd6 13 ..tb5+! (the point!) 13 ... .l:id7 14 ..txb8 ..tg7 15 d6! (saving

124

A ITA CK WITH BLA CK

the pawn) 15 ... 0-0 16 i..xd7 'i¥xd7 17 dxe7 .li!.e8 18 0-0 l:he7 19 'i¥c2 ± with an extra exchange. 1 1 tLlg5 ! ? White stops ... tLld7-e5, but Black has an excellent way to give the pawn back. Alternatives: a) 1 1 0-0 (S.Ivanov-Pisulinski, Cheliabinsk 1991) should be met by l l ...tLld7!, transposing to Line A3. It's important for the knight to pass through the d7-square as quickly as possible since while it is on this square, it disrupts the c8-bishop's control of the e6-square. b) 1 1 h4?! tLld7! (again this move is more precise than l l ...i..g7?! be­ cause of 12 tLlg5, when Black cannot reply 12 ... tLld7) and now: b 1 ) 12 tLlg5 tLle5 ! 13 tLlxe4 fxe4 14 'it'xe4? (14 0-0 is a bit more stubborn, though still bad for White: 14 ... i..g7 +) 14 ... tLlxc4 15 'ii'xc4 �b6 -+ and ... i..d7 wins the knight. b2) 12 h5 i..g7 13 hxg6 hxg6 14 .li!.xh8+ i..xh8 + Efimov-Shytaj, Italian Team Ch, Senigallia 2009. l l ... i.. g7 (D) l l ...tLlxg5?! is not so convincing because after 12 i..xg5 i.. g7 13 0-0 Black has to make some move like 1 3 ... 'it>f7. 12 0-0! It is best for White not to rush with the capture on e4, since while the knight remains on g5 it is hard for Black to play his desired ... tLld7-e5 manoeuvre. But even here Black can claim an edge. Other moves: a) 12 f3?! is simply too slow. After 12 ... tLlxg5 13 i..xg5 h6! White has a

choice between sharpening the game or simply playing on a pawn down; in both cases Black is better. 14 i.. xe7 (14 .th4 g5 15 .tf2 0-0 + followed by ... tLld7-e5 is miserable for White, who should prove he has at least something for the pawn) 14 ...'i¥xe7 15 tLlc7+ (the only move; 15 tLlxd6+? 'it>d8 16 'i¥xe7+ 'it>xe7 17 tLlxc8+ .l:i.xc8 -+) 15 ... xe7 17 tLlxa8 i..xb2 18 l:.b 1 i.. c 3+ 19 'it>e2 lDd7 + (Black has a pawn for the exchange and strong play on the queenside; besides, the knight has real problems getting back from a8) 20 tLlc7 tLlb6 21 'it>d3 f4! 22 .l:.be1+ and now 22 ... i..e5 ! -+, leav­ ing the f5-square available for the queen's bishop, is even better than 22 ... 'it>d7?! + Etchegaray-Pap, San Sebastian 201 1 . b) 1 2 tLlxe4 fxe4 13 �xe4 0-0 + leaves Black better since his pieces are more harmonious: ... i..f5, ...tLld7-b6, etc. 12 ... 0-0 13 .l:.e1 13 tLlxe4? is still not good: 13 ... fxe4 14 �xe4 tLld7 +. 13 ...�b6 14 a4

BENKO: ZAITSEV, DLUGY AND MODERN LINES

125

Supporting the knight. After 14 tLlxe4 fxe4 15 'ii'xe4 .l:If7 + Black's ideas include ... .l:!.a5, ... .lta6, ... .ltf5 and ... tLld7-e5. 14 bxa3 15 .Uxa3 lixa3 (D) •••

w

16 bxa3

16 tLlxa3? ..td4 17 .l:!.fl tLld7 18 tLlxe4 fxe4 19 'ii'xe4 tLlf6 20 'ife2 (20 'fixe??? .l:!.e8 -+) 20 ... .ltg4 21 'fic2 ..td7 + intending ... tLlg4. 16 ..td7 17 a4 'ii'a5 ! The queen attacks a4 and eyes the el-rook. 18 tLlxe4 fxe4 19 'fidl (19 'fixe4 �xa4 + with an extra pawn) 19 ... 'ii'b4 20 .l:r.xe4 .ltxb5 21 axb5 ..td4 22 ..td3 .l:!.xf2 23 �h 1 .U.f7 + followed by ... tLld7-e5. ••.

A3 ) 8

.ltc4 (D) This is a more subtle way to prepare e5, but also potentially the most vio­ lent. The bishop takes aim at the f7square, planning to smash open the a2-g8 diagonal with a double pawn sacrifice. 8 tLlbd7 •••

White's main idea is shown by 8 ... g6 9 e5 dxe5 10 d6 exd6 1 1 .ltg5, whip­ ping up a dangerous initiative. It's ac­ tually not so clear that this is good for White, but if you play this way as Black, you are walking through a mine­ field that your opponent will definitely have studied. White has serious compensation after the immediate capture on e4: 8 ... tLlxe4?! 9 'fie2 tLlf6?! (9 ... f5 leaves e6 totally weak: 10 f3 tLlf6 1 1 tLlh3 ;\; intending tLlf4-e6) 10 .ltf4 l:!.a6 (Black looks superficially solid, but an effec­ tive tactic wins an exchange for White) 1 1 tLlxd6+! l:.xd6 12 .ltb5+ :d7 ( 12 ... ..td7 13 ..txd6 'fib6 14 .ltxd7 + tLlbxd7 15 ..te5 ±; 12 ... tLlbd7 13 .ltxd6 tLlxd5 14 0-0-0 +-) 13 ..txb8 ..tb7 (Black does not solve his problems in the case of 13 ... tLlxd5?! 14 0-0-0! ±) 14 ..tg3 .ltxd5 15 tLlf3 ±. Black simply does not have enough for the material damage. The text-move reinforces the e5square, but does not abandon the idea of taking on e4 either. 9 tLlf3

126

A ITA CK WITH BLA CK

White again wishes to play e5 (e.g., 9 ...g6 10 e5 ! with d6 to follow), but the knight blocks the f-pawn, which has consequences ... 9 lt:Jxe4! ? This move has been very rare in practice, but after considerable analy­ sis I feel that it is good. It's under­ standable that players are reluctant to grab a 'hot' pawn in such a position with no preliminary preparation, espe­ cially since 9 ... lt:Jb6 (the most popular move) gives Black a safe position; af­ ter 1 0 Ji.d3 g6 1 1 b3 he is not going to be overrun by a e5 pawn sacrifice. The game is equal after l l ...J.. g7 12 Ji.b2 0-0 13 0-0 e6! (Black has developed all his pieces and now is ready to fight for the centre) 14 dxe6 Ji.xe6 Breuti­ gam-Fedorowicz, German Cup 1990. 10 �e2 f5! (D) This move is untested in a high­ level game. After 10 ...lt:Jdf6 (Tancik­ Vucinic, Senta 201 1) 1 1 0-0 the posi­ tion remains more complicated. .•.

=

e4-pawn once White has put his knight on f3, and then defends the e4knight with ... f5. He will give the pawn back by letting White exchange on e4, and quickly finish his develop­ ment. The difference is that Black has played ... lt:Jd7 instead of ... g6, which gives White an important extra option. 11 0-0?! Of course, there is no point in tak­ ing on e4 with the bishop: 1 1 Ji.d3? g6 12 Ji.xe4 fxe4 13 lt:Jg5 lt:Je5 14 lt:Jxe4 Ji.g7 + with ... 0-0 and ... J..a6 to follow. But the critical test of Black's idea is 1 1 g4!?. Black shouldn't be in real danger, but must be precise: l l ...li:Jdf6 12 gxf5 (12 g5? is bad because of the simple 12 ... lt:Jd7, when e4 remains a good outpost for the black knight) 12 ... g6! (12 ... Ji.xf5? is bad due to 13 lt:Jh4 followed by f3, winning the e4knight) 13 li:Jh4! (the natural 13 fxg6?! hxg6 is in Black's favour since his bishop gets the f5-square) 13 ... g5 14 f3 li:Jd2! (avoiding fxe4, which would improve White's pawn-structure) 15 Ji.xd2 gxh4 with a very complicated position. ll g6 12 lt:Jg5 lt:Je5! 13 lt:Je6 In this line, this move makes some sense. After 13 lt:Jxe4 fxe4 14 �xe4 Ji.g7 + followed by ... 0-0 and ... Ji.f5 Black is typically better. ...

13 ... Ji.xe6 14 dxe6 lt:Jxc4 15 fixc4 Ji.g7 16 l:td1 !

A prophylactic move against ... 0-0. 16 ...'ii'b 6 16 ... 0-0? loses a pawn to 17 f3 li:Jf6 18 .l:txd6! 'iVa5 19 lldl . For the mo­ We saw a similar idea in Line A2 ment it seems Black can win the b5(i.e. 8 li:Jf3 lt:Jxe4): Black takes the knight, but it's not so: 19 .. Jifb8 20 a4

BENKO: ZAITSEV, DLUGY AND MODERN LINES

l:ixb5? (20 ... ltJe8!? 21 i.f4 is less clear) 21 'i!Vxb5 'i!Vxb5 22 axb5 l:txa1 23 l:td8+ i.f8 24 b6 .l:!.xc1 + 25 'iiif2 l:tc2+ 26 'ito>el ! ± queening the pawn. 17 f3 tLlf6 18 i.f4 0-0! (D) w

127

b1) 23 .lite1 'ii'e7 24 .l:.adl tt:lcH ! (protecting the black king and prepar­ ing for the centralizing ... i.d4+) 25 'i!Vxe7+ (after 25 'i!Vc4 i.d4+ 26 'iith I l:ta8 + White can't activate his pieces - the d4-bishop controls everything) 25 .. .':Ji;xe7 26 .l:td7+ �f8 (the only move; 26 ... 'iitf6?? walks into a mat­ ing-net: 27 h4! g5 28 h5 g4 29 f4 +-) 27 li!.f7+ Wg8 +. Black's pieces are ready to realize his advantage. b2) 23 'ii'c4 'ii'b5 24 'ii'h4 (or 24 'i!fc2?! ltJe8! + and ... i.d4+) 24 ... liaS + followed by surrounding the e6pawn. 19 exd6 20 ltJxd6! Surprisingly, the idea of playing e7+ is very serious and Black must be extremely accurate. The immediate 20 e7+?? fails to 20 ... d5 21 exf8'i¥+ lihf8 -+, when ... c4+ wins the b5-knight. 20 'i!Vc6! Black wants to play ... 'i!Vd5. 21 e7+ 21 l:.d1 leaves the a2-pawn unde­ fended and Black can exploit this by 21...'i¥a6 22 e7+ (the only move) 22 ...Vxc4 23 exf8V+ ..ixf8 24 tLlxc4 l:.xa2 +. Black has won a pawn and has chances to realize it. 21. 'i!Vd5 22 exfS'ili'+ i.xf8 Black is slightly better because with queens off the board, White's vulnera­ ble queenside pawns become a more significant factor. •.•

Black is ready to cope with all possible captures on d6. 19 llxd6! ? Or: a) 19 tLlxd6? allows Black a strong zwischenzug: 19 ...ltJh5 ! 20 i.g5 exd6 21 e7+ l:!.f7 22 'iYe6 J:le8 23 l:!.xd6 i.d4+ 24 'ifilh1 'ii'b 5. Black has an ex­ tra piece while White is unable to cre­ ate any real threats. b) 19 i.xd6 Iia5 ! (this exact move gives Black an advantage; 19 ... exd6? 20 l:hd6 'ii'a5 21 e7+ wins for White) 20 i.xe7 (White can insert 20 a4 be­ fore capturing on e7, but it doesn't hold the balance either: 20 ... bxa3 21 i.xe7 'ii'xb5 22 'i¥xb5 .litxb5 23 i.xf8 'iii xf8 24 .l:!.xa3 .litb7! + and ... 'iite 7) 8) 5 f3 (D) 20 .. .'i!Vxb5 21 'ii'h4 (21 'i¥xb5? .l:r.xb5 With this move, White adopts a very 22 i.xf8 'ifilxf8 +) 21 ... 'i¥e8 22 iLxf8 'iiixf8 and White needs to defend the aggressive stance, but with a clear posi­ tional aim: to capture space in the e6-pawn. .•.

•.

128

A TTA CK WITH BLA CK

centre by means of the e4 advance. It became prominent in the 1980s, with Max Dlugy a major protagonist on the white side. 5 ... axb5 After the natural-looking 5 ... g6?! White continues 6 e4 d6 and here one knight goes to a3 and the other to c3: 7 lba3 ! i..g7 8 lbe2 0-0 9 lbc3 ;!;; with an unpleasant position for Black. In gen­ eral, if White establishes an outpost on b5, Black can expect trouble. There is a simple explanation for this: in the Benko, Black sacrifices a pawn to get half-open a- and b-files; if he doesn't have access to both these files, then he doesn't achieve full compensation ei­ ther! That's why in all lines I recom­ mend, I seek to avoid this type of problem. After 5 ... e6 the most critical lines start with 6 e4 exd5 7 e5 Wle7 8 'iVe2 ltJg8 9 lbc3 i.b7 10 lbh3 c4 and here White has a choice between 1 1 i..e 3, with positional compensation for the pawn, and 1 1 ltJf4 'iVc5 12 ltJfxd5 ! ..txd5 13 ..te3 'iib4 14 a3 'i¥a5 15 i.d2 i..e6 16 ltJd5 1'Hd8 17 'iixc4. This is

the end of a forced line. The position is very complicated, but it seems that Black is running much greater risks here - any inaccurate move can lead to a lost position because of his uncastled king and White's passed pawns on the queenside. That's why I recommend avoiding these variations by playing 5 ... axb5 instead. 6 e4 White threatens both i.. xb5 and e5. Black has only one answer. 6 ... 'i¥a5+ 7 i.. d 2 White also can try a gambit ap­ proach with 7 b4!?, but Black is OK after 7 ... cxb4!: a) 8 ..tb2? is not good at all be­ cause after 8 ... e5 ! (a new move, im­ proving over Csiszar-Mede, Zalakaros 2003) White has no time to prevent ... ..tc5. For example, 9 i..xe5 i.c5 10 ltJd2 Wlb6 1 1 ltJh3 d6 12 i.. xf6 (12 i..b 2?? i.. xh3 13 gxh3 i..f2+ 14 'it>e2 'i!r'e3#; 12 i.. g 3? i..xh3 13 gxh3 lbbd7 + followed by ... 0-0, ...ltJh5 and ...ltJe5 leaves White with a terrible structure) 12 ... gxf6 (Black's pawn-structure is destroyed but White has a lot of prob­ lems to solve; first of all, there is the threat of ... i.. xh3 and ... i..f2+ leading to mate) 13 'il¥e2 i.xh3 14 gxh3 ltJd7 15 'ii'xb5 'ii'a7 +. The difference be­ tween the c5-bishop and its white counterpart on f1 is huge and deter­ mines the evaluation of the position. b) 8 lbd2 (heading for b3) 8 ... e5 ! 9 lbb3 'i!r'b6 10 'i!r'd3 lba6! (there is no point defending the b5-pawn: after 10 ... i.. a6? 1 1 i..e3 'i!r'd8 12 lbh3 ;!;; fol­ lowed by i.e2 and 0-0 Black's pieces are awkward) 1 1 'i¥xb5 (White should

BENKO: ZAITSEV, DLUGY AND MODERN LINES

take this pawn immediately since after 1 1 i..e3?! tbc5 12 'ife2 { 12 'ifxb5?? 'ifxb5 13 .ixb5 tbxb3 -+} 12 ... d6 13 'ifxb5+ tbfd7 + Black has a strong ini­ tiative on the queenside) 1 l ...'ifxb5 12 .ixb5 (Ksieski-Komljenovic, Issy-les­ Moulineaux 1997) 12...i..d6!? (a new move, leaving the e7 -square for the king) 13 tbe2 tbc7 14 .id3 i.. a6 15 j.xa6 l:!.xa6, planning ...xg7 reaches an equal endgame. •••

1 63

Or 15 i.g5 .i.xc3 16 bxc3 f6! (the most solid option for Black; he closes the centre and now all the play will be focused on the queenside, where he is clearly better) 17 i.d2 'ii'a4 18 .tel 1h- 1h Almasi-Delchev, Croatian Team Ch, Pula 2000. Play could continue 18 ...'iVxd1 19 .l::!.xd1 .li!.a4 =. 15 tt::lb6 16 'i!Ve2 'i!Va6! (D) •••

w

•••

8 1 2)

14 .li!.e2 tt::le8 The knight still heads for b5. 15 l:tc2 This is a typical idea for White: bringing the rook to c2 to defend the critical squares c3, c4 and b2. But here Black easily gets good counter­ play.

It's more difficult for White to keep control of d3 and c4 without queens on the board. 17 'i!Vxa6 Or: a) 17 'ii'e3?! tt::la4! (this looks more aggressive than 17 ...tt::lc4, which was chosen in Wilhelmi-Mainka, Bad Wor­ ishofen 2000) 18 tt::lxa4 'iVxa4 19 'ii'd3 .li!.b4 20 a3! (this is a nice idea to reduce the tension, but it does not solve all White's problems) 20 ... .li!.xe4 21 b3 'ii'd7 22 'i!kxe4 i.xal + followed by ...i.f6 and ...tt::lc7. Black is better due to his more solid pawn-structure. b) 17 e5?! weakens the d5-pawn for no good reason. 17 ... 'i/xe2 18 .Uxe2 tt::lc4! 19 exd6 exd6 20 a4 tt::lc7 21 :e7

A ITA CK WITH BLACK

1 64

i.xc3 22 .l:!.xc7 i.xb2 23 i.xb2 .l:!xb2 + S.Ivanov-Sivokho, St Petersburg Ch 1999. c) 17 i.g5?! also fails to counter Black's play: 17 . ."ilixe2 18 lhe2 ltJc4! (this is even better than equalizing by 18 ... i.xc3 19 bxc3 f6 20 i.cl ltJc4 A.Graf-Ziiger, Geneva 1999) 19 i.xe7 ltJxb2 20 e5 liJd3 21 l:r.dl and now 21 ...ltJxe5?! 22 ltJxe5 i.xe5+ 23 l:txe5! dxe5 24 d6 ltJg7 25 d7 lDe6 26 i.f6 gives White sufficient compensation for the exchange thanks to the d7pawn. However, the e5-pawn can be taken later, and Black should play 21 ...c4! +. 17 .l:!. a6 (D) =

•.•

x

18 l:tb1!? White's idea is to play b4. Or 1 8 i.g5?! ltJa4! (18 ... ltJc4!? is no doubt viable, but ...ltJa4 is almost always more effective, since exchang­ ing off White's c3-knight weakens a2, e4 and a4, while the greater exposure ofb2 is also useful for Black), and then: a) 19 i.cl?! (such moves can't even maintain equality) 19 ...ltJc7 20 liJd2

f5! 21 f3 e6! (improving over 21....l:.b4, Becking-Van der Weide, Saarlouis 2002) 22 dxe6 ltJxe6 with a slight ad­ vantage for Black. b) 19 ltJxa4 .l:.xa4 (Black has opened all the lines he needs) 20 e5 (there is no other way to defend both b2 and e4; 20 i.xe7? .l:lxe4 21 i.g5 i.xb2 22 J:r.d 1 i.f6 + leaves White with weaknesses on a2 and d5) 20 ... f6! 21 exf6 lDxf6 22 l:tel .l:!.xa2 23 l:he7 ltJxd5 24 l:r.d7 and now 24 ...i.f8! + is best. Black will take on b2 for free; instead, 24 ... l:taxb2? only led to an equal position after 25 .:txb2 i.xb2 26 .Uxd6 liJb4 in Golod­ Van der Weide, Hoogeveen 1998. 18...ltJa4!? This is untried but logical. Black makes a planned move and prevents b4. Instead, 1 8 ... ltJc7?! allows White to demonstrate his idea: 19 b4! c4 20 b5 ! .Ua5 21 i.e3 f5? (allowing a typi­ cal idea: 2l...i.xc3 is more resilient, though after 22 lhc3 lha2 23 liJd4 ;!;, lDc6 is coming anyway) 22 liJd4! i.xd4 23 i.xd4 fxe4 24 i.xb6! .l:.xb6 25 a4 with a lost endgame for Black, Poobesh Anand-R.Scherbakov, Pune 2004. 19 liJd1 Or 19 liJd2 f5 ! 20 ltJxa4 l:!.xa4 21 b3 l:!.a7 22 f3 fxe4 23 fxe4 ltJc7, with compensation. 19 ltJc7 20 b3 f5! Again ... f5 works! After 21 exf5 ltJxd5 ! 22 fxg6 ltJac3 23 gxh7+ 'ittxh7 24 ltJg5+ 'iti>g8 25 ltJxc3 ltJxc3 26 .l:.bb2 Black can, if he wishes, take a draw by repetition with 26 ...liJdl 27 �bl ltJc3. ..•

BENKO A CCEPTED: KING WALK

1 65

82)

10 g3 l2Jbd7 It's also possible for Black to start with 1 0 ... 0-0 1 1 'it>g2 lDbd7, reaching the same position. 11 'it>g2 0-0 (D)

w

w

a1) 16 l::tc 1 l2Jc4 17 '1We2 '1Wb4 18 a3 '1Wb3 19 lDd 1 l:!.a4 20 '1Wc2 l:!.b8 21 '1Wxb3 .l::!.xb3 Van der Sterren-Van Wely, Dutch Team Ch, Breda 1997. White can't escape from Black's pres­ sure, while Black has no way to turn the screw at the moment. a2) 16 '1We2 .:!.fb8 17 l:tacl '1Wa6! (the threat of ... l2Jd3 virtually forces White to take on e5) 18 i.xe5 .i.xe5 19 .:.c2 '1Wxe2 20 l:texe2 .l:ib4 21 a3 l:.b3 22 l2Jd1 .i.d4! 23 l2Je3 .i.xe3! (preventing l2Jc4) 24 .l::!.xe3 l:.xe3 25 fxe3 l:tb8 26 'it>f3 .l:!.b3 Karpov­ Salov, Belgrade 1996. White is a pawn up, but the activity of Black's rook provides full compensation. b) 13 '1Wc2 '1Wb6 14 a4!? (this logi­ cal move, seeking to place the knight on b5, has not been suggested before; Black got sufficient compensation in Spiess-Vogt, Leipzig 1997 after 14 .i.f4 l:!.fb8 15 .l::!.ab 1 '1Wa6 16 h3 l2Jge5 17 l2Jxe5 l2Jxe5 1 8 .i.xe5 .i.xe5 19 b3 c4) 14 ... '1Wb4! 15 .i.d2 (or 15 l2Jb5 l:tfb8 16 .l:i.a3 l2Jb6 {attacking a4} 17 b3 c4 with compensation) 15 ... �fb8 16 l2Jd1 '1Wb7 (16 ...'1Wb3?! is not good =

Here White has two main moves: B21: 12 .l:i.e1 165 168 B22: 12 h3 The latter prevents ... l2Jg4, while the former just ignores this idea. 821)

12 .:!.e 1 l2Jg4 Now White has many possibilities. 13 '1We2 Or: a) 13 i.f4 is a strange move. The best squares for this bishop are g5 (to attack e7) and d2 (to prepare b3 by de­ fending the c3-knight). There is not much for it to do on f4 with knights on g4 and d7. It can potentially take a black knight on e5, but such an ex­ change rarely gives White much hope of any serious advantage. 13 ...'1Wa5 14 h3 l2Jge5 15 lDxe5 l2Jxe5 (D) and now:

=

1 66

ATTA CK WITH BLACK

since after 17 'i¥xb3 .l:r.xb3 18 a5 ;t Black can't do anything real; White plays ll.c3 and .l::te2, and enjoys his ma­ terial advantage) 17 h3 tLlge5 18 tLlxe5 .txe5 19 il.c3 (after 19 l:ta2 f5! the d5pawn becomes vulnerable) 19 ... .txc3 and here: b1 ) 20 �xc3 'i!Va6 21 tLle3 l:lb4 22 b3 'i¥b7 (not 22 . Jhe4? 23 tLlg4! .l:r.xg4 {23 ....l:r.xe1 ?? 24 tLlh6+ 'iii>f8 25 'iVh8#} 24 hxg4 ±) 23 tLlg4 f6 intending ... h5 and ... l:tb8. Black is OK. b2) 20 bxc3 .l:r.a5 followed by ... 'i¥a6, ... tLle5, etc. The black pieces domi­ nate. c) The idea of advancing with h3 and f4 should not be good since White weakens his kingside too much with all these pawn moves: 13 h3 tLlge5 14 tLlxe5 tLlxe5 15 f4 (15 'i¥e2?! doesn't halt Black's plans because of 15 ... c4! followed by ... tiJd3) 15 ... tiJd7 (D) (in case of 15 ... tLlc4 the black knight is pushed back by 16 �d3).

with good compensation, while 16 a4 has never been played but looks a logi­ cal try, as White hopes to post his knight firmly on b5. 16 ...'itb8! looks the most convincing reply: the f8-rook will go to c8 to support the ... c4 ad­ vance, while the knight will jump to c5 and the queen is ready to be de­ ployed to various squares on the b­ file. Then: cl) Trying to block the b-file by 17 tiJb5 does not work due to 17 ... tiJb6!, intending to take on a4 with this knight (that's why the queen moved to b8 - the b6-square was reserved for the knight). After 1 8 'i¥b3 'itb7 Black intends .. Jlfb8 and ... tLlxa4. c2) 17 �c2 c4 18 .te3 �b4 19 l:tecl l:tfc8 gives Black sufficient com­ pensation. d) 13 .l:.e2 �b6 (13 ... 'ili'a5 !? is also fine for Black: 14 .l:tc2 tLlge5 15 tLlxe5 tLlxe5 16 �e2 �a6! 17 'itxa6 .l:!.xa6 18 il.d2 .l:tb8 with compensation, Gavri­ lov-Andreikin, Moscow 2012) 14 .tf4 l:.fb8 15 l:tc1 'iVa6 16 :tcc2 (the white rooks have safely defended the b2and c3-squares) 16 .. J::tb4 17 �d2 (Mil­ w anovic-Pap, Belgrade 2001; I don't see any reason to force the g4-knight away: 17 h3?! tLlge5 18 tLlxe5 tLlxe5 19 .txe5 .i.xe5 20 b3 c4 gave Black good compensation in Kaidanov-Wang Zili, Erevan Olympiad 1996) 17 ....l:!.ab8 (l7 ...tLlge5 !? is also possible) 18 b3 c4 19 bxc4 .l:hc4 leaves Black's pieces active enough to compensate for the material imbalance. e) 13 tiJd2 tLlge5 14 'ife2 (threaten­ Now both 16 'itc2 'itb6 17 a4!? (a new move) 17 ....l:r.fc8 1 8 il.d2 c4 and ing to trap the knight by f4) 14 ... tiJb6 16 .l:.e2 �b6 17 'ilfc2 l:tfb8 leave Black 15 f4 tLled7 (D) and then:

BENKO A CCEPTED: KING WALK

167

on the black king by playing e6. Now the e6 advance will be met by ... f5. e2) After 16 a4 (stopping ... lt::la4) 16 ... J.xc3!? Black regains the pawn and while the long dark-square diago­ nal may look worrying, he has ideas of blocking it by playing ... e5 (twice, if necessary!), and other resources too. 17 bxc3 lt::lxa4 18 c4 and now: e21) 18 ... lt::lab6 is possible, though it is not the safest move. Then: e21 1) 19 J.b2 lt::la4 20 l:.abl !? (20 .tel lt::lab6 Lalic-Bures, Pardubice 2010) 20 ... e5 ! 21 dxe6 fxe6 22 J.al lt::lab6 23 'ii'd3 Vi'e7 White can't profit from his control of the long di­ agonal since Black is ready to block it by ... e5. Moreover, Black controls the a-file. e212) 19 .ta3 e5 ! 20 dxe6 fxe6 followed by ... 'ike7 and ... e5. e22) I agree with Alterman's ad­ vice to play 18 ... lt::lc 3 !, as from here the knight can transfer to b4 via a2, se­ curing an equal position: 19 I!xa8 'i'Vxa8 20 'ife3 lt::la2 21 J.b2 l:.b8 22 .tal lt::lb4 23 'iVc3 f6 24 e5 fxe5 25 fxe5 lt::lxd5 ! (the only move, but good enough) 26 cxd5 Vi'xd5+ and ... lt::lxe5 with full equality; White seemingly can't remove the knight from this cen­ tral outpost. 13 'iVa5! A strong move; Black wants to take on c3. 14 ltld2 After 14 J.g5, the move 14 ...J.xc3!? (rather than 14 ... .l:r.fe8, Lalic-Aikhoje, British Ch, Scarborough 1999) has not yet been tested, but it is a typical strategic idea: Black exchanges his =

el) 16 ltlf3 lt::la4! (usually, if Black plays ...lt::la4 there are no problems for him any more, and this position is not an exception) 17 lt::ld 1 (Black is OK af­ ter 17 lt::lxa4 .Uxa4 18 a3 'ika8 19 l:bl c4 followed by ... lt::lc5-d3, ....:tb8, etc.) 17 ...'ii'b6 and here: el l) 18 e5 is a weighty decision as it both blocks the g7-bishop's diago­ nal and weakens the white d-pawn. 18 ...'ii'b7 19 'ii'e4 .lir.fc8!? (this is a new move, intending ... c4 and bringing one of the knights to c5; 19 ... h8 was played in Yusupov-Baklan, Bundes­ liga 1997/8) 20 b3 (20 e6? is not good in view of 20 ... f5 ! 21 Vi'd3 lt::lf6 +) 20 ... lt::lab6 2l ltlc3 dxe5 22 fxe5 c4! 23 b4 lt::la4! (one knight removes the c3knight, while the other will come to b6 to attack the d5-pawn) 24 lt::lxa4 l:.xa4 with sufficient compensation. el2) 18 l:tbl permits a rather un­ usual tactic: 18 ... lt::lc 3! (18 .. .'ili'a6 was chosen in M.Gurevich-Banikas, Kor­ inthos rapid 2002) 19 lt::lxc3 J.xc3 (the a2-pawn falls) 20 .l:r.dl :xa2 21 e5 and now 21.. .lt::lb8 ! is an important move as White intended to launch an attack =

=.

=

•.•

A ITA CK WITH BLACK

168

dark-squared bishop and blocks the long diagonal by ... f6. 15 bxc3 f6 16 il.d2 'i!fa4! (securing the c4-square for one of the knights; the natural 16 . J�tb8?! leads to an unpleasant po­ sition with an awkward knight on f7 after 17 liJh4! li:\ge5 18 f4 liJf7 19 a4 1) 17 lii.ebl lHb8 followed by ...li:\ge5c4. White has no plan for improving his position. 14 .Jtxc3 15 li:\c4 15 'ifxg4 is not dangerous for Black: 15 ... .txd2 16 il.xd2 Vi'xd2 17 'iYxd7 lilfe8 18 b3 'iiff8 19 a4 llab8 (leaving the e8-rook free to fend off the white queen; 19 ... lieb8 was played in Tal­ laksen-Haugli, Norwegian Team Ch 2010/1 1) 20 l:ad1 'ilfc2 21 e5 (the only way to seek any advantage) 21...l:ted8 22 'ili'c6 'ili'xb3 leaves the position equal since after 23 .l:!.b1 Vi'xb1 24 llxbl .l:txbl the black rooks easily contain White's a-pawn. 15 .'�a6 16 bxc3 li:\ge5 17 li:\xe5 li:\xe5! (D) =

•••

=

18 .th6 Now: a) 18 ... l:.tb8?! loses control of the b-file: 19 �xa6 l:txa6 20 .:!.ebl t in­ tending l:!.b7. b) The natural 18 ...'ifxe2 19 .:!.xe2 litb8 allows White to secure some chances by 20 a4 li:\c4 21 a5 lixa5 22 l:txa5 li:\xa5 23 e5, though this isn't too much, Moiseenko-Tukhaev, Ukrainian Ch, Kharkov 2004. c) I propose the new continuation 18 ... .l:tfe8!?. Then 19 a4? liJd3 ! costs White the exchange since 20 .lir.edl?? loses the queen to 20 ... liJf4+, while 19 'iWxa6 .:!.xa6 leaves Black with no problems since he controls c4, d3 and a4 (after .. J:Iea8). After 20 l:teb1 f6! Black intends to use his king to de­ fend e7. This is the point of playing 18 ....l:!.fe8 instead of 1 8 ... .l:!.tb8 - there after 20 .l:!.ebl Black had no time for 20 ... f6. =

••

w

B22)

12 h3 'ifa5 (D) This and 12 ....l:!.a6 are Black's two most popular choices after 12 h3. w

It's better for Black when White takes on a6 himself.

BENKO A CCEPTED: KING WALK

Before we start to look at theoretical variations, let's consider some general themes that will help us to understand events on the board. It's clear that Black's main idea remains the same ­ pressure on the a- and b-files from his rooks and queen backed up by the g7bishop and both knights (mostly via c4 but possibly from other squares). How should White counter his opponent's plan? There are two main ideas. The first is 13 �e1 followed by e5. How­ ever, almost all the variations have al­ ready been analysed in depth and often end with drawish positions. An­ other option is to create a blockade on the queenside light squares with pawns on a4 and b3 and pieces defending them and preventing ... c4. This is a similar concept to the plans we see White adopting in Chapters 14 and 15. Black naturally tries to disrupt this plan; possible ways to do so are ...lt:Jc4, exchanging queens (usually in Black's favour), setting up a blockade on the a3-square after b3 is played, etc. 13 lilel White's usual plan with this move is the e5 advance; he can also follow up with l:te2 (and even llc2), defend­ ing b2 so that the queen's bishop can be developed. In my opinion, Black should be quite safe in these lines, since the e5 idea is not dangerous, while l:te2 looks awkward (or at least not aggressive) and Black can simply get on with his own plans and secure good play. Moreover, it very often turns out that the rook would be more useful on other squares (d 1, c 1 or even b1). My main reason for having 13

1 69

l:lel as the main line is that it is the most popular in practice. White's principal alternatives both fit in with his queenside plans: a) 13 ..td2 (D) and now:

a1) 13 ... li:Je8?! is unconvincing as it lets White continue his own plan: 14 iic2 lt:Jc7 (Kelecevic-Hertneck, Aus­ trian Team Ch 199617) 15 a4!? ;;!; fol­ lowed by ti:Jb5 (or ti:Jd1-e3), ..tc3, etc. a2) 13 ... li:Jb6 (the knight immedi­ ately sets off to cause trouble for White on the queenside) 14 'i!Ve2 (stopping ... lt:Jc4; another way to do this is 14 b3, but after 14 ... lifc8 15 'i!Ve2 'i!Va6! the exchange of queens will give Black long-lasting compensation) 14 ...'ti'a6! 15 �xa6 l:.xa6 16 b3 ti:Jfd7 intending ... f5, ... c4, ....l:.fa8, etc., gave Black compensation in Tyda-Jaroch, Gdansk 2009. b) 13 'ifc2 and now: bl) Here is a good illustration of White's dream in this line: 13 ... .lil.fb8 14 i.d2 li:Je8 15 llhb1 ! (that's why White left his rook on hl !) 15 ....l:'tb7 16 a4 'ti'a6 17 ti:Jdl ! (preparing i.c3)

A TTACK WITH BLACK

1 70

17 ... .:labS 1 S i.c3 i.xc3?! (Black could try 1S ... i.h6!?, to avoid bring­ ing the d1-knight to an active post) 19 tt:lxc3 .l:tb4 20 b3 tt:lef6 21 tt:ld2 ± Gligoric-Tseshkovsky, Bled/Portoroz 1979. Black achieved nothing on the queenside. b2) 13 ...tt:lb6 and here: b21) 14 .:!.b1 tt:lfd7 (Black has pre­ vented b3, and exerts pressure on White's queenside; now Black can gradually improve his pieces) 15 i.g5 .l:.feS 16 l:.hc 1 i.xc3! 17 bxc3 ( 17 �xc3?! 'ii'xc3 1S l:txc3 l:txa2 19 b4 tt:la4 20 :tee 1 f6 21 .i.e3 .:taS + Ada­ mantidis-Pikula, Zurich 201 1) 17 ... f6 1S .i.h6 'it'xa2 19 �b2 'iVa4 with equal­ ity. b22) 14 11e1 (D) and now: B

... .

•

•

� � -··

•

-·-· - - -·· � � � �----� � • -�· • ·�· • • � .ttJD �

(/.' "'"'

Jo', ..

?.

, ·

"'

b2221) White can't push the knight back by 15 1Wb3 because of another queen-exchange idea: 15 ...'iVb4!? 16 'ii'xb4 cxb4 17 tt:lb5 tt:ld7 and Black's pieces (the aS-rook, g7-bishop, c4knight and the b4-pawn) are perfectly placed. b2222) After 15 a4 l:tfbS 16 tt:lb5 (an important tactical motif is re­ vealed by 16 e5 tt:leS 17 exd6 tt:lxb2! 1 S .i.xb2 l:hb2 19 'ii'xb2 .i.xc3 20 'ii'b7 .i.xa1 21 Iha1 exd6 with equal chances) 16 ... .l:!xb5 17 axb5 'i!i'xa1 1 S 'ii'xc4 tt:ld7 all the black pieces are more active than their white counter­ parts and Black has sufficient com­ pensation. 13...l:.tb8 (D) w

� �

'

,,

� DiY.� Dhl .l:.fl+ with a per­ petual check, or 26 'iVc4 'i!i'a8 27 �xc5 h5 20 .'iVa6 A good square for the queen: it at­ tacks e2, defends b7 (important in case White plays d6) and controls the d6-square. 21 a4 .tf6 (D)

22 'iVf3 Another possibility is the untried 22 l:tc2 attacking c5. But then the b7rook plays a major role: 22 ....l:tb4 23 'ii'e3 lid4! and then: a) 24 l:!.xc5 l:idl (intending ...'i!Vfl+) 25 11b5 'i!Vc8! (planning ...'i!Vc4-fl+; also, the c 1-bishop is attacked) 26 .:tc5 (after 26 'ikc5? 'ikf5 Black threatens mate by ...'iVe4+, and 27 'iVe3 :c8 in­ tending . .'ifc2 leaves White worse) 26 ...'ii'a6 27 Ilb5 with a practically forced repetition. b) 24 �f3 .l:.d3 25 .te3 (25 'ii'e4 .l:.d4 26 'ilff3 1ld3 is equal) 25 ... .td4 leaves Black with enough compensa­ tion. 22 ... l:td8 23 .l:tc2 .td4 Now the bishop defends the c5pawn, while the d5-pawn starts to look beleaguered. 24 a5 After 24 .tg5 (A.Shneider-Laza­ rev, Paris 1998) 24 ...11bd7 intending ...f6 (to defend e7) and then ...11xd5, the position is equal. 24 ... :bd7 25 .te3 Or: 25 :e2 .tf6 26 :d2 .td4 ; 25 W.Ve4 :xd5 26 'ii'xe7 .i.e5 ! intending .. .'ita8 and ...:dl. 25 �xd5 Now: a) 26 .txd4?! is dubious since it creates a well-supported passed pawn for Black: 26 ...cxd4 27 b4 (27 .:!.c7!?) The e7-pawn is securely defended 27 ... d3 + Beliavsky-Leko, Cacak 1996. and the b7-rook is free to move along b) 26 �h2!? e5 keeps the position the b-file. equal. =

=.

.•

=

...

1 4 Ben ko Accepted : Fia nchetto Li nes 1 d4 li:Jf6 2 c4 cS 3 dS bS 4 cxbS a6 14 Jtb2 0-0 15 li:Jh4 'i!Vh5 16 e4 �xdl 17 .l:tfxdl li:Jfd7 intending ... c4 or 5 bxa6 g6 6 lt'lc3 Jtxa6 7 g3 (D) ... .i.xc3 and ...l:txa2. b) 1 0 0-0 lt'lb6 (heading for c4) 1 1 l:tel 0-0 1 2 e4 lt'lfd7 1 3 'i!Vc2 lt'lc4 (D) and now:

We now begin our coverage of lines where White fianchettoes his king's bishop. They are considered the most dangerous for Black in the Benko Gambit, so we need to pay close attention to them. 7 d6 8 ..tg2 White can play an 'extended fian­ chetto' by 8 li:Jf3 ..tg7 9 ..th3, when Black should play by analogy with the main lines: 9 ... lt'lbd7 and now: a) I 0 .l:.bl is untried, but as in Chap­ ter 15, the idea is to meet 10 ...lt'lb6 with 1 1 b3. However, with the bishop on h3, this is less effective: l l...Jtc8! 12 ..txc8 'ii'xc8 13 0-0 'i!Vf5 (attacking d5) ••.

bl) 14 ..tn transposes to Line B43. b2) 14 i..xd7 has scored well for White, but I find it hard to understand this exchange when White has a bunch of weak light squares. 14...'i!Vxd7 15 b3 lt'le5 (a good alternative is 15 ...lt'lb6!? 16 i..b2 'i!Vg4 {the f3-knight feels un­ comfortable} 17 �g2 'ili'h5 intending ... ..tc8 or . .f5) 16 lt'lxe5 i..xe5 17 .i.b2 f5! (17 ...l:.fd8? gives White time to make an important exchange: 18 lt'ldl ! i..xb2 19 lt'lxb2 ;!; Beliavsky-Palatnik,

AITA CK WITH BLA CK

1 78

Kiev 1978) 18 exf5 (after 18 t'Dd1 i.xb2 19 t'Dxb2 f4 20 t'Dc4 'i!ih3 21 'ii'd 3, as in Kakageldiev-Alburt, Ash­ khabad 1978, Black can reach a type of endgame where he has no problems by 21 ...f3 22 'ili'fl 'ii'xfl + 23 'iitxfl i.xc4+ 24 bxc4 .:!.a4 ) 18 ...l:txf5 19 t'Da4 i.xb2 20 'i!i'xb2 l:.af8 21 f4 and then: b21) 2l...i.b7?! 22 l:tadl e6! (not 22 ... i.xd5? 23 t'Db6 'itb7 24 t'Dxd5 ::r.xd5 25 .l:!.xd5 'ifxd5 26 :xe7 +-) gave Black counterplay in Darznieks­ Shereshevsky, Daugavpils 1973, but it's objectively just enough compensa­ tion. b22) 21...g5! is a new move that my editor Graham Burgess pointed out. It gives Black a strong initiative: 22 l:te4 (22 fxg5? li!.f2 23 'ifxf2 l:!.xf2 24 'iitxf2 'ii'f5+ 25 'iitg l 'ii'xd5 -+) 22 ... gxf4 23 l;lae1 :gS! (23 ...:8f7!? is also worth a try) 24 l:txe7 "ii'g4 25 "ii'c 2 l:tg6 with the better game for Black. 8 ...i.g7 (D) =

178 A: 9 t'Dh3 1 80 B: 9 t'Df3 A) 9 t'Dh3 Of course, the knight on h3 looks a bit ugly but it has its logic: the bishop defends d5 while the knight aims to go to f4. The main disadvantage is that the knight is a long way from the c6square, and this makes Black's coun­ terplay with ... c4 much easier and ef­ fective. 9 t'Dbd7 10 0-0 0-0 (D) •.•

Here White has plenty of possibili­ ties, but his main idea is to play l:tbl, "ii'c2, b3 and i.b2. Black's task is to prevent this arrangement, and his prin­ cipal tools in this effort are the ... c4 advance, putting pressure on the c3knight with .. .'�aS and by unmasking the g7-bishop, and bringing a knight to c4. ll l:lbl Or: Now we consider two knight moves, of which the latter is by far the more a) 1 1 �c2 'iVa5 12 l:tbl is another important: way to reach the main line.

BENKO A CCEPTED: FIANCHETTO LINES

b) 1 1 i..d2 'i!ka5 12 'i!kc2 :ttb8 13 b3 (for 13 l:.abl see the main line) 13 ... c4 14 .l:.abl transposes to the note to White's 13th move below. c) 1 1 lLlf4 does not assist with White's main idea and Black can continue developing his counterplay: 1 1...1i'c7 12 iic2 l:!tb8 13 l:!.bl (after 13 b3 c4! 14 i.. d2 cxb3 15 axb3 'i¥b6 16 l:tabl lLle5 17 l:.fcl l:.c8 we have a picture of typical compensation for the pawn: Black's pieces are so active that White can't free his game; how­ ever, at the moment it's no more than sufficient compensation) 13 ... c4 and then: cl) 14 lLlh3?! lLlc5 15 lLlg5 is an at­ tempt to bring the knight back to a normal square, but it is thwarted by 15 ... i..c 8 ! (preventing lLlf3 due to ... i..f5) 16 f3? (this allows a typical manoeuvre in untypical form; even af­ ter the superior 16 l:.dl i..f5 17 e4 i..d7 + Black is still better, since d3 is weak) 16 ... lLla4! (the d5- and a2-pawns are in danger) 17 lLlxa4 'i!ka7+ 1 8 'it>hI 'ii'xa4 + Yanofsky-Gheorghiu, Siegen Olympiad 1970. c2) 14 i.. d2 (intending life I and an advance by the b-pawn) 14 ... lLle5 15 l:!fcl i..c 8!? (Black aims to weaken the d3-square by ... i..f5) 16 b3 i..f5 17 e4 i..d7 18 lLldl �a7! (counterattack­ ing a2) 19 bxc4 lhbl 20 'iixbl 'i!kxa2 21 'ifxa2 :xa2 22 i..c 3 i..h6! 23 lLle3 i.. xf4 24 gxf4 lLld3 25 :tal .l:hal + 26 i.. x al lLlxf4 with an equal endgame. c3) With 14 i..e3 (Koploy-Cusi, Californian Ch 1996), White wants to place his bishop on d4. Black should naturally disrupt this plan: 14 ... lLlg4!?

1 79

15 i.. d2 .l:ta7 (preparing to double rooks and put pressure on the b2pawn) 16 lLle4 (this is why White pro­ voked ... lLlg4; now the lLle4 and i..c 3 idea is available) 16 ... l:.ab7 17 i..c 3 lLlc5 18 i..xg7 fl In the game Black played a typical move, 24 f5, with a pleasant but not necessarily decisive advantage. He could instead have targeted the h6bishop by 24 ... g5 ! 25 hxg5 'ittg6 26 f4 lt:Jd3 27 f5+ 'ifi>f7 28 l:.dl lt:Jc2 +, win­ ning an exchange.

Again we see a picture of domina­ tion: the a2-pawn is fixed and weak, while the c4-square is guaranteed for one of the black knights. 19 lt:Jd3 i.a6 20 h3 lt:Jge5 21 lt:Jxe5 lt:Jxe5 22 f4 lt:Jc4 23 e4 lt:Ja3 24 net i.d3 25 lt:Jf2 ..ic4 26 e5 .:Z.b2 Black has good practical winning chances.

•..

DREAM POSITIONS FOR BLA CK

1 0)

205

1 1)

Plachetka - Peev

Alekseev - Grishchuk

Stary Smokovec 1974

FIDE Grand Prix, Elista 2008

The a2-pawn is safety blockaded and ... c4 is coming inevitably. 22 l:tacl c4 23 bxc4 i.xc4 24 e4 .l:.b2 25 i.fl tbc2 26 l:f.edl i.xfl 27 l:.d2? White could maintain the balance by 27 �xfl .l:.a3 28 .l:.bl l:hbl 29 lbxbl l:!.xa2 30 .Ud2! 27 .l:r.c8! 28 :dxc2 lixc2 29 l:.xc2 i.d3 30 l:tcl i.xe4 Black has the advantage thanks to his better pawn-structure.

This is a typical situation where Black has exchanged his dark-squared bishop for a knight in order to win his pawn back. He has the advantage thanks to his better pawn-structure, while White's bishop-pair is not a ma­ jor factor here. 25 f3 25 b4?! doesn't help because after 25 ... cxb4 26 :!xb4 l:!.xb4 27 ..ixb4 i.b5 + the f2-pawn falls. 25 tbf6 26 l:.f2 26 b4 tbxe4 27 fxe4 l::tc2 28 .U.f3 i.h3 + intending ... :a8-a2. Now Black allowed White to main­ tain the material balance by 26 Jla3 27 i.xf6 exf6 28 :tb2, although Black was still better and later won. Instead, he could have won a pawn: 26 ...tbxe4! 27 fxe4 .Ua3 28 l:l.fb2 ..ib5 (intending ... ..id3) 29 l:tel ..ia6 + and the b3pawn drops off.

=.

••.

•••

•.

A TTACK WITH BLA CK

2fJ6

1 2)

13)

w

w

Andonovski - Sta nojoski

Skopje 201 1

Without his dark-squared bishop, White rarely has much hope of any serious advantage; usually, when he makes this exchange, he is just seek­ ing to build a solid defensive fortress. 18 .l:tec1 l:!.b4 19 a3 l:l.b7 20 l:tab1?! 20 b4! cxb4 21 axb4 'iVxa1 22 .l:.xa1 .l:txa1 20 .l:!.ab8 21 lLld1 White's plan is to bring the knight to c4, where it will securely defend b2, and Black should prevent this with 2l .. . ..id4! + followed by ... :b3 and ... 'iVb5. In the game he played 21. .l:tb3? but White squandered his chance by playing 22 'iVe2? (rather than 22 lLle3 ! .l:txb2 23 lLlc4 l:hc2 24 l:.xb8+ Wg7 25 lLlxa5 .:xc 1 26 lLlc6 with an equal endgame) when Black now saw 22 i.d4, with an advan­ tage. =.

.•.

••

•..

M . Socko - J . Houska European Women 's Ch, Plovdiv 2008

Black's pieces are dancing on the weak squares: d3, c4, d4, etc. 18 lLlab5 lLld3 19 �e2 lLld7 20 �b1 .l:Ia5 21 i.g5 'ii;lf8 22 a4 c4! Fixing the weaknesses on b3 and d3. 23 l:.c2 lLl7c5 24 il.d2 Now one of the black rooks must move, and it should be the one on b4: 24 ... l:!.b3 ! and after 25 lLla3 il.xc3 26 i.xc3 Black can simply take on a4 with total domination: 26 .. Jha4 +. In the game Black chose the less accurate 24 .l:!.a8?! 25 lLla3, when 25 ... i.xc3 26 il.xc3 .l:!.bxa4 would not have left Black so much better. ..•

1 7 Positions to Avoid

This is the flipside of the previous chapter: here we see what White is aiming for, and what Black should therefore avoid. This is very useful knowledge for Black, since if you can't steer the game towards a dream position, then maybe you can at least evade the opposite fate. 1)

Strategically, the main scenario Black needs to avoid is one where White has captured the light squares on the queenside by putting pawns on a4 and b3, and a knight on b5 and/or c4. This set-up blocks three files on the queenside and Black can only wait for his opponent to break through in the centre. That's why Black must pre­ vent this set-up before it is too late. Here l...c4 does not help because of 2 b4!, since the pawn is untouchable: 2 ...ltJxa4 3 b5 'fla7 4 l:t.al costs Black material.

2)

Ba rsov

-

S . Kaspa rov

Cairo 2001

Again White has managed to block the queenside. Black has no counter­ play since 24 ... c4? does not work: 25 b4! .l:.xb4 26 l:t.xb4 l::txb4 27 lDb5 wins an exchange. White will play lDa2, i..c 3, i..f3-e2 and slowly put his extra pawn to use.

A TTA CK WITH BLACK

2()8

4)

3)

B

Kota nj ian - Averianov

Cernousek - Krivoruchko

Moscow 2004

Olomouc 2006

With his last move, 18 a3!, White stopped the ... c4 idea for a long time. Black's queenside play has got bogged down; he can achieve nothing real. White has a lot of time to manoeuvre his pieces to better squares. 18 ..tf5 19 e4 ..tg4 20 h3 ..txf3 21 .il.xf3 tDc7 22 ..te2 .l:!.b7 23 'iti>g2 .l:!ab8 24 a4 liJcS 25 .tal White has defended b3 and is ready to occupy b5 and c4.

Even with a knight oddly placed on f4, putting pawns on a4 and b3 can sometimes provide White with an ad­ vantage. 17 JUb8 After 17 ... c4? 18 bxc4 ±, the b7bishop is hanging. 18 .il.d2 "ila7 19 "ilc2 White prevents ... c4, and once again Black can't do anything. 19 .il.c8 20 .il.c3 .il.xc3 21 'i!Vxc3 f6 22 tDd3 White has a large advantage.

•.•

••

•••

POSITIONS TO A VOID

5)

209

6)

Mala khatko - Avesku lov

Alushta 2001

With his last move, 25 lLlf3-d2, White defended b3, prevented ... c4 and prepared a pawn-storm in the cen­ tre. Black's main problem is that he lacks any active play. 25 �d4 26 f4 �f8 27 eS �g7 28 lLldl �xb2 29 lL!xb2 lL!b6 30 lL!bc4 lL!xc4 31 bxc4 The a6-knight is especially misera­ ble. 31...l:.ab7 32 �c3 lL!c7 33 .l:.abl White is much better. •••

This is a variation arising from the game Wojtaszek-Rasulov, European Ch, Budva 2009. White wants to play lLld3 and �c3 and turn his extra pawn to account. Black's only chance is to play ... c4 right now but it is not fully satisfactory: 2l ...c4 22 bxc4 .l:.xbl 23 'iVxbl 'iVxa4 (23 ... �xc4 24 lL!d3 ± and .l:r.c 1) 24 lL!d3 �f6 25 c5 ± with a clear extra pawn.

A TTA CK WITH BlA CK

210

8)

7)

Kota nj ian - C . Balogh

Ca rlsen - Calzetta Ruiz

Moscow 2008

Taormina 2003

Black has no pieces in the area bounded by b3-b4-d4-d3, and very lit­ tle activity. White simply puts his own pieces on desirable squares and real­ izes his advantage. 17 tl:lc4 tl:ld7 18 i.d2 tl:le5 19 b3 :bs 20 l:tbl 'i¥b7 Now 21 b4! would be the shortest path to victory: 21...cxb4 22 l:txb4 'i!i'c8 23 .l:.xb8 �xb8 24 tl:lxe5 i.xe5 25 l:tb3 'i!i'd8 26 a5 +-. In the game White played 21 aS and won in more drawn-out fashion.

Again White has played a4 and b3 and is well on the way to his ideal set­ up. 18 c4 Black tries the standard disruptive thrust, but White has a strong reply: 19 e5! cxb3 20 exd6 exd6 In the game Carlsen played 21 l:.e7, which is strong, but less convincing than 21 tl:lb5 ! tl:lc4 (21 ...tl:lxb5? 22 l:te8+ i.f8 23 �h6 with mate next move) 22 �g5 i.d7 23 i.xg7 �xg7 24 tl:lxa7 'i¥xa7 25 .l::.e7 +-. ...

POSITIONS TO A VOID

9)

21 1

1 0)

B

Here we see a situation where Black Ku ljasevic - Daskevics made the ...c4 advance, but it had little World Under- 1 8 Ch, Kallithea 2003 impact on the game. White's knight is ready to occupy the c6-square, the b3To be completely happy, White only pawn is safe enough, and Black has no needs to gain control of the b5-square. other sources of counterplay. Even For this purpose, 1 8 ..tfl ! ± is ideal. though it is Black to play, he is worse. After the exchange of the a6-bishop, White's knight will be free to move to b5. In the game White allowed some counterplay by 18 lt:lb5 ..txb5 19 axb5 Ita2, with a less clear position.

1 8 Tacti ca l Exe rc ises

Now it is your tu m to practice. Most of these positions are of a tactical nature (1-19), but it is up to you to decide whether you are looking for a concrete way to win, secure an advantage or maintain sufficient counterplay. Exer­ cises 20-25 feature positional ideas that are typical for the Benko Gambit. In every case, Black is to play. Please note that these exercises don't necessarily have neat and tidy solutions like you will find in tactical textbooks. In many cases White has counterplay and the outcome may not be clear, so just try to choose the best option for Black. If you wish, simply study each position and decide what you think about it, and then compare your answer with mine. All that mat­ ters is how much you learn that you can use in your future games. I hope that these last three chapters have helped to deepen your under­ standing of the Benko Gambit and that your Benko games will include posi­ tions suitable only for the 'dream' chapter!

2

B

TA CTICAL EXERCISES

213

13

B

TACTICAL EXERCISES

17

B

215

216

A TTA CK WITH BLA CK

21 7

TA CTICAL EXERCISES

Solutions 1) This is one of the most common tactical motifs in the Benko Gambit: 1 ltlxb2! 2 .ixb2 :xb2 3 'ifxb2 .ixc3 4 'ii'c2 .ixa1 5 .ixa6 l:.xa6 6 l::lxa1 ltlb6 After ... �b4, the a4-pawn falls, leaving Black with the advantage. 2) This is another version of the same idea. 15 l:txb2! 16 'ii'xb2 .ixc3 17 �b7 'ifa7! After this important move, White is really in trouble. 18 �xa7 18 �c6 f6! (to stop e5 forever) 19 .if4 �a4! +. Black forces an ex­ change of queens and a much better endgame. 18 J:lxa7 It's hard to find a good move for White. 19 a4 19 .ixe7? f6! 20 e5 fxe5 21 .ig5 .ixe1 22 ltlxe1 ltlb6 -+. 19 ....ixe1 20 ltlxe1 ltlb6 21 a5 ltlc4 22 a6 f6 After ... ltlc7 and ... .l:txa6 Black has a clear extra pawn and a technically winning position. •••

••.

••

3)

Tri ka l iotis - Toran Al bero

4)

This was the first game played by my South African student Fran�ois Oberholzer in the Benko Gambit. 18 ltlxd2 19 'ifxd2 l::lxb2! 20 'ii'xb2 .ixc3 21 'ii'c l .ixe1 22 'ifxe1 'ifxe1 + 23 l:.xe1 .ic4 24 .ifl .ixfl 25 �xfl 11xa2 Black has a practically winning rook endgame. Well done, Fran�ois! •.•

Ma rkos - Ponizil

5)

Czech Team Ch 201 1112

Sometimes even GMs are trapped by standard Benko tactics. 17....l:!.xb2! 18 .l:!.xb2 'ii'xc3 19 �xc3 .ixc3 20 .l:!.ab1 .ixb2 21 .l:!.xb2 f6 22 .id2 White has defended the a2-pawn and it might seem that a long endgame lies ahead. But Black denies him any hope: 22 .l:!.a4 23 �f3 �f7 23 ... f5? loses the lion's share of Black's advantage: 24 exf5 gxf5 25 .ig5 �f7 26 .l:le2! ltlf6 27 .ixf6 �xf6 +. 24 g4 White tries to prevent .. .f5 but it's already too late: 24 f5! 25 gxf5 gxf5 The d5-pawn is doomed and Black gets connected passed pawns and an easily winning position. •.•

•••

Averki n - Alburt

6)

Odessa 1974

Siegen Olympiad 1 970

21 .'i6xc2! The e2-rook is defended twice, but Black can also get a big advantage this is not enough to keep it safe! 19 ... 1ixb2! 20 �xb2 .ixc3 21 'itxc3 by 21 ...ltlxb2 but it provides White with an extra option: 22 :xb2 (22 �xe2 22 .:tel �xa2 .ixb2 'ifxc2 23 .ixc2 transposes to Black has a useful extra pawn. .•

A ITA CK WITH BLA CK

218

the game) 22 ... �xc3 23 �xc3 i.. xc3 24 .l:.c2 i..g7 25 i..d3 + and White re­ tains some defensive chances. 22 i..xc2 ltlxb2! 23 i..xb2 23 .l:.'!.xb2 i..xc3 24 .l:.xb8+ l:lxb8 25 i..d3 c4 26 i..c2 ltlc5 -+ with total domination. 23...i..c4 24 .l:tda1 .l::r.xb2! Clearer than 24 ... i..xa2 25 l:.xa2 l:!.xb2 26 .l:.xb2 .l:.xa3 !. 25 .l:.xb2 i..xc3 26 .l:!.b7 i..xa1 27 l:txd7 i..xe2 Black has an extra pawn and a win­ ning position. Laver - Benko

7)

Santa Monica 1 967

Black has a very pleasant game; the h4-rook is out of play, there is strong pressure on b2, but a concrete idea is needed for how to proceed. 27 i..a2! Black wins the pawn back with a better position. 28 ltlxa2 .l:bb2 29 .l:.xb2 llxb2 Objectively, Black's advantage may not be overwhelming, but in practice White's defensive task is rather un­ pleasant. 8) Black's pieces are well placed but it is not obvious how he should make progress. 18 'ii'xb2! Yes, sometimes even Black's queen takes on b2! 19 'ifxb2 Iixb2 20 l:txb2 i..xc3 21 :b7! The best try: this leads to a rook end­ game with some saving chances. 21 .l:.abl? is worse: 2l...i..xb2 22 .l:txb2 •.•

•.•

f5 ! (it often happens that a tactical se­ quence finishes with this pawn move) 23 e5 l:.a6! followed by ...ltlb6xd5 with a big advantage for Black. 21. i..xa1 22 l:.xd7 .l:!.xa2+ 23 'iii>f3 .l:l.a3 24 'iii>e2 'iii>f8 25 .l:!.d8+ 'iii>g7 26 �d7 26 .l::te 8? .l:[a7 +. 26 ... i..d4 27 i..xd4+ cxd4 28 l:r.xe7 'iitf8! Forcing the rook to leave the e4pawn unprotected. 29 .l:f.d7 l:txg3 30 �d2 .l:!.e3 31 .l:!.xd6 Uxe4 Black is much better; the only ques­ tion is whether he can win. .•

Averki n - M i les

9)

Dubna 1976

15 ltla3! This little piece of tactics does not regain the pawn, but changes the struc­ ture in a vital way: Black obtains a passed pawn. 16 bxa3 'ii'xc3 17 �xc3 i..xc3 18 l:.b1 c4 Black has the advantage. He plans ... l:tfc8 followed by a retreat of the c3-bishop and ... c3, etc. .•.

1 0) Ga rza M a rco - Perez M itjans Zaragoza 201 1

White had just played 26 i.e l-c3? to provide the b3-knight with a way to reach the c6-square. However, this nat­ ural strategic plan has a tactical refuta­ tion: 26 ltle3! 27 fxe3 'ikxe3+ 28 �f2 �xc3 Black has won back the pawn with a decisive advantage. •••

21 9

TA CTICAL EXERCISES

1 1 ) This is a variation from the game Polugaevsky-Seirawan, Amber Rapid, Monte Carlo 1993. The position could have occurred if White had tried a logi­ cal-looking simplifying manoeuvre that fails because of the following com­ bination: 22....1l.xb2 23 l:txb2 lDxa4! 24 bxa4 .i.xa4 25 'i¥xa4 l:[xb2 Already White must be thinking about his own safety. Zilberma n - G reenfeld

1 2)

Israel 1 995

White has just attacked the e5knight by 20 f2-f4?. However, this weakens the e3-square, which Black exploits in masterful fashion: 20 .l:.xb2! 20 ... ltJd7? would maintain merely sufficient compensation. 21 fxe5 This zwischenzug turns out not to help White. Or 21 .:t.xb2 l:.xb2 22 'i¥a4 'ifxc3 ! 23 fxe5 lDxe5 -+. 21. .1l.xe5 Not 2l ...l:i.xc2?? 22 l:[xb8+ .i.f8 23 .i.h6 +-. 22 l:[xb2 l:.xb2 23 'i!Vd3 ltJd2! This precise move practically finishes the game. 24 'ifxd2 24 'i¥e3 lDxfl 25 �xfl .l:1c2 -+. 24 Jbd2 25 il.xd2 Jl.xe2 26 lDxe2 'tWa4 Black is winning. ••.

•.

.•

least some advantage. But Black has different opinion on this: 15 lDd3! 16 exd3 'tWxc3 17 'i¥xc3 .i.xc3 18 ltJc4 In the game Black played 18 0-0?, allowing White to defend the a2-pawn by 19 .i.b2 i..f6 20 i..xf6 lDxf6 21 l:[fe1 l:[fe8 22 l:[b2 ;!;. The correct path is 18 ... .i.xc4 19 bxc4 lixa2 20 l:[b7 �d8 = intending ... �c8. .•.

.••

Hort - Alburt

1 4)

DeCin 1977

With his last move, 16 lDd2-fl, White attacked the d5-knight. But Black has no need to defend it: 16 ltJc3! A typical and in this case neces­ sary combination. The more natural 16 ... e6?! lets White win a pawn for insufficient compensation: 17 ltJe3 .i.b5 18 lDxd5 exd5 19 i.. xd5 .i.c6 20 i.. xc6 'i¥xc6 21 �c2 ;!;. 17 .i.xb7 .i.xb7 The point is that Black wins the b1-rook for free now. 18 'i'd3 il.e4 19 'i'e3 .i.d4 20 'ifh6 .i.xb1 Black has enough compensation for the queen. Later he even managed to win the game. 1 5) White simply wants to pick up the c4-pawn, so Black needs something radical in reply, and the following fits the bill: 1 .1l.xb2 2 lDxb2 After 2 'i'xb2? lDc5 + Black wins the pawn back and retains some pres­ sure. 2 lixa2! .•.

.••

1 3)

Neverov - Baklan

Ordzhonikidze Zonal 2000

White needs one more move (.i.b2) to cement his position and secure at

•.•

22()

A TTACK WITH BLA CK

I don't recall seeing this tactical motif in practice, but I'm sure it must have occurred somewhere. 3 .l:i.xa2 3 'iVxc4 does not give Black any problems: 3 ... .:1.xa1 4 .:1.xa1 tbc5 5 .l:i.a7 'iVb6 6 .l:i.a3 'iVb4 7 .l:i.a7 3 ...cxb3 4 'iVxc7 bxa2 Black is a knight down but his pawn is just one move away from queening. My analysis of this position didn't re­ veal any advantage for White. 5 tt:Jd3 Or 5 'iVxd7 'iVxb2 6 'iVxe7 'iVxf2 7 .l:i.c I 'iVe3 8 'iVxd6 (forced) 8 ...'iVxcl 9 'iVxb8+ r#ilg7 10 'iVe5+ Wh6 1 1 tbg5 ! 'iVxg5 12 'iVb2 with a drawish queen endgame. 5 'iVb5 6 tt:Jf4 'iVa4 7 tt:Je6!? a1'iV 8 .l:.xa1 'iVxa1 9 'iVxd7 fxe6 10 'iVxe7 exd5 11 'iVxd6 'iVa7 12 'iVxd5+ 'iVf7 This ending should be a draw. =.

..•

1 6)

Arlandi - Va"isser

Athens rapid 1 997

22 l:txb4? White immediately loses control of the situation, whereas a few moves earlier his game looked trouble-free. Better is 22 ..tc 1 ! .i.d4 followed by ... e5 with a rather unclear position. 22 tt:Jxb4 23 'iVb5 tt:Jc2! Black intends ...tbe1+. 24 'it>g1 24 'iVfl e5 25 ..te3 'iVe4 -+. 24 tt:Ja3 Black is winning. .•.

•••

1 7) Brun ner - Landen bergue Bie/ 1 994

White just needs to make practi­ cally any reasonable move to stabilize the situation. But it is Black to play, and in the Benko Gambit very often one move decides everything. 15 tt:Jxe4! 16 tbxe4 tbxd5! The point of Black's combination! Now there are many threats and White can't fend them all off. 17 tt:Jed2 17 tbel tbb4 18 'iVe2 .i.xa1 19 l:ha1 l:.xa2 20 l::txa2 'iVxa2 +. 17 ...tt:Jb4 18 'iVc4 e5! 19 .i.g5 e4 20 tt:Je1 d5 21 'iVe2 21 'iVc 1 tbxa2 22 'iVc2 tbc3 23 I!.xa5 'iVxa5 -+. 21. tbxa2! Black totally dominates. .•.

White's position looks favourable: everything is defended and he has the typical plan to strengthen his control of the queenside by a4, .i.d2-c3, tbb2, etc. But if you look more carefully, you see that the white queen and e2-rook are overloaded by the need to defend the a2- and e4-pawns. That just leaves the question of how to exploit this. 19 ...tt:Jxe4! 20 I!.xe4 I!.xa2 21 'iVd3 1 8) Aleksa nd rov - P. Cra m l i ng And what now? Black is a piece Erevan Olympiad 1 996 down, after all. It's easy to see that the last moves 21...t2Jxd5! were ... c4 and b4. It might appear that Black simply takes a second pawn there is no scope for tactics, but Black and thanks to his better coordination does have a way to smash through. maintains dynamic equality. 18 tt:Jd3! 19 .i.d2? ••

•..

221

TA CTICAL EXERCISES

Even after the superior 19 exd3 Black is better: 19 ... cxd3 20 'i!Vxd3 l:.xc3 21 'i:Ve2 'iVb7 + intending ... e5!. 19 tt::ldxf2 20 a4 i.xc3! 21 tt::lxg6 A desperate throw. 21 'iWxc3 tt::le4 22 'i!Vc2 tt::lxd2 23 'i!Vxd2 tt::le 3+ -+. 21 i.f6 Black is winning. •••

•••

1 9) Korchnoi - Baklan European Clubs Cup, Neum 2000

White's position looks a bit strange (more often a knight is on h3 and a bishop on g2) but it still appears solid enough. However, the following com­ bination totally destroys the position of a legendary chess-player. 17 i.xe2! 18 tt::lxe2 'i!Vxa2 The loose minor pieces on d2 and e2 are the problem for White. 19 'iWxa2 19 .l:tbcl tt::lf3+ 20 Whl tt::lxd2 -+. 19 .l:txa2 20 f4 llxd2 21 Wf2 21 fxe5 l:.xe2 22 exd6 exd6 -+ and ... tt::lxd5 with two extra pawns. 21...tt::ld3+ 22 'it>e3 .l::!.b2 23 llxb2 tt::lxb2 Black has a decisive advantage. 20) Black can't make serious prog­ ress on the queenside since White has blocked it with the a4 and b3 set-up. So instead he exploits another stan­ dard source of counterplay: undermin­ ing the d5-pawn. l...tt::le 8! 2 tt::le2 Or: a) 2 tt::lb5 f5 ! 3 exf5 gxf5 + and ... tt::lxd5. b) 2 i.d2 i.d4+ (2 ... f5 + followed by ... tt::lf6 is also possible, but Black's .••

•.•

idea is more ambitious - he wants to pick up both the e4-and the d5-pawns) 3 'iii>h l tt::lf6 4 g3 i.xc3 5 i.xc3 tt::lxe4 6 i.xb4 cxb4 7 l:r.a2 tt::lc 3 8 .l:!.d2 .U.a5 + and the d5-pawn falls. 2 f5 3 exf5 gxf5 4 tt::lg3 i.d4+ 5 'iii>h 1 tt::lg7! That's why the knight moved to e8 and not d7. The next move is ... tt::lxd5, and Black is somewhat better. •••

2 1 ) G rablia uskas - Khal ifman Vilnius 1 997

Here Khalifman applies a typical idea: he exchanges on c3, closes the centre and occupies the weak squares on White's queenside. 15 i.xc3! 16 bxc3 f6! This is an important move. Now White can forget about the e5 ad­ vance. 17 i.d2 'i!Va4! Black has good squares for his pieces: a4, b5 and c4; also, the a2pawn is extremely weak. White has no active plan, which is why I prefer Black here. 18 'i!Ve1 'i!Vc4 19 i.cl tt::lc7 20 a3 tt::lb6 21 h4 tt::lb5 22 lie3 'iWb3 23 h5 tt::lc4 Black is much better. In the subse­ quent play, White sacrificed material for insufficient compensation, but went on to draw thanks to his opponent's in­ accuracies. .•.

22)

Beliavsky - Khal ifman

Linares 1 995

This is another example from Alex­ ander Khalifman - one of the best Benko players. Black's pieces are

A TTACK WITH BLA CK

222

clearly active enough to compensate for the gambit pawn, but it is always important to choose the right moment for concrete measures. He could now play 20 ... .l:.b3, maintaining the pres­ sure, but the Russian GM decided on a plan to extend the g7-bishop's diago­ nal. 20 f5! First the e4-pawn is weakened. 21 f3 fxe4 22 fxe4 lLlxc3 23 lLlxc3 lLlb5! 24 lLlxb5 %1xb5 Now Black has strong pressure on b2. White must make a tricky deci­ sion: continue to defend passively, or give the pawn back and seek drawing chances in the rook endgame. 25 l:tc4?! It seems that passive defence was a better option. After 25 l:ie2 l:.b3 26 l:ta2 l:ta4 27 .ltf4 Black has merely sufficient compensation, and White should survive. 25 .txb2 26 .Ubl l:.ab6 27 a4 Ub3 28 aS?! 28 l:txb2 is better: 28 ....l:!.xb2+ 29 ..txb2 l:.xb2+ 30