Nicolai V. Pedersen
play the Benko gambit
EVERYMAN CHESS
Gloucester Publishers plc www.everymanchess.com First published in 2011 by Gloucester Publishers plc (formerly Everyman Publishers plc), Northburgh House, 10 Northburgh Street, London EC1V OAT
Copyright © 2011 Nicolai V. Pedersen
The right of Nicolai V. Pedersen to be identified as the author of this work has been asserted in accordance with the Copyrights, Designs and Patents Act 1988.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher.
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
ISBN: 978 1 85744 634 0
Distributed in North America by The Globe Pequot Press, P.O Box 480, 246 Goose Lane, Guilford, CT 06437-0480.
All other sales enquiries should be directed to Everyman Chess, Northburgh House, 10 Northburgh Street, London EC1V OAT tel: 020 7253 7887 fax: 020 7490 3708 email: [email protected]; website: www.everymanchess.com
Everyman is the registered trade mark of Random House Ine. and is used in this work under licence from Random House Inc.
Everyman Chess Series Chief advisor: Byron Jacobs Commissioning editor: John Emms Assistant editor: Richard Palliser
Typeset and edited by First Rank Publishing, Brighton. Cover design by Horatio Monteverde. Printed and bound in the US by Versa Press. Contents
Bibliography 4 Introduction 5
Benko Accepted 1 The Fianchetto Variation 12 2 The Classical Variation 33 3 The 'Quiet' 5 e3 73
Benko Declined 4 The 5 f3 Variation 85 5 White plays bS-b6 99 6 Other Options for White 114
Anti-Benko 7 The Kasparov Gambit 149 8 The Topalov Variation 1 73
Index of Variations 202 Index of Complete Games 207 Bibliography
Books 1 d4 Volume Two, Boris Avrukh (Quality Chess 2010) The Benko Gambit, Pal Benko (Batsford 1974) Dangerous Weapons: Flank Openings, Richard Palliser et al (Everyman Chess 2008) The Gambit Guide to the Benko Gambit, Steffen Pedersen (Gambit 1999) Play the Benko Gambit, Vaidyanathan Ravikumar (Cadogan Chess 1992) Squeezing the Gambits, Kiril Georgiev (Chess Stars 2010) Winning with the Benko, Byron Jacobs (Batsford Chess Library 1995)
Websites and periodicals Chess Informants 1-109 New in Chess Yearbooks 1-98
Databases/DVDs MegaBase 2011 The ABC of the Benko Gambit, Andrew Martin (ChessBase DVD)
4 I ntrod uction
The Benko Gambit is an aggressive and Gambit has been my own main reply to sound pawn sacrifice which has been 1 d4, and with decent results. The played by such prominent grandmas reader can, for instance, see my wins ters as Garry Kasparov, Veselin Topalov, against grandmasters P.H.Nielsen and Viswanathan Anand, Alexander Khalif V.Petrov, as well as the recent game man, to name but a few. Significantly, against V.Neverov in the Fianchetto Topalov has used it in many games, line. which says something about its Now I can claim that the Benko is soundness. It was the American sound. Unlike other more dubious grandmaster, Pal Benko, who found gambits (I could mention many here), it many new ideas in the 1970s and is based on long-term compensation. played the gambit successfully, after The old maxim that acceptance is the which it increased in popularity. Al critical test is probably especially true though there were others who ven for unsound gambits, however it still tured it earlier, they were not the main has truth in it and Black has to play inspiration for me; whereas a lot of the with some precision. ideas in this book are inspired by Benko and Khalifman in particular. Some typical ideas I have faced the Benko myself many and themes in the Benko times and feel that it is one of the more To justify the sacrifice of a pawn Black dangerous replies to 1 d4, as White has aims to seize the initiative and it is of to be careful not to cede the initiative ten a feature that Black has a slight to Black. I have played around 10 lead in development, which can lead to games with tt::lh3 in the Fianchetto material gain or some weaknesses for Variation, and maybe 20 games (if one White. In some cases further sacrifices includes rapid games) with the fash are called for, of which there are many ionable 10 l:[bl. My experiences also examples in this book. Indeed, the include the Benko Declined with 5 b6. Benko Gambit is a good way for Black Over the last few years the Benko to get the chance to win - and to win in
5 Play the Benko Gambit
style. Furthermore, endgames can •.• C5-C4 sometimes be quite good for Black, too, Although this is a committal move, it if there is still an initiative on the can be lethal if it supports a knight on queen side for instance. d3 after... ltJg4 -e5 or ...ltJd7 -C5 (or ... ltJd7- e5). The obvious drawback is that it Themes for Black weakens the d4-square; in particular ltJf3-d4-c6 is usually not a manoeuvre to ...e7-e6 allow. So after ltJf3-d2 or ltJh 3, for in The move ...e7-e6 is a common theme, stance, ...c5- c4 is more likely to be suc trying to break up the white centre. It is cessful. As examples I could mention usually more effective if White has Burmakin-5.Kasparov (Game 1), Aseev played committal moves such as f2- Ponomariov (Game 7) and Mancini f3,e2- e3 or f2-f4. Pal Benko often Degraeve (Game 18), and it is often seen played ...e7-e6 after some preparation in the analysis too. So ... c5-c4 is a theme in the middlegame, with both the idea which can be effective under the right of attacking the centre and sometimes circumstances. It can also be a way to of a later ...d6-d5 starting a rolling cen open filesafter a white b2-b3. However, tre of his own. So it is a theme even Black should be careful that White can when White hasn't created substantial not simply answer ...c5- c4 by b3-b4, fol weaknesses. There are many instances lowed by a2-a4, with two strong con of this in the book. It should be noted nected pawns on the queen side. that thanks to the missing white c4- pawn the breaks ... e7-e6, as well as ...f7 - An example is Aseev-Ponomariov f5, are more likely to be effective. (Game 7):
One example, from Comas-Topalov (Game 13):
Here the white knight on h3 can't reach c6 sufficiently quicKly and there is no b3-b4.
6 Introduction
... tt:Jd7-e5 (or ...tt:Jg4-e5), seen many times in the games and
••• tt:Jd7-C5-d3, or •••tt:Je8-c7-b5-d4 analysis as leading to at least equality In some case"S the knight on f6 is hin for Black, it should be considered very dering the dark-squared bishop and carefully, with the possible gain might seek another post via d7 or g4, weighed against the weaknesses it while the thematic ... ttJe8-c7-bS(-d4) is leaves around the black king. also sometimes effective. Some of the key squares which can be good for the From Neverov-Pedersen (Game 3): knight are eS and (especially) d4, c4 and d3 .
.. :iVa5, ••. 'iVb6, •..'iVC7, or .•• 'iVa7 The move ...1Wa s is especially thematic, followed by . ..l::tfb8. Sometimes the queen might go to b6 where it is sup porting the d6-pawn after the ... e7-e6 break, while in some cases f2 might be weak. The squares c7 and a7 are usu ally less ideal, but in the Benko De clined (with bS-b6) the queen often The a2-pawn is won and there is ends up here because of ttJd2-C4. further pressure on the light squares: Obviously the various pawn moves ...i.b 3 as well as ...ttJb 6-a4 (or ...ttJx dS) and knight and queen manoeuvres are possibilities. should be combined appropriately and require some timing and precision. ••.f7 -f5 The .. .f7-fS break can be quite risky,
••• .i.xc3 weakening the black king and the e6- This move does indeed seem very square in particular. So in the middle committal, giving up the good bishop game circumstances often have to be on g7. In some cases, however, it en favourable for Black to play ...f 7-fS. In ables Black to regain the white a-pawn the endgame, on the other hand, it is (usually on a2) with further threats. very thematic and is seen quite often, Some examples are seen in the games especially when Black is pressing for a and notes to Neverov-Pedersen (Game win. Then the ds-pawn can be weak as 3), Cossin-Vachier (Game 8), Lugovoi it already lacks the support of a pawn Khalifman (Game 9), and Piket-Topalov on c4. (Game 10). Essentially, ...i.xc3 is a very An example is Cossin-Vachier concrete move and, while it can be (Game 8):
7 Play the Benko Gambit
A classical formation or Themes for White a more rapid approach The thematic and slightly slow set-up The blockade: a2-a4, b2-b3 with ...liJb d7, .. .'iWas, ...0-0 and ... .l:.fb8, and/or a piece on bS possibly followed by ... liJc7-bs, is often In the Benko Accepted one of the cen a good plan. However, I recommend tral themes is a white blockade on the that Black keeps a good degree of flexi queenside. A big part of Black's strat bility in the Benko. For instance, egy is counterplay on the queen side, so against the Fianchetto Variation, one of White's plans is to play a2-a4 9 ...liJfd7 with the idea of ...liJb 6-C4 and and plant a piece on bs, often prepared .. .'�as (and ...liJd7-b6) is more effective by moves like "C2 and b2-b3. Naturally than the classical formation. Against Black will try to hinder these plans. For White's plan with f2-f3 I suggest a instance, the move b2-b3 weakens the quick ...e7-e6 in reply, which leads to a a3-square, while ...cs -c4 is sometimes very sharp and unclear game. And more dangerous (this is mostly charac against some of the less correct lines, teristic of the Fianchetto Variation). such as 4 �gs in Grivas-Khalifman The theme also occurs now and then (Game 32), Black plays ...liJe4 an d ..."as in the Benko Declined, for instance in and even ... g7-gs (see fo l1 owing dia the f2-f3 and e2-e3 lines if Black plays a gram), gaining time, and is already bet later ...a6x bs. On the other hand, in the ter from the early opening phase: bs-b6 lines I mostly do not recommend With the gambit Black is trying to playing ... a6-as, as the bs-square is then obtain a lead in development (amongst a serious weakness. other things), so slow moves by White In general the blockade on bs com are often a good reason to sharpen the bined with the e4-e5 break (see below) play. is something Black should make a seri ous effortto prevent.
8 Introduction
M.Gurevich-Z.Azmaiparashvili, Stras bourg 1994. White has a comfortable advantage (see the notes to Game 10).
i.g5, i.f4 As with .if4 in the previous note, .ig5 often leaves the b2-pawn weak. There fore .ig5 makes particular sense when it attacks an undefended e7-pawn, af ter ...ltJe8 for example. On the other hand, under some circumstances Black An example of a good blockade for can consider leaving the pawn en prise White. This can arise after inaccurate if there's the possibility of ...f7-f6, trap play by Black in the notes to Game 14. ping the bishop on e7, in which case .ig5 might just be met be ... h7-h6. The e4-e5 advance Another key theme for White is the e4- From Campos-Topalov (Game 25) e5 advance. Sometimes this needs to be prepared, as in the Fianchetto Variation, whereas in the Classical main line it can often be played sooner. The move e4-e5 does have drawbacks; for instance the d5-pawn may become weak, while pre paring it with .icl-f4 can leave the b2- pawn loose. However, if White has good control, then e4-e5 can be effective.
White has just answered Black's mistimed ...ltJf 6-e8 with .icl-g5, at tacking e7, and Black's best idea may be simply to put the knight back on f6.
Prophylactic moves (such as l:.e2, �C2, 'ii'c2, h2-h3, b2-b3 or a2-a4) In the Benko, as in chess in general, a prophylactic approach is often worth This position arose in the game considering and l:te2, l:tc2, (etc)'iVc2 are
9 Play the Benko Gambit
all moves which Karpov has used with Benko nonetheless with 2 ...cS 3 dS bS, success in the Classical main line. How intending 4 c4 g6 S cxbS a6. If White ever, I think the recommendations in plays 4 �gs instead we have a system Chapter Two show how to deal with which Topalov has used with success. these ideas, so the Classical is no longer Since Black has played the ambitious that popular. ... b7-bS, I am not in favour of allowing The modern .l:.bl in the Fianchetto �gsxf6, compromising the pawn Variation is also prophylactic in nature, structure further. So here the recom as White takes steps (with b2-b3 and mendation is 4 ...ttJe4, involving ideas �b2) to prevent a later ...cS -c4, ...'iWa3 like ... 'iYb6 (or .. "iVas) and ... g7-g6, or ... ttJC4 arising from Black's classical ...�g7. In general the concept remains formation with ...ttJb d7, ...0-0, ...'iWas active play. The struggle is often quite and ... l:r.fb8. In fact this idea has been positional, but it can quickly turn and the most serious threat to the Benko in be very tactical. In either case Black is the last decade or so, which is why in fineas long as he knows what he is do Chapter One I suggest that Black take ing. preventative measures in turn by play Another anti-Benko system is 1 d4 ing 9 ...ttJfd7 and 10... ttJb 6. ttJf6 2 c4 cS 3 ttJf3, refusing to push the d-pawn, and transposes to a line of the English Opening. This time Black can sharpen the play with the Kasparov Gambit: 3 ... cxd4 4 ttJxd4 es!? S ttJbS dS 6 cxdS �cS. I think this is a sound gam bit as well. One of the critical variations is 7 ttJSC3 0-0 8 g3, as played by the new guy in the 2800+ club, Lev Aronian, and which he used to beat 5utovsky. In re sponse I recommend that Black follow 5utovsky with 8 ..."i!Vb 6!?, forcing 9 e3, White avoids the Benko when the follow-up 9 ...Sl.g4 more or Although there are other more or less less prevents the plan of �g2 and qualified ways to test the Gambit, my leaves White with weaknesses on the conclusion is that Black is doing fine light squares (especially c4, d3, f3 and from a theoretical perspective. So what h3). Furthermore, Black has a lead in if White avoids the Benko? development which can be used for the One approach is with 1 d4 ttJf6 2 piece sacrifice ...�d 4!? as -a way of at ttJf3. In this case Black can aim for the tacking the exposed ds-pawn and the
10 Introduction
white king, as in Matlakov-Evdokimov ttJf6 2 ttJf3, but these cannot be re (Game 41): garded as within the scope of this book, particularly if White is aiming more for a flank approach than direct occupa tion of the centre. In conclusion, I hope that readers new to the opening will be inspired to try the Benko Gambit for themselves, while existing Benko players will find some new ideas enabling them to play the Gambit more successfully. The Anti-Benko chapters might even prove useful as part of another repertoire, There are many new ideas in these such as the Benoni, Nimzo-Indian, lines and I think Black is doing fine Queen's Indian, Bogo-Indian or Grun against g2-g3. White's other options feld, for instance. are less critical, so Black should be able to prove compensation there too. Nicolai Pedersen, Other ways for White to avoid the Aarhus, Denmark, Benko include 1 c4 and 2 g3, or 1 c4 June 2011.
11 Play the Benko Gambit
Chapter One The Fianchetto Variation
1 d4 tLlf6 2 c4 cS 3 ds bs 4 cxbs a6 S strong bishop on g7, and the other bxa6 g6 6 tLlC3 .1i.xa6 7 g3 d6 8 ..tg2 pieces have good potential to be very ..tg7 active on the queen side, particularly on the half-open a- and b-files. Mostly Black has followed the standard plan in this position: ... 0-0, ...'iYa s, ...l:tf b8, ... tLlg4-es (or ... tLle8-c7-bS), etc. How ever, 10 l:tbl and the ideas behind it (which are typical for the variation) have proved quite effective against this set-up. And other books which cover the same subject have failed to show solutions for Black. 50 the reader might ask, what This is the critical position for the should Black do against 10 :bl - ? Well, Fianchetto Variation, which is itself one my suggestion is 9 ...tLlfd7, aiming to of the most important lines of the gain counterplay earlier in the middle Benko Gambit from a theoretical view game with moves such as ...tLlb 6, point. Now 9 tLlf3 tLlbd7 10 l1bl brings ...tLl8d7, ...0-0, ...tLlC4 and ..:iWa s. Th e about perhaps the most serious chal main idea is to plant a knight on the lenge to the Benko. Here White is aim c4-square before White has managed ing for two main strategic plans: a to prevent it by 'iVC2, ..td2an d b2-b3. blockade on the queen side with moves The conclusion is that Black is like 'iYC2, b2-b3, a2-a4, and placing a achieving an unclear game here. This knight or bishop on bS; and the central line has been played many times with advance e4-es. success by GM 5ergey Kasparov, who From the other side, Black has a wrote an article on it in New in Chess
12 The Fianchetto Va riation
Yearbook #94. In my recent game Benko Gambit, so much so that I think against Neverov (Game 3) I was also Black should avoid it altogether. happy to test this idea and obtained, first, an unclear game from the open ing, and then the better position, all without too much effort at the board (as most of the moves were prepared). In general, it is said that acceptance is the critical test of a gambit, and with the Fianchetto Variation that is some how confirmed, as this has been the main theoretical focus in recent years, and will probably continue to be so. With this in mind, and as theory will With 9 ...tDfd7 Black intends to play inevitably continue to evolve, the ...tDb 6, from where it can quickly jump
reader is advised to follow develop to c4 (10... tDb 6, 11 ... tD8d7is the typical ments in this variation carefully. At the follow-up). By sending the king's knight moment, however, it is up to White to on this journey, Black uncovers the fi come up with something new, and anchettoed bishop on g7, which means that, I think, makes quite a change. that White's prophylactic move b2-b3 As the variation with 9 tDf3 tDfd7 is is itself prevented. White will usually the most important in this chapter, we try to prepare it anyway with Wc2 and will examine it straightaway in the first Jt.d2,but as we will see, this can be an four games below, before going back swered by ... tDC4 and ...Wa s. and considering earlier moves and pos Sergey Kasparov wrote an opening sibilities in Games S-7. survey of this variation in New in Chess Yearbook #94. I will use a few of his games in this chapter, while some of Game 1 the analysis is also inspired by his arti V.Burmakin-S.Kasparov cle. Deizisau 2003 10 0-0 The immediate 10 'iYc2 is seen in 1 d4 tDf6 2 tDf3 cS 3 ds d6 4 c4 bs S the next game, while other moves are cxbs a6 6 bxa6 Jt.xa6 7 tDC3 g6 8 g3 covered in the notes to Game 4.
Jt.g79 Jt.g2tDfd 7 10... tDb6 To reiterate what I've already said, Before castling Black continues to
9. . . tDbd7 10 .l:f.bl is one of the most - if mobilize his pieces on the queen side. not the most - critical tests of the 11 'iYC2
1 3 Play the Benko Gambit
The queen is usually placed on c2 in still seems to be equalizing. For in this line. stance: 13 ...liJC4 (a standard move in 11... liJ Sd7 this variation; the knight is quite well placed here) 14 lbcl 'ifas 15 lLld2 liJdb6 16 liJxC4 liJxc4 17 b3 liJa3 (show ing another difference from the 9 ...liJbd7 10 l:tbl main line: White's plan of b2-b3, a2-a4 and blocking the queen side is more easily prevented in the 9 ... liJfd7 line) 18 'ii'd2 c4 19 .i.h6 l:tfb8 20 .i.xg7 'it>xg7 21 bXc4 liJxC4 22 'iWd4+ liJes 23 'ife3 11c8, when Black is quite active and has enough compen sation for the pawn. A basic position for the 9 ...liJfd7 line. Alternatively, 13 b3 is covered in the Here White has a few different paths. next game, while 13 .l:t.bl is examined In the current game Burmakin tries in Games 3 and 4. going forward in the centre, but I don't 13 .••liJ C4 think this should cause Black any prob lems. 12 :d1 0-0
14 liJd2 14 .i.f4 can be tried here too, though Black is again doing fine. A pos 13 e4?1 sible continuation is 14 .. :iVas 15 .l:[acl This allows Black to get a rather i..xc3! 16 'ifxC3 -.xC3 17 bxC3 liJb2 18 comfortable game by simply playing .l:t.d2 liJd3 19 .l:t.cdl liJxf4 20 gxf4 liJf6, ...liJC4. when Black is slightly better and is set Instead, 13 .i.f4 might be a better to win at least one of the weak pawns attempt to get an edge, although Black on e4, f4, a2 or ds. So it seems that 13
14 The Fianchetto Variation e4, allowing the knight to jump to c4, Stronger was 22 ..:ii' a5! and Black means a comfortable game for Black. keeps a slight advantage - White is 14 ...tDxd 2 quite uncoordinated. Black is also fine after 14 ...tDd e5!? 23 Iidbl? intending 15 tDxc4 tDxc4 and ...Va5 23 l:.xb8! equalizes. and sometimes ...l1fb8 with pressure 23.. Jlxb4 24 l:ixb4'iVC 51 25 tDb5 on the dark squares, or just 14 ... tDce5 15 tDf3 tDC4with a repetition of moves. 15 .i.xd2 tDe5
25 ...tDd 3? Overlooking a simple combination: 25... .i.xds 26 'iVxc5 dxc5, regaining the Black has solved all his opening pawn with good winning chances. problems and it is White who should Black is more active and the c-pawn is think about equalizing. Notice that more dangerous than the white a ...c5 -c4 and ...tDd3 is now an effective pawn. plan. Often in the fianchetto variation 26 Ubi tDxel 27 .l:txel .i.xb5 28 'iixC5 White can respond with tDf3-d4-c6, but lixC5 29 axb5 l:r.xb5 30 .i.f3 l:tb2Vz-Vz obviously that is not a theme here. 16 .i.el'ili'b6 17 l:tablc41 18 b4 cxb3 19 Game summary 'iVxb3'iVa 7 20 a4 .i.C421 'iVc2?1 Burmakin played unambitiously with 21 'iVb7 would maintain the bal the white pieces and Sergey Kasparov ance. managed to equalize. Later on Black 21... .:tfc 8 even had the slightly better position, Now Black has the better position, though after inaccuracies the game due to his good activity and well placed ended in a draw. In conclusion, the line pieces. The bishop on g7 and the rook with .lir.dlan d e2-e4 is not a theoretical on the open c-file are working espe problem for Black, whose knight cially well together. shouldn't be allowed to go to c4 so eas 22 lib4l:!a b8?1 ily.
1 5 Play the Benko Gambit
played in lVeng-R.5kytte, Copenhagen Came 2 2010, when 15 ... 'ii'a5 16 ltJxC4.txC 3+ 17 A.Goganov-S.Kasparov .td2 ii.xd2+ 18 ltJxd2 l::tb8 19 b3 ltJe5 Peterhof 2008 gives Black compensation - Skytte) 15 ...'i1t'a5 16 a3, when Avrukh rates 1 d4 ltJf6 2 c4 cS 3 ds bs 4 cxbs a6 S White's chances as preferable, men bxa6 g6 6 ltJC3 ii.xa6 7 g3 d6 8 .tg2 tioning the plan of l:ia2,mo ving the c3- .tg7g ltJf3ltJfd 7 10 'iWc2 knight, followed by b2-b3. But I don't Since Goganov follows this with 11 agree White is better here! After 0-0, it just leads to the same position as 16 ...ltJd b6, I think Black has definite in the previous game. However, the compensation whether White plays current move order does give White Iita2 or any other move: another option (see the next note).
10... ltJb6
a) 17 :a2?! (this move is dubious, in my opinion, as the rook is offside here) 11 0-0 17 ...ii.xc3 ! 18 'ii'xc3 (or 18 bxc3 ltJa4 19 In his book 1 d4 Volume Two Avrukh l1al .tc8 20 ltJe3 ltJxe3 21 .txe3 'iixc3 recommends 11 h4 here. Black has to 22 'iWe4 'iWb2 and Black is slightly bet react precisely as there is some danger ter) 18 .. :iVxc3 19 bxc3 f5 20 ltJe3 ltJxe3 on the king side with the knight away 21 .txe3 .tC4 and Black regains the on the other side of the board: 11 ...h6! pawn with a slight advantage. 12 ltJh2 ltJ8d7 13 hS (after 13 ltJg4h5 b) 17 :bl (a better move than l:.a2) 14 ltJe3 .td4 15 .td2 ltJf6 "Black has 17 ... .txc3 18 bxc3 .tc819 ltJe3ltJxe3 20 nice play" - Avrukh; a possible con ii.xe3 ltJc4 21 ii.cl ltJxa3 22 ii.xa3'ikxa3 tinuation is 16 0-0 0-0 17 b3 ltJbd7with 23 f4 (or 23 l:tb60-0 24 .:r.fbl 'iWa4is just compensation for the pawn) 13 ...g5 14 equal) 23 ...'iIi'a2 24 .l:tb2 'ii'C4 with an ltJg4ltJC4 (continuing with the plan of unclear game. ... ltJC4 and ..:tIi'a 5) 15 0-0 (15 ltJe3 was c) 17 ltJe3 .txc3 18 bXc3 'iia4 19
16 The Fianchetto Variation
'iVb 1 ttJxe3 20 i..xe3 litb8 21 liIe1 i..C4 with compensation. 11... ttJ8d 7 12 �d1 0-0 13 b3
19 'ii'd2 llfa8 And here 19 ...lIa3 ! with the better chances. A straightforward attempt to pre 20 a4 c4 21 'iid4 'iid8 22 bxc4 lba4 23 vent the further ... ttJc4. But now we see lIxa4 ]:txa424 litb1ttJbd7 one of the reasons for preparing b2-b3 Despite the reduction in forces with litb1 in these lines. Black's game is still favourable, due to 13... i.. xe2! the weak c4-pawn. A tactical trick, exploiting White's 25 l:Ia1?! 'ii'b6 26 'iVC3 ':'xa1+ 27 'iixa1 vulnerability on the long diagona1. ttJe5 Black equalizes quite easily. 14 'ii'xe2 i..xc3 The position is roughly balanced. Black has the superior structure, a strong dark-squared bishop, and good chances of activity in the middlegame. White has the bishop pair. 15 i..b2?! See the note just above. It is more logical to play 15 .l:.b1, keeping the bishop pair; e.g. 15... i.. g7 16 a4 lite8 17 i..f4 ttJf6 with chances for both sides. Now Black is clearly better with his 15 ...i.. xb2 16 'ii'xb2 ttJf6 17 ttJg5'iVd 7 18 well-placed and active pieces, while the h4?! l:la7 white c-pawn can no longer be de Stronger was the alternative fended. 18 ... l:tfb8!, when White has a difficult 28 'ii'e1 ttJfg4 29 'ii'e2 'ii'd4 30 i..h3 h6 position. 31 ttJe4f5 32 i..xg4 fxe4 33 i..e6+ 'ii;>g7
1 7 Play the Benko Gambit
34 'ii'u?? strong opponent. A blunder in a difficult position. 10 0-0 tbb611 �C2 tbSd712 l::td1 0-0 13 34... 'ii' d1+ 35 �g2 tbf3 36 'iVb2+�h7 37 :b1 tbc4 'iVC3'iVg1+ 0-1 This position can also be reached via 9 ...tbbd7 10 .l:tbl 0-0 11 0-0 'ii'as 12 'iVc2 Game summary tbb6 13 .l:tdl tbc4, but this move order White played the set-up with 'iVc2 and allows White the strong alternative 12 .l:r.dl and we quickly saw why this �d2. Hence I recommend 9 ... tbfd7. should be followed by 13 l:lb1. The 14 tbd2 game continuation 13 b3 allowed I don't think this move is critical 13 ... �xe2!, after which Black has no from a theoretical standpoint. More problems. Once Goganov exchanged interesting is 14 b3, which is examined the bishop on b2, he could no longer in the next game (Gallardo-Ortega). count on the bishop pair to balance 14... 'iVa s 15 tbxc4�xC4 16 �d2 Black's superior structure. After a fur ther few inaccuracies and a final blun der Sergey Kasparov won the rather one-sided game. A more critical continuation was seen in the notes to White's 11th move: Avrukh's recommendation of 11 h4 - though, as shown in the analysis, Black should still be able to obtain an unclear game at least.
Here the game C.Horvath-M.Pap, Game 3 Pula 2003, ended abruptly 16 �gs .l:r.fe8 V.Neverov-N.V.Pedersen and a draw was agreed in this unclear Dieren 2010 position. I had also looked at the text move and concluded that it is okay for 1 d4 tbf6 2 c4 cS 3 ds bs 4 cxbs a6 5 Black. bxa6 g6 6 tbc3 �xa6 7 tbf3 d6 S g3 �g7 16 ...i.x c31 9 �g2 tbfd7 I saw in my preparation that this is I was happy with Neverov's choice unclear. It might look odd to give up of variation as I had already made the g7-bishop, but it is justifiedby the quite some study of 10 .l:tbl ideas and gain of the a2-pawn and further 9 ...tbf d7, and could now show in prac threats against dS, with' ideas like tice that Black is fine, even against a ...�b3 and ...tba4.
18 The Fianchetto Variation
Instead, E.Bacrot-P.Tregubov, Paris position via the 9 ...ttJbd7 move order 2006, continued 16 .. :iVa6!? 17 a4! i.xe2 given above) saw 19 h4 f6 20 h5 g5 21 18 ttJxe2 'ii'xe2 19 'uel'ii' a6 (19 ..:ji'g 4?! ..te4i.b 3, when White elected to force 20 b4 was good for White in a draw by 22 i.xh7+!? 'ittxh7 23 'iVd3+ H.5tefansson-R.Lagerman, Icelandic �h8 24 'ii'g6 i.xdl 25 'iVh6+ 'ittg8 26 Championship, Reykjavik 2007) 20 'iVg6+ 'itth8 27 'iVh6+ 'ittg8 28 'iVg6+ Y2-Y2 I!xe7ttJe5 21 b4 i.f6 22 b5 (more accu A.Yusupov-P.Tregubov, German League rate was 22 %:tC7! i.d8 23 b5 'ii'xa4 24 2006.
'ii'xa4 l:txa4 25 l:tb7, when White is 19•.. f6 20 'ii'e3 I!f7 21 b3!? i.xb3 22 slightly better due to the bishop pair lildclttJxd S 23 'iVf3 and the strong b-pawn) 22 ..:ii' xa4 23 So Neverov seems to be planning to 'ii'xa4 l:txa4 24 l:tb7 l:fa8 (now Black win some material based on a weak has some activity and the position is ness on the light squares. I was not too unclear) 25 i.C3lla3 26 ..txe5 ..txe5 27 worried about this during the game b6 I!al 28 I!xalI!xal+ 29 i.fl l:lbl 30 though. ':'b8+ 'ittg7 31 b7 c4 32 'ittg2 i.d4 33 23... ttJx C31 24 'it'xC3 'ub82S 11b2'ii' a4 26 ..txc4i.a7 34 I!e8I!xb7 35 I!e2Y2-Y2. l:tcbl?! 17 i.xc3'ii' xa2 18 'iVd2 ttJb6 26 i.d5 i.xd5 27 l:txb8+ �g7 was perhaps a better way to win the ex change, although White has no advan tage. 26... C4
The position is unclear. The bishop on c3 pointing towards the black king might seem threatening, but 'iVh6 can always be answered by ...f7-f6 . Mean while Black has queen side counterplay, Now Black is slightly better with particularly against the weak pawn on two pawns for the exchange (after d5 and the threat of ... ttJa4. i.d5Xf7), while the white rooks are 19 'ii'h6 quite passive. In addition Neverov was An earlier game (which reached this in time trouble. So quite a good situa-
1 9 Play the Benko Gambit
tion for me; the main worry here was a 62 fxg4 'ifxd5+ 63 �g3 'ii'e5+ 64
3S ...'iVb4 might provide more win 68 . . . es. Then it would be possible in ning chances, but the risk is also a bit many cases to place the queen on f4 greater. I saw this during the game, but and use the eS- and d4-pawns as shel thought that the queen endgame of ter from the checks. fered reasonable winning chances too. 69 'ii'd5 Ve3 70 �f3 Ve2 71 Vg5+ �f7 36 'ii'xe3 .txd5 37 exd5 Vb1 72 'iff5+ �e8 73 �e3 'ife2+ 74 �d4 e5+ 75 �d5 'ifU+ 76 �e4 d2 77 'ii'xe5+ �f7 78 'ii'e7+ �g6 79 Vd6+ �f7 80 "fid7+ �f6 81 �e3 Ve2 82 'ii'xd2 'ilkxd2+ 83 �xd2 �g5 Yz-Yz
Game summary With 9 ...liJfd7 I had the chance to test my recommendation against a strong grandmaster. In the opening phase Neverov didn't play the theoretically most testing line, and the thematic
Unfortunately I didn't win manage 16. ...lixc3 ! led to an unclear position in to win this endgame, despite being a which Black was not worse. In the mid clear pawn up. Since the opening dlegame Neverov played sharply, trying phase is the main focus here, I will go for an advantage after all, but an inac quickly through the rest of the game. curacy gave me the slightly better posi 38 'ii'd4 Ve1 39 h4 'ii'e5 40 'ifd2 �g7 41 tion. Eventually I obtained good win
20 The Fianchetto Variation bxa6 g6 6 tDc3 .ixa6 7 g3 d6 8 .ig2 .i.g7 9 tDf3tDfd 7 10 0-0 Apart from 10 0-0 and 10 'iWC2, White has tried two other moves in this position: a) 10 .i.d2 is adifferent way to aim for the desired set-up with b2-b3, a2- a4. The drawback is that the d5-pawn is temporarily vulnerable, which Black can exploit by playing the sequence 10 ...tDb6 11 b3 .i.b7!, and if White de fends the pawn with 12 e4, then the all) 17 f3 (defending the pawn) very thematic return 12 ....i.a6 hinders 17 ...'ii' d7 18 ..t>g2 h6 19 tDh3 e6 20 dxe6 his castling short. tDxe6 21 'ii'bl tDd4 22 'iWd3 g5 and Black's activity, together with the poor white knight on h3 and exposed king on g2, gives him enough compensation for the pawn. a12) 17 'ii'e2 'ii'd7 18 a4 h6!? (alter natively, 18 ....i.x c3!? 19 .i.xc3 fxe4 20 'ii'xe4 tDbxd5 transposes to line 'a3' below) 19 as tDbxd5 20 tDxd5 tDxd5 21 exd5 hxg5 leads to an unclear position.
Now White has to play rather artifi cially, returning his fi anchettoed bishop to its original square, simply in order to get the king to a safe location away from the middle of the board: al) 13 .i.fl0-0 14 .i.xa6 tDxa6 15 0-0 f5 ! (afterth is White has some difficulty defending the d5-pawni instead 15 ...c4 16 .l:.bl tDC5 17 'ii'e2 cxb3 18 axb3 was slightly better for White in D.Hou For instance, 22 1:[a4 1:[fb8 23 'ii'C4 sieaux-V.5olodovnichenko, Condom g4 24 l:tel :a7 and 22 .:ta3 f4! both of 2005) 16 tDg5 tDC7 with the following fer Black counterplay and/or equality, possibilities: so White might be tempted by the ex-
21 Play the Benko Gambit change sacrifice 22 .i.xgS!? i.xal 23 ':xal, when he has a strong a-pawn and possible attacking chances against the king on g8. On the other hand, after 23 ...l:tf e8, intending ...e7-e6(es), Black can look for counterplay in the centre, while White also has a slight problem on the back rank. A possible continua tion is 24 a6 eS 25 dxe6 ':xe6 26 'iia2 dS 27 a7 d4 28 'iVas'iVc6 29 h4 and the game remains unclear. a13) 17 'ii'C2 (this gives Black fewer Black should reply 10 ...tiJb 6 11 'ii'cl options than 'iVe2) 17 ... 'iVd7 18 a4 fxe4 h6 (weakening, but safeguarding the (18 ... h6?! would be dubious here, due g7-bishop) 12 0-0 (as in Y.Yakovich to 19 as tiJbxds 20 tiJxdstiJxdS 21 exds S.Kasparov, Russian Team Champion hxgs 22 i.c3 and White's chances are ship 2004), and now 12 ...gS! with com preferable) 19 "ii'xe4 i.xc3 20 i.xC3 pensation. For instance: 13 i.d2 tiJ8d7 tiJbxds 21 'iVh4 hS! 22 i.d2 "g4 23 14 l:tdl (or 14 'iVC2tiJC4 15 i.el'iVb6 16 'iixg4 hxg4 with equal chances in the tiJa4 'iWb7 17 i.c3 tiJf6 18 %bcl 0-0 19 endgame. ':'fdl lIfb8) 14 ... tiJC4 15 i.el 0-0 16 b3 a2) 13 litcl (a waiting move) tiJces 17 'iVc2 tiJxf3+ 18 i.xf3 fS 19 13 ... tiJ8d7 (Black develops and keeps llacl 'iVe8 20 i.d2 tiJes 21 i.g2 'iWhs the balance) 14 a4 (14 i.fl i.xfl 15 and Black has a strong initiative. 'iii'xfl 0-0 16 a4 comes to the same 10 ...tiJb6 11 "iVc20- 0 thing) 14... 0-0 15 i.fl i.xfl 16 'iii'xfl 11 ...tiJ8d7 12 l:tdl 0-0 is the move was played in V.Epishin-M.lvanov, Arco order I recommend. 2002, and now instead of 16 ...tiJf 6?! 12 ':'d1tiJSd7 13 l1b1tiJC4 Black should prefer 16... C4 17 b4 i.xC3 18 i.xc3 tiJxa4 regaining the pawn. Fo11owing 19 i.al White's bishop mig ht look powerful, but Black can block it off with the c-pawn if neces sary, so the game is just unclear after 19 ...'iVb 6. b) 10 i.f4 is another possibility, which has the idea of continuing "ii'cl and i.h6, exchanging the strong bishop on g7.
22 The Fianchetto Variation
14 b3 l:tdcl l:ta3 19 .i.xe7l:te8 20 .i.xd6.i.xd5 As we saw in the previous game, 14 21 e3 with a slight plus for White in his ttJd2is less critical for Black. New in Chess article. I agree with this Instead, 14 .i.h3 was tried in evaluation, but Black might prefer L.Schan dorff-R.5kytte, Danish Team 17 ....i.f6 ! 18 .i.xf6 ttJxf6 19 ttJd2 .i.c8 20 Championship 2011. From h3 the .l:1b2 l:ta321 e4 .i.d7with a likely draw. bishop exerts pressure against d7, hin 15... Ji.b5 dering Black's intended ...'iWa 5. The Now Black is very active and White drawback is that d5 becomes weaker, has some tactical problems. This is in so a logical reply is 14 ... ttJce5. For in deed something to aim for from a Black stance, 15 ttJxe5 ttJxe5 16 b3 'iWa5 17 perspective and shows that White of iob2 ioc818 .i.xc8lIf xc8 19 .i.alc4 and ten doesn't know to react to an early Black has compensation with good, ...ttJd7-b6-C 4. active play. 16 ttJd2 The text asks the question of the After 16 a3 l::tfb8 17 .i.fl ttJxa3! 18 knight on c4, noting that its retreat .i.xa3c4 Black wins the piece back with would mean the failure of Black's en a slight advantage. Again he is very tire strategy, and the triumph of active. White's. Fortunately, the knight does 16... ttJcb6 17 ttJxb6 ttJxb6 18 a4 .i.xe2 not have to retreat. 19 l:tel'iWa6 14 ...'iWa 5!
Black has regained the pawn with a Now White's plan of .i.d2,a2-a 4, etc good game. It is clear that White has is prevented. failed in the opening phase. 15 ttJa4?! 20 ttJC4 .i.xC4 21 bxc4 'ili'xC4 22 'iWxC4 White should follow through; i.e. 15 ttJxc4 23 l:txe7 l::txa4 24 .i.flttJe5 bXC4! "tIi'xC3 16 'i!VxC3 .i.xc3, when Ser Now Black is a pawn up and stands gey Kasparov gives 17 .i.g5! .i.xC4 18 clearly better. Again, I will just give the
2 3 Play the Benko Gambit
rest in brief as it is not relevant to the Having now dealt with the most opening discussion. critical line in the Fianchetto, we will 25 �f4 11d4 turn our attention to the earlier moves Or 25 ... g5! 26 �xe5 �xe5. and examine other White possibilities. 26 �xe5 �xe5 27 f4 �f6 28 J::td7 l::txd5 29 l:tbb7J::t d1 30 'it>g2 d5 31 �C4 dXC41? 32 ':xd1 c3 GameS S.Pavlovic-T .Giorgadze lugano 1985
1 d4 ttJf6 2 c4 cS 3 d5 b5 4 cxb5 a6 5 bxa6 g6
White faces a difficult defensive task, as one of his rooks rendered quite passive by the need to take care of the advanced black c-pawn. 33 �f3 .l::[e8 34 .l::[c1 �d4 35 lic2 .l::[e3+ 36 �g4 litd3 37 lia7 h5+ 38 �h3 �g7 This move requires a little explana 39 :e2 �e3 40 l1a6 �xf4 41 lIa4 lId2 tion. Very often Black plays 5 ...�xa6 42 lie1 l:id443 :a5 �d2 44 .:tf1 ':'C445 here, when 6 ttJc3 d6 7 g3 g6 8 �g2 l:ta7 c20-1 �g7 reaches the initial tabiya for this chapter. However, via that move order Game summary there is a serious alternative in 6 g3 d6 In the early part of the game Gallardo 7 �g2 g6 8 b3 �g7 9 �b2, and White played inaccurately. Instead of 15 has scored well from that position. ttJa4?!, the logical 15 bXc4 would have By playing 5 ...g6 first and delaying been a stern er test of Black's concept, the capture on a6, Black neutralizes the although Black should be doing alright possible double fian ch etto, since if here too. After White's inaccuracy Or White goes ahead with it anyway, the tega was already slightly better, and he pawns on d5 and a2 can find them kept the advantage and converted it to selves weak. This is because Black still a win without major problems. has the option of ...ttJxa6, an d can then
2 4 The Fianchetto Variation
apply pressure with ...ltJb4 and possibly Again Black is preparing for ...ltJb4, ... �b7 or ... �f5 (as in the current game). followed by ... �b7 or ...�f5 , highlight In fact, after Tb3?! it is White who must ing White's weaknesses on the light think about equalizing! For this reason, squares. Both the pawns on a2 and d5 the double fianchetto is rarely seen after are vulnerable, while the C2-square has 5 ... g6 in modem practice. also to be defended somehow after 6g3 . .. �f5. After 6 b3?! ltJxa6 Black is already 9 �g2 0-0 10 'iVd2?! slightly better; e.g. 7 �b2 �g7 8 g3 10 ltJc3�f5 also gives Black the ini ltJb4 9 �g2 l:txa2, and the b3-pawn, tiative (11 e4? fails to 11... ltJxe 4!). In too, will be weak in the middlegame. A stead, 10 e4, trying to build a strong logical continuation could be 10 li[xa2 centre, seems like a more logical plan ltJxa2 11 'iWd2 ltJb412 ltJh3 0-0 13 ltJC3 to me. A possible continuation is d6 14 0-0 'iWb6 15 ltJf4 �d7 16 ltJe4 10 ...�b7 (10... c4 !?, intending ...ltJc 5, is ltJxe4 17 �xe4 �xb2 18 'iWxb2 l'1a8 19 another good move) 11 ltJe2 e6,an d if l:[al'iWb 7 and Black has good winning 12 dxe6 fxe6 13 0-0 �xe4 14 �xf6 chances. 'iWxf6 15 �xe4, then 15... d5 16 �xd5
6 ... d6 7 b3?1 l::tfd8 17 ltJbc3 exd5 and Black is slightly Black gets a lead in development better.
and good play on the light squares af 10 •..�f 5! l1 ltJf3? ter this move, so it seems that the plan White had to play 11 ltJc3ltJb 4! and of b2-b3 and �b2 is a luxury White then: cannot afford; whereas if Black had a} 12 e4? still fails to 12... ltJxe4! 13 played 5 ...�xa6, instead of 5 ...�g7, the �xe4 (or 13 ltJxe4 �xb2 14 l:tbl ':'xa2 double fianchetto would have been with a decisive attack) 13 ...�xe4 and more challenging. Black is winning, based on the simple 7 ...�g7 8 �b2 ltJxa61 tactic 14 ltJxe4�xb2 15 'iWxb2?? ltJd3+. b} 12 'iitf1! looks a bit artificial but it is the best move in the circumstances (i.e. following the dubious moves 7 b3 and 10 'iVd2). Nevertheless, after 12 ... 'iVa5 13 ltJf3 �e4 Black was better in p.spacek-R.Mainka, Luxembourg 1988. He is about to regain his pawn (on d5) and has a good lead in devel opment. Black went on to win in 43 moves. 11... ltJb4!
25 Play the Benko Gambit
White resigned without waiting for a reply, since his position is hopeless after 21 ...li'lx d5, and 21 i.xf3 li'lxd5! was winning for Black as well.
Game summary The idea of the double fianchetto sim ply does not work after 5 ...g6; indeed Benko player should be very happy to face it, as White is struggling to stay in the game right from the opening Pinpointing the light-squared weak phase. This is main reason why 5 ...g6 is ness on the c2-square. Now Black is the most accurate move in practice. In clearly better. the game White could have kept his 12 0-0 disadvantage to a minimum with 10 The attempt to defend tactically by e4. Instead, 10 'ii'd2?! lost more time 12 li'ld4!? is met by 12 ...i.e4! 13 0-0 and after 10 ... i.f5 ! it was no longer i.xg2 14 �xg 2 li'le4etc. possible for White to guard against the 12 ...li'lc2 13 li'lh4 li'lxal 14li'lxfS?! threat of ...li'lb4- C2 in any satisfactory 14 i.xal might have been a better way. try. 14 ...li'lxb3 14 ... l:txa2! looks even stronger than Game 6 the text. S.Furman-E.Cieller lS axb3 gxfs 16 'ii'gs h6 17 'ii'xfs 'iid7 USSR Team 18 'ii'f3 .l:ta2 19 i.C3 Championship 1975
1 d4 li'lf6 2 c4 cS 3 ds bs 4 cxbs a6 S bxa6 i.xa6 To repeat the lesson fromthe previ ous game, I recommend 5 ... g6 6 li'lc3 i.xa6as a more accurate move order. 6li'lC3g6 7 li'lf3 i.g7 8 g3 0-0 This is also inaccurate, as White could now play 9 i.g2 d6 10 l:tbl! reaching his favourable set-up. Correct was 8 ... d6, and if 9 i.h3, lhen 9 ...li'lbd7 19 ...fl'g4! 20 ttJd2"iVxf3 21 ttJxf3 0-1 (or just 9 ... 0-0) 10 0-0 li'lb6 (or again
26 The Fianchetto Variation
10 ...0-0) 11 l':.el 0-0 transposes below. 11... �C4 While on the subject of move orders, it As the bishop doesn't cover dS should be noted that 9 e4 �xfl 10 pawn from h3, it is quite logical for 'it>xfl 0-0 11 'it>g2 lLlbd7 here would Black to target the d-pawn directly. Of transpose to the Classical main line in course with the bishop on g2 this move the next chapter. could just be answered by 12 lLld2�a6 9�h3 13 b3. 12 11dl:ta7 13 e4 'ii'a8
Preventing ...lLlg4 -eS which, as we will see in the next chapter, is a com The set-up with ...l:1a7 (or l:1a6),... mon manoeuvre in the Benko. Also, if ...'iWa8 an d ...l:1b8 is one of the standard White is going to play e2-e4 the bishop plans the Benko player should consider. looks more active on h3 than ob Although it is rarely seen in the Fi structed by the pawn chain. On the anchetto Variation, here it makes sense other hand it doesn't support the ad as White has less control over the cen vance e4-es so well. tral light squares with the bishop on 9 ... d6 10 0-0 lLlbd7 11 Wc2 h3. For instance, 14 b3 would be met by After 11 .l:.el lLlb6 Black is doing 14 ...lLlxe4! 15 lLlxe4 (or 15 'ii'xe4 �xc3) fine. For instance, 12 e4 lLlC4 13 'ii'c2 lS... �xdS 16 l:txdS WxdS 17 l:tbl fS 'ii'as 14 eS dxes 15 lLlxes lLlxdS! 16 lLlc6 with good play for Black. In the next lLlb417 lLlxe7+'it>h 8 18 'ii'e4 l:[ad8with game we will see this set-up being ef good activity for the pawn, or 12 �f4 fective for another reason. lLlC4 with an unclear game. Notice that 14 a3 the c4-square is an excellent square for White is usually cautious about the knight, while �f4 is not usually playing a2-a3 in these positions, as the played in this line as it weakens b2. b3-square is now weak, but here it is White should consider defending the hard to avoid it. 14 b3 was given in the c3-knight, preparing b2-b3. previous note, while if 14 �f4 l:tb8 15
2 7 Play the Benko Gambit lLld2 ..ta6 16 l1abl, then 16... lLlh S 17 24 �g2 f5?! 25 ..tb2 Itb4 26 Wc2 fxe4 ..te3 lLles probes the vulnerable light 27 fxe4 'iVe2+2s lLlf2 ..te3?! squares, and after18 ..tfl lLlg4 ..txa619 With this move Black is overplaying .l:lxa6 20 a3 lLlxe3 21 fxe3 lLlf6 Black is his hand; the simple 28... ..txb2 fol lowed doing fine. White has numerous weak by ...lLle s would still have led to an un nesses, and the likely exchange of the clear position. f6-knight for the knight on d2 (arising 29 ..tc3 lIfS 30 lLlC4 'ii'f3+ 31 �gl from lLlC4 and ...lLlg4 -es) will leave ..txf2+ 32 1:txf2 'iVxf2+ 33 'iVxf2 :xf2 34 Black with active play on the b-file: the �xf2 1:txb3 heavy pieces are working well with the dark-squared bishop.
14 .•.1ibS 15 lLld2 ..ta6 16 ..tfl l1ab7 17 :a2 ..txfl lS l:txfl lLleS Another standard move in the Benko, the knight going on to c7 and sometimes further to bS and d4. 19 b3 Wa6 With the idea of ... lLles-d3. 20 lLldl'ife 2 21 WC4 Wh5
A string of inaccurate moves have led Geller from an advantageous posi tion (after the move 22 f3) to one in which he is clearly worse, and he now had to fight for the draw - a task in which he was ultimately successful. 35 lIa3litbl 36 �e3 �7 37 ]:tal .l:r.b3 3S �d2 lLla6 39 �C2 :bS 40 lLla5 lla8 41 lLlc6�e8 42 a5 lLldbS43 lLlxbS l:txbS 44 �d3 �d7 45 �C4 e6 46 :fl exd5+ 47 2213?! exd5 �e7 4S h4 lLlc7 49 h5 gxh5 50 llf5 22 ..tb2 was a better way to keep .l:tgS 51 ..tf6+ �d7 52 ..tg5 lIg7 53 ..tf4 the balance. h4 54 gxh4 litg4 55 :f7+ �cS 56 h5 h6
22•.. lLlc7 23 a4 ..td4+?! 57 a6 lLlxa6 58 �b5 lLlb4 59 ..txd6 Here Geller starts to go astray. After lLlxd5 60 lIh7 ':g5 61 l1xh6 �d7 62 the thematic 23 ...e6! 24 dxe6 lLlxe6 ..txc5 �eS 63 �C4 �f7 64 ..td4lLle7 65 Black is very active with a strong initia �d3 lLlf5 66 lIh7+ �e6 61 ..tb2:g4 6S tive for the pawn. ..tCl :h4 69 ..tg5lIh3 + 70 �d2 Yz-Yz
28 The Fianchetto Variation
Game summary White plans to blockade the queenside, The line with 9 .i.h3 is largely forgotten with moves like .i.d2, 'iVC2, l::tbl, b2-b3, nowadays. The idea is prophylactic and a2-a4, etc; the well-timed break with e4- is directed against ...tZJg4-e s, while eS will also come into consideration, White reasons that the d-pawn is well while the knight itself will oftengo to f4 defended anyway. However, with to reinforce the ds-pawn. 11 ... .i.c4 followed by ..J:ta7 and ...'iVa8 Geller showed that the absence of a defender of the ds-pawn does have some relevance. This gave him some initiative from the opening, especially after Furman let slip the chance to equalize. But later Geller went wrong in turn, seeming somehow to miss White's defensive resources, and in the resulting endgame he had to fight fora draw. Nevertheless, the conclusion must This system had some popularity be that 9 .i.h3 is hardly a serious test of around the millennium when it was the Benko from a theoretical stand used by GMs Aseev, Aleksandrov, lonov point, and that Black is doing fine as and Se.lvanov. I, too, have played it long as he knows what he's doing. For with White in about ten games, and it White the main advantage of this line can be quite dangerous if Black doesn't may be that the opponent has forgot watch out. However, the tZJh3 line is ten about it, or else that it is so old they currently unfashionable, as Black has never knew about it at all. mostly sortedout how to neutralize it. Ironically, the knight's absence from f3 is a significant factor, as we will see.
Game 7 9 .••0- 0 10 0-0 tZJbd711 'ikC2 K.Aseev-R.Ponomariov ll .i.d2 l:ta7 leads to similar play.
European Championship, 11 •..:a7 Oh rid 2001 I like the idea of ... l:ta7, ... 'iVa8, and ....l:r.b8 here. The reason is that Black can 1 d4 tZJf6 2 c4 cS 3 d5 b5 4 cxb5 a6 5 later play ... cS-c4 without worrying bxa6 g6 6 tZJC3 .i.xa6 7g3 d6 8 .i.g2 about the possibility of tZJf3-d4-c6, .i.g7 9 tZJh3 since the white knight is not on f3. It is The ideas behind this move are quite this crucial difference that justifies similar to those in the 9 tZJf3 main lines: Black's play in this line.
2 9 Play the Benko Gambit
12 b3 "a8 13 l:tb1l:tb8 14 .l:.d1!? �xe4 23 i..xe4 i..xh3 with equal 14 i..d2 .l:r.ab7 15 l:tfcl would saved chances. White a tempo on the game, but I think 17 bXc4 i..xC4 18 �f4 Black has compensation nevertheless.
14 •..l:tab7 15 i..d2 c4
18 .....a6?1 And here 18... �f d7! was more accu This thrust is well timed and effec rate, supporting the knight on cS and tive. Now most of Black's pieces will opening for the bishop on g7, so that become active, giving him good com 19 ':xb7 ':xb7 20 �a4 (as in the next pensation for the pawn. As already note) can be met by 20 ...i.. xa2. mentioned there is no disturbing �f3- 19 ':'xb7 .l:txb720 :b1?1 d4-c6 for him to worry about, and no Now 20 �a4! would have been b3-b4 closing the queen side either. slightly better for White, since 16 l:tdc1 20 ...i.. xa2? fails to 21 �xcS, while after As it happens I had reached this po 20 .. :ifxa4 (or 20 ... �xa4 21 "xc4) 21 sition myself as White in an earlier "xa4 �xc4 22 ':xc4White would have game. N.V.Pedersen-J.Carrasco Marti more chances of achieving something nez, World Junior Championship, Za with his a-pawn and bishop pair. gan 1997, continued 16 i..e3 cxb3 17 20 .••':xb1 + 21 "xb1 �fd7 22 "C2 'i'a3 axb3 l1C7 18 i..d4 'iWb7?! (18 ..J:�b4 was Rather than regain the pawn by correct) 19 "d2 (here 19 b4! would 22... i.. xC3 23 i..xc3 "xa2, which would have given White an edge) 19 .. :ifb4, mean ceding the bishop pair, Black when Black had reasonable compensa keeps things fluid and is soon re tion. warded with a mistake. 16 ...�C5?1 23 h4 �e5 24 i..c1 "a7 25 i..h3 "a5 26 Better was 16... cxb 3 17 l:.xb3 %bb3 i..d2 "a3 27 'iii>f1? �f3! 18 axb3 �CS 19 l:tbl i..c8, and if 20 b4, Oops! Now Black gain-s the bishop then 20 ... i..fS! 21 e4 �cxe4 22 �xe4 pair and is clearly better.
3 0 The Fianchetto Variation
manages to free his pieces, so that Po nomariov only manages to win the game due to a rather big mistake from Aseev at the end. 45 .ig2 .ic8 46 ltJe4 .ig7 47 ltJxC3 .ixc348 ltJC1�7 49 .ie4h6 50 ltJd3 g5 51 hxg5 hxg5 52 �1 �e8 53 �e2 �d8 54 f4 g4 55 �e3 �C7 56 .ig6 �b6 57 .ie8.if5 58 .if7 �b5 59 .ie6 .ixe6 60 dxe6 �c4
28 ltJd1 ltJxd2+ 29 'iWxd2 .ixa2 30 .ig2 .ib331 ltJe3 .ia4 32 'iWa5 'iWa1+ 33 'iWe1 'iWb2 34 .if3 ltJb3 35�g2 ltJd2 36 ltJd3 'ii'C3 37 'iWc1 'iWxc1 38 ltJxC1.ib5 39 ltJf1 ltJb1 40 ltJb3 ltJC3 41 e4 .id3 42 ltJfd2 .ih643 �gl f5
61 f51 Here 61 �e4 would have held; e.g. 61... dS+ (not 61... .id4?62 ltJb4!) 62 �e3 .id4+ 63 �e2 �bS 64 fS and Black can make no progress. Whereas now the black king invades. 61... .id4 + 62 'ifole2�d5 63 ltJf4+1 Thematic in a Benko endgame, 63 ltJb4+was a better try. Black plays ...f7 -fS to undermine the 63... �e5 0-1 centre. Note that in the middlegame, Since 64 ltJg6+ 'ifolf6 6S �d3 .if2 is on the other hand, this move is usually winning easily for Black. a bit risky, and circumstances have to be favourable for it to work. Game summary 44 exf5 .ixf5?! The line with ltJh3 is not so popular More accurate was 44 ...gx fS!, with these days. Many of the ideas for White good chances of winning the ds-pawn are the same as in the 9 ltJf3 main and the game. After the text White lines. However, by choosing the plan
3 1 Play the Benko Gambit
with ...l:ta7, ..-'.a8, ...l:.b8 and a later in Game 1), the knight goes to c4 at ... c5-c4, Black can exploit the fact that once and Black is doing fine. the knight can't go from f3 to d4 and White's most consistent plan is to c6. From the opening phase the game continue with moves like 'ii'c2, l:tdl, l:tbl is unclear. and b2-b3 anyway, though I don't think In the middlegame Aseev made a Black has a theoretical problem here tactical blunder, which allowed Black either. Avrukh's suggestion of 'iWc2 and to gain an edge. His technical realiza h2-h4 is covered in the notes to Game 2; tion of this in the endgame was less as long as Black responds with ...h7 -h6, than convincing, however, and it was this leads to an unclear game. The con only thanks to a final mistake from his clusion is that Black seems to be doing opponent that Ponomariov managed fineafter 9 ...liJfd7, and it is up to White to win. to try and findsom ething new. The conclusion is that liJh3 is not Regarding White's other possibili too difficult to neutralize. It is charac ties: 9 liJh3 can be met by ...l:Ia7, ... 'ii' a8, teristic for the variation that it is Black ..JIb8 and ... c5-c4 (as Ponomariov who is pressing and has the winning played in Game 7). 9 .i.h3 is a forgotten chances. move, which prevents ...liJg4 and puts some pressure on d7, but in Game 6 Summary Geller showed that Black is doing fine This chapter examined one of the most after 11... .i.c4. Finally, the proposed critical and fashionable ways of playing move order with 5 ... g6 (rather than against the Benko, where White aims 5 ... .i.xa6) is specifically designed to to neutralize Black's queenside initia counter the double fianchetto with g2- tive, starting with 9 liJf3liJbd7 10 :tbl. g3 and b2-b3. Should White proceed To counter this, my recommendation is nonetheless, Black responds with the 9 ...liJfd7 !, by which Black goes for more active ...liJxa6 !, ...liJb4, and possibly rapid counterplay with ...liJb6-C4 and ....i.f5 , and even comes out of the open .. :iVa5. If White essays a quick e2-e4 (as ing with some advantage.
3 2 Chapter Two The Classical Variation
1 d4 t2Jf6 2 c4 cS 3 d5 b5 4 cxb5 a6 5 bxa6 This is one of the sounder ways to g6 6 t2JC3.i.xa6 7 e4 .i.xfl 8 'it>xfld6 meet the Benko Gambit. Karpov played it as White with quite good results. Typical moves for White include .:tel, followed by 'iYc2 or 'iVe2, or l:te2 and 'iVel, or :e2-c2 and 'ii'e2. In any case White defends the second rank and the knight on c3 while considering a well timed e4-eS. Sometimes .i.gs is played, especially if the black queen and f6- knight have moved away. Typical plans for Black include ...t2Jg4-e s or ...t2Je8-c7- bS, ...'iVb 6/'iVas Th e starting point for the Classical and ...:f b8, or ... .l:ta6/.l:ta7, ... Was and Variation, the main line continuing ...... e7-e6. In some cases the committal 9 g3 .i.g710 ..ti>g2 0-0 11 t2Jf3 t2Jbd7 move ...cS -c4, intending ...t2Jc s-d3, also comes into consideration. The first question here is: should White play h2-h3 or not? Is ...t2Jg4 th at dangerous a threat that White should use a tempo preventing it? The conclu sion used to be "yes", but now White often prefers 12 :el, since if Black opts then for the plan of ... .l:ta6, .. :ii'a8, ... e7- e6, White may save a tempo. Therefore against 12 l:1el I recommend the the matic 12 ...t2Jg4 , when Black is doing
33 Play the Benko Gambit
fine;wh ile against 12 h3 the plan with So White allows 12... liJg4. Th e alter 12 .. .lla6 and ...Va8 is seen to be effec native 12 h3 is covered in Games 10-13. tive, and 12 ... 'i¥b6 is possible too. Either 12... liJg4 way Black has sufficientcom pensation for the pawn. The Classical has declined in popu larity somewhat, probably because, as shown by the ideas and analysis in this chapter, Black seems to be doing fine. Incidentally, there are many novelties in this book and in this chapter in par ticular. I don't point out each time a novelty appears as I only found around 10%-30% of them; the rest were found by the computer. However, the conse This is the principled reply and the quence is that some of the lines which one I recommend. The knight will often were previously regarded as problem go to eS, and from there, if Black man atic for Black are in fact seen to be at ages to justify it tactically with precise least equal. play, to the good squares d3 and c4. The first six games below examine Alternatively, it may retreat back to d7 the main line with g2-g3 and 'ittg2. Af after the other knight goes to b6, ter that, in Games 14-16, we look at should White prepare f2-f4 with liJf3- variations in which the white king d2. takes another route, via h2-h3 and 13 'iWe2 c;t>gl-h2. The final two games in this One of the critical tests of 12 ... liJg4. chapter see White playing more ag Others are are 13 h3, 13 l:!e2 and 13 gressively on the kingside, with 9 g4 liJd2. I will cover the first two of these and 7 f4 respectively. here, and 13 liJd2in the next game. a) 13 h3 (by chasing the knight at once with h2-h3 and f2-f4, White will Game 8 gain some space in addition to the S.Cossin-M.Va chier Lagrave pawn; but moving forwards can also Chalons (rapid) 2008 backfire for White in the Benko) 13 ... liJges 14 liJxes liJxes 15 f4 liJd7 16 1 d4 liJf6 2 c4 cS 3 ds bs 4 cxbs a6 5 VC2 1:.a6 17 a4 'ii'a8 18 liJbs 1:.c8 19 bxa6 g6 6 liJC3 .i.xa6 7 e4 .i.xfl 8c;t>xfl i..d2 c4 (clearing the path for the d6 9 g3 .i.g7 10 'ittg2 0-0 11 liJf3 liJbd7 knight to cS and d3, b3- or a4) 20 .i.C3 12 l:.el (better was 20 .i.e3! liJcS 21 i..xcs l:txcS
34 The Classical Variation
with a balanced position: it is difficult which means Black can usually meet for either side to make any progress) .i.g5 with ...h7 -h6; it also prevents .i.d2, 20 ....i.xc3 21 ttJ)(c3 ttJC5 22 lle2llb8 23 as the b2-pawn would be hanging . .l:[a3 .l:[b4 24 'ifilh2 "ili'b7 25 as ttJd3 and Black had enough compensation for the pawn in V.Neverov-V.Bologan, Ni kolaev 1995. b) 13 Iite2 covers the second rank. White has a good score with this move, though Black is doing fine, if he plays correctly:
For example: b21) 14 .i.g5 h6
bl) 13 ...'ii' a5 14 .i.g5 (attacking the pawn on e7; after 14 ':'c2 .l:lfb8 Black has compensation with the idea of ...'ii' a6 and ...ttJe 5-d3) 14 ....i.xc3 !? (14... .l:.fb 8? 15 .i.xe7is clearly better for White; while 14 ....l:tf e8?! is a bit slow 15 .i.cl (after 15 .i.xe7?! l:.fe8 16 and also offers White a slight advan .i.h4g5 17 ttJxg5 hxg5 18 .i.xg5 ttJge5 tage in my opinion) 15 bxC3 f6 16 .i.d2 Black is doing fine; 15 h3?! ttJxf2 16 'ii'a4 17 .i.f4 g5 (V.Gavrikov-Y.Meshkov, l:txf2 hxg5 17 ttJxg5 ttJe5 18 'ii'fl c4 USSR 1981) is given as unclear by a few gives Black a very comfortable position sources. It is certainly playable for too) 15 ... l:.fb8 16 ':c2 'ii'a6 17 'ii'e2 Black, but I would say it is rather ttJge5 (heading for d3; Black has some slightly better for White. pressure in the endgame) 18 'ii'xa6 b2) 13 ..:iV b6! is the move I would ':xa6 19 ttJelc4 20 .i.e3ttJd3 21 ttJxd3 prefer. Although it doesn't attack c3, cxd3 22 l:td2 ':'xb2 23 Ibb2 .i.xC3 24 from b6 the queen covers the d6-pawn, litabl .i.xb2 25 l:txb2 ttJe5 and Black
35 Play the Benko Gambit was slightly better in F.Fleury a) 14 h3 (not the most critical move) S.Andrianoho, correspondence 1997. 14 ...tDge5 15 tDxe5 .txe5 16 f4?! .txC3 b22) 14 .tf4'i¥a6 15 I!c2(or 15 'i¥c2 17 bXc3 'ili'xC3 18 .tb2'iWb 4. 1:[fb8 16 h3 tDge5 17 tDxe5 tDxe5 18 .txe5 .txe5 with compensation) 15 ..Jitfb 8 16 litacl:b4 17 'iVe2 .txc3 18 bXc3 'i¥xe2 19 ':'xe2 :ba4 20 litcc2tDgf 6 21 c4 (or 21 .tg5 1:[c4 22 tDd2 ':ca4 23 tDf3 11C4 with a repetition) 21...tDb6 22 e5 tDh 5 and Black regains the pawn with an unclear game.
13 •.•'iWa5
Given the unopposed bishop on b2 one might think this position is good for White. But in fact Black is slightly better here with the a- and b-files for the rooks, while the king on g2 is not too safe either. S.Bilsel-F.Buettner, cor respondence 1998, continued 19 a3 'iWb3 20 lIabl l:tfb8! 21 1i'd2?! (but if 21 .tf6, then the queen sacrifice 21...tDxf6 ! Attacking the knight on c3, so that 22 l:txb3 l:txb3 shows the point of Black can just win the pawn back in 20 ...I:tfb 8 - Black will soon have rook, some lines. Otherwise ..JIfb 8 and knight and pawn for the queen, and .. :�a6 (as in the game) is an interesting plans ...tDh 5, ...llaxa3 and Itxg3 with a plan. As in many Benko positions, Black clear advantage; e.g. 23 e5 tDh 5 24 is doing alright in the endgame despite exd6 ':'xg3+ 25 �h 2 exd6 26 'fie7 being a pawn down. The bishop on g7 Itaxa3 27 'i¥xd6 l:txh3+ 28 �gl l:tag3+ is strong, Black has the nice a- and b 29 �fl l:tf3+ followed by ...tDxf4 and files for the rooks, while a knight can wins) 21...'iVa2 22 'iVCl ':xa3 23 l:1e2 go to an active square such as d3, c4 or 'iWb3 24 .txa3 'i¥xbl 25 'i¥xbl ':'xbl a4. with a decisive advantage for Black, 14 .td2 who went on to win in 34 moves. White has tried several other moves b) 14 .tg5 .txc3! (Black doesn't here, of which 14 tDd2 (line 'c') is the want to spend time covering the e7- most significant: pawn) 15 bxc3 f6 16 .td2 l:tfb8 with
36 The Classical Variation
compensation. The position is similar indeed sufficientto draw in the game to that in Game 10 (Piket-Topalov). after 29 lldl 'ili'a5 (the immediate c) 14 liJd2 - �xc3! (the immediate 29 ... e5 may improve) 30 h4 e5 31 l:tfl 14 ... liJge5?? would lose a piece to 15 f4, 'ika3 32 l:te2 'fixc3 33 J:[e3 'iVC2 34 llf3 while 14 ...liJgf 6? would also be a hu liJC5 35 llel'ii xa2 36 11ee3'ii a8 37 l:tf6 miliating retreat, leaving White clearly l:tb3 38 %le2l:tb 8 39 lIe3 Yz -Yz. better after 15 liJc4) and now: c2) 15 liJC4 'iia6 16 bxc3 liJge5 and now White has two sound choices, 17 liJb2 and 17 liJxe5, though in both cases Black will obtain good compensa tion for the pawn:
Cl) 15 liJb3 'ii'a6 16 bxC3 liJge5 is unclear. Again there are no worries in the endgame: with the open lines for the rooks and the squares d3 and c4 for the knights, Black has good compensa 17 liJxe5 (or if 17 liJb2, intending f2- tion for the pawn. One example: 17 f4, a2-a4 and liJc4 again, then 17 ...c4! �h 6 l:tfb8 18 'iiC2?! (or 18 'iixa6 l:txa6 leads to an unclear position) 17... liJxe 5 19 l:tebl l::ta3 with sufficient compen 18 'ii'xa6 (more accurate was 18 �h6! sation) 18 ... liJd3 19 lle2 liJ7e5 20 lldl l:tfb8 19 'iixa6 1txa6 20 llebl l:tba8, c4 (Black has the initiative so White M.Krallmann-M.Kopylov, Dortmund gives up the queen for rook, knight and 1999, though Black should be fine pawn) 21 liJd4 litb2 22 'ii'xb2 liJxb2 23 anyway; e.g. 21 l:tb7 liJg422 �g5 f6 or l:txb2 liJd3 24 1:1C2 l:te8 25 llbl e6 26 22 �f4 �f8) 18... 11xa6 19 i..h6 (or 19 dxe6 fxe6 27 liJf3 'ili'a8 28 liJg5 l:1b8an d �f4 liJd3 20 .l:.ebl f6 with compensa Black had good winning chances in tion) 19 ..JU a8, when 20 l:te2 is just K.Frey Beckham-D.Chan, correspon equal; whereas after 20 l:tebl?! (as in dence 2000. Nevertheless, White has C.Becker-M.Burschowsky, Austrian some hopes of making a fortress, while Team Championship 1997) 20 .. .f6! Black should watch out for a TOok or would even give Black the edge in the two appearing on the f-file, which was endgame. Some ideas are ...g6 -g5,
37 Play the Benko Gambit
...cj;; f7, ..J!a4, ...lla3, an d ...l:txa2 (aslong Black wants to keep things fluid, but as there is no mate on f8). the more concrete ls... lLlg es 16 lLlxes c3) 15 'iVxg4 .i.xd2 16 .i.xd2 'ii'xd2 .i.xes 17 l:tecl .i.xC3 18 .i.xc3'iVxa2 was 17 'iWxd7 lilfe818 b3 (or 18 a4 'iVxb2 19 correct, with an equal position. Black as c4 20 a6 Y2-Y2 H.Horta-A.Tsvetkov, can neutralize the bishop with .. .f7-f6, correspondence 1999, since the posi if necessary, or sometimes with ...lLlb 6- tion is equal; e.g. 20... c3 21 l:tebl 'iVe2 a4, and target the pawns on b2 and 22 l1el'iVb 2 or 21 'iiC7 (2) 18... cj;; f8! (de possibly dS (if White plays e4-es). fending e7 and e8 and thus threaten 16 b3 .i.XC3? 17 .i.xc3'iWxa 2 18 h3? ing ...l:.xa2, as well as enabling Black Here 18 lLld2!, with the threats of next move) 19 a4 (19 'iVh 3?! cj;;g8 20 'ii'xg4 and l:.al, would have given 'iVd7 cj;;f8 led to another draw in White a clear advantage; e.g. 18 ... lLlgf6 S.B.Jensen-R.5kytte, Danish Champion (unfortunately 18 ...lLlg es is no longer ship, Odense 2011, though here possible, due to 19 f4) 19 .:tal 'iic2 20 19 .. J:txa2 20 liba2 'iVxel 21 'iVxh7 l:.ecl'ilkx c3 21 l:.xC3:'xal 22 lLlc4. 'iVxe4+ 22 f3 'iVes would have been 18 ... 'iVxe2 19 :'xe2 lLlgf6 20 lLld2 lLle8 slightly better for Black) 19 ...l:.eb8 20 21 lLlc4 l:tadl'iWc 2 21 eS l:.d8 22 'iWC7 llac8 23 21 li:teelfol lowed by li:tal is okay for 'ii'b6 llb824 'ii'C7 l:.bc8Y2-Y2 R.Ruether White. F.Buettner, correspondence 1998. 21... lLlc7 22 ':'eb2f5 23 f3 White might have tried 24 'iVc6,ho ping Or 23 .i.aslLlb S 24 exfs gxfs 25 l:[el for 24... l:tbc8 ?! 25 exd6! exd6 26 'iVb7 with an unclear game. with the advantage, but 24 .. :iVxb3 25 23... lLlb5 24 .i.d2 fxe4 25 fxe4 exd6 exd6 26 as li:tbc8 would still be equal. 14 ...l:tfb8 15 nabl
Black has no problems here; indeed after 2s ...lLla3 26 lLlxa3 .uxa3 27 b4 cxb4 28 li:txb4li:txb4 29 .i.xb4li:td3 he is
15 ••.'iWa 6?1 even on the favourable side of a draw,
3 8 The Classical Variation as his rook is more active, while the 34.. .'� g7 35 lta7+�h6 36 11ee7ttJf5 37 white pawns e4 and dS are slightly .l:txh7+�g 5 38 l:thd7?1 more vulnerable. Instead, Black plays And here 38 �2 lld3 39 lla4l:txdS for more and should have gotten less. 40 h4+ �f6 41 hS offered more chances 25... ttJf6 26 11e1ttJa 3 27 ttJa51 l':txa5? of saving the game. A faulty combination. No better is 38... l1d 3 39 l:ta4Iixd 5 27 .. J:tbS? 28 l:ta2l:tb xas 29 �xas l:lxas 30 es! ttJxdS 31 l:tea1, when White re gains the piece with the advantage. Black should admit his mistake and play 27 .. J:tb6 28 �C3 (not 28 ttJc6?! ttJc4!) 28... ttJb S still with a fairly level position. 28 �xa5 ttJC429 11a2? Missing his chance with 29 �c7!, af ter which White is clearly better; e.g. 29 ...lk7 30 l:tc2 ttJa3 31 lIa2 Iitxb3 32 es!. Now White has a difficult task, de 29 ...lix b3 30 �d2 ttJxd2 31 ':'xd2 ttJh5 fending against the connected c- and d-pawns. 40 �3 �6 41 �2 ltd2+42 �3 l:td3+ 43 �g2 c4 44 l:tC7 d5 45 lla6+ �g5 46 ':'cc6 �4 47 :C5 c3 48 �2 l:tf3+ 49 'ii;>e2 d4 50 llxg6d3 + 0-1
Game summary When White tries to do without play ing h2-h3, Black should exploit this by playing ...ttJg4 , when the knight is ready to jump to eS. With some knowl Now Black gains another pawn and edge of the position, Black should it is White who must try to draw. I will equalize without major problems. give the rest with just light comments Some of the ideas are ....l:Ub8, 'ifb ... 6/as as it is not relevant for this opening 'ilfa6, ...ttJeS-d3 /c4 and ...ttJb 6-c4/a4. book. Another recurring theme is ... �xc3 32 e5 ttJxg3 33 exd6 exd6 34 lla2?1 with the idea of winning the a-pawn, Taking the seventh rank with 34 which often works surprisingly well l:te7 wasstron ger. despite its risky appearance. Here it is
39 Play the Benko Gambit suggested as an improvement for Black precision not to get in trouble. White's on move 15; we saw another example last move offers to return the pawn in Game 3 (Neverov-Pedersen). (after ...1.xc 3, etc). Here it is considered In the game Cossin had two oppor unfavourable for Black to accept it, but tunities to obtain the better position: I think Black is doing alright after when Vachier didn't play ls... 1.x c3, ...1.x c3. and later when he sacrificed the ex Another option is 16 tDf3, aiming change. Once White had spurned those for e4-es, when play can continue chances (in his defence, it was a rapid 16 ... tDa4 17 tDdl (or 17 'ii'c2 tDxC3 18 game) he found himself in a difficult bxc3 Was 19 1.d2'iVa4 with compensa endgame and was outplayed by his tion) and now: higher rated opponent.
Game 9 A.LuIOVoi-A.Khalifman St Petersburg 1996
1 d4 tDf6 2 c4 cS 3 ds bs 4 cxbs a6 S bxa6 g6 6 tDC3 1.xa6 7 e4 1.xfl 8
4 0 The Classical Variation
'iVxdS+ 22 'iif3 lilff8 23 ':'xe7 lUe8 with 21 �xa3 'fIC7with a balanced position. compensation) 19 .. 5�h8 20 b3 ttJab6 21 b) 18 'iid3 liJab6 19 �b2 'ii'C7 20 c4 ttJC3 fS 22 1Vd3 was A.Yusupov liJa4 21 �C3 was A.5trunsky-M.Vasilev, V.Baklan, German League 1998, when Neuhausen 2008, and now 21 ...ttJxc3 22 ...dxe s 23 fxes e6 24 �f4 ttJxdS 2S 22 'iixc3 lilxal 23lilxal lilb8woul d lead liJxdS exds would have given Black a to an equal game. slight edge according to Baklan. c) 18 c4 liJC3! (Black is trying to use 16 ... .l:ta7?1 the slight lead in development, in some There has been some debate about cases playing ... ttJa2 and ... �fb8, re this position over the years, as it is stricting the bishop on cl; instead quite logical to chase the black knights. 18 ... e6?! 19 dxe6 fxe6 20 liJf3 was Indeed, while the game is very unclear slightly better for White in Y.Pelletier after 16 ...lila7 ?! 17 as, it seems that l.Leventic, Buekfuerdo 1995) 19 lilxa8 White can get a slight advantage. 'iWxa8 20 'ii'e3 ttJa2 21 �b2 (21 i..a3 Instead, Black should prefer liJb4is unclear) 21..Jlb8 22 litbl'ii' a4 23 16 ...�x c3! 17 bXC3. This used to re eS 'iWc2 24 exd6 exd6 2S �f6! (a tactical garded as good for White. However, I trick based on the back rank mate) think - as I hope the annotations will 2s ... l:ta8with equality. show - that Black is not worse after 17 as 17 ... liJxa4! (17 ... lilxa4?! 18 ':xa4 liJxa4 19 c4 is slightly better for White) and then:
17... 'fIa8 17 ... liJa8? is waytoo passive and 18 ttJC4is clearly better for White, as Black a) 18 J:ta3ttJa b6 19 c4 (19 'iWhsl:txa3 lacks any compensation for the pawn. 20 �xa3 'i'a8 21 .:tal 'iWa4 22 'iWd3 c4 18 'iWbs 23 'iibl lila8 24 liJf3 'iWas 2S 'iWel ttJf6 With the threat of 19 axb6 J:[xal 20 and 19 lIb3 lila2 20 c4 'iWa8 21 'ifd3 .:tb8 'iWxd7; the answer is another develop are both unclear) 19 ...f6 20 �b2 lilxa3 ing move. Alternatively:
4 1 Play the Benko Gambit
a) 18 liJb3? is a tactical mistake: i.d2 l:ta4 27 'itte2 l::txc4 28 'ittd3 l:txd2+ 18... i.xc3 19 bxc3 liJxd5 and Black is 29 'ittxd2 and White is slightly better. In better, since the b3-knight is hanging the rook endgame there are ideas of after20 exd5?! 'iWxd5+. e4-e5, as well as 'ittd3 and litbl-b7. The b) 18 a6 (advancing the pawn, with pawn endgame after 29... :a4 30 �d3 the threat of liJb5, is another logical ':a2 31 l:!e2 :'xe2 32 'ittxe2 is also prob move) 18 ...l:tb 8! (18 .. :.c8?! has also lematic for Black, since the white king been played, but leads to a worse posi is far more active. tion for Black) 19 liJb5 l:txa6 20 liJc7 ':'xal 21 liJxa8 l:tbxa8 22 liJf3 c4 and Black had some compensation for the queen in C.Olausson-R.Skytte, Stock holm 2000, though objectively White must be slightly better here. lS ..JlbS ! 19 axb6 The only sensible move.
19 .•Jlxa l 20 'iWxd7
21 liJdbl?? Missing his chance. Lugovoi later found the correct line 21 'iVxe7! iLxc3 22 bxc3 l:!xcl 23 l:[xcl llb2 24 'iWxd6 l:lxd2+ 25 'itth3 and White is winning; e.g. 25 ...'iWa2 (or 25.. :ifc8+ 26 fS ! gxfs 27 l:.bl) 26 ':'hl! (stronger than Lugo voi's 26 'itth4, which allows 26 ... h6!) 26 .. :ifc4 27 Wb8+ (or just 27 'iVes)
20•.. ':xb6?? 27... 'itt g7 28 'fie5+ 'itth6 29 'fig 5+'itt g7 30 This should lose. Black had to play f5 'iWxe4 31 f6+ 'ittg8 32 'iVxd2 'ifxh l 33 20 ...iLx c3! 21 bxC3 'iWa2 22 'iWa7! l:txb6 'iWd3with a decisive advantage.
23 'iVxa2 (23 'ii'xb6?! l:txcl leads to 21••• 'iWbS 22 e5 .l:ta7! equality, as after24 l:te2 l:tC2Bl ack wins Better than 22... l:tb 7?! 23 'iWg4 dxe5 the knight on d2; e.g. 25 'iVd8+'itt g7 26 24 fxe5 i.xe5 25 l:txes 'iWxes 26 'ii'c8+ 'fixe7 ':'xd2 27 ':'xd2 'iVxd2+ 28 'itth3 'ittg7 27 'iVxb7 l:txbl 28 i.d2! (not 28 'iWe2 with a draw in S.Bekker Jensen liJxbl?! 'iVe2+ with perpetual check) S.Petrosian, Norderstedt 2003) 28 ... .l:.al 29 'ilfb5, when the minor 23... .l:txa2 24 'ittfl :c2 25 liJC4 l:!a6 26 pieces are somewhat superior to
4 2 The Classical Variation
Black's rook and pawn. errors, with White missing a clear win 23 "iVg4 dxe5 24 fxe5 .i.xe5 25 l:txe5 and Black obtaining some chances in 'ii'xe5 26 'ifc8+ 'iti>g7xc5 27"ii' l:.ab7 28 the ending. 'ikf2 I don't think we will see that much of 16 a4 in the future, as Black has a relatively easy game, and White should search for a theoretical advantage in other lines.
Game 10 J.Piket-V. Topaloy Monte Carlo (rapid) 1997
1 d4 liJf6 2 c4 cS 3 d5 b5 4 cxb5 a6 5 28.. .'itg8 bxa6 g6 6 liJC3 .i.xa6 7 e4 .i.xfl 8'iti>xfl 28 ...e6! might give Black a slight d6 9 g3 .i.g7 10 'iti>g2 0-0 11 liJf3 liJbd7 edge. 12 h3 29 'ife2 'ii'xe2+ 30 liJxe2'iti>g 7 Here, and in the next three games, And here 30 ...l:ta6 31 liJbc3 1:tal 32 White opts for this prophylactic move, 'it>f3 l:tb4 perhaps offered more win preventing the knight manoeuvre to ning chances, though it is still drawish. g4.
31 liJbc3 'iti>f6 32 'iti>f3 Itxb2 33 .i.xb2 12 .•."ii' b6 .!:[xb2 34 'iti>f4 g5+ 35 'iti>e4 h6 36 g4 'iti>g6 37 h3 h5 38 liJd4 hxg4 39 hxg4 'it>f6 40 liJde2Yz-Yz
Game summary Th e plan of chasing the knights on g4 and b6 used to be regarded as prob lematic for Black. However, with the thematic 16 ....i.xc3 followed by
17 ... liJxa4, Black seems to equalize, which I think is new. When Khalifman didn't play this idea, Lugovoi could In my opinion this is the soundest have gained a nice advantage. Instead, choice for Black in the Classical main the game continuation was very un line. From b6 the queen prevents the clear and ended in a draw after mutual cl-bishop moving for the time being,
43 Play the Benko Gambit and Black has a reasonable chance to a) 15 eS is a good choice if White is gain good counterplay with typical trying for a draw. Then ls ...liJxe s 16 moves such as .. JUb8, ...Wa6, ...liJe8-C7- liJxes dxes 17 ':xes 'ii'b7 transposes to bS and ...liJb 6-C4. The drawback is that 14 eS (see the next game), albeit with White can adopt a plan with l:.el and each side having made an extra move. e4-eS which leads to a drawish posi b) 15 :bl (planning b2-b3 and �b2) tion. If the Benko player wants to avoid ls... liJe8 (the logical response, after this and is ready to take some risks, the which White has to spend a tempo alternative idea of 12 ...l:.a6 fol lowed by covering the knight on c3) 16 �d2 and: ...Was can be considered, for which see Games 12 and 13. 13 l:!el.l:tfb8 14 l:!e2 The immediate 14 eS is examined in the next game.
14... 'ii' a6
bl) 16 ... C4?! was played in M.Gure vich-Z.Azmaiparashvili, European Cup, Strasbourg 1994, but it's not a good move. After 17 �e3 l:.b7 18 �d4 White had a comfortable edge. The game con tinued 18 ...�h 6 (18 ...�xd4 19 'ii'xd4 A good square for the queen. Now litab8 20 eS dxes 21 liJxes liJxes 22 litxes one of the main priorities is to find is no better) 19 eS dxes 20 liJxes liJxes, good posts for the knights, such as d3, and now 21 �xes :d8 22 :e4! liJd6 (or c4 0r a4. 22 .. .f6 23 �d4) 23 �xd6 followed by 24 15 'ii'el We2 would have maintained White's Mirroring the similar idea of ... :ta6, plus, according to Gurevich . ... Was as Black can play in this line. The b2) 16 ...�x c3! is preferable. It might threat is now e4-eS with pressure on look dangerous for Black to take the a2- the e-file. pawn with e4-eS coming, but the dS White has quite a few alternatives pawn is weak after that, so basically it here, though Black is doing fine against leads to an unclear position. For exam all of them: ple: 17 �xc3 Wxa2 18 b3 (the immedi-
44 The Classical Variation ate 18 eS could be met by 18 .. :iVb3) to C4 or sometimes a4) 17 1:[cl(White 18 ...'iWa6 19 es?! (Gurevich's intention; can't play 17 b3? due to 17 ... ttJfxdS! 18 instead 19 ttJd2 'iVd3 20 1:[e31Wb s is un exds i.xC3 19 i.xh6 ttJxdS and Black clear after 21 ttJC4 ttJb6 or 21 b4 %:ta2) was clearly better in M.Gurevich-J.Koch, 19 ...ttJC7 ! (targeting the now vulnerable European Team Championship, Haifa ds-pawn) 20 e6 (20 l::td2 "iVb7 21 b4 c4 1989) 17 ...ttJC4 18 b3 (otherwise Black 22 lIcl ttJb6 is also finefor Black, or 20 will get at least a slight edge) exd6 'ii'xd6 21 .l:.d2 f6! 22 1:[cl 1:[c8 23 18 ... ttJxe3+ 19 ':xe3ttJd7 and Black had .Ilb2 ttJb6 and Black wins the ds-pawn excellent compensation for the pawn with a good position) 20.. .fxe6 21 dxe6 in J.Hjartarson-L.Alburt, Reykjavik 1986. ttJf8 22 'iWel (or if 22 ttJgs, then C2) 16 i.f4 gs 17 i.e3ttJh 7 (an unor 22..."ir'c6+ 23 'it;>gl 1i'ds with a clear ad thodox move with ideas of ... ttJhf8-g6, vantage) 22 ... 1i'c6 23 i.al (now the ....l:txb2 and sometimes ...i.x (3); e.g. 18 threat of "ir'C3 looks dangerous, but...) :cl i.xC3! 19 l:txC3 (19 bXc3 ttJhf6 20 23 ...'iWd s (a cool reaction - the ideas are h4 g4 21 ttJel ttJxe4 intending ..:.C4 ...ttJcxe6 (-d4) and ...l:.xb 3) 24 'iVC3 ttJcxe6 and Black is doing fine) 19 ...ttJhf6 20 25 1i'h8+ 'it;>f7 26 1:[bel ':xb3 27 i.C3 l:ta3 'iWbs with compensation. I1aa3 28 1:[e3 ':xc3 29 ':'xC3 ':'xc3 30 C3) 16 i.xf6 i.xf6 17 :'Cl and now 'ii'xC3 ttJd4 and Black is better with two Black has three possibilities. As they all strong pawns for the exchange. lead to a balanced position, it's a mat c) 15 i.gs is a logical move with the ter of style which one to play: b2-pawn defended, and the bishop is well placed on gS. So Black should play lS ...h6, when White has three sound choices, though none is sufficient for more than an unclear game:
c31) 17... i.xC3 18 bxc3 (18 l:txc3?! 'ii'xa2, regaining the pawn, is fine for Black) 18 ...'iVC4 19 l:tcc2 lIa3 20 "iVd2 'it;>g7 21 1i'e3 1:[ba8 22 ttJd2"iVb s 23 ttJb3 1:[3a4 24 1:[b2 ttJes 25 f4 ttJC4 26 'ii'd3 Cl) 16 i.e3ttJb6 (prep aring to jump 'ii'd7 with compensation for the pawn.
45 Play the Benko Gambit
C32) 17... lLles 18 lLlxesi.xes 19 'iVd2 in B.5passky-1Koch, French Champion i.g7 is roughly equal, since any at ship, Montpellier 1991) 18 bXc3 lLlXa4 tempt by White to make progress will 19 eS (Stohl) 19 ...lLlxe s 20 lLlxes dxes allow Black counterplay. L.Schandorff 21 l:[xes 'iVC4 22 .l:txe7 'iVxdS+ 23 'i¥i>h2 A.Khalifman, Aarhus 1997, ended: 20 (better than 23 f3 ?! lLlb6 with compen b3 c4 21 bxc4 'i!VxC422 lLldlY2-Y2. sation) 23... lLlb6 24 l:txa8 l:txa8 2S 'iWe2 C33) 17... l:tb4 (the sharpest choice, with some initiative for White. leading to an unclear game) 18 l1cc2 (a d2) ls ... lLle8! is correct. Then 16 a4 thematic move in this line, this was the :b4 maintains Black's queenside coun continuation in G.Danner-R.Mainka, terplay, while either 16 i.gs or 16 i.f4 Dortmund 1993) 18 ... .l:tab8 (Mainka runs into 16... l:!.x b2; i.e. 16 i.f4?! ':'xb2 played 18... 'i¥i>h 7 here) 19 'iVd2 and now 17 'ii'xb2 i.xC3 18'ivxc3 'iVxe 2 and Black White might soon play h2-h4, so I has a slight advantage. would suggest 19 ...i.g7, trying to 1s... lLle8 minimize the slight weakness on h6; ls ... lLlb6is a simple way to get com e.g. 20 h4 lLlb6 (this was obviously the pensation for the pawn, and I would main plan here) 21 'ii'cl lLlC4 22 hS gs recommend this move active move; e.g. with a sharp, unclear game, though I 16 i.gs h6 17 i.xf6 i.xf6 and now we would findit easier for Black to play. have: d) 15 'ii'C2 is a little tricky:
a) 18 eS (one should always watch dl) ls... lLlb 6?! is now met by 16 a4! out for this move in the Classical main lLlfd7 17 lita2, when White has a good line, but Black is okay here, thanks to chance of obtaining an advantage; e.g. the pressure against the ds-pawn) 17... i.x c3?! (but not 17... lLlc4? 18 b3 18 ...dx es 19 lLlxes (19 l:.dl lLld7 is un lLlces 19 lLlxes lLlxes 20 lLlbS! c4 21 clear) 19 ...'iVb 7 (notice the unfortu i.f4! :c8 22 i.xes cxb3 23 'iVxb3 i.xes nately placed king on g2) 20 lldl :d8 24 l:[ec2an d White was a safe pawn up 21 lLlg4 lLlxdSan d Black is doing fine.
46 Th e Classical Variation
b) 18 :cl lbd7 19 b3 (or 19 .:tcc2 White cannot make progress, so the .Jlxc3 and if White captures with rook position is drawn. or queen, then -20 .. :ifxa2; otherwise 20 30 �f3 'it;e8 31 lbf1 'it;d7 32 lbg3 l1b8 bxc3 'iVc4 is unclear) 19 ...c4 20 bXc4 33 lbfS lbxfS 34 exfs ,Ub3 35 l:ta3 .llb8 'ii'xC4 21 lbbl 'iWa6 22 .:tec2 was 36 e4 'it;d6 1i13a237 1Ib338 :a3 'it;cs39 M.Bertok-D.Gliksan, Zagreb 1971, and
Game 11 V.Kachar-V.Arbakoy Russian Cup, Moscow 1999
1 d4 lbf6 2 c4 cS 3 ds bs 4 cxbs a6 5 bxa6 g6 6 lbC3 .Jlxa6 7 lbf3 d6 8 e4 i..xf1 9 'it;xf1 lbbd710 h3 If White intends to play the king to g2 it is more usual to play 10 g3 first, although there is nothing wrong with
47 Play the Benko Gambit the current move order. trap - see the note to White's next move. 17 'iVf3 Instead: a) 17 'iVe2? falls into the trap with 17 ...li:JxdS !,
10 .••..tg7 11 g3 0-0 12 �g2 'ilkb6 13 l:[e1 ':fb8 14 e5 A more straightforward approach than the 14 .l:r.e2 of the previous game.
14 •..dxe5 15 li:Jxe5 li:Jxe5 16 l:[xe5 as occurred in G.Rechlis-J.Plachetka, Ostrava 1981. Probably GM Rechlis simply overlooked this move, since the position after 18 lIxdS e6 19 'iVf3 exds 20 li:JxdS doesn't look like it offers White sufficient compensation for the exchange. Indeed, 20 ... l:ta7 21 a4 (not 21 ..tgS? 'iVxb2) 21... .l:r.d8 would lead to a clear plus for Black. b) 17 'iitgl removes the king from the long diagonal, but is a rather pas
16 ..•'ilkb71 sive move, and White must be careful Inferior is 16... l:tb 7?! 17 'iVf3 li:Je8 18 not to get the worse position. For ex l:te2 li:Jd6 (the knight would be great ample, 17 ..JId8 18 'iVe2 (not now 18 after ...li:Jf S-d4, but this is easily pre 'ilkf3 ?! e6 and Black is better, as 19 vented) 19 g4 li:JC420 ..tf4 with the bet ..tgS?! can be met by 19 ... li:JxdS) ter position for White in N.Diedam 18... li:Jxds 19 l:.xdS l:txdS 20 li:Jxds R.Mainka, Liechtenstein 1993, a game 'iVxds 21 'iVxe7 'iVdl+ 22 'iitg2 'ilkdS+ which Diedam quite surprisingly won when Black is certainly equal and against his GM opponent. might try for more with -... ..td4, ...l:.xa2 The text involves a rather simple or ..:iVf3 (if allowed), hoping for play on
48 The Classical Variation the dark squares, although 23 �gl ...s.. xc3 and ...'ii' xd5, without allowing s..d4 24 'ii'e2 (or 24 a4 'ii'f3 25 s..e3 l::txe7) has the drawback that, com i.xb2) 24... :xa2 25 ':xa2 'ii'xa2 is still pared with the text, it takes an extra drawish. tempo. After 20 l:te2 f5 21 J:[ael White 17 .. Jld8 is slightly better, since 21 ... s..xc3 22 bXc3 'ili'xd5?? would now lose to 23 l::te8+. 20 %:te2 s..XC3 21 bXc3 'ili'xd5 22 'ili'xd5 ':'xd5 23 lIxe7':'U
18 s..g5 18 a4 is more ambitious, since the a-pawn can be annoying if Black doesn't play precisely: 18 ...l:ta5 ! (from as the rook can attack the d5-pawn Black is winning the pawn back, af after ...c5 -c4 in some lines; 18 ... l:.a6?! 19 ter which the rook endgame is very lIe2 left White with a slight edge in likely to end in a draw. R.Akesson-S.Pinchuk, Pardubice 2008) 24 l:te3 19 s..g5 h6 20 s..xf6 s..xf6 21 l:te2 c4 22 24 C4 l:td4 25 lIbl lIa6 also ended tLle4(after 22 'ii'e4 'ii'b4 or 22 l::td2 'ii'b4 peacefully in W.lclicki-1Bellon Lopez, Black should also keep the balance) Brussels 1987. 22 ...l:td xd5 with equality in I.Saaed 24... lId2 25 :f3 lIaxa2 26 lIel lIe2 27 K.5eegert, World Junior Champion l::tbl .l:tab228 l::tdl 1/2 -1/2 ships, Kemer 2008. 18 ... h6 Game summary 18 ...'ii' xb2?? is a huge blunder, due The e4-e5 line is known as a way for to 19 l:tbl �C2 (or 19 ...'ii' a3 20 lIxe7) 20 White to try and force a draw. Black l:[cl! 'iVb2 21 l:lxe7 and White is win may obtain some chances against inac ning. curate play, but the game was a logical 19 i.xf6 s..xf6 consequence of White's peaceful atti The attempt to create some compli tude, and is perhaps the main draw cations by 19 ...exf6 (planning .. .f6-f5, back to 12 ... 'ii'b6.
49 Play the Benko Gambit
Therefore, if Black is in a must win a6 also covers the d6-pawn. So this can situation, or facing a much lower rated be a quite a dynamic plan. For White, a opponent, he might try the alternative set-up with lbb5 and a2-a4 looks to be plan of 12 .. Jla6 and ... 'ita8, which we the most effective response. will examine in the next two games. 13 liel'ii' a8 Objectively White is slightly better but, In addition to preparing the ...e7-e6 as we will see, it is still interesting from break, Black's 'iVa8/l1a6form ation also a practical viewpoint. exerts some pressure on the queenside. 14 'tiC2
Game 12 J.Aagaard,.L.D.Nisipeanu Berlin 1997
1 d4 lbf6 2 c4 cS 3 ds bs 4 cxbs a6 S bxa6 iLxa6 Please remember that 5 ... g6 is our preferred move order, after which 6 lbc3 .i.xa6 7 e4 .i.xfl 8 'ifolxfl d6 trans poses below. 6 lbc3 d6 7 e4 .i.xfl 8 'ifolxfl g6 9 lbf3 White has several ideas here: .i.gs, iLg7 10 g3 lbbd711 h3 0-0 12 'ifolg2 lia6 preventing ...lbe8, while if Black plays ... lbb6 White might consider .i.xf6. An other move is l:te2 covering the second rank, followed sometimes by :ec2 or :cl-c2. The plan of a2-a4 and lbbS should also be mentioned, threatening lbC7as well as lbxd6in some cases. 14 ':'e2 is examined in the next game. After 14 .i.g5 Black should probably play 14 ...h6 (even though this weakens the g6-pawn somewhat) and if 15 iLd2 then 15 .. J:tb8, intending A move played by Topalov and ...lbe8-c7 with compensation. Instead, Kh alifm an. It is usually followed by 15 ...e6?! 16 dxe6 fxe6 17 'iVc2 is slightly ...'i!i'a8, putting pressure on the long better for White. diagonal after the thematic ...e7-e6 14 ...e6?! break, when we notice that the TOok on Here 14 ..:ij'b 7! is a better try, avoid-
50 The Classical Variation ing the idea of d5xe6 and liJb5 (see the a2} 15... e6 is another option; the note to 15 i..f4 in the game). idea of .. .f7xe6, d6-d5 and play in the centre is interesting from a practical perspective, but I think the position slightly favours White here: 16 dxe6 fxe6 17 i..gs d5 18 i..xf6 liJxf6 19 exd5 exd5 20 l:tdl'it>h 8 21 liJa4 liJe4 22 liJd2 l:txa4!? 23 'ii'xa4 'ii'xb2 24 liJxe4?? (a blunder; 24 :'xe4! dxe4 25 liJxe4 i..d4 26 l:td2'iVb7 27 f3 would still be a little better for White) 24 ...'ii' xe2 25 l:Ixd5 c4 0-1 D.Fuhrmann-H.Ellers, German league 1999, since White cannot de Now White has tried: fend both e4 and f2. a} 15 11e2?! and then: b} 15 b3! e6 16 dxe6 fxe6 and now:
al} 15... 'it'b 4! - I like this move.
bl} 17 i..b2 d5 with a further split: The idea is ...l:tfb8 and ... liJe8-(c7- bll} 18 exd5?! exds (now the posi b5}, while, as we have already seen a tion is unclear, though White must be few times, ...i.. xc3 is also a possibility careful not to get in trouble since Black with an unclear game; for example, 16 has a good centre and active play) 19 i..d2 (or 16 i..g5 l:[b8 17 J:tdlh6 18 i..cl 'it>gl?! (tactics ensure that this neither liJe8 with compensation) 16.. JUa8 17 wins nor loses material, but it cedes the .l:!.dl liJe8 18 i..f4 i..xc3 19 bxc3 'iiC4 20 initiative) 19 ...d4 20 liJdl'it'xf3 21 'it'C4+ 'ii'd3 'it'xd3 21 llxd3 f6 22 litdd2 g5 23 (White's point) 22 ... 'iid5 22 'iixa6 liJe5 .i.e3 liJb6 24 ':c2 liJC4 25 i..cl liJC7 26 (the counter-point: Black wins the ex liJd2 liJe5 27 liJb3 l::ta4 Yz -Yz M.Marin change back, and there is the threat of N.Nedobora, Saragossa 1994. ... liJe4 amongst others) 23 l:[xe5 'ii'xe5
5 1 Play the Benko Gambit
24 tiJe3tiJe4 25 f4 'iie7 26 tiJfl tiJC3! 27 Black has excellent compensation ac 'ii'c6 d3 28 .1i.xC3 .1i.xC3 29 1:[cl.1i.d 4+ 30 cording to West; for example, 22 ':'acl �hl 'iVe2, when Black had a strong ini (or 22 .1i.xd4!? cxd4 23 ':'acl tiJcS 24 tiative for the pawn and went on to l:tedl .1i.h6 25 l:[xcS 'iVxcs 26 'iVxe6+ win in J.Dorner-G.West, correspon �h8 with roughly equal chances) dence 1997. 22 ... tiJxes 23 ':xes .1i.xes 24 l:!txcS"ili'xc s b12) 18 tiJd2(Wes t gave this an ex 25 'i!Vxe6+ �g7 26 tiJxes ':'xf2+! 27 �xf2 clamation mark without further analy "ili'c2+ with a draw. sis) b2) 17 tiJgS!? dS is very unclear too.
18... C4 ! (the threat is now ...dS -d4 Black has a strong centre, while the and c4-c3) 19 bXc4 tiJxe4 (this is the king on g2 is in some danger. Therefore point; the bishop on b2 is hanging in White must play with some precision: some lines) 20 tiJdxe4dxe4 21 l:[e2e3+ 18 .l::r.bl! (West shows that other moves (21... tiJcS is also possible) 22 'iie4 11c6 are good for Black; e.g. 18 es?! d4+ 19 with an unclear game, the threat is f3 tiJdsor 18 f4?! h6 19 eS hxgs 20 exf6 now ...tiJc S. ':'xf6 with a clear advantage) 19 ...d4 19 b13) 18 eS tiJhs 19 tiJa4(or 19 a4 d4 tiJa4 'ii'bs 20 tiJb2 (or 20 .1i.a3 ':'c8 21f4 20 tiJe4'ii ds with compensation: Black es) 20... 1txa2 21 tiJxe6 ':'b8 22 tiJxg7 has a strong centre and pressure on the �xg7 23 eS tiJds 24 e6 tiJ7f6 25 .1i.gs light squares and against the es-pawn) tiJC3 26 J:[al J:[xal 27 1:[xal 'iWc6+ and 19 ....:xa 4!? (19 .. JUS is also possible; e.g. Black regains the pawn with an unclear 20 .l::r.acl c4 21 .1i.d4 na8 intending game. ....:.a f8, or 20 tiJxcs tiJxcs 21 tiJd4 .1i.f8 15 .1i.f4?1 22 tiJxfs exfs with an unclear game, The critical test of Black's play is 15 and not 21 'iixcS?? because of 21... d4 dxe6! fxe6 16 tiJbs, when White looks to winning the knight on f3 and the have a slight advantage, -although the game) 20 bxa4 d4 21 'iib3 'iic6 and game remains very complicated:
5 2 The Classical Variation
16 ...'iVb 7! (if 16 ..."iYc 6?! 17 a4 dS 18 exds 22 "iYe3) 22 ttJxdS ttJxdS(or 22... l:txdS 23 ttJxdS, as in Y.Yakovich-P.Cramling, 'ike1) 23 l:td7, when the reduced mate Stockholm 1999, then 19 'i'e4 is good rial gives him good hopes of a draw. for White) 17 a4 dS 18 exds exds (again 20... ttJfxdS 21 l::te2? ttJxc3 18 ... ttJxdS?! 19 "iYe4gives White a com fortable advantage) 19 ttJgs d4+, and now White should play 20 53 Play the Benko Gambit bxa6 g6 6 ltJC3 i..xa6 7 e4 i..xf1 8 �xf1 llxd3 ltJds 27 l1adl i..d4 with an un d6 9 g3 i..g7 10 �g2 ltJbd7 11 ltJf3 0-0 clear position. 12 h3 l:ta613 l:le1 'ita8 14 l:.te2?1 18 f4 A thematic move in the Classical: Quite ambitious, but also weaken- the rook makes room for the queen to ing. double up behind it on el, while cover 18 .••c41 ing the second rank and enabling the This move is especially effective cl-bishop to develop gs or f4 if desired. when ...ltJcs-d3 is a possibility. The In this precise position, however, Benko player should always be aware White's plan is a bit slow, and I would of this theme, and it can be found quite recommend the previous game's 14 a few times in this book. Nevertheless, 'iWc2 as a more serious test of Black's other moves were good too, such as set-up. 18... h6 19 ltJf3 c4, or 18... dS, now that 14 ...l:tb8 the white king has been further ex The immediate 14... e6 is also play posed by f2-f4. able. With the text Black maintains all his options, such as ...ltJb 6-a4, ...ltJe8- c7-cS or ....:.b4, depending on what White does next. 15 'iVe1e6 With the white queen having left the d-file this move is even stronger. 16 dxe6 fxe6 17 ltJg5lite8 17 ... dS!? is also interesting; for in stance, 18 exds exds 19 i..f4 d4+ 20 �h2 :bb6 21 i..C7 .l:r.b422 ltJce4 ltJxe4 23 %:txe4 ltJf6 24 ':'e2 d3 25 ':'e3 ':'xb2 26 19 i..e3 54 The Classical Variation Preventing the black knight ma (not 30 19 ...e5 20 litdl116 21 liJf3 exf4?! the strong threats of ...l:tf2 + or ...litxh 3+, Correct was 21...liJxe4 22 liJxe4 whereas White's attack is stymied. 'iVxe4 23 �C5 'i!Vb7 24 �xd6 e4 25 liJe5 28... liJxd6 29 "C3 'ili'f6 30 "xf6 liJxf6 ttJxe5 26 �xe5 (or 26 fxe5 litxa2) 26 .....txe 5 27 fxe5 .l:Ixa2with a level po sition. 22 gxf4? Returning the favour. 22 �xf4 would have been good for White; e.g. 22 ...liJC5 23 l:td4 liJh5 24 litxC4liJd3 25 'ii'd2 liJdxf4+ 26 gxf4 �xC3 27 bXc3 d5 28 l:.d4dxe4 29 11exe4. 22 ...liJxe4 23 liJd5? Of course not 23 liJxe4 'iVxe4 24 �C5?? "xe2+ (this was the point 31 liJe5? 21...exf4), but 23 �d4 would only fa 31 l:tb6woul d offermo re chances of vour Black slightly. holding. 23 .....b7 24 �d4 "xd5 31... liJfe4 32 h4 c3 33 liIc2 liJf5 The twin threats of ...liJxh 6 and ...liJe 3+ are decisive. 34 35.•. liJxh6 36 �e2 liJc40-1 Game summary The plan with 14 l:le2 and 'ili'e1 is not effective here. Topalov gets an unclear Now Black is clearly better, due to position and wins a nice (if imperfect) the strong knight on e4. game, illustrating many of the themes 25 ..txg7 'ii'f5?! for Black: such as the ...e7-e6 and ...c5- 25 ...'ii b7! was more accurate. c4 breaks and play against the exposed 26 �xh6 l:lxa2 27 ':xd6 l:txb228 l:txb2 white king on g2. Nevertheless, I would Or 28 l:1dd2l:tb 3 with the initiative; only recommend ...l:ta6 and .. :�a8 for for example, 29 ':xd7 ':'xf3! 30 1:[g7+ the Benko player who is desperate to 55 Play the Benko Gambit avoid a draw. The theoretical status After d6 glDf3 �g7 10 h3 10 •••lDbd7 My recommendation is to play the classical set-up with ...lDbd7, ... 0-0, ... "as and ... l:tfbB, after which when Black's various ideas include ...l:tb4, ...lIa bB, ...lDb 6-c4/a4, and sometimes ... cS-c4 and ... lDcS-d3. However, 10 ...0-0 11 'itgl 'irb6 12 'ith2 lDa6is an interesting alternative. This is playable and can be quite dan gerous, so I will show a few ideas with this line; but as White can probably In the first six games of this chapter achieve a small advantage, I think the White consolidated his king with 10 g3 game continuation of ... lDbd7 should and 'itg2. With the text White prepares be preferred. a different bolthole for his majesty. While the manoeuvre h2-h3 and 'itgl h2 takes one move longer, on this square the king is a bit more secure; and as we have seen, White often plays h2-h3 in any case in order to prevent ... lDg4, so the tempo is not wasted. Fur thermore, Black's plan with ... e7-e6 is less effective without the white king on the long diagonal. On the other hand, the thematic e4-eS is now more diffi cult for White to achieve. From this position White has two 56 The Classical Variation main moves: 2004 - a good illustration of how a) 13 J:tel ttJb4 (adirect approach, things can go wrong for White after a Black is planningto place the knight on few inaccuracies. d3 after ...'ii' a6 - it works well here) 14 b) 13 'ike2! ttJb4 14 a4 e6! (the the :e2?! (a thematic move in the iLxfl matic 14 ...'iVa6 ?! is dubious here be lines, but ineffective here; then again, cause of 15 ttJb5! .l:r.fb8?! 16 I1a3 and after 14 Ve2 'iVa6! the endgame is fine White is better - Black should be care for Black) 14... 'iVa6 15 cJilgl ttJd716 iLg5 ful to avoid a blockade with a2-a4 and lUe8 17 ttJel ttJe5 18 a4 h6 19 iLh4?! ttJb5, especially after the light-squared (preferable was 19 iLd2,though Black bishops have been traded; 15 ...ttJd7 16 still has good compensation) 19 ...l:teb8 iLg5 is also good for White) 15 lldl (now Black has a slight advantage) 20 exd5 16 exd5 lUe8 17 'iVC4 'iVa6 18 l:ta3 (20 iLxe7 runs into 20 ... g5 and Vxa6 :txa6 19 %:tbl ttJe420 ttJxe4':'xe4 21 ...ttJg6, picking up the trapped 21 b3 with a slight advantage to White. bishop) 20... :b7 (or 20... ttJc 4!) 21 b3 11 cJilg1 %:tab8 22 :e3 c4 It is not too late for White to revert to the main lines with 11 g3 and 12 cJilg2,sh ould he wish to do so. Indeed, we saw this move order used in Game 11. 11..• 0-0 12 cJilh2Va s 13 l:[e1l:tfb8 23 'iVe2? (more resilient was 23 ttJbl, when 23... cxb 3 24 l:taxb3 ttJC4 is only clearly better for Black, whereas after the text Black is winning) 23 ...ttJed3 (the black pieces are much more active than their white counter This position can be compared with parts) 24 'iVd2 ttJxel 25 :xel ttJd3 26 the main line 9 g3 iLg7 10 cJilg2 0-0 11 %:tdl cxb3 (or 26... .l:1x b3! 27 l:txb3 ':'xb3 ttJf3 ttJbd7 12 h3 'iVb6 13 :el J:tfb8, 28 ttJe2 l:tb2 29 'ife3 'iVxa4 with adeci seen in Games 10 and 11. The only sive advantage) 27 ttJb5 ttJC5 28 'iVb4 difference is that the black queen is on %:ta8 0-1 V.Jianu-M.Ghaem, Istanbul as rather than b6, while White has 57 Play the Benko Gambit played �gl-h2 instead of g2-g3 and �g2. 14 lie2 As played by Piket in the similar po sition from Game 10. The main alterna tive, 14 'ifc2, is examined in the next game. But note that 14 es?! (as in Game 11) would be a mistake here due to 14 ...dxe s lS ttJxes?? ttJxes 16 lIxeS ttJg4+!, winning material. 14... ttJe8 14... �b4 is another option, prevent 16 'ife2 ing lS :C2 (as the e-pawn would be Black is okay after the alternatives hanging) and leading to an unclear as well. game. For example: lS 'ifel (after lS a) 16 i.gS?! leaves White weak on i.gs the logical follow-up is ls... lla b8, b2 and c3, so that 16 ...ttJC4 is then good creating pressure on the b-file, and if 16 for Black. b3 then 16... 'ifa6, inten ding ...cS -c4 with b) 16 a4?! is met by 16 ... 'i!Vb4 and compensation) ls ...ttJe8 (giving the g7- Black regains the pawn with a slight bishop more scope, and preparing advantage . ...ttJc7-b S) 16 ttJdl (notice that the oth c) 16 l:tbl ttJc4 leads to an unclear erwise thematic 16 i.gS?! is dubious game; e.g. 17 'iVe2 ttJes 18 a3 (or 18 here due to 16 . .J:lxb2! 17 l:txb2 'ifxc3 ttJelttJd7) 18 ... ttJxf3 + 19 Wxf3 l:1b3with and Black regains the exchange with the compensation, since the move a3 a2- better position) 16... 'iVa7 17 i.d2 i.xb2 has created a substantial weakness on (17 .. Jia4?! 18 a3 was good for White in b3. C.Garcia Palerrno-J.Nogueiras Santiago, 16.. .'iVa6 17 'ifxa6 .l:txa6 Cienfuegos 1984) 18 ttJxb2 :lxb2 19 a4 'ifa6 20 as ttJb6 21 'ifdl ttJf6 22 �gl ttJC4 with compensation. 15 liC2ttJb61 lS ...ttJC 7?!, preparing ...ttJb S, is now a little slow and allows White to gain an edge after 16 .td2 'ifa6 17 �acl i.xc3 (somewhat risky, but other moves are good for White too) 18 .txC3 ttJbS 19 'ilfd2 'ifxa2 20 b4 'ifa7 21 i.al and White is slightly better. 58 The Classical Variation This endgame is quite thematic for 22..• l2Jxe 5 23 .l:!.el l2Jxf3+24 gxf3 f5?! the Classical main line. Thanks to the Giving away most of the advantage. activity of his pieces, Black is perfectly Instead, after 24... .ie S+ 25 Wg2 ':xa2 fine, and often has greater winning or 25 l:[xes dxes 26 d6 l2Jg7 27 d7 l2Je6 chances in practical play. 2S dS'iV+l2JxdS 29 .ixdS l:txa2Bl ack has 18 .ig5 l2JC419 .ixe7?! good winning chances with his strong Safer was 19 ...l2Jxb2 20 %lb1 (but not passed c-pawn. 20 l2Jxb2 .ixb2 21 l:tb1? .ies+) 25 l:te6 l:txa2 26 'iStg2 .id4 27 .ih4cM7 20 ....l:Iab6 21 l2Jxb2 .ixb222 .ixe7:6b7 28 f4 l:ta6 29 1;:[e7+ 'iStf8 (now if 22... .ieS+, th en 23 l2Jxes l:lxb1 24 l2Jc4) 23 a4 l:lxe7 24 1kxb2 lIxb2 25 .u.xb2 lIxe4 and a draw is the most likely outcome; e.g. 26 as l:.a4 27 l:tb7 ':'xas 2S l2Jgs f6 29 l2Jf7 l:ta2 30 l2Jh6+ with perpetual check. 19... .l:!.xb2 l:txb2 20 l2Jxb2 21 e5? Correct was 21 l2JbS l2Jd322 J:.b1 f6 23 l2Jd2 cM7 24 .idS .l:txa2 25 l2JC4 .ifS 26 f3 :a4, when the game is in the bal ance, even if Black has some pressure. 21... l2Jd3 30 1:tb7?! After 30 :xeS+ WxeS 31 l2JC7+'iStd7 32 l2Jxa6 'iStcS 33 .ie7 Wb7 34 .ixd6 'iStxa6 35 .ies, the game would proba bly be drawn. 30 .•..if6 31 .ixf6? A definite mistake, after which Black is clearly better again, whereas 31 .igs would keep his advantage to a minimum. The rest of the game was probably played in time trouble, as it was not conducted with any great ac Now Black is clearly better. The a2- curacy, but Black triumphed in the end. and eS-pawns, as well as the pieces on 31..• l2Jxf6 32 l2JC7l:tu 33 l2Jb5 l2Je4 34 a1 and c3, are all vulnerable. l:txh7 l2Jxf2 35 Wf3 l2Je4 36 l2JC7'iStg8 37 22 l2Jb5?! l:te7 l:ta3+ 38 'iStg2 l:.a7 39'iStf3 c4 40 22 l:.b1! l2Jxes 23 l:tbS l2Jxf3+ 24 gxf3 We3 c3 41 Wd3 .:tb7 42 l:1e8+ Wf7 43 i.xc3 25 'iStg3was a stronger defence. .l:!.c8 l:tbl0-1 59 Play the Benko Gambit Game summary would be thematic, but then White can After White played the prophylactic gain an advantage with 15 ttJdl, pre rook moves J:[el-e2-c2, covering c3, paring .itd2. For example, lS ... C4!7 16 Greenfeld equalized with the thematic a4! (16 .ite3 ttJfd7 17 l:[e2 ttJcS 18 ttJd2 ...'Wa s, ... l1fb8, ...ttJb6 and ...'Wa6. Black ttJb3 19 ttJxb3 cxb3 20 'Wxb3 ttJa4 was even got some pressure in the end unclear in P.Haba-M.Prusikin, German game and won aftermi stakes by both League 2003) 16 ...ttJxa4 (or 16 ...ttJfd7 17 sides. The knight going to b6 and c4 .itd2 intending a4-as) 17 ttJC3 ltb4 18 was quite thematic. So again the con .itd2with an annoying pin on the a-file clusion is that Black is doing fine. and a clear advantage for White. 15 .itd2ttJC7 16ttJd l If 16 a4 'ifb6 17 l:[ebl, then 17 .. :iVb3 Game 15 18 'Wxb3 .l:.xb3 19 as .l:.ab8 20 ttJdl ttJes C.Gabriel-D.Sermek 21 ttJxes .itxes+ 22 c;t>gl .itxb2 23 ttJxb2 Pula 2000 l:[xb2 24 l:[xb2 l:[xb2 and the endgame is level. 1 d4 ttJf6 2 c4 cS 3 d5 b5 4 cxb5 a6 5 16 .. :iia4 bxa6 g6 6 ttJC3 .itxa6 7 e4 .itxfl S c;t>xfl d6 9 ttJf3 .itg7 10 h3 0-0 11 c;t>gl ttJbd7 12 'ifolh2'ii' a5 13 .l:[ell:.fbS 14 'ii'C2 Black has sufficient compensation. As usual the endgame is perfectly ac ceptable for Black with his active The second of White's two usual pieces, especially the rooks on the a choices here. The alternative, 14 l:[e2 and b-filessuppor ted by the bishop on intending ':'C2,was covered in the pre g7. vious game. 17 'WCl 14 ...ttJeS ! Or 17 'Wxa4 l:[xa4 18 .itC3 ttJbS 19 The most accurate move here. In .itxg7 c;t>xg7 20 eS ttJd4an d Black is do stead, 14... ttJb6, heading for c4 or a4, ing fine. 60 The Classical Variation 17 ...tt:JbS 18 tt:Jc3 'iVa6 19 a4 has a dangerous initiative) 26 tt:Jxe3 19 tt:JxbS 'iVxbs 20 �C3 �xC3 21 l:.lb2with a drawish position. bxc3 'iVb2 22 -'ifxb2 l:Ixb2 shows an 24•.. c4 2S .l:r.a2 other drawback to ..to>gl-h2: the f2- Here 2S tt:Jd2! tt:JcS 26 'ii'e2 would be pawn is en prise. slightly better for White due to the 19 ...tt:JC7 weak c4-pawn. Black is ready to double the rooks on 2S.•• I1bl 26 l:r.d2 'ii'xe3 27 tt:Jxe3 .l:r.xel the b-file and, by retreating his knight 28 tt:Jxel.l:r.a8 Vz-Vz to c7 again, prevents White from set ting up a blockade on bS. More risky is Game summary 19... tt:Jd4 20 tt:Jxd4 �xd4 21 tt:JbS �xf2 With 14 1WC2, followed by �d2 and 22 l:tf1 �d4 23 tt:Jxd4cxd4 24 b4, when tt:Jd1,White aimed to secure his queen White might hope to make something side against opposing counterplay. of his passed pawns. However, after Black's thematic and 20 l:ta3l:tb 6 21 tt:Jdl l:tab822 �C3 �xC3 accurate ...tt:Je8-c7, White found it diffi 23 bxc3 cult to undertake anything active ei ther. In the game Black equalized. Game 16 Y.Shulman-A.Khalifman FIDE World Cup (blitz playoff)., Khanty Mansiysk 2005 1 d4 tt:Jf6 2 c4 cS 3 ds bs 4 cxb5 a6 S bxa6 g6 6 tt:JC3 �xa6 7 e4 �xfl 8 ..to>xfl d6 9 tt:Jge2 23... 'ifd 3?! With his knight no longer required on the queenside, Black might prefer 23 ...tt:Je8 and ...tt:Jef6 , when the game remains unclear. 24 'iVe3 If 24 c4, then 24 ... l:.b1 2S 'ilt'e3 'ii'xe3 (2S ...'ii' C2?! is quite risky; e.g. 26 tt:JC3 l:.xe1 27 tt:Jxe1 'iVb2 28 tt:JbS tt:JxbS 29 axbs tt:Jb6 30 1;k3 J:.a8 31 l:tc2 'ili'b1 32 es l:ta4 33 exd6 exd6 34 .l:te2an d White 61 Play the Benko Gambit A solid move; the knight supports its 25 �hl �d4 26 l:tffl lDe2 0-1 R.Gross colleague on c3 and therefore hinders P.Benko, US Open, Aspen 1965. Black's counterplay somewhat. On the b) 13 a3! is more interesting. This other hand, the advance e4-e5 is now move restricts the knight on a6, since if not really a theme. The classical plan of it continues to c7 White can play b2-b4 ...lDbd7, .. :iVa5, .. .l::tfbS, followed by showing the black queen to be less for ...lDb6-a 4 (as played in the game) gives tunately placed on b6, while 13 ...c4 al Black enough compensation for the lows 14 �e3 lDc5 15 lDd4 and lDc6. pawn. Although Shulman has scored Therefore Black returns to the original quite well with 9 lDge2, in my opinion knight manoeuvre seen in Games S the main lines with 9 lDf3 offer White and 9 of this chapter: 13 ...lDg4 14 h3 more chances to play for an advantage. lDe515 f4 (or 15 Wc2 lUcS with com 9 .••�g7 10 h3 pensation) 15... lDc4 16 'fia4 lDb4 17 10 g3 is less effective with the Wh3 Wa6 lS axb4 cxb4 19 lDdl lDd2 knight on e2, as the white king is then (19... lDa3 !? may be even better) 20 even more exposed on g2. Benko him �xd2 'iVxe2+ 21 lDf2 'iVxd2 22 l:thdl self won a nice game as Black in this We2 23 'iixb4 litfbS and Black has com line: 10... 0-0 11 �g2 'iVb6 12 :bl lDa6 pensation for the pawn. (intending ... lDC7an d ... e7-e6, attacking the centre; in the case of d5xe6 Black can play ...d6-d5, which is not without venom): 10 ..• 0-0 11 �gl lDbd712 �h2 Was Probably the most common set-up for Black in the Benko. As in the previ ous two games (where White played a) 13 b3?! lDC714 f3 e6 15 dxe6 fxe6 �gl-h2 in conjunction with lDf3), the 16 �e3 'iVc6 17'iVd2 ?! d5! (Black is al plan of .. Jita6 and ...WaS is less effective ready clearly better) lS exd5 exd5 19 with the white king off the long light �f4 d4 20 lDa4lDfd5 21 lthfl lDe6 22 squared diagonal. .l:r.f2 d3 23 lDgl lDexf4+ 24 gxf4 lDxf4+ 13 1i'c2 lDb6 62 The Classical Va riation the immediate 16... ttJd7 was more ac curate, so as to create the possibility of ... i.xc3at once. 17 Iihel?1 Here White could have played 17 b3, and if 17... ttJd7 then 18 a4 and 19 ttJb5 with a solid advantage. 17 ...ttJd7 18 i.g5'iit f8 19 'iii'd2 i.xC3 This idea occurs quite a few times in this book. Here it is well timed: Black gets the a2-pawn, and the king is safe On its way to a4, seeking to under enough on f8. mine White's solid defence of c3. 20 bXc3 ':'xb1 21 l:txb1 'ili'xa2 22 llb2 14 l:tbl 'iVC4 Deviating from the two players' Now the position is unclear. rapid playoff game: 14 l:td1 ttJa4 15 23 l::tb7 'iite8 24 'it'f4 i.d2 ttJxc3 16 ttJxc3 l:tfb8 17 l:telttJd7 There was no need to give up the c 18 b3 'ita6 19 l:tabl ttJe5 20 lle3 c4 21 pawn, but Black was okay in any case. 'iitgl, which Shulman in fact went on to For instance, 24 'ifb2(pl anning l:txd7)is win, although Khalifman would natu easily met by 24... 'iWa6 an d 25... .:.a7. rally have had an improvement ready 24... 'i!t'xC3 somewhere. 14 ...ttJa4 15 i.d2ttJX C3 16 ttJxC3l:tfb8 25 'iWh4?? 25 i.xe7 'iitxe7 26 .l:txd7+ 'iitxd7 27 After the exchange on c3 Black is 'iVxf7+ 'iitc8 28 'iVe8+'iit b7 29 'iVc6+with ready to send his other knight round to a perpetual was called for. the queenside via ...ttJd7 -e5 (or b6) and 25•.. h5 ... ttJd3 or sometimes ...ttJC4. However, It seems Shulman didn't see this 6 3 Play the Benko Gambit possibility, which I suppose is forgiv bxa6 g6 6 ttJc3 .txa6 7 e4 iLxf1 8 �xf1 able in a blitz game. But Black is now d6 9g4 just a pawn up with the initiative, whereas the white pieces are not play ing together. The pawn on e7 can no longer be captured because of 26 iLxe7 'it'eS+, winning a piece. 26 iLe3?1 This and White's next move only make things worse. 26... 'it'es+ 27 g3 l:.a4 28 llC7 �d8 29 l:tc6 :xe4 30 .td2 f6 31 .tf4 'iVxds 32 l:ta6l:te1 0-1 To his credit Shulman recovered Instead of the standard positional from this defeat to win the second blitz g2-g3, White uses the g-pawn far more game (as Black), and then progress to aggressively. This idea was semi the next round in sudden death. popularized by Yasser Seirawan, who won some nice games with it in the Game summary 1980s (he later switched to S b6). Al Shulman has played 9 ttJge2nu merous though 9 g4 is a rare continuation times with good results, though it is nowadays, it can still be dangerous for otherwise relatively uncommon. For the unprepared Benko player. One of Black the plan with ... 'iIi'as and ... ttJb6- the main ideas is to continue with g4- a4 is quite effective and should guar gs, attacking the knight on f6, and then antee sufficient compensation for h2-h4-hS in order to attack the king equal chances. In the game Benko ex after Black has castled kingside. Typi pert Khalifman played the thematic cally White will play 'it>g2,ttJe2-g 3 and ...iLx c3, regaining the pawn satisfacto aim for a mate on the h-file. rily. Then, probably due to the fast time The best replies, in my opinion, are limit, Shulman blundered and Khalif 9 ...'iYc8 and 9 ...hS! , when Black man won in a rather simple way. shouldn't have any problems achieving equality. After 9 ...'iVc8 Black has good potential counterplay on the queenside Game 17 too. However, I would recommend I.Sakovich-Z.Lanka 9 ... hS, because the white king on f1 can Riga 1980 then be in some danger in the middle game. 1 d4 ttJf6 2 c4 cS 3 ds bs 4 cxbs a6 S 9 ... hSI 6 4 The Classical Variation In the spirit of the Benko Gambit, 12... 0-0 13 �f4 tiJa614 'iVe2 (or 14 'iVd2 Black simply attacks the pawn. After c4! intending ... tiJacs with an unclear g4-gs, there is no h4-hs anymore, game) 14 .. :ii'as 15 :hCl :fb8 (the clas wh ereas Black will have chances for sical way of playing the Benko) 16 l:tabl good play on the queen side. tiJCl (heading for bs) 17 es e6! (a cool 10 g5 reply) 18 a3 (or 18 dxe6 tiJxe6 19 �g3 10 gxhs?! is not the most logical tiJxes) 18 ...dx es 19 tiJxes tiJxes 20 move: opening the file for the black �xes �xes 21 'ii'xes C4 22 tiJe4(a fter rook on h8 and creating weaknesses on 22 l:tdl l:td8 23 tiJe4tiJe8 Black is ready the king side seems a little unwise with for ...l:txd s, so White should probably the white king sitting on fl, so Black opt for 24 tiJC3 tiJC7 25 tiJe4 with a has good chances to gain an advantage repetition) 22... 'ii' xds 23 'ii'f4 tiJe8 and in practice. Nevertheless, it seems the game is roughly equal. White can keep the balance with pre 12 ..•tiJa6 cise play: 10 ... tiJxhs 11 �g2 �g7 12 On the way to Cl and bs, or some tiJf3! (after tiJge2 tiJd7 Black is slightly times d3 after ... cs-c4 and ... tiJcs. better, since the king on g2 is none too 13 tiJf3 0-0 14 'ii'e2 'ii'C7 safe) 12 ... tiJd7 13 a4 tiJes 14 h3 and Preventing e4-es at the cost of a Black has sufficient compensation for tempo. Black might also consider the the pawn: there are fair prospects for direct 14... 'iVas !? 15 es tiJb4 which an active game, whereas White has leads to a sharp, unclear game, though several weaknesses. it still requires some testing in practice: 10... tiJfd 7 11 �g2 �g7 16 :dl! (the most dangerous move; 12 f4 16 e6 tiJb6 17 exf7+ l:txf718 'iVe6fo rces Consistent, but also risky. 12 tiJf3 is Black to defend accurately, but a quieter alternative, though Black is 18 ...'ilia6 ! 19 fs 'ii'd3 20 tiJh4 'iWc2+ 21 doing alright here too. For example: �h3 'ii'd3+ draws by perpetual check) 65 Play the Benko Gambit 16... c4 17 e6 ttJC5! (17... ttJb 6? 18 f5 gives 18 a3?! ttJa2! White a dangerous initiative; e.g. 18... gx f5 19 exf7+ .l:txf7 20 g6 ':'f6 21 ttJh4 ttJ6xd522 ttJxd5 ttJxd523 'ikxh5 e6 24 'ilVh7+ �f8 25 �h6 �xh6 26 'iixh6+ �e8 27 ttJf3 and White is clearly better) 18 exf7+ :xf7 19 'ikxC4 ttJC2 20 .:tbl �xC3 21 bxc3 Wxa2 22 'iixa2 1:[xa2 23 1 ttJe424 l:td3ttJC5 25 .:tdlttJe4 with a repetition of moves. 15 �e3 .l:lfb8 A nice trick, probably overlooked by Sakovich. White is now clearly worse and exacerbates this with his next. 19 ttJxa2? 19 1:[c2 wasne cessary. 19 ...,Uxb2 20 :C2 ':XC2 21 "ii'xc2 "iWxa3 The point: the knight on a2 and the bishop on e3 are hanging. 22 �2 'ikxa2 Black has enough compensation for the pawn. 16 .ll.adl?! 16 l:ihcl would be more solid. 16... 'ii a5 With White's bishop obstructing his play on the e-file,the black queen goes to as after all. Actually, White ought to consider 17 e5 anyway; e.g. 17 ... dxe5 18 �Cl exf4 19 'iVxe7 l:.b7 with equality, though I would prefer Black over the White is lost. board as his pieces are quite active. 23 :cl �b2 17 l:tclttJb4 23 .. .f5! is even more convincing. Now Black is slightly better, primar 24 lidl ttJb6 25 ttJd2 lira3 26 :bl ttJa4 ily because White's reply does not 27 'iYC4 ttJC3 28'iYxa2 :xa2 29 .l:.el�a 3 work. 30 �f3 �b4 31 :cl? ttJxd5 0-1 66 The Classical Variation Game summary For example, 7 dxe6 (7 ttJC3?! ttJxdS The variation with 9 g4 can be quite 8 ttJxdS exds 9 'ilfxds ttJc6 gives Black dangerous if White manages to build good compensation, due to the lead in up an initiative with h2-h4, ttJge2-g3 development, the white weaknesses on and h4-hS. However, if Black responds e4, f2 and c4, and the queen on dS with 9 ... 'iVc8 or 9 ... hS (my recommenda which can be chased) 7 ...fx e6 8 ttJC3 tion) he should equalize comfortably; c4!? (alternatives are okay too, but this indeed, the latter neutralizes the threat move seems the most principled to me) of h4-hS altogether. In the middlegame 9 e3 (9 e4?! is risky, weakening the dark Sakovich played inaccurately and al squares and the king on el; Black must lowed Lanka a nice win. For White I be slightly better after9 ... i.b4, and can think g2-g4 should probably only be follow with ... 0-0, ... ttJc6, ...'ilb 6, or pos- used as a surprise weapon. sibly ... d7-dS) 9 ... i.cS 10 ttJf3 0-0 11 ttJd4�6 12 ttJa4'iVas+ 13 i.d2i.b4 14 ttJC3, when Black can choose between Game 18 14 ... eS, 14 ... 'ilfb6, or 14 ... 'ilfc6 followed M.Mancini-J.M.Degraeve by ... ttJc6 and ... d7-dS, with compensa- French Team tion for the pawn in all cases. Championship 2004 But note that this is only possible if Black plays an early ... ii.xa6, whereas I 1 d4 ttJf6 2 c4 cS 3 ds bs 4 cxbs a6 5 am recommending s ...g6 as the most bxa6 i.xa66 f4 accurate move order (for the explana The usual move order is 6 ttJC3 and tion see the notes to Game S in the then f2-f4. previous chapter). Then 6 ttJc3 i.xa6 7 6 ...g6 f4 reaches the position below. Here 6 ...e6 !? is an interesting at 7 ttJc3 tempt to exploit White's early f2-f4. White's idea is to continue with 6 7 Play the Benko Gambit tiJf3, e2-e4 and J::txfl, most likely fol tinued 9 ...0-0 (instead, 9 ...tiJg4 +! is lowed by e4-eS, looking for attacking quite an annoying check: Black must be chances. This direct approach was slightly better after 10 'iitg3 i.xc3 11 played now and then in the late 1980s, 'iitxg4 i.g7with the exposed king now in particular by GM Jacob Murey, Now on g4, or 11 bxc3 tiJf6- I would be very it is out of fashion, and for good rea happy if anyone played like that son: Black should have few problems if against me in the Benko) 10 h3 (pre he plays correctly, whereas the white venting ...tiJg4+) 10 ...e6! (opening the king will often feel quite vulnerable; centre; a logical strategy with the white and Black will generally have his king on f2) 11 e4?! (11 dxe6 followed by queenside counterplay as well. e2-e3 was less ambitious, but superior 7 ••:i!Va sl to the move played) 11 ... exds 12 eS Preventing the immediate e2-e4 by tiJe4+ 13 tiJxe4 dxe4 14 tiJgs c4 15 pinning the knight, and enabling Black tiJxe4dS! to recapture with the queen after a later exchange on a6. The straightfor ward 7 ...i.g7 8 tiJf3 0-0 9 e4 gives White some chances to prove an open ing advantage. 16 tiJgs (no better is 16 exd6 :te817 tiJc3 tiJd7 18 i.e2 tiJcS, when Black's activity gives him huge compensation for the pawns; or 16 tiJf6+i.xf6 17 exf6 'iVb6+ and Black is very much on top of 8i.d2 things after 18 'iitg3 d4 19 i.e2tiJd7 or The most sensible move, preparing 18 i.e3 'ilVxb2+ 19 i.e2 (3) 16 .. .f6 17 e2-e4 again. White has a few other op exf6 i.xf6 18 tiJf3 tiJc6 19 'iitg3 �cS 20 tions, but none of them should pose l:tbl?! (20 'iith2) 20... i.c8 21 i.e3? (a any problems for Black: tactical trick which doesn't work) a) 8 tiJf3 i.g7 9 68 The Classical Variation four after 25 g3 l:r.xf4+. Quite a nice 0-0 14 .i.e3liJC5 15 0-0 liJg4an d Black game by Fedorowicz. is at least slightly better, with good b) 8 'ii'd2 al so prepares e2-e4 but prospects of increasing his advantage. looks a bit artificial, while the queen 10 e4 d6 11 .i.xa6'iVxa 6 12 <;i;>f2?! doesn't support e2-e4-e5 so well from Stronger was 12 'i'e2! liJbd7 13 d2, so Black should equalize quite eas 'ii'xa6 l:txa6, although Black has suffi ily. For example, 8 ... .i.g7 9 liJf3 (allow cient compensation for the pawn. ing White to recapture on fl with the rook; after 9 e4 .i.xfl 10 <;i;>xfl d6 Black has the usual long-term compensation for the sacrificed pawn) 9 ... d6 10 e4 .ixf1 11 1'bfl 0-0 12 e5 liJfd7 and Black is doing fine. 8 ...i.. g7 9 liJf3 In many Benko endgames, as we have seen several times already, Black's pieces are so active that he is hardly ever worse, despite being a pawn down. Indeed, from a practical view point, I would generally prefer Black. For instance: 9 ...0-0 a) 14 a-a-a? (as in E.Bayer Black can also try 9 ...d6 and answer lFedorowicz, Porz 1988) is just a mis e2-e4 with ... c5-c4. This is a sharper take: after 14... liJg4 Black threatens option than the text and I would say is ....i.x c3, ...liJf2 and ...l:.xa2 and is clearly a matter of taste which one to play. better. After 10 e4 c4 Black has excellent play b) 14 <;i;>e2! l:r.b8(br inging the rook to on the dark squares and a good lead in the open file seems the correct move development for the pawn; e.g. 11 'iVc2 here) 15 b3 (or 15 l:r.abl liJb6 16 b4 (or 11 .i.e2 0-0 12 0-0 �6+ 13 <;i;>hl liJfd7 17 bXc5 liJxC5 18 e5 liJcd7 19 liJbd7 14 l:tbl liJg4 15 'ili'el liJe3 16 J:thcl l:td820 .i.e3 liJc4 21 exd6 exd6 22 .i.xe3 'i'xe3 17 'i'd2 'i'a7with compen i..d4 liJC5with compensation; one idea sation) 11 ... liJbd7 12 .i.e2 �6 (pre is ...l:r.e8+ and ...liJd3 xf4) 15 ...h6! (the venting castling for a while) 13 liJdl immediate 15... liJxe4 is quite possible, 6 9 Play the Benko Gambit but as White's best move is to develop tage) 20 ...l:txb l+ 21 �xbl l:txa4 22 �C2 the hl-rook anyway, Black plays a use fs 23 tDegs h6 24 tDe6.i.f6 , when Black fulwaiti ng move, covering gs in case of is slightly better with active, well a later ...f 7-fS) 16 l:thcl tDxe4 17 tDxe4 coordinated pieces and three strong .i.xal 18 ':xal nxb3 with an unclear pawns for the knight. game. 12 ••.Wb6 c) 14 �dl?! was played in 12 ... tDbd7! 13 l:tel c4 seems more G.Andruet-lFedorowicz, Wijk aan Zee precise. Afterth e text White might play 1989. Now rather than Fedorowicz's 13 'ifh3!? 'ii'a7 14 'iWC4 tDbd7 15 .l:I.hbl 14 ...tDg4 , I think 14... l:tb 8! again looks and the d7-knight cannot reach d3, best, intending 15 �C2 tDxe4! 16 tDxe4 though Black should still have enough ':xb2+ and then: compensation. 13 lIb1?!c4+ 14 �g3 Or 14 �1 tDa6 15 'ii'e2 tDb4 and Black is slightly better, one idea being 16 'ii'xc4 tDxa2! 17 tDxa2 lUc8 and Black's activity has increased. 14 ...tDa6 Cl) 17 �d3? c4+! 18 �e2 (or 18 �e3 tDb6, followed by ...tDxd s+ and ...c4-c 3) 18... C3 and Black wins the piece back with a big advantage. C2) 17 �dl?! tDb6 18 a4 tDxds19 as fs 20 tDg3 tDC3+ 21 �el tDbl 22 tDe2 ds and Black is clearly in the lead. Heading for d3. c3) 17 �Cl (the best move) 17 ... tDb6 15 We2 tDC516 �h3 18 l:tel tDxds 19 a4 J:tb8 20 l:tbl (20 16 'ii'xc4? runs into 16 ..:iVa6 ! 17 l:ta2?! is met by 20... C4 !, when the Wxa6 tDfxe4+ 18 tDxe4 tDxe4+ 19 �h 3 threat of ...c4 -c3 is quite serious; e.g. 21 tDf2+ 20 �h4 .i.f6+ and White loses �dl .i.b2! 22 �e2?! c3 23 liitbl liitb4 24 material. Or if 16 es, then 16... tDh s+ 17 tDf2 liitaxa4 or 22 �C2 c3 23 .i.xc3 .i.xc3 �h3 tDd3 18 g3 .i.h6 and Black is bet 24 tDxC3 ':'c8 25 l:ta3 l:.ac6 and Black ter; the threats are ....i.xf4! and ... tDf2+. regains the piece with a clear advan- 16 ...tDd3 17 .i.e3?1 70 The Classical Va riation 17 l:thfl followed by 18 ttJel, chal and ...c5 -c4, and obtained compensa lenging the d3-knight, was more resil tion for the pawn without too much ient. effort. On move 12, Mancini should 17 .. :iWb718 g3 e6! probably have preferred 12 'ife2, aim A thematic attack on the white cen ing for the endgame in which he at tre. least has equal chances. Instead, after 19 dxe6 ttJxe4 20 ttJxe4 'ili'xe4 21 ttJg5 12 'iitf2?! the white king became rather 'iYf5+ 22 g4 'iVd5 23 exf7+ 'iith8 24 b3 h6 exposed and Degraeve won nicely. Summary The Classical has lost some of its popu larity nowadays. White's long-term plan is simply to neutralize Black's ini tiative and then play e4-e5, or perhaps set up a blockade on the queenside. However, as we have seen in this chap ter, Black can equalize quite comforta bly, and I don't think too many theo retical discussions will occur here in Here 25 'iVf3 ! was forced, when the foreseeable future. Black is at least a little better after If White allows 12 ... ttJg4, Black has 25 ... 'ii'xf3 26 ttJxf3 llxf7i e.g. 27 bXc4 ideas like ....ix C3, ...ttJg4-e 5(-c4), ....:tfb 8 .:te7! 28 .id2 ttJf2+ or 28 .ib6 ttJxf4+. and ...'iVa6, after wh ich White has failed Instead, White blundered with: to demonstrate any advantage. Even if 25 l:r.hdl??hx g5 he plays 13 'ife2an d delays chasing the Now Black just wins a piece. knight, Black is still doing fine after 26 bxc4 'ii'e4 27 .l:f.xd3 13 .. :iWa5, as shown by Vachier Lagrave Unfortunately 27 'iVxd3 fails to in Game 8. 27 .. :iIi'f3 mate. In the main line White inserts the 27... gxf4 28 l:tbb3 lIxf7 29 'ii'f2 .if6 30 prophylactic 12 h3, preventing ...ttJg4 , 'ili'xf4 l:1h7+ 0-1 but then 12 ...'iVb6 13 l:te1 l:tfb8 gives Black a relatively easy game, the only Game summary drawback being that 14 e5 leads to a Th e idea of 6 ttJC3 and 7 f4 (or vice very drawish position (as seen in Game versa) is not a critical line against the 11). If Black wants to play for a win, he Benko. Degraeve responded with the might consider the dynamic 12... .l:f.a6, accurate 7 ...'iVa5 !, followed by the followed by ...'ifa8 and ... e7-e6 (as in standard moves ....ig7, ... d7-d6, ...ttJbd7 Games 12 and 13), although in this 7 1 Play the Benko Gambit case White probably can achieve a usual compensation for the pawn, slight advantage. which continues into the endgame. With h2-h3 the white king can also This set-up is also appropriate against shuffleover to h2 (instead of the more Shulman's line with 9 lLlge2. common g2), when there are some Finally, the aggressive options of 9 slight differences in the play. Here g4 or 7 f4 are not to be feared either. Black should opt for the standard Indeed, these can often backfire on ...lLlb d7, .. :ii'as, .. J:tfb8, and possibly White, whose own king position is then ...lLle8-c7- bS, when he obtains the more vulnerable in the middlegame. 72 Chapter Three 5 The 'O- uiet' e3 1 d4 tDf6 2 C4 cS 3 d5 b5 4 cxb5 a6 5 e3 Game 19 A.Karpov-A.Miles Tilburg 1986 1 d4 tDf6 2 c4 cS 3 d5 b5 4 cxb5 a6 5 e3 i..b7! 5 ...axb 5 6 i..xb5 'iWa5+ 7 tDC3 i..b7 8 i..d2 is a more common move order, but via this route White also has 8 tDge2tDxd5 (or 8 ...i.. xd5) 9 0-0 which is both critical and, in my opinion, fre This system doesn't have an estab quently quite difficult for Black. For lished name. Despite its quiet appear that reason I prefer the alternative ance 5 e3 is actually a sharp line, as 5 ...i.. b7, which has the advantage of White sacrifices a pawn for the initia avoiding this line. tive in many variations; and I am not 6 tDC3 suggesting a quiet approach for Black 6 bxa6 i..xd5 7 tDc3 i..c6 is roughly either. equal. For example, 8 tDf3 tDxa6 9 tDe5 After 5 e3 Black has three main (aiming for the bishop pair; in return choices: the classical 5 ...g6 (which is Black has an extra central pawn) probably okay), 5 ...axb 5 (see the first 9 ... tDb4!? (sharper than 9 ...tDb 8 which, note below), and 5 ...i.. b7, which is the however, is just as good since White move I am recommending. However, will mostly take on c6 anyway; e.g. 10 some of the gambit lines can be quite i..C4 e6 11 tDxc6 tDxc6, transposing to dangerous, so Black has to play with line 'c') and now we have the following some precision. possibilities: 73 Play the Benko Gambit In comparison to s ... axbs, White gains the extra option here of taking the a-pawn: 7 bxa6!? is examined in Game 21. 7 ...axbs 8 i..xbs 'ilVb6 a) 10 ttJxc6 ttJxc6 11 .i.bs 'ilb6 with an equal game. b) 10 a3 'ilb8 11 f4 (after 11 ttJxc6 ttJxc6 12 .i.bs 'ilb6 the position is basi cally the same as in line 'a') 11... 'ilb 7 12 ttJxc6 ttJxc6 13 i..d3 dS, followed by As noted above, this position often 14 ... e6 and Black is doing fine. arises via s ... axbs, or sometimes c) 10 .i.C4 e6 11 ttJxc6 ttJxc6 12 e4 s ..:ii' as+. Whichever way, one of the i..e7 is also TOughly equal. One theme main ideas behind ... .i.b7 is to attack for Black is a well timed ...d7 (d6)-dS; the ds-pawn, and hopefully win it, in e.g. 13 .i.f4 0-0 14 0-0 d6 15 a4 'ili'as 16 return for the b- or a-pawn. i..d2 'ifb6 17 b3 J::tfd8 18 l:[cl l1ac8 (or 9 'i'b3 18 ... dS) 19 l':tel ttJxe4 20 ttJxe4 dS with Defending the pawn. Alternatives to balanced chances. this move, in particular 9 .i.c4, will be 6 •••'iWas discussed in the next game. 9 ..•e 6 Trying to open the position for the light-squared bishop. 10 e4 10 i..C4 also transposes to Game 20. The text allows Black to achieve his ob jective by means of a small combina tion. 10 ...ttJxe41 11 ttJxe4 .i.xds 12 'ilVd3 'i!i'b7 The safest move. 12 ...fS is also pos sible and leads to sharp play after 13 7.i.d2 ttJg3 .i.xg2 14 ttJle2 .i.xhl 15 ttJxhl 74 The 'Quiet' 5 e3 i..e7 16 tDhg3 (or 16 .i.c3 O-O!, since 17 13 ... c41 14 Ji.xc4 .i.xc4 15 'iVxC4 d5 16 .i.xd7?? litd8 18 'ifg3 fails to 18... .i.f8, 'ifc2 dxe4 when the d7-btshop was unpinned but still trapped in O.Almeida Quintana M.Leon HOYos, Havana 2011) 16 ... 0-0 17 'iVxe4 Unsurprisingly, Karpov aims for the clarity of a simplifiedpositi on, though it 17 a4 (17 tDh5?! was played in may not be not the best move. others: I.Koga-Dunning, USA 1985, but then a) 17 1:[cl tDd7 18 'iVxe4 'iVxe4+ 19 17 ...'ifb 7! is clearly better for Black) fxe4 tDC520 'iii>e2 l::txa2 21 tDf3l::t xb2 22 17 ...tDc6 18 .i.C3 d5 19 'iVe3 tDd8 20 llc4Ji.e7 23 ':'al 0-024 11a7 .i.d6 25 e5 .i.xg7 (or 20 tDh S?! d4 21 tDxd4 cxd4 22 Ji.b8 was obviously fine for Black in �xd4 'ifb7, D.Gurevich-J.Nicholson, J.Litvinchuk-P.Wolff, USA 1985. Lugano 1985, when 23 'iii>e2 'iVg2 24 b) 17 fxe4 leads to a more unbal tDxg7 �g5 25 'iff3 'ifxf3+ 26 'iii>xf3 .i.f6 anced position, in which Black's active 27 .l:tgl 75 Play the Benko Gambit Now Black cannot be prevented from regaining his pawn with equality. 19 liJe2 Worse is 19 .i.e3?! �a4 20 liJf3 l:txe4 21 'itd2 .i.e7 22 a3 0-0 23 b4 .l:td8 24 'ite2 liJb6 and Black has a slight edge, which led to a win after his opponent later blundered in H.Kummer-L.Milov, Oberwart 1994. 19 ...liJcS 20 liJc3 liJd3+ 21 'ite2 liJxb2 22 liJbS .l::[a4 23 liJC3.l:ta8 24 liJbS l:!.a425 For example, 18 �bl (18 'il'xa8 liJc3]::t a8 26 liJbsYz-Yz 'iVxal+ 19 'ite2 .i.d6 20 1Ii'a4+ 'ite7 is worse for White, who has more diffi Ciamesummary culty developing his kingside) 18 ...1Ii'xa2 The S e3 line can be quite dangerous if 19 liJe2 (19 .i.c3 is met by 19 ... .i.cS! 20 Black is not careful . With S ....i.b7 Black .i.xg7 fS! 21 'iVb7 'ii'f2+ 22 'itdl 'ii'fl+ attacks the centre straight away, rather and ... �a2+ with advantage) 19 ... liJd7 (or than trying to win a pawn with the 19 ...fS !?) 20 liJC3 'iVa6 21 liJbs ':'c8 chance of White seizing the initiative. (Wolff), when Black has slightly the bet In the resulting position after 6 liJC3 ter of things with his extra pawn, al 'iVas 7 .i.d2 axbs 8 .i.xbS 'iVb6, Karpov though Nguyen Anh Dung-Z.Medvegy, defended the ds-pawn with 9 'ifb3 e6 Budapest 1999, was soon drawn after 10 e4, which allowed Miles to eliminate 22 'il'd3 'iVc6 23 'ite2 liJes (23 ....i.cS !?) 24 the centre by means of a combination. 'iVd4 'iVds 2S 'iVxds exds 26 .l:.hcl ':'xcl With White having failed to gain any 27 �xCl liJC4%-V 2. advantage, Black even had two oppor 18 fxe4 liJd7 tunities to sharpen the play (12 .. .fS and 17 ...'iIi'x b2). However, Miles was content to play safely as Black, after which a draw was the logical outcome. Game 20 D.Gurevich-J.Benjamin Chicago 1986 1 d4 liJf6 2 c4 cS 3 ds bs 4 cxbs a6 5 e3 .i.b76liJ C3 axbs 7 6 The 'Quiet' 5 e3 This move order allows White the 17 .Jtgs .Jte7Bl ack is slightly better; e.g. possibility of 8 tiJge2 again. Instead, 18 l:.dlh6 19 .i.xf6 .Jtxf6 20 'ii'g3 hS 21 Black should prefer 6 ... 'iVas, when 7 es h4 22 'iVf4 .Jte7 23 tiJe4 h3 24 g3 .Jtd2 axbs 8 .JtxbS reaches the game ifb8) 14... .i.e7 15 l:tadl tiJxbs 16 axbs position. was played in A.Graf-S.Halkias, Fuerth 7 .JtxbS 'iVas 8 .Jtd2'ilVb6 2002, and here 16 ...'ilVC7 is about equal. White has a small lead in development, but Black has the bishop pair and a su perior pawn structure. 9 ...e6 9 .JtC4 Alternatives include: a) 9 'iVb3 was Karpov-Miles (Game 19). b) 9 tiJge2tiJxds 10 0-0 tiJf6 11 tiJf4 10 'ilVb3 e6 12 'iie2 .Jte7with an unclear game. Going for the endgame - or perhaps c) 9 'ili'f3 (Graf's speciality) more accurately, queen less middle game - but Black should have no prob lems in that case. Sharper possibilities are: a) 10 e4 tiJxe4 11 tiJf3 ! (after 11 tiJxe4? exds 12 .JtxdS .JtxdS 13 'iVe2 'ii'e6, as in K.Zietak-K.Miton, Warsaw rapid 2005, or 11 dxe6?! fxe6 Black is clearly better; while 11 tiJge2 tiJf6 12 tiJf4 .Jtd6 is quite okay for Black too) 11 ... 'iVb4!? (the sharpest move, leading to an unclear position; otherwise the 9 ... tiJa6 10 tiJge2 tiJb4 11 0-0 tiJbxdS quieter 11... tiJf6 12 0-0 .i.e713 .Jtf4 0-0 12 e4 tiJC7 13 a4 e6 14 .Jtgs (after 14 looks sufficient for equality) 12 tiJes .Jte3tiJxb S 15 axbs 'ii'd8 16 :xa8 .Jtxa8 'ii'xb2! (since 13 tiJxe4?? now loses to 77 Play the Benko Gambit 13... 'ii' xe5; instead 12... tiJxd2 'ili'xd2 13 transposes according to where White 'ii'xd2 �d6 14 tiJxf7! rj;xf7 15 dxe6+ develops his king's knight. gave White a strong attack in C.Csiszar 12 tiJf3 M.Morvay, Hungarian Team Champi Instead: onship 1995) 13 J:.bl 'iixd2+ 14 'ii'xd2 a) 12 tiJge2 exd5 returns to 10 tiJge2 tiJxd2 15 11xb7 (15 'ii;Jxd2?! is dubious in above, or else Black can just play view of simply 15... �c8, when Black is 12 ...tiJb4 13 0-0 tiJbxd5 with equality. slightly better) 15... �d6 16 tiJxf7 tiJxc4 b) 12 f3?! is too slow, due to 17 tiJxh8 �e5 18 tiJb5 (not 18 tiJe4? 12 ...tiJb4 13 dxe6 fxe6 14 tiJb5 (or 14 exd5 19 tiJxc5 g6 and Black picks up the tiJge2 tiJd3+ 15 rj;fl �a6) 14... c4 15 knight on h8 with a clear advantage) �xc4 tiJc2+ 16 rj;dltiJxal 17 tiJC7+rj;f7 18 ... tiJa5 (18 ... exd5 is met by 19 f4!) 19 18 tiJxa8 �xa8 and Black was slightly lIb6 tiJc420 l:tb7 with a draw. better in Y.Zimmerman-A.Chernin, b) 10 tiJge2 exd5 11 'ii'b3 'ii'xb3 12 Hungarian Team Championship 1999. �xb3 tiJa613 0-0 tiJC714 .l:1fdl �d6 15 12 . ..exd 5 13 0-0 �e7 14 :fdl �el 0-0 16 f3 lUe8 17 �f2 �e5 (17... �f8 and 17.. Jlab8 also come into consideration) 18 lIacl�c6 was played in A.Beliavsky-D.5ermek, Terme Zrece 2003. White is trying to put some pres sure on the centre, but it is quite well defended, so Black has more space and can look for counterplay on the queen side. Black has no real problems. 10 ...'ii' xb3 11 �xb3 tiJa6 The position is roughly equal. Black has some nice central pawns and the possibility of play on the queen side; while White can try to put some pres sure on the centre, especially d5. 14 ...0-0 15 �el tiJC7 16 tiJh4?1g6 17 g3 l:tfb8 18 tiJg2 �c6 19 l:tabl C4 20 .ic2 tiJe6 21 tiJf4 The point of the White's long winded knight manoeuvre. Unfortu The immediate 11... exd5 followed nately, Black next move renders it by 12 ... tiJa6 is also fine, when play worthless. 78 The 'Quiet' 5 e3 Here the deflection 42 ...g4! 43 i.xg4 l1c6! was decisive, when White loses his passed pawns. 43 bS g4 44 b61 A great try. 44 ... gxf3+ 4S �xf3 21 ...d41 22 exd4 i.f3 23 tiJfe2 23 Ud2 tiJxf4 24 gxf4 i.b4 is also very good for Black. 23 ...i.b 4 24 dS tiJgs 2S h4 i.xe2 26 hxgs i.xd1 27 :xd1 tiJe8 4S ...:b2?1 The computer finds a win with 45 ...l:txa4! 46 b7 tiJg4, intending 47 b8'iWl:txf2 mate, or 47 'itg2 lIb4! 48 litb5 l:1xf2+ 49 �gl l::lbb2! 50 b8'iW':'g 2+ and mates again; but of course that would be very difficult to see over the board, especially afterth e shock of 44 b6. 46 b7 .llb3+ 47 �g2 lIxb748 tiJxb7l:ixa4 49 tiJd6 1ld4SO tiJbS l:tdSl:txd S1 S tiJxds Black is the exchange up, stands clearly better and eventually won. I will go quickly through the rest as it is not relevant for an opening book; though, as you will see, White puts up a good fi ght with his queenside pawns. 28 a4 i.e7 29 tiJbs i.xgs 30 i.C3 i.f6 31 i.xf6 tiJxf6 32 .l:td4 Ue8 33 l:[xC4 tiJxds 34 ':'d4 l:e1+ 3S �h2 tiJf6 36 i.d1 �g7 37 b4 ':c838 tiJd6 l:tc139 i.f3 :C2 40 �g2 lla1 41:f4 gs 42 l:tfS h6?1 79 Play the Benko Gambit Gurevich has done well to get this oftenpl ays 7 �d2 here. far, but Benjamin won the knight end game anyway. 52 'iii>f3 'iii>f6 53 'iii>e4 liJe7 54 liJC7 h5 55 liJe8+'iii> g5 56 f4+ 'iii>g6 57 'iii>e5 liJf5 58 liJC7 liJxg3 59 liJd5 liJf5 60 'iii>e4 d6 61 'iW3liJh6 62 'iii>e4 liJg8 63 liJe3liJf6+ 64 �3 liJg4 65 liJC4 d5 66 liJb6 liJf6 67 liJc8�5 68 liJe7+'iii> e6 69 liJc6 liJe470 liJd4+'iii> d6 71 liJf5+ 'iii>C5 72 liJg7 h4 73 liJf5 h3 74 liJh6f6 75 liJf5 'iii>C4 76 liJe3+ 'iii>d4 77 liJf1'iii> d3 78 liJh2f5 0-1 7 ...liJxa 6 Game summary Not 7 ...�xd5 ?! 8 �d2 �c6 9 liJf3 In the game White gets the queens off liJxa6 10 liJe5 'iWC7 11 liJxc6 'ii'xc6 12 early, hoping to target the black central 'ii'f3, when White has some advantage pawns, but is outplayed in the due to the bishop pair. queenless middlegame. Although 8�d21 Gurevich fights hard in the endgame, The strongest continuation. 8 .llC4 eventually he is unable to hold on. liJb4 is comfortable for Black; for ex If White wants to play this way, ample, 9 liJge2 (now if 9 .lld2?!, then then 10 liJge2ma y be a better try, even 9 ....ll a6! and Black is already slightly if objectively Black has no problems better) and then: there either; nor after Graf's more complicated 9 ..wf3. Both these lines are seen in the notes. Game 21 R�Mohr-J�Polgar Amsterdam 1989 1 d4 liJf6 2 C4 C5 3 d5 b5 4 cxb5 a6 5 e3 �b7 6liJc3 'iWa5 7 bxa6 Of course this is not possible after a) 9 ... liJfxd5 10 0-0 e6 11 e4 liJb6 12 5 ... axb5, so it must be regarded as an �b5 �c6 13 �xc6 liJxc6 14 �f4 �e7 15 important test of Black's move order b3 (15 'iWd3 0-0 16 .l:tad1f5 gives Black with 5 ... �b7, even if White far more the edge) 15... 0-0 16 .lld6 l:tfd8 and the 80 The 'Quiet' 5 e3 position is about equal. Black has fairly J.Polgar, European Team Champion good central control though, and if ship, Haifa 1989 . ...d7-ds -d4 can be achieved Black will Instead, Black should prefer have decent chances of obtaining an 10... .i.e 7! 11 dxe6 (11 d6?! fails to advantage. 11 ... 'ii'xd6! 12 es .i.xf3 !, while 11 .i.C4?! b) 9 ...liJbxds is also possible; e.g. 10 runs into 11...liJxe4! 12 liJxe4 exds 13 0-0 e6 11 .i.d2 'ifb6 with an unclear .i.xa6 .i.xa6 14 liJC3'iWe6+ and Black is position; or if 10 .i.d2, then 10 ...'iWC7 clearly better after either 15 .i.e3d4 or and Black is again comfortable due a ls liJe2d4) 11...fxe6 and then: greater influence in the centre. S ...'ii' b6 9 e4 e6 a) 12 .i.d3!? liJb4 13 .i.bl 0-0, when the open f-file and active pieces gives Assaulting the white centre in order Black sufficient compensation for the to open the long diagonal for the light pawn; e.g. 14 0-0 liJc6 15 .i.d3 liJg4 16 squared bishop. At the moment, Black .i.e2 liJf6 17 .i.d3 liJg4 with a repeti has sacrificed a pawn, but she has a tion. good lead in development and is about b) 12 es liJe4 13 .i.e2 0-0 14 0-0 to win the pawn back. liJxd2 15 'iWxd2 (15 liJxd2?! allows 10 .i.C4 ls ...'iWx b2!) ls ... .i.c6 and in my opinion 10 liJf3 ! is a critical test of s ....i.b 7. Black has enough compensation. For Now it is dangerous to take the pawn instance, 16 a4 liJb4 17 b3 (or 17 l::tfdl with 10 ...exds ?! because of 11 exds 'iVb8 18 lIa3 'iWe8 19 liJbs liJds with liJxds 12 .i.C4 'iVe6+ 13 liJe2, when compensation) 17 ...'iVd8 18 lIfdl 'ii'e8 White has a lead in development and a 19 as 'ii'g6 20 �fl 'ii'hs 21 �gl 'ii'g6 is dangerous initiative; e.g. 13... 'ii' g4? another repetition. (13 ...liJac7 is more solid) 14 'iVb3 liJab4 10 •..liJb41 15 0-0 liJb6 16 .i.xf7+ �d8 17 liJg3 and Putting further pressure on the White was already winning in J.Lautier- lig ht squares. 81 Play the Benko Gambit 0-0 and White has to start thinking about equalizing. 11•.. liJxe4! 12 dxe6 fxe6 13 liJxe4JLxe4 14 0-0 d5 14... JLe7 ! promises Black at least a slight edge after ls liJg3 (15 JLc3?! ds is much better for Black better due to the strong centre) ls ... JLg6 16 'iWf3 'iWc6. 15 JLxb4'ii' xb4 16 JLd3JLxd 3 17 'iWxd3 Now the position is unclear. White has a lead in development, whereas This time the tempting 10... liJxe4? ! is Black has a potentially strong centre. not as good, because White still has the option of liJge2 to enable castling; i.e. 11 liJxe4exds 12 JLxa6 JLxa6 13 liJg3 i.d6 14 liJle2! 0-0 15 0-0 JLes 16 I1el 'iixb2 17 J:tcl and White had a definite advantage in lFedorowicz-L.Alburt, US Championship, Cambridge Springs 19BB. ll liJge2?! 17 •••JLe7 18 'iWe3'iWg4 ?! lB.. .'it>f7 ! was a better way to defend the pawn, and if 19 liJf4, then 19 .. :i'e4 20 'iixe4 dxe4 21 l:tfel l:thbB, wh en Black will win a pawn on the queenside (22 ':e2JLf6 ) with a good game. 19 liJf4 .l:i.a4!? An interesting attempt to play for the win. The straightforward 19 ...d4 20 But if 11 dxe6 fxe6 12 a3 liJxe4! 13 'iie4 (or 20 'ii'es 'iifs 21 'iVxg7 'iitd7 22 liJf3 liJd6 14 i.e2i.xf3 15 JLxf3 (or 15 'ifh6 i.fB 23 'ifh4 i.d6 24 liJe2 hs is gxf3 liJc6) ls... liJd3+ 16 'ittfl :cB and more complicated) 20... 0-0 21 'ilxe6+ Black is clearly better; or 11 a3 exds 12 'iWxe6 22 liJxe6 lU6 would give Black exds liJbxds 13 liJf3 'ike6+! 14 'iVe2 JLe7 enough compensation for the pawn, 15 0-0 liJxc3 16 i.xc3 'iWxe2 17 JLxe2 but few winning chances. 82 The 'Quiet' 5 e3 20 ltJxe6 l:te421 ltJC7+�d7 22 'ii'C3 ':'c4 ing 3s... d2 36 ltJg2. After the text the d 23 'ife1 ..td624 h3 'ii'f4 25 g3 'iVg5?1 pawn becomes too strong and Black is Still playing for the win, although winning. this is no longer objectively correct. 35.•. d2 36 l:tbb1 lIe1+37 �h2 Instead, 2s ... 'ii'fs 26 ltJbs 'ii'xh3 27 ltJxd6 llh4! 28 gxh4 'iVg4+ would force a draw. 26 ltJb5..tb8 27 l:td1 l:te8 37.. J:txf4! A nice finishing touch; White can do nothing while Black brings up her king. 38 gxf4 �d5 39 �g2 �e4 40 b4 .l:txd1 28 1i'd2? 41 l:txd1 ..txb40-1 As we have seen many times, enter ing the endgame does not reduce Game summary Black's initiative in the Benko. Instead, With s .....tb7 6 ltJC3 1i'as Black contin the best defence was to counterattack ues to offer the gambit pawn, and in with 28 'iVas! ..txg3 29 'iVa7+ ..tC7+ (or this game White decided finally to ac 29 ...�d8 30 'ii'a8+ ..tb8+ 31 �hl �e7 cept it with 7 bxa6, after which Black 32 l:tfe1+ l1e4 33 .l:txe4+ wins a pawn) obtained compensation by attacking 30 �h l 'iVes 31 ltJxC7 'iVxC7 32 l:txds+ on the light squares. The moves ...'ifb 6, ..t>c6 33 .l:txcs+! .l:txcs 34 'ii'a4+ and 35 ...e7-e6, ...ltJb4 and ...ltJxe4 are quite 'ii'xe8, when White clearly has all the thematic for this line. In a sharp mid chances. dlegame battle, Judit Polgar took a few 28.. Ji'xd 2 29 l:txd2 �c6 30 ltJc3?! risks in the quest for a win, but she was 30 b3! l:tce4 31 a4 would still have ultimately rewarded as Mohr incor kept the balance, but perhaps White rectly went into an endgame and then was in time trouble. defended it poorly. In the opening, 10 30... d4 31 ltJe2..tC7 32 b3?! ..ta533 l:tb2 ltJf3 would have been more testing, d3 34 ltJf4 l:td435 l:td1? though after 10... ..te7 ! Black looks to The last hope was 35 .l:tbbl, intend- have sufficientpl ay all the same. 83 Play the Benko Gambit Summary broke through in the centre to reach a The 5 e3 line is quite tricky and can be winning position. dangerous against an unprepared op Finally, should White belatedly ac ponent. With S ... ..tb7,6 ... 'i¥as, and then cept the gambit pawn with 7 bxa6 (as 7 ... axbs, Black responds in a sharp way in Game 21), Black can follow the same by targeting the central light squares. strategy, attacking on the light squares As we saw in Game 19, White cannot with the thematic moves ...'i1ia s, simply defend by 'ilfb3 and e3-e4, since ...liJxa6, ...'i¥b 6, ...liJb4 and ...liJxe4. Black has the combination 10.. .i"t:'lxe4 !, Again Black achieves a fine game. destroying the centre after all with at Summing up, 5 e3 leads to double least equality. edged play where a mistake by either In Game 20, White opted to give up side can easily mean a rapid change in the centre, hoping to apply pressure in the evaluation. Although the line is not turn in a queen less middlegame. How played so often these days, since the ever, Black's pieces are well placed to theoretical chances of an advantage defend it, and he has good prospects are not that serious, at a practical level for counterplay on the queen side. After Black has to know what to do if he is to White tried a faulty plan, Benjamin avoid coming out worse. 84 Chapter Four The 5 f3 Variation 1 d4 liJf6 2 c4 cS 3 ds bs 4 cxbs a6 S f3 manoeuvre one of the knights to c3. The second option is 5 ...axb 5, but after 6 e4 'iVa5+ 7 ..td2 b4S liJa3! and 9 liJC4, I prefer White here too. Therefore I prefer the natural 5 ...e6, attacking the centre, when the critical line runs 6 e4 exd5 7 e5 'iVe7 S 'ii'e2 liJgS. With 5 f3 White begins an ambi tious attempt to build a strong centre. This was a popular variation in the old Soviet Union, but as we shall see, Black has since found some nice ideas against it. Nevertheless, the f2-f3 varia tion can be a dangerous White weapon, so good preparation is rec In this position Black has two nice ommended. central pawns, while White intends to After 5 f3 Black has three main develop rapidly and target them with choices. The standard Benko set-up moves like liJc3, liJh3-f4 and often with 5 ... g6, in my opinion, allows White 0-0-0. Sometimes White can even sacri to gain an edge after 6 e4, as he has fice a knight on d5, as Nakamura did in good chances to create a blockade on a recent game (see the notes to Game the queen side, later play a2-a4, and 24). In this variation both sides need to 85 Play the Benko Gambit know what they are doing, as a single variation and, as I mentioned above, it mistake can quickly lead to a loss. is crucial to have concrete knowledge Practice has mostly shown that in this line, since the sharp character of Black is okay, with ideas like ...c5 -c4, the play often makes even a single, ...Wb4 and, in reply to llxd5,sacrif icing slight mistake fatal. the queen on c3 for long-term compen 6e4 exds sation (as in the first game below). In 6 ...C4? ! is a sharp sacrifice of an response Nakamura's new plan (or else other pawn, based on rapid develop his computer's) has been to play liJh3- ment and counterplay on the dark f4xd5 without even waiting for the squares; e.g. after 7 i.xc4 axb5 8 i.xb5 rook to come to dl, after which White i.C5. However, I won't be covering this gets a lead in development and a dan line as I don't believe 6 ...C4 is fully gerous initiative for the sacrificed sound. If Black wants to try it anyway, I piece. Although the game is very un would suggest restricting it to games clear and further tests are required, I at fast time limits or else as a surprise, think that Black should be fine. when it may not be so easy for White to prove an advantage. 7 eS Game 22 B.Lalic-A.Khalifman Linares Open 1997 1 d4 liJf6 2 c4 cS 3 ds bs 4 cxbs a6 S f3 e6! Again, the most principled and sharpest reply. In the following varia tion Black will obtain some nice central pawns, but falls behind in develop ment and so has to play precisely. Afterth e less popular alternative, 7 The principled reply, immediately exd5, Black can equalize more easily attacking the white centre. Both sides with 7 ... i.d6 (a logical developing are fighting for the initiative in this move, which reminds a bit of the so- 86 The 5 /3 Variation called Snake Benoni; here White has .i.es 13 ttJh3 .i.xbs 14 .i.xbs 'iVd6 15 the weakening move f2-f3 included ttJgs ttJa6 16 ttJe4ttJxe4 17 fxe4 f6 with but, on the other hand, is now a pawn compensation) 12 ....i.es 13 d6 'fie6 14 up) ttJC7 'fifs 15 �el .i.xfl 16 ttJxa8 .i.d3 17 ttJh3 ttJc6 18 ttJf2 .i.C4 and Black has excellent play for the exchange with two active bishops and the exposed king on el. a2) 11 ttJh3 ttJc6! 12 'fid2 (after 12 dxc6 dxc6 13 .i.f4! .i.xbs 14 'ii'd2 .i.xf4 15 ttJxf4 gs! 16 'iVxd8+ lIxd8 17 .i.xbs cxbs 18 ttJe2 �g7, White has to think about how to equalize) 12... .i.xb s 13 .i.xbs ttJd4 14 .i.C4 'iliC7 and Black has more than enough for the pawn, when 8 'iVe2+ �8 (Black has to forgo cas you take into account the strong knight tling since the exchange of queens on d4, play on the dark squares, the would only benefit White, due to the f3 rook on its way to e8 and the weak weakness and lack of development) 9 white knight on h3. ttJC3 axbs! (9 ... .i.b7 10 'ii'd2?! .i.es 11 b) 10 .i.e3c4 ! 11 �f2 .l:[as12 g3 bxa6 ttJxa6 12 .i.xa6?! .i.xa6turn ed out well for Black in S.Volkov-A.l1jushin, internet 2004; but White can improve with 10 ttJh 3!, when I think he is slightly better) and now: 12... ttJa6 ! (Bareev also suggested 12 ... .i.b7 as good for Black, but after 13 a4! the game is more unclear, so I would prefer the text) 13 a4 (E.Bareev A.Khalifman, Russian Team Champion a) 10 ttJxbs .i.a6with a further split: ship 1995) and now 13... ttJC7 ! (Bareev) al) 11 'iVd2 .e7+ 12 �f2 (or 12 �dl 14 .i.b6 (14 ttJe4?! .i.es is slightly better 87 Play the Benko Gambit for Black) 14.. .l:1a6 15 i.xC7 i.xC7 16 This has been the main move here l:tel and although this position is also since Hertneck introduced it in 1991. unclear, it seems to me that Black (Until then the passive 10 ...'iIi'd8 was would have more winning chances in a thought to be best.) After the text both practical game. (But not 16 lDxbS? .. :ifcs and ...'ifb4 are made possible, i.b6+,wh en Black is clearly better with while Black is planning ... a6xbs in some ...lDxdS and ...lDe 3 coming.) variations. For instance, with Black to 7 ..:ile 7 S 'ike2 move in this position, 11 ...axb s 12 lDxbS?? would run in to 12 ...'ifb4+ 13 lDc3 d4, winning a piece. Another main idea is the thematic queen sacrifice .. :ilb4xC3+ (in response to White play ing 0-0-0 and l:txdS); we will see this both in the current game and in the notes below. 11 i.e31 The strongest move in this position. White develops, enabling queen side castling, while preventing ... 'iVcs or S •••lDgS ...dS -d4. We will examine other moves There is no other choice, as in Game 24. 8 ...lDh S?? loses a piece to 9 g4. 11••• axbs 12 0-0-0 9 lDc3i. b7 10 lDh3 The alternative 12 lDxbS is seen in The knight heads for f4 and then on the next game. to dS, while lDhS can also be a danger 12 lDf4? is mistimed because it al ous idea in some lines. lows Black to take the nice central eS 10 •••c41 pawn: 12 ..:ifxe s! 13 0-0-0 (13 lDxbS lDa6! is also slightly better for Black, who has ideas of ...i.c6, or ...i.b4+ fol lowed by ...lDe7) 13 ...lDf6 ! (instead, 13 ... lDe7 14 lDxbS d4! 15 lDxd4 l1xa2 or 15 'iixc4 'iVxe3+ is good for Black; but 14 'iVd2 b4 15 lDbS l:txa2 16 i.xc4! (V.lvanov) or 14 ... i.c6 15 i.cS!? is very unclear) 14 i.b6 (or 14 i.d4 'iVxe2 15 i.xe2 i.d6! with a clear advantage) 14 ... i.d6 15 lDfxdS lDxdS 16 lDxdS 0-0 and White was struggling to show any 88 The 5 13 Variation compensation for the pawn in .1i.d2) 14... 'i!las 15 .1i.b6 (not 15 ttJfxdS E.Dalsrud-J.Skjoldborg, Prague 2005. ttJxdS 16 ttJxdS? .1i.xdS 17 :xdS "xa2 12•• JlVb4 and Black is clearly better, while 16 .li1xdS cannow be met by 16... .1i.xds 17 iLb6'ii' xa2! 18 ttJxa2 :xa2,again with good compensation) ls .. :ika6 16 l:txdS (or 16 a4!? iLc6 17 as 'ii'b7 gives Black the edge, whereas 16 .1i.cS .1i.c6 or 16 ...'ikh 6!? is unclear) 16 ....1i.c6 ! (prepar ing ...bS- b4 by taking the bs-square away fromthe white knight) and now: A cool way of clearing the ranks; whereas taking the e-pawn now would give White a very dangerous attack after 12 .. :ii'xes?! 13 f4 "e7 (or 13 .....e6 14 ttJgS) 14 ':xdS!, as in M.Dlugy L.Alburt, US Championship, Los Angeles 1991. 13 ttJf4 Now 13 l:txdS can be met by the a) 17 a4 b4 (17 ...ttJc8 18 .l:.xbS! J..e7 thematic 13 .. :ii'xC3+! 14 bxc3 .1i.xds 15 19 ttJfdS .1i.xdS 20 ttJxdS "xa2 is also ttJf4 .1i.a3+16 �bl ttJe7wh en Black has very unclear) 18 .l:.as 'Wb7 19 l:txa8 excellent compensation because of the 'i!lxa8 20 ttJbl ttJdS 21 'ikd4 ttJxb6 22 exposed king on bl. E.Klein-J.Waitzkin, 'ii'xb6 .1i.e7 23 .1i.xC4 .1i.gs 24 g3 0-0, USA 1993, continued 17 'ii'd2 .1i.e6 18 when the king has finally escaped the ttJh S l:tg819 ttJf4 ttJbc620 ttJxe6fx e6 21 centre and Black can look to regain his f4 cJ;;f7 and Black went on to win; while pawn next move with a good position. 21... ttJdS !?, planning a well-timed ...bS b) 17 :xbS!? (it seems like the b4, also came into consideration. lower-rated player prepared a draw 13 ...ttJe7 with the computer) 17 ....1i.xb s 18 ttJe4 Covering the ds-pawn. (now White has sacrificed a rook, but 14 .1i.b6 he has the threat of ttJd6 mate as well 14 l:txdS could still be answered by as a weakness on C7 to aim for) 14 ... 'i!lxc3+!. However, a noteworthy 18 ...ttJc8 (18 ...ttJfS !? might be a better alternative is 14 "f2 !? (threatening 15 try) 19 ttJdS ttJxb6?! (now White gets 8 9 Play the Benko Gambit the advantage; but 19 ...l:ta7 leads to a loses to 20 ...:cl mate! draw after 20 ttJef6+ gxf6 21 ttJxf6+ b) is llxdS Ji..c6 (covering the bS �e7 22 ttJdS+etc, or 20 Ji..xa7 'ii'xa7 21 pawn; of course lS... 'ii xC3+ is still an 'iixa7 ttJxa7 22 ttJC7+ �d8 23 ttJxbS option) 16 g3 (after 16 �bl Ji..h6 17 ttJxbS 24 Ji..xC4 ttJd4 2S ttJgs �C7 26 lld4 Ji..xf4 18 l:txf4 ttJdS 19ttJxdS Ji..xds ttJxf7 ':'g8 27ttJg s with a repetition) 20 Black is slightly better) 16 ... Ji..h6 and ttJC7+�d8 21 ttJxa6l:.xa6 (B lack doesn't the position is unclear; for instance, really have enough for the queen here, after 17 Ji..cs 'iVas 18 Ji..xe7 Ji..xf4+ 19 though he can play on) 22 ttJc3Ji.. c6 23 gxf4 �xe7 Black will either take on dS Ji..xC4 ttJxC4 24 'ii'h4+ Ji..e7 2S 'iVxc4.l:.f8 next move, or if 20 l:td4 then 20 ...b4 26 b4 .l:tb6 27 a3 ttJa6 28 ttJdSJi.. xds 29 with the initiative. 'ii'xds ttJC7 30 'ii'd3 �c8 31 �b2 f6 32 c) is �bl Ji..h6 16 l:txdS Ji..c6! (not 'iVxh7 fxes 33 'ii'xg7 l:te634 �b3 ttJbS now 16 ...'ivx c3?! 17 bxc3 Ji..xds 18 Ji..e3! 3S :el ttJd4+36 �a2 llff6 37 'ii'h8+ llf8 and White is better) 17 lld4 Ji..xf4 18 38 'ii'g7 l:.ff6 39 �8+ �-� l:txf4 0-0 19 'iie3 l:.e8 with an unclear A.Bitalzadeh-M.Vachier lagrave, EU game. Championship, Liverpool 2008. lS llxdS 14... hS!? Preparing to swing the king's TOok into the attack on the queen side via h6. A good alternative is 14... g6, intend ing ...Ji.. h6, which gives Black a satisfac tory position in all lines: The less direct is 'iVf2 !? lith6 16 �bl was tried in M.Lindinger-M.Kopylov, German league 200S, when Black should have played 16 ...gS! 17 ttJfxdS (or 17 a3 l:txa3! 18 Ji..cs l:tal+! 19 �xal 'iias+ 20 �bl gxf4 with compensation) a) is ttJfxdS?! Ji..xds 16 l:txdSJi.. h6+ 17 ... Ji..xds 18 Ji..cs 'iVas 19 Ji..xe7 Ji..e6 20 puts White in difficulties; e.g. 17 �bl Ji..xf8 �xf8 and the game is still very liIa3! 18 Ji..d4 ttJbc6 19 l:lxbS I:[xc3! and unclear. Black wins material, since 20 l:txb4 lS •••'ii' XC3+! 9 0 The 5 /3 Variation Again the thematic queen sacrifice. 'irxb6 tDc6,inten ding ....:f a8, ...l:tb 2 etc, 16 bXc3 tDxd5 when the vulnerability of the white king appears to provide sufficient counterplay. For example, 24 'irC7 (other lines include 24 'iVxb5 l::tb8, or 24 'it>cl Ilfa8 25 'iixb5 .llf2 26 'iWbl tDxe5, or 24 f4 l:.fa8 25 'iWb7 %:t8a3! 26 'iVxd7 :b2) 24... :fa8 25 'it>el :c2! 26 .i.e2 lilaa2 27 .i.dl .llxc3 28 'irxd7 tDb4 29 'it>fl:d3 30 'iie8+ 'it>h7 31 .i.e2l:Idd2 32 'ij'xf7 ':xe2 and White has to take the draw. 17... .i.xd s 17 tDxdS?1 White has done badly after this natural move. Instead, 17 'ire4! is a key improvement, which forces Black to be very accurate: 17 ....i.a3+ 18 'it>d2 tDxb6 19 'iixb7 .i.C5 20 tDd5 ':'xa2+ 21 'it>dl! 0-0 22 tDxb6 Black only has rook and knight for the queen, but he has serious counter play on the queen side, often simply by ... tDc6 and ... b5-b4, attacking the white king. In practice, the position is quite difficult to play for White and hence, in my opinion, slightly better for Black. and now the game A.Allemann 18 'ii'd2 .i.e619 .i.e2 R.Costantini, Biel 2001, concluded in It is indicative of White's problems White's favour after 22.. J�al+? 23 'it>c2 in this line that he doesn't really have .i.a3 24 f4 b4 25 tDxC4 :cl+ 26 'it>d2 time to develop his pieces. Stohl sug bxc3+ 27 'it>d3 .i.C5 28 g3 tDc6 29 .i.g2 gests 19 .i.C7!? tDc6 20 .i.d6 .i.xd6 21 tDb4+ 30 'ii'xb4! 1-0. exd6 0-0 as critical, but it is hard to see Black had to find 22 ... .i.xb6! 23 how White should continue; for exam- 9 1 Play the Benko Gambit ple, 22 g4!? can still be met by 22... b4! In my opinion, the 5 f3 e6 line is one 23 cxb4 :fb8 with a very strong attack. of the most dangerous for Black in the 19 .•.lDc6 20 f4 b4 21 fS bxc3 22 'ii'xc3 Benko, and it is characteristic that both l:ta323 'iWb2? sides have to play with precision. My This loses quickly. White had to play advice, therefore, is to prepare better the unappealing 23 Wal, though after than your opponent! 23 ....i.x f5 24 .i.xC4 .i.b4 Black is still clearly better. 23 ... c3 24 'ifbs.i.xfs 2S l:[fl Game 23 Se.lvanov-A.Khalifman St Petersburg 1997 1 d4 lDf6 2 C4 CS 3 ds bs 4 cxbs a6 5 f3 e6 6 e4 exds 7 eS 'ife7 8 'ife2 lDg8 9 lDC3 .i.b7 10 lDh3 c4 11 .i.e3 axbs 12 lDxbS This is the main alternative to 12 0-0-0 as seen in the previous game. 12 ...'ii' b4+ One of the points of ...c5 -c4 and 2S.•. .l:txa 21 26 .l:txfs lDb41 ...a7x b5. Threatening ... l:[alma te. 13 lDc3 27 'iVas g61 0-1 Black doesn't even need the queen back. Game summary A good win by Khalifman, utilizing the thematic queen sacrifice on C3. In the game Lalic could have improved with 17 'ilke4, after which Black's attack is less straightforward, even if still suffi cient. Therefore Black might prefer 14 ... g6 as given in the notes. It is logical 13 ••. lDe71? to develop the bishop to h6, especially The sharpest move in this position. as the white king is sitting on Cl, and Khalifman gives it an "!" in Informant Black can reach an unclear game in a #69, claiming that White is slightly bet nice way as shown in the analysis. ter after 13 .. :iVas or 13 ... .i.c5.Ho wever, 9 2 The 5 13 Variation I don't agree with the latter evaluation, tiJf4 iLc6 20 iLe2tiJa6 21 l:ld2 0-0 22 as I think Black is at least equal after J:.hdl tiJC7 23 h4 (taking some space on 13 ... iLcS, so that the choice between the kingside in the absence of better that move and the text is largely a mat plans) 23 .. J�b8 24 g4 g6 with an equal ter of taste. game. After 13 ... iLcs 14 a3l1 14 O-O-O! is a more logical move, fighting for dS, without wasting time on luxuries like a2-a3. 14 a3 'iWas 15 iLxcs 'iWxcs 16 'iWf2 ! (the queen exchange is the best here; otherwise the strong central c4- and ds-pawns will cause White some prob Now Black has: lems in the middlegame) 16 ... 'iWxf2+ a) 14 ...tiJfs was given by Khalifman (worse is ...d4 17 0-0-0 tiJc6 18 tiJe4 and is quite sharp but, in my opinion, 'ii'xes 19 'ii'g3 'ii'xg3 20 hxg3 - Khalif leads to a slightly better position for man, when White is a bit better; e.g. White if he plays precisely: 15 iLf2 20 ...iLa6 21 tiJd6+ �e7 22 tiJxc4) 17 'iWas! 16 tiJf4 (after 16 g4 d4! 17 iLxd4 'ittxf2 tiJe7 18 l:r.dl Khalifman claims a tiJxd4 18 litxd4 iLb4 Black has reason slight advantage for White, probably able compensation for the pawn) because the black pawns are on cur 16 ...d4! 17 iLxd4 iLb4 (Khalifman),an d rently stuck on the light squares. How now 18 'iWxC4! (18 'iWf2 tiJxd4 19 'iWxd4 ever, Black can obtain his usual queen 0-0 20 tiJhs iLxC3 21 'ii'xC3 ':'c8 is un side counterplay with ...:as and ...l:1b 8, clear) 18 ...tiJc6 (or 18 ...tiJxd 4 19 litxd4 and can fi ght for d4 with the manoeu iLxC3 20 'iwxc3! 'iWxa2 21 tiJh S! 0-0 22 vre ... tiJa6-C7-e6. In my opinion Black :g4 g6 23 e6! and wins) 19 tiJfdS has at least equal chances, and some tiJcxd4 20 tiJC7+ is good for White; e.g. times even winning chances if his 20... 'itt e7 21 'iWxb4+! 'iWxb4 22 tiJ7dS+ knight can get to e6 without him losing iLxdS 23 tiJxdS+ 'itte6 24 tiJxb4with an control of dS. For example, 18... l:tas 19 extra pawn. 93 Play the Benko Gambit b) 14 ...ltJa6 ! is stronger, the main 15..• l:.a61 idea being to support the ds-pawn with ...ltJC7, and then: A nice and exact move, taking time to prevent .i.e3-b6 before continuing bl) 15 a3 'ifb3 16 ltJf4 ltJC7 17 'ii'C2 with ...ltJfS . Now Black is slightly better. "ii'XC2+ 18 'iti'xc2 g6 (the simple The central pawns on dS and c4 can be 18... .l:la 6!? also comes into considera quite dangerous. tion) 19 .i.cs (or 19 �b6 �h6!) 19 ... ltJfS! 16 'ii'd2?1 20 �b6 d4! 21 .i.xd4 .i.h6 22 ltJfe2 ltJe6 This is quite slow and allows Black 23 .i.f2 .i.e3with compensation for the to increase his advantage. However, sacrificed pawn. Black would play ...ltJfs after other b2) 15 ltJf4 ltJcS(lS ... ltJc7 is still pos moves too; for instance, 16 ltJf4 ltJfS 17 sible) 16 ltJfxds (or 16 "ii'd2 nas, indi ltdl (or 17 0-0-0 ltJxe3 18 Wxe3 �cS 19 rectly covering the ds-pawn with an 'iVel .i.xa3!) 17 ...ltJxe3 18 'ii'xe3 :b6! 19 unclear game) 16 ...ltJxds 17 :xdS (the 'ii'd2 d4 with a strong initiative; or if 16 only move) 17 ...ltJa4! (undermining the �d2 ltJfs 17 ltJf4, as in S.Uchitel rook on dS) 18 .i.d4 .i.e7 19 :d6!? .i.c6 V.Marinin, St Petersburg 1997, then 20 ltJe4c3 ! 21 ltJxC30-0 and in practice 17 ... .i.cS 18 'ii'e2 ltJd4 19 'iVdl 'ii'C7, fol this position would be easier for Black lowed by ...'iVxe s with a big advantage. to play, even if objectively unclear. Finally, the preventative 16 g4 is met 14 •.JWa 5 15 'iVf2 simply by 16 ...ltJg 6 and Black is clearly After 15 O-O-O?! ltJfS Black is very better (Khalifman). comfortable; or 15 ltJf4 ltJbc6 16 .l:tdl 16... ltJf5 17 .i.f2.i.C5 18 l1dl ltJxes 17 .i.d4ltJSg6 18 ltJxg6 hxg6 19 Or 18 0-0-0 d4, followed by ...c4-c 3, "ii'f2 l:ta6! 20 'iVg3 l:te6+21 .i.e2(Z.Z hao and White is close to losing. L.Moylan, Brisbane 2005) and 21 ... ltJc6! 18... .i.xf2+ 19 ltJxf2 0-0 20 ltJe2'iVc 7 would have been very good for Black; Avoiding the queen exchange, since e.g. 22 �f2 .i.d6 23 f4 ltJe724 0-0 gS. White is behind in development. 94 Th e 5 /3 Variation e6 6 e4 exds 7 eS 'ii'e7 8 'ii'e2 t'iJg8 9 t'iJC3.i.b7 10 t'iJh3c4 11 t'iJf4 Attacking the ds-pawn immediately. Instead, 11 .i.gS?! is a waste of time as the black queen often goes to cS any way. M.Notkin-V.L.lvanov, Moscow 1994, saw 11 .. :iVcS 12 0-0-0 axbs 13 .i.e3 and now, rather than 13 ...'iWb4 transposing to Game 22, Black can play 13.. :iVc 6! with a slight advantage. For example, 14 t'iJf4 (or 14 :xdS b4 15 t'iJbs l:1xa2) 21 'iWf4 t'iJe7 22t'iJg4t'iJg6 23 'iVg3 d6 14... t'iJe7 15 'itJbl t'iJa6, when 16 t'iJfxdS? Khalifman wasn't sure whether this t'iJxdS 17 t'iJxdS fails to 17... 'iVg 6+! 18 was the most precise move, but it's 'itJal (or 18 'it>cl .i.xds 19 l:1xdS t'iJc7) good enough to win. 18 ... .i.xds19 1:txdSt'iJb4 an d wins. 24 exd6 .l:t.xd62S t'iJd4fS 26 t'iJf2 t'iJc627 11... 'ii' cS .i.e2 t'iJxd4 28 l1xd4 f4 29 'iVgs .l:t.e8 30 t'iJg4':de6 31 Itd2 c3 32 bXc3 'ili'xC3 0-1 Game summary In gaining a llawn with 12 t'iJxbs White falls behind in development and Black reaches an unclear position with good compensation. After Ivanov's further inaccuracies (at moves 14 and 16), Khalifman obtains a dangerous initia tive, with the white king stuck in the centre. The result is a nice game by 12 a4? Kh alifman, who was one of the leading Too slow. No better is 12 .i.e3?! d4 Benko theoreticians at that time. 13 .i.f2 axbs, when Black also has a clear advantage. The critical response is the danger Game 24 ous sacrifice 12 t'iJfxdS!, which again U.Adianto-J.Polgar looks like a computer suggestion. In Jakarta (rapid match, deed, the computer is quite good in the game 2) 1996 5 f3 e6 variation! This forcing sequence continues 12 ... .i.xds 13 .i.e3 'iWb414 a3 1 d4 t'iJf6 2 C4 cs 3 ds bs 4 cxbs a6 S f3 'iVas 15 .i.d2.i.e6 16 t'iJdSan d now: 95 Play the Benko Gambit bl) 18 ... axb5?! 19 &i:JC7+ llxC7 20 'ili'xC7 &i:Jc6 21 'ili'xd8+ 'itxd8 22 �xb5 gave White a favourable endgame in H.Nakamura-M.Vachier Lagrave, Cap d'Agde 2008. White has a rook and two strong passed pawns for the two knights, whereas Black is yet to develop his kingside. The game continued 22 ... &i:Jxe5? (a mistake; but even after 22... &i:Jge7 23 f4 g5 24 fxg5 �g7 25 0-0 �xe5 26 �C3 11g8 27 h4 �g3 28 �el a) 16... 'ii' a4?! was played in a recent .te5 29 b4 White has the better game: 17 &i:Ja+ 'itd8 18 &i:Jxa8 axb5 19 chances with his passed pawns) 23 &i:Jb6 'ii'a6 20 �e3 �b4+ 21 axb4 'ili'xal+ �a5+ 'ite724 ltc8(n ow Black is losing 22 'it>f2 and White was clearly better in as he is too far behind in development, I.Nyzhnyk-S.5iebrecht, Wijk aan Zee though his next move doesn't help 2011, due to the vulnerability of the matters) 24.. .f 5?! 25 �b4+ d6 26 ltxf8 black king. Indeed, the game ended in 'itxf8 27 �xd6+ 'itf7 28 �xe5 &i:Jf6 29 an amusing king hunt: 22... &i:Jc6 23 Wd2! 0-0 1:1c8 30 a4 &i:Jd5 31 .l:tal .l:tc2 32 �d3 &i:Jxe5 24 'ii'd6! 'ii'xb2+ 25 'itgl &i:Jc6 26 .l:td2 33 �c4 ltc2 34 b3 &i:Je3 35 .txe6+ 'ii'f8+ 'ita 27 'iVc8+ 'itd6 28 �C5+ 'ite5 'itxe6 36 �d4 f4 37 �xe3 fxe3 38 as 29 &i:Jxd7+ �xd7 30 'iVxd7&i:Jf6 31 Wxc6 'itd539 a6 11c840 a7 J:1a841 'it>fl l-0. c3 32 'iWd6+'it>f5 33 �d3+ 'itg5 34 �e3+ Vachier was out-prepared by Na 1-0. It seems that Siebrecht hadn't stud kamura's novel knight sacrifice. How ied the Nakamura game quoted below. ever, if Black reacts precisely he can b) 16 ... 'ii'd8 17 'ii'xc4 lla7 18 1:[cl reach an unclear game: reaches a critical position: b2) 18... �x d5! 19 Wxd5 axb5, keep ing the extra material, is not conspicu ously bad for Black. For instance, after 20 �xb5 &i:Je7 21 'ii'd3 &i:Jbc6 22 f4 g6 23 0-0 llb7 24 �e3 'ii'b8 25 a4 �g7 26 i.C5 0-0 White has definite compensation for the piece due to the bishop pair, more space, and two passed pawns, but Black can look to unravel with moves like ...1:1e8, ...�f8, and ...&i:Jf5 or ...&i:Jc8 an d should not stand worse. 12 •••d4! 9 6 The 5/3 Variation 19 ttJxg7+�d8 20 'iiel 'iVb6 Keeping the queens on. otherwise 20 ... 'iVxel+ 21 'ittxel ttJb3 22 .l:tbl 1:[g8 23 ttJh s ttJds was certainly winning for Black. 21 .i.e3 21 c;t>gl :g8 22 ttJhs ttJds is no bet- ter. 21... :g8 White should be careful not to allow this pawn to start advancing. In this variation time is crucial. Black is al ready clearly better. 13 bxa6 ttJxa614 ttJe4'iVb4+ 14 .. .'ii'xes 15 'iVxC4 fs looks even stronger. 15 �f2 d31 16 'iVe3 i.cs17 ttJxcS ttJxcS 22 ttJhS?1 White had to try 22 b4!? cxb3 23 ttJh s just to break up the black pawn duo. After the text he is pretty much lost. 'ii'xb2+ 23 'it>gl 'ili'xes 24 ttJg3 f6 25 f4 'i!kds 26 'ii'b4 ttJe4 27 i.b6+c;t>e8 28 lle1 fS 29 as c;t>f730 :c1 :ac8 31 .i.e3 hS 32 ttJxhS c3 33 ttJg3 d2 34 l:Ia1 'ii'd1 35 l:txd1 'ii'xd1 36 'iixb7 c2 37 'iVb3+dS 0-1 Now Black is a pawn up and very comfortable, with good squares for her Game summary knights on b3 and e6; whereas White This game is a good example of what has an exposed king and is finding the can happen if White plays inaccurately c4- and d3-pawns quite annoying. in the opening. Adianto's 12 a4?! was 18 ttJhSttJe7 not in the spirit of this line: White 18 ... ttJe6 was solid, but Polgar just should develop rapidly. Instead, Polgar wants to get her pieces out and attack. was able to use her two central pawns 97 Play the Benko Gambit and lead in development to win con the bs-pawn instead, but in so doing vincingly. falls behind in development, enabling Far more critical is Nakamura's 12 Black again to generate strong coun lLlfxdS!, which he used to defeat terplay. Vachier Lagrave. However, by playing Finally, Game 24 is a good example to keep the sacrificed material, Black of how things can go wrong for White can reach an unclear game with accu if he plays too slowly and allows the dS rate defence. Nevertheless, there is still pawn to advance. Instead, Nakamura's plenty of room for practical tests here. dangerous knight sacrifice, seen in the notes, leads to very interesting play. Summary With the suggested improvement for The 5 f3 variation is both uncompro Black the game can be regarded as un mising and double-edged. With s ... e6 clear, but I think more theoretical dis Black responds in the most principled putes in this line are required before way, and after 6 e4 exdS 7 eS 'fie7 8 any definitive conclusion can be We2 lLlg8 9 lLlc3 i.b7 10 lLlh3 c4! the reached. battle becomes very complicated, My overall conclusion is that Black is where a single mistake can easily mean doing fine after the sequence 5 f3 e6. a loss for either side. However, it remains a critical line in In Game 22 it was seen how a direct the Benko, and one that requires solid 0-0-0 and l1xdS can be met by the the preparation from both sides. The game matic queen sacrifice ..:ii xc3!, after quickly becomes extremely sharp and a which Black obtains excellent compen single mistake can alter the evaluation sation. In the next game White takes of the position dramatically. 98 Chapter Five White Plays bS-b6 1 d4 lL'lf6 2c4 cS 3 ds bs 4 cxbs a6 5 b6 whereas if Black plays ... d7-d6 and ...lL'lbd7x b6, White will oftenpl ay a4-aS, followed by lL'ld2-C4 and lL'lc3-a4-b6. Black should be careful not to allow such a blockade. After S b6 Black has several options. The straightforward S ...'i¥xb 6 is the move I am recommending, though S ... g6 and S ... d6 can lead to the same positions if Black plays ... 'i¥xb6la ter on. Instead, s ... e6 is unclear, but White can avoid this by playing lL'lf3bef ore taking White returns the pawn in order to on bS (as he in fact did in Games 26 interfere with Black's usual queenside and 28 of this chapter). s ...aS was a counterplay. The plan of bS-b6 used trend for some time, but I regard it not to be regarded as too dangerous move as dubious as it weakens the bS for Black. However, in the early nineties square. Shirov won some impressive attacking After S ... 'i¥xb6 the two sides engage games with it, and the line became in a positional battle, where White pro more popular. GM Vadim Milov also ceeds with the plans outlined above. won some games in a more positional Black, on the other hand, will aim for way. play on the b-file and an active game, White's main idea is to seize control with moves like ....ltg4 (if allowed), over the c4-square, and he will usually ...l::t b8 and ... lL'les, or sometimes ..,'¥la7 follow up with lL'lC3, e2-e4 and lL'lf3-d2- and ... lL'lb6 to challenge a white knight c4. Th en if Black recaptures on b6 with on the c4-square. In some positions the queen, lL'lC4 will gain a tempo; Black can even play ...g6 -gS to secure 99 Play the Benko Gambit the es-square as a knight outpost, al his dark-squared bishop, plus counter though this is naturally somewhat play on the b-file, even though White risky. Often the queen will be attacked hasn't accepted the offered pawn. In on b6, either by a knight on c4 or a my opinion, with dynamic and precise pawn or as and, depending on the cir play, Black can obtain at least an un cumstances, my recommendation is to clear game. place the queen on a7, c7 or d8! The 7 e4 d6 8 a4 .i.g79 as "iib4! main consideration, as in many Benko lines, is that Black should play actively. I used the ... 'iIi'xb6 line to win a nice game against Peter Heine Nielsen (Game 28). He played the opening inac curately, allowing a strong and unex pected knight sacrifice, after which there was no defence. So what looks like a positional struggle can quickly become sharpened, and Black has to keep an eye for tactical opportunities. Another well-known example is Now it is not so easy for White to that an early a4-aS can be answered by defend the pawn in a natural way. Al ...'iWb4, attacking the e4-pawn, while if though the position is unclear from an l:ta4there is ...JLd7- bS. This is one of the objective point of view, in a practical reasons, for instance, why I prefer a game I think Black has a good chance quick .....xb6 to ...d7-d6 and to get the advantage, since White can ...liJbd7xb 6. easily end up in a passive position. So the early a4-aS cannot be recom mended for White. Indeed, it seems Game 25 that nowadays it mostly occurs wh en J.Campos Moreno-V.Topa lov White is not too well up on the theo Aviles 1992 retical nuances of the bS-b6 line. lo l:ta4 1 d4 liJf6 2 c4 cS 3 ds bs 4 cxbs a6 5 b6 Virtually the only move. 10 'iWc2? 'i!Vxb6 and 10 .i.d3? both fail to 10... liJXe4, s ... d6 6 liJc3 "xb6 is another move winning at least a pawn; while 10 f3 order which often reaches the same does not fit in at all with White's in position; e.g. after7 e4 g6. tended set-up, as he now has no mean 6 liJC3g6 ingful way to use his pieces. Black now has a good diagonal for lo ...Wb7 100 White Plays bS-b6 Planning ...�d7-b 5, taking advan axb5 16 'iixb5 "fiC7 17 "fia4 l:ta7?! 18 tage of the white rook's unfortunate �b5 .l::[b8 19 0-0 in L.Van Wely placement, whkh is one of the main L.Riemersma, Dutch Championship, points of 9 .. :iWb4. Amsterdam 1995. In this line 17... e6! l1 lLlf3 would have offered more counterplay, The main alternative is 11 �C4 �d7 but in that case Black might prefer 12 l:ta3 'ii'b4 13 'ii'e2, when Black can 12 ...e6, attacking the centre at once. For choose between: example, 13 lLlC4 exd5 14 lLlxd6 (or 14 exd5 .tg4 15 .te2�xe2 16 'it>xe2'ike 7+ 17 11••• 0-0 12 �e2 in the Benko in general. The knight is White has also tried 12 lLld2, when on its way to c7, and sometimes to b5 12 ... �d7 13 %:.a3 �b5 turned out well and d4 (although the bishop on b5 is in for him after 14 'ii'h3! lLlbd7 15 lLlxb5 the way here), or else to support the 101 Play the Benko Gambit attack on the centre with ...e7-e6 (as in tiJxbs tiJxbs 24 l1xbS 'it'xds 2S 'it'xds the current game), or even ...f 7-fS. tiJxds 26 i.d2 would give him a clear Here, however, the knight manoeuvre advantage with the bishop pair and the is mistimed and premature. dangerous passed a-pawn. Better was lS ... l:tfcS 16 h3 (prevent- 23... l:txel+ 24 'ifxel tiJfxdS 2S i.a4?! ing ... tiJg4-eS) 16 ... i.xe2 17 'ifxe2 litabS Keeping control of eS was less im is tiJd2,an d now lS ... tiJeS 19 tiJC4tiJC7 portant than the time White wastes 20 i.gs liteSwith an unclear game. Two with this bishop. Better was 2S i.c4, ideas are ...tiJb S, and ...'ila7 fol lowed by when 2s... 'it'aS ?! 26 tiJxds tiJxds 27'iWd2 ..Jlb4. .l:teS 2S l:td3 tiJb4? 29 ':'xd6'it'xas runs 16 i.gstiJdf 6?! into 30 i.xf7+! 'iii>xf7 31 l::td7+ 'iii>gS 32 Covering the e7-pawn and thus in l:txg7+! 'iii>xg7 33 'it'd7+, while 2S... tiJb4 sisting on his erroneous knight ma 26 'fie7 'ifc6 27 i.f4 dS 2S 'fixC7 'it'xC7 noeuvre. It was not too late to go into 29 i.xC7 dXc4 30 l:ta3is just equal. reverse and play 16... i.xe2 17 'ifxe2 2s... 'ifa S 26 tiJxds tiJxds 27 i.bstiJb4 tiJef6, when the game is still finely bal anced. Black has the potential for good play on the queen side with ... .:tfbS, ..:flIe 7 and ...l:tb4, wh ile White has some squares to play for too. 17 l:.b3tiJC7 lS tiJd2 l::tadS 19 tiJC4e6?! 19 ...'ii' a7 would restrict White to a slight advantage. 20 tiJa3 exds 21 exds l:.deS 22 tiJaxbs axbs 2S i.d2? Allowing the knight to switch to d4 gives Black a definite advantage. White had to try 2S .te7, and if 2S ... :cS (or 2S ...tiJC2 29 'iWdl) 29 i.xd6tiJc2, then 30 'it'e7! tiJd4 31 i.C4! tiJxb3 32 i.xf7+'iii> hS 33 i.es 'ii'xas 34 i.xg7+ leads to a draw, but 29 .. .'iMS! is still good for Black. 2S ... tiJC2 29 'iWfl tiJd4 23 .txbS? A great square for the knight. Now White begins to go wrong. 23 30 l:1g3tiJxbs 31 'iWxbs 1:.bS 102 White Plays b5 -b6 Instead, 4 cxbS a6 S b6 "ii'xb6 6 liJC3 g6 7 liJf3 i.g7 8 e4 d6 9 liJd2 0-0 10 i.e2liJbd7 11 0-0 l'1b812 a4 would be a more standard route to the position at move 12. White's move order with 4 liJf3 is significant in that it enables him to avoid the line 4 cxbS a6 S b6 e6!? True, he must then be prepared for any of 4 ...b4, 4 ...i.b 7, 4 ...bxC4 and 4 ...e6, which all have their adherents. But none of After Black wins the b2-pawn the that matters here as I am advising rest is relatively easy. The c- and d Black to play the consistent 4 ...g6 in pawns are ready to advance, and White any case, aiming for a normal Benko has no real counterplay. afterS cxbS a6. 32 'ild3 l:txb2 33 h4 d5 34 i.C3 C4 35 4 ...g6 5 cxb5 a6 6 b6 d6 7 liJC3 ii'd1 i.xC3 36 Ibc3 'ii'xa5 37 :e3 'i¥d2 38 'ikf3 'ii'd4 39 g3 llb1+0-1 Game summary In the opening White plays a4-aS too early, as 9 ...'iWb4 fo rces him to cover the e4-pawn in a rather awkward way with lIa4, after which Black has pretty much equalized. Nevertheless, Black's posi tion doesn't just play itself, and Topa lov's inaccurate handling of the early middlegame allowed White to gain the 7 ...liJbd7 advantage after all. It was only when It doesn't really matter that Black Campos also went wrong that Topalov has delayed taking on b6 up to now, managed to come out on top. but I think he should definitely play . .. 'ii'xb6 here. Partly because it is more flexible and gives Black the option of Game 26 ... i.g4 if White delays liJd2 for too long; V.Milov-V.Topalov but more because it avoids the difficul Prague (rapid) 2002 ties noted at move 10 below. After 7 ... 'ii'xb6! play can continue 8 1 d4 liJf6 2 C4 cs 3 d5 b5 4 liJf3 e4 i.g7(n ot yet 8 ...i.g4? ! because of 9 103 Play the Benko Gambit 'ii'a4+ .i.d710 'illb3 'ii'C7 11 .i.e2.i.g7 12 11 lIa4 'iVb7, he would be unable to tiJd2an d White is slightly better) 9 i.e2 continue with the thematic ....i.d7-b 5 0-0 10 0-0 (here 10 tiJd2 is more accu since his knight is already occupying rate, when 10... tiJbd7 11 0-0 l:tb8 12 a4 the d7-square. transposes to the game) 10... i.g4 11 h3 10 ... 0-0 11 .i.e2.l:tb8 12 0-0 (or 11 tiJd2 .i.xe2 12 'iVxe2 tiJbd7) 11... .i.xf3 12 .i.xf3 tiJbd7an d Black has equalized. 8 e4 .i.g7 9 a4 'iYxb6 12 •..Wa 7 Black will have to retreat his queen in any case after White plays a4-aS or tiJC4,an d so elects to does so at once, in 10 tiJd2 order to keep the options open for his Milov could have exploited Black's other pieces. The a7-square is chosen move order by playing the sequence 10 so as to leave c7 free fora knight after as! 'ii'a7 11 tiJd2 0-0 12 tiJC4 tiJe8 13 ...tiJe8-c7, or even ...tiJb 6-a8-c7 (as in i.e2 l:tb8 14 0-0, transposing to his the note move 14). 12 ..."ii' d8 and later games against Anand and Naka 12 .. .'ii'C7 are also played, and the for mura, given in the note with 13 as tiJe8 mer is seen (via another move order) in below. the next game; the latter makes less Instead, Black might retreat the sense here, however, as after 13 as the queen to d8, aiming to transpose to the queen can get in the way and often has next game after 10 ...Wd8 11 i.e2 0-0 to move again. 12 0-0 l::tb8 13 tiJd2, but in that case 13 tiJC4 White has other options such as 11 13 as is a major alternative here; h3!? 0-0 12 i.C4tiJe8 13 0-0 tiJC714 .i.f4 and indeed, as we saw in the note to with a space advantage. move 10, White may already have Of course in the previous game played this move en rou�e to a transpo (Campos-Topalov) we saw Black answer sition below. From this exact position a4-aS by ... 'iVb4, but here after 10 ... 'iVb4 Black has a few choices: 104 White Plays b5-b6 ttJxb6 f5 (17... i.. d7 might improve but it doesn't equalize; e.g. 18 i..d3 e6 19 ,Ubl lUe8 20 i..e3 exd5 21 exd5 ttJd422 l:tel and the b6-knight still gives White an edge) 18 ttJxc8(after 18 exf5?! i..xf5 19 i..g4 �7 20 l:ta4 ttJd4Bl ack was finein V.Milov-V.Anand, Bastia rapid 2005, and later went on to win) 18... l:tb xc8 19 exf5 gxf5 (19 ...l:txf5 20 i..g4 skewers the rooks) 20 i..C4 'it>h8 21 lIel i..f6 22 'iVh 5 a) 13 .. J:tb4?! 14 ttJC4! is good for White. For example, after 14... ttJe8 15 i..d2 ttJe5 16 ttJb6 White has achieved one of the main strategical goals in this line: to make a blockade on b6 with one of the knights. b) 13... ttJe8 14 ttJC4 ttJC7 15 ttJa4 again intends to plant a knight on b6 and block the b-file. (Milov's improvement has yielded him a definite advantage; also good was 22 'ikd3 llb8 23 I:tbl - Milov) 22 ...ttJd4 23 lIa3 ttJC2 24 11h3 e5 25 .l:.fl ttJd4 26 i..xa6 l:tC7 27 i..d3 e4 28 i..h6 ':'g829 ':g3 lIcc8 30 i..C4 i..e5 31 a6 f4 32 lIxg8+ ':'xg8 33 �hl 'ilkd7 34 g3 ttJf3 35 i..e2 ttJd4 36 gxf4 i..g7 37 i..g4 'fIe7 38 i..xg7+ 'iWxg7 39 h3 ttJe2 40 'iWf5 e3 41 'iWe4h5 42 'ilkxe3 hxg4 43 'ili'xe2 'iVh6 Anand managed to win from here 44 f3 gxh 3 45 ':'gl ':'xgl+ 46 'it>xgl as Black with the plan of ... ttJb5-d4 and 'ii'xf4 47 a7 'ii'd4+ 48 �hl 'iWa4 49 'ilke7 .. .f7-f5. However, Milov later improved Wdl+ 50 'it>h2 WC2+ 51 �xh3 'ii'f5+ 52 on that game to beat Nakamura, and 'it>g3 'iWg6+ 53 'it>h4 1-0 V.Milov the current status is that 13 ...ttJe8 al H.Nakamura, Philadelphia 2006. This lows White a small advantage. For in game is a good illustration of what stance, 15 ...ttJb 5 16 ttJab6 ttJxb6 17 Black should avoid in this line. 105 Play the Benko Gambit c) 13... liJe s! is my recommendation 24 1:[c2 dS 25 liJb6 'ii'd6 and in this un here. By moving his knight to eS Black clear position Black should not be pre-empts his opponent's knight ma worse. noeuvre to c4 and b6, which would c3) 14 <;i(h lgS!? (risky but necessary; now simply be swapped off. The obvi Black maintains the knight on the eS at ous problem is that the knight might the cost of weakening his kingside) 15 be chased away with f2-f4, but at the liJc4 (after 15 'ii'c2 'ii'b7 16 b3 e6! 17 moment it is safeguarded by tactics on dxe6 fxe6 18 .l:.bl "fie7 19 liJC4 liJxC4 20 a7-g1 diagonal (one of the reasons be bxc4 l:txbl 21 'ii'xbl liJd7 22 liJdl liJes hind 12 ... 'ii'a7), and otherwise Black Black is doing fine with another strong gains the time to attack the centre with knight on es) lS... liJxc4 16 i.xC4 l:tb4 ...e7-e6. With 5 b6 White tries to play in 17 'ii'e2 h6 18 liJa4 i.d7 19 liJb6 i.bS a positional way, but tactics still play with an unclear position; e.g. 20 f4 an essential role. 'ikxb6!? 21 axb6 i.xc4 22 'ii'el i.xfl 23 'ii'xfl l:txb6 24 fxg s hxgs and Black has enough compensation for the queen. So 13 ...liJes is a sharp, uncompro mising move which leads to an unclear game. 13 ...liJb6 Cl) 14 f4 liJd3! 15 �xd3 c4+ 16 <;i(hl cxd3, and even if the d3-pawn drops off Black will obtain excellent compensa tion with the two bishops and rooks on the open b- and c-files. (2) 14 h3 e6 15 f4 (or lS liJC4?! liJxC4 16 i.xc4 exds 17 exds J:.b4 18 b3 liJd7 This move is a useful option when and Black is doing fine with his active White hasn't played M-aS. pieces) lS ... liJd3! (a nice move to play 14 liJa5 and not really something White should Alternatively: allow) 16 dxe6 i.xe6 17 fS liJxcl 18 a) 14 liJe3 lle8 15 a� liJa8 (heading lIxclgx fs 19 exfs i.ds 20 liJxdsliJxds for C7-bS-d4) 16 liJC4 liJd7 17 f4 liJC7 18 21 liJC4i.d4+ 22 <;i(hl liJf6 23 i.f3'ilVC7 <;i(hl liJbs 19 liJa4 liJd420 liJab6 liJxb6 106 White Plays b 5 -b6 21 liJxb6 liJxe2 22 'ii'xe2 V2 -V2 A thematic break in the centre. V.Dimitrov-M.Perez Fungueiro, Ponte Unlike in many queen's pawn open vedra 2003, though I think Black is ings, the ds-pawn is not supported by a slightly better after 22 ...e6, with an ac colleague on c4, while if White takes on tive position and the bishop pair e6 the black pieces become very active, (unless White wishes to give up his especially after ... d6-ds opens up the blockading knight). position further. This is what Black is b) 14 liJa3 liJa8 15 liJc4 .:tb4 16 liJa2 aiming for, whereas White has not ':xC4!? (otherwise 16.. J::t b8 is fine) 17 managed to make a blockade on b6 i.xc4 liJxe4 with compensation in and his knight on as is now misplaced. N.V.Pedersen-L.Van Wely, Amsterdam 17 dxe6 i.xe618 b3 2002, as it hard for White to do very Not 18 i.xa6 1i'a7 19 1i'd3 c4 and much; indeed, Black later won after I White loses a piece. overplayed my hand. At that time I was 18 ... dS 19 exds liJfxds 20 liJxds liJxds inspired by the games of Gleizerov and Black is slightly better, due to his Dreev, who are still playing 5 b6 suc more active pieces and the offside cessfully as White, but I went back to white knight. the Fianchetto Variation. 21 l::tacl i.d422 liJC4 liJf4 23 i.xf4'iWxf4 14 ...i.d7 lS 'ii'c2 15 liJc6 i.xc6 16 dxc6 :fc8 17 as liJa8 18 i.f3 liJc7 19 liJa4 liJe6 20 i.e3 liJe8is also unclear; e.g. 21 es i.xes 22 liJb6 lIC7 (orjust 22.. Jbb6 !? with com pensation) 23 %:tbl liJd4 24 i.xd4 cxd4 with ideas of ...liJf6, ...e7-e6 and ... d6- ds. lS ...'ii' C7 16 i.d2e6 Now Black has the bishop pair and an active position, and Topalov goes on to convert his advantage into a win. The b3-pawn and the white king are obvious targets here. 24 i.d3�g7 2S :cel 'iWf6 26 i.e4hS Aiming to create weaknesses on the white king side. 27 �hl .l:r.b428 f4 i.xc429 bXc4 J:te8 30 107 Play the Benko Gambit fS gs 31 �dS ':'xe1 32 ':'xe1 lib2 33 ing the a-pawn is recommended by GM 'iWe4:f2 34 'iVd3 :xfS 3S 'iWd1g4 36 g3 Kiril Georgiev in Squeezing the Gambits. liIf2 37 �e4 �c3 38 l1g1 �d4 39 �e1 Note that Topalov's ...'iVa7 is less 'iigs 40 :f11 appropriate against a quick ttJd2 and Losing. 40 'iVcl was necessary. a4-as, as after 11 .. .'ii'a7 12 ttJC4 White 40 .. :ii'd21 41 'iWxd2 l:txf1+0-1 has already achieved base camp to wards his desired blockade; whereas Game summary Black is unable to respond with either With 12... 'iWa7 and 13... ttJb 6 Topalov ... ttJb6 or ...ttJe s, and so has nothing obtained good play and won a nice better than 12... .l:tb8 13 0-0 ttJe814 0-0, game. White could have prevented reaching the slightly unfavourable po ... ttJb6with 13 as, but then 13 ... ttJes! is sitions from Milov-Nakamura in the also unclear, as seen to the notes notes to the previous game. above. It is characteristic of this line 8 e4 g6 9 SLe2 that Black can achieve at least an un Or 9 ttJd2 �g7 10 a4 0-0 11 as 'iVd8 clear game with active play. as in the previous note. Note that In the opening, rather than commit 11... 'ilVb4 12 1:ta4 'iVb7 is again ineffec the knight to d7 so early, I would prefer tive, because the black knight is ob to play ...'ii' xb6. Then Black keeps the structing the d7-square, thus prevent option of answering a quick a4-aS with ing ...�d7-b S. ...'iVb 4, ...'iVb7 and ....i.d7 (as in Game 9 ...�g7 10 ttJd2'ii' d8! 25), while in some cases the bishop can also go to g4. Game 27 J.Hellsten-V.Baklan German League 1997 1 d4 ttJf6 2 c4 cS 3 ds bs 4 cxbs a6 S b6 d6 Or S ...'ilVxb6 fol lowed by ...g7 -g6 and ...d7-d6. Another prophylactic retreat, pre 6 ttJC3'ii' xb6 7 ttJf3 ttJbd7 empting ttJC4 or a4-aS. The natural Here 7 ... g6 8 e4 �g7 9 ttJd2 0-0 10 10 ... 0-0 would give White the option of a4 ttJbd7 11 as 'iVd8woul d be my move 11 ttJC4 followed by 12 �f4, whereas order, when 12 �e2 l:.b8 13 0-0 trans now 11 ttJC4might be met by 11 ... ttJb6. poses to the game. The plan of advanc- 11 a4 0-0 12 0-0 lIb813 as 108 White Plays bS-b6 Now there is no ... liJb6 in answer to Not now 16... liJd4? 17 liJxd6!. But liJC4, so Black must play for ideas like 16 .. .fs! works especially well with the ... liJe8-C7-bS-d4 -and sometimes ... e7-e6 bishop on gs as .. .l::txfS gains a tempo. or .. .f7-fS. For instance, 17 exfs (17 f3 h6 18 i..e3 13 ...liJe8 14 liJc4liJC7 15 liJa4 liJb5 fxe4 19 fxe4 :xfl+ 20 'ili'xfl liJd4is also unclear) 17.. .l::txfS 18 h4 liJes 19 liJab6 liJxC4 20 liJxc4 liJC7 21 liJb6 i..xb2 22 ]::ta2 .i.d4 23 .i.g4l:tf7 24 liJxc8.l:lxc8 2S .i.xc8 .xc8 and Black has good com pensation for the exchange with his active pieces and the weak white pawns at as, dS and h4. 17 .l::tb1 The immediate 17 liJab6 seems more accurate. White is now ready to plant a knight on b6, while Black is ready to put his on d4, after which the advance ...f7-fS is more effective as the ds-pawn undermined. Obviously having an op posing knight on such a great square as d4 can quite dangerous for White. Again active play is the way for Black. 16 i..g5 16 liJab6 should be met by 16... liJd4 ! (Georgiev only covers 16 ...liJx b6?!, which 17 .. :f1C7 may well offer White a slight advan Here Black might try 17 ... h6 18 .i.f4 tage) 17 .i.d3 liJxb6 18 liJxb6 e6 19 .i.e3 gs 19 .i.e3 e6 20 dxe6 ':'xe6 21 f3 liJd4 (19 .i.C4'it>h 8 20 'ili'd3 exds 21 i..xds fS! 22 liJab6 .i.b7 with an unclear game. gave Black good play in I.Golichenko Black has good activity and can even V.Baklan, Ukrainian Championship, Kiev answer i..d3 with ... gS-g4. 2008) 19 ... exds 20 exds l:te8 21 :bl 18 liJab6liJf6 19 liJxc8 :bxc8 20i.. d3?! 'ifh4 22 liJxc8 lIbxc8 23 .i.xa6 :a8 24 Correct was 20 .i.xf6, intending liJb6 'it'd3 l1e4 with an unclear game; one and a later b2-b4 with a slight advan idea is 2S... liJf3 + 26 gxf3 :g4+ 27 fxg4 tage. The a6-pawn is a long-term iixg4+with a draw by perpetual check. weakness and it is not easy for Black to 16 ...l:te8 obtain counterplay here. 109 Play the Benko Gambit 20... ltJd7 Losing at once, though White was Now the position is unclear again. much worse in any case. 21 ..td2e6 2S... ltJe c31 29 bxc3 ltJxC330 ..tf3 Faced with threats at d1 and e2 White elects to give up his queen, but his slight initiative is easily neutralized. 30... ltJxd 1 31 ..txdS+ �hS 32 nb7 'ii'xas 33 llxd1 ':'C7 34 ..tf7 �e1+ 35 �h2 :xb7 36 :dS+ 'ii'xdS 37 ..txdS':'xf7 0-1 Game summary This game shows some nice ideas against an early a4-aS, the primary one being ... ltJc7-bS-d4, exploiting the fact The thematic break. As the game that the white infantry can no longer continuation shows, the pawn roller control bS. Other themes include ...e7- starting with ... e7-e6 ancl ...d6-dS can e6, or ...f7-fS (as seen in the note to be quite dangerous for White. move 16), while ...ltJd7 -es is sometimes 22 dxe6 fxe6 23 h4?! possible too. White should keep an eye on the The position from the opening centre; e.g. 23 i.f4 'ii'c6 24 ltJe3 ..td4 phase was unclear. In the early middle and the game is still in the balance. game Hellsten made a few inaccura 23 ••.dS cies, allowing Black to obtain a strong Now Black's advancing pawns give initiative with ...e7-e6 and ...d6-dS, af him a definite advantage. ter which it was all one-way traffic. 24 exds exds 2s ltJe3c4 26 ..te2 ltJf6 27 ltJg4 ltJe42S i.gS? Game 28 P .H.Nlelsen-N.V .Pedersen Danish Team Championship 2006 1 d4 ltJf6 2 ltJf3 cS 3 ds bs 4 c4 Note the rather unusual route to the Benko. White's other fourth moves, of which 4 ..tgs is the most significant, will be examined in Chapter Eight. 4 ...g6 5 cxbS a6 6 b6 110 White Plays bS-b6 I used to play the bs-b6 line myself queen then covers d6 and es, as well as as White (see the notes to Game 26 for as. Importantly, Black cannot be pre example). In this game I ventured the vented here from playing ... lDb6 to Benko as a surprise weapon and it challenge the knight on c4, so White turned out very well. Since then I have has more difficulty in blockading the b been playing the Benko more regularly, file. generally also with good results. 12 �f4 :b8 6 .. :ifxb6 7 lDC3d6 8 e4 �g7 9 lDd2 A thematic move, bringing the rook If White is intending to play this to the half-open b-file. At this point I way he might as well do so at once, saw that the forthcoming sacrifice was ruling out any ideas of ... �g4. probably problematic for White. 9 ...0-0 10 �e2 lDbd7l1 lDc4 13 'ili'd2?1 13 0-0 is sounder, when 13 ...lDb6 can be met safely by 14 lDe3or even 14 es!? Instead, I like 13 ...lDe s!?, which may not seem that logical, but it's based on concrete tactics (the b2-pawn is hanging) and leads to an unclear position. For example: 14 lDxes (or 14 �xes dxes intending ...lDe8-d6) 14 ...dxe s 15 .te3 (or 15 �Cl e6 with counterplay against the centre) ls ... .:xb2 16 lDa4 J:[xe2!? 17 'i¥xe2 11 •.. 'i¥C7 lDxe418 lDxcs lDC3 19 'i¥d2 e4 20 lilacl Both 11...'i¥a7 and 11...'i¥d8 are pos 'ifes 21 lDb3 lDxds 22 �d4 was seen in sible (in the style of the previous two L.Seres-L.D.Nisipeanu, Eger Agria 1995, games), but I prefer 11...'i¥C7 as the and now 22... 'i¥fs 23 �xg7 Wxg7 24 111 Play the Benko Gambit 'iWd4+ (or 24 11C5 �b7) 24... tiJf6 25 tiJC5 18... 'iWxa4 19 'iWxe4�f5 .i.e6 26 a4 :a8 27 tiJxe6+'ilixe6 looks to give Black reasonable compensation with two pawns for the exchange. 13... tiJb6 14 tiJe3? Now White can't cover the knight on c4 in a good way. 2o 'iWxe7 If 20 'iWe3 1Ixc4 21 b3, then 21 .. :iVd7! Other moves allow Black to equal 22 �xc4 �xal and Black is better. ize; e.g. 14 0-0 tiJxC4 15 �xC4 tiJg4an d 20 ••. .:te821 'iWxd6 �xb2! 22 tiJxb2 Black is doing fine. 22 .l:tel would be met by 22 ... �c3 14... tiJa4! ! with a decisive advantage. The idea is to take the pawn on M, 22.•• l:txb2 23 �e3 check on as, and play ...llb4. Then This fails spectacularly to defend White has weaknesses at a4, b2, f2 and the e-file. f4, while his king is prevented from 23 .. Jlxe2! castling. Back in 2006, the engines I tested this position on after the game were unable to see that 14... tiJa4 is good for Black. 15 tiJxa4 tiJxe4 16 'iWC2 'iVa5+ 17 �f1 l:tb4 Attacking the knight on a4, and with a veiled threat against the bishop on f4. 18 tiJC4 Not 18 tiJC3? tiJxC3 19 bxc3 lIxf4, while after 18 �C4 'ii'xa4 19 'ili'xe4 �f5 24 'it>xe2 20 tiJxf5 gxf5 Black regains the piece Or 24 'iWc6 �d3! 25 'ifx a4 !;t2xe3+26 with a clear advantage. 'it>gl l:tel+an d mates. 112 White Plays bS-b6 24 ...'i¥C4+ 25 �e1 'ifc3+ 26 �e2 iLg4+ reasonable defence is the artificial 27 f3 'ii'xe3+ 28 �f1 iLf5 0-1 looking :a4, but then, after retreating There is no serious defence to the queen, Black can play ...iLd7-b S and 29 ... iLd3+an d mates. equalizes, as seen in Game 25. Conse quently, this variation is no longer so Game summary popular for White. In the opening White played an early In the main lines White mostly plays tLld2-C4, to which I replied ...'ifC 7. Then tLld2-C4 and/or a2-a4-as, so that Black 13 0-0 would have been interesting has to remove the queen from b6. Just from a theoretical perspective, when where the queen should go depends on the unorthodox-looking 13 ... tLles!? the precise circumstances. leads to an unclear game. Instead, Niel In Game 26 White played a2-a4, sen mistakenly delayed castling, allow tLld2, iLe2 and 0-0. Here ... '¥Ia7is a good ing the dangerous, if unexpected, move, intending either 13 tLlC4tLlb6 or knight sacrifice with 14 ... tLla4!!, after 13 as tLles!. The latter is supported by which I don't think there is really a the tactical idea of answering f2-f4 good defence for White. The game was with ...tLld3 !, as the knight is indirectly a good illustration of the activity Black defended by the ... cS-c4 check! is aiming for in this line, as well as In Game 27 White plays a quick showing how a couple of inaccuracies tLld2 and a2-a4-aS (the most accurate by White can sometimes be disastrous. move order is given in the notes). In this case the queen does better to drop Summary back to d8, when Black can generate In the 5 b6 line White is trying to play counterplay by ...tLle8-c7- bS-d4, exploit in a more positional way against the ing the hole on bS, with further ideas Benko Gambit, using the b6- and c4- of ... e7-e6 or ...f7-fS . squares in particular to try and restrict Finally, in Game 28, White played a Black's counterplay and prove an ad quick tLld2-c4 without a2-a4. Here it vantage; the advance e4-eS is some makes sense to play ... 'ifC7,from where times a theme too. However, after the queen controls more squares, while ...'iVxb6 Black can still aim for active the white knight on C4 can be chal play on the b-file, with a rook on b8 lenged by ...tLlb6 oreven ...tLle s!? supported by the bishop on g7. Often Overall, my evaluation is that the Black is able to obtain at least equality bS-b6 line offers chances for both sides, by tactical means. and as long as Black plays with due White should take care not to play care and precision, I don't think that a4-aS too soon, as ...'ii' b4 gives him White should really be able to prove a problems with the e-pawn. The only theoretical advantage here. 113 Chapter Six Other Options for White 1 d4 lbf6 2 c4 cS 3 dS bS 6 e3 and 6lbc3 (see Game 31). The rest of the chapter covers White's various declinations on move 4: the Trompowsky-like 4 .i.g5 (Game 32), the unthreatening 4lbd2 bXc4 5 e4 (Games 33 and 34), the unambitious 4 a4 (Games 35 and 36), the unchalleng ing 4 f3 (Game 37), and the rather du bious 4 e4 (Game 38). As the reader will gather from my choice of descriptors, none of these lines offers White a seri ous chance of a theoretical advantage. Here we round up White's remain ing ways of playing against the Benko. First is the Zaitsev Variation, 4 cxb5 Game 29 a6 5 lbc3axb 5 6 e4 (Game 29), which is M.Hochstrasser-D.Pikula sharp but not theoretically dangerous Pontresina 2000 for Black. Indeed, it is White who needs to think about equalizing here. 1 d4 lbf6 2 c4 cS 3 ds bs 4 cxbs a6 5 The next two games feature 4 lbf3, lbC3 which is a significant alternative to 4 The so-called Zaitsev Variation, with cxb5 - for White, that is. For Black I which White aims for rapid develop suggest the consistent 4 ...g6 5 cxb5 a6, ment and tactical play from the open offering to transpose to the main lines ing. It is not so popular these days and after 6 bxa6 or 6 b6. Other ideas for should perhaps be con�ined to rapid or White include an early 'ilVc2 (examined blitz games as, in many cases, Black in Game 30) and the tricky move orders takes the initiative and the question is 114 Other Options fo r White whether White can equalize. a) 7 dxe6 (giving up the centre like this is not usually advisable for White) 7 .. .fxe6 8 e4!? (the sharp option; in stead, 8 e3 ds is just equal: Black has a nice centre with ideas of ...lLlc6, ...i.e7, and possibly ...e6-e s, ...lLle4, or some- times ...l:tb 8) 8 ...lLlxe4 (Black takes up the challenge; otherwise 8 ...lLlc6 is more quiet, though still unclear) 9 'ilVhs+ (critical; after 9 lLle2 i.e7 10 lLlec3 lLlxc3 11 bxc3 0-0 12 i.d3 .l:tf7 and ...d7-ds Black is doing fine, with his S ...axbs 6 e4 control of the centre) 9 ...g6 10 'ifes This line is not critical. Instead, 6 'ii'as+ 11 'it>dl (or l1 lLlc3lLlxC3 12 i.d2 lLlxbs i.a6 7 lLlC3 g6 would transpose :g8) 11 ... lLlxf2+ 12 'it>e2lLlxh l 13 lLlC7+ to the main lines if Black is willing to 'it>f7 14 i.d2"iVa4 15 'ii'xh8 with an un do that. However, there is an additional clear game in O.Uwira-L.Milov, Bad option in 6 ... e6!, which I would prefer Homburg 2009. as it equalizes more simply; indeed, b) 7lLlc3 exds 8lLlxds i.b7 White must be careful not to come out worse from the opening. After 6 ...e6 Black might be regarded as having an improved version of the 5 b6 e6 line. True, the white knight has some active possibilities on bs, but Black has more control over ds. There can follow: 9 e4?! (but after 9 lLlxf6+ 'iVxf6 10 'ii'b3 i.c6 11 lLlf3 ds 12 g3 i.d6 Black has good compensation with his strong centre and chances n the queenside) 9 ... lLlxe4 10 i.C4 'ifas+ 11 'it>fl lLlc6 12 i.f4 0-0-0 13 lLlf3 i.d6 14 "iVcl?! (14 i.e3 is still unclear) 14... i.xf4 15 'iVxf4 'uhe8 and Black was slightly better in 115 Play the Benko Gambit G.5osonko-H.5chussler, German League position has been played. However, 1982. Black has an interesting novelty which 6 ...b4 7 1ZJbsd6 my friend IM Rasmus Skytte has found for this book: 11 ...lZJxe 4!? 12 i.xe4 i.g7, playing against White's weak points at al, a2 and bS. Taking the e-pawn this time would be a terrible mistake, due to 7 ...lZJxe 4?? 8 'ii'e2 1ZJf6 9 1ZJd6ma te! 8 i.f4?1 For example, 13 lIbl I:l.xa2 14 0-0 The main line. White has ideas of i.a6 15 i.d3 0-0 16 i.C4 'ilfd7 17 'ii'd3 e4-eS, while still hoping for 8 ... lZJxe4 9 l:ta8 18 i.f4 l:tas 19 lZJxd6 exd6 with a 'ii'e2 lZJf6?? 10 lZJxd6+ and wins, al clear advantage to Black, or 13 i.e3 though 9 ...gS! here is now unclear - 'ii'd7! (the knight on bS is toast) 14 and Black does even better to play that lZJbd4 (14 a4?! bxa3 15 lZJxa3 'i!i'g4 16 move straight away. Therefore, instead i.d3 l:txa3! 17 l:txa3 'ii'b4+ is even of the game continuation, White worse) 14 ...'i!i'g 4 15 lZJgs 'iWxdl+ 16 should probably try: l:txdl cxd4 17 i.xd4 i.xd4 18 l:txd4 a) 8 i.d3 (or 8 lZJf3 lZJbd7 9 i.d3) ':'xa2 19 0-0 h6 20 l:txb4 1ZJxds 21 i.xds 8 ...lZJbd7 9 1ZJf3 g6 10 0-0 i.g711 a4 0-0 hxgs and Black is a pawn up. White's 12 ':'el lZJg413 h3 lZJges 14 1ZJxes lZJxes best reply may be 13 i.d2, intending 15 i.fli.b7 with a level position. 13 ...'iWd7 14 a4 bxa3 15 i.c3!, but after b) 8 i.C4 lZJbd7 (avoiding the sharp 13 ...i.xal 14 'ili'xal :g8 15 i.d3 'iWd7 16 lines after 8 ...g6 9 eS dxes 10 d6 etc is a4 bxa3 17 lZJxa3 i.a6 18 i.xa6lha6 clearly Black's strongest option, espe Black looks slightly better even then, cially as these kinds of tactics are what due to the exposed knight on a3 and White is aiming for with the Zaitsev weak pawn on dS. Variation) 91ZJf3 lZJb6 10 i.d3 g6 11 b3; 8 ...gsl this position has occurred many times Although there are- other options, and so far only 11... i.g7 with an equal this is the strongest move here. Now 116 Other Options fo r White Black is slightly better, due to the vul 1:txa6 20 Wxgs l:tb6 21 l:txb6 Wxb6 with nerable white knight on bS and pawns good winning chances for Black due to on e4 and dS. - the weaknesses at a3 and dS; e.g. 22 liJe2"iWb l+ 23 liJclliJd7 24 0-0 liJf6 25 l:tel"iVb 2 26 liJe2:g 8 27 WfS 'itf8. 9 ...liJxe4 10 i..f4 liJf61 9 iLxg5 White has also tried: a) 9 es?! gxf4 10 exf6 liJd7! 11 Wf3 (no better is 11 fxe7 Wxe7+ followed by The simplest continuation. There is 12 ...liJe s, or ll liJh 3 liJxf6 12 liJxf4iLh 6) no need at all to enter the complica 11... liJxf6 12 iLc4iLg 7 13 liJe2(V.Milo v- tions of 10 ... i..g7 11 'ii'e2 liJf6 12 liJxd6+ A.Tzermiadianos, Heraklio 1993) 'itf8 13 liJxc8 Wxc8, even if Shirov and 13 ...liJd7 ! with a clear advantage to Khalifman have both played this way. Black. After the text Black has a slight ad b) 9 iLe3 liJxe4 10 iLd3 "iWb6 (saving vantage: the f8-bishop can safely take the piece tactically) 11 'ii'e2 l:.as 12 a4 up residence on the long diagonal, (after 12 i..xe4 'ii'xbs 13 WxbS+ :xbS whereas White has problems with both 14 i..xgs :as 15 liJe2 i..g7 Black is the knight on bS and the weak dS clearly better due to his activity and pawn. queenside pressure) 12 ...bxa3 13 bxa3 ll .i.C4 i..g7 14 lilbl i..a6 15 f3 (15 liJf3 allows Now 11 'ili'e2 lla6 gets White no Black to swap down to a favourable where, while after ll liJf3 i..g7 12 i..d3 endgame with lS... liJC3 16 liJxC3i.. xC3+ l:tas13 0-0 liJxds14 i..g3 0-0 15 'ili'C2 h6 17 i..d2 'ii'xbl+ 18 i..xbl i..xe2 19 i..xC3 16 l:tfdli.. e6 Black is clearly better with i..xf3 20 gxf3 l:tbS 21 i..c2 0-0 22 h4 g4 his active pieces, greater central con 23 :gl f6, when White's bishops are trol and extra pawn. insufficient compensation for the ex 11 ...i.. g7 12 liJe2 change) lS ... liJC3 16 liJxC3 i..xC3+ 17 Or 12 liJf3liJbd7 and Black is doing i..d2 i..xd2+ 18 'ii'xd2 'ii'C7 19 i..xa6 fine,since the attacking try 13 liJxd6+!? 117 Play the Benko Gambit exd6 14 �xd6 should ultimately prove against the black king. unsuccessful after 14... tDe4 15 'ii'e2 f5 19... tDxC4 20 bXc4 tDe521 f4? 16 tDd4 (or 16 tDd2 'ii'f6 17 tDxe4 fxe4 21 'ii'e2 was called for; it was no good 18 'iVxe4+�d8) 16 ... 'ii'f6 17 tDb5 �f7 18 letting the c4-pawn go. �C7 �g6, as in B.5alvermoser 21... tDxC4 22 f5 tDxd2 23 'ii'xd2 �e5 S.Camilleri, Bad Worishofen 2005. 12•.. tDbd7 The b6-square is great for a black knight here - attacking the bishop on c4 and the pawn on d5. 13 tDg3?! It is natural that White should want to place his knight on f5 or h5, but this isn't really an option unless Black is especially careless. 13 ...tDb6 White's try at an attack has been easily parried and Black now has a de cisive advantage. 24 :f3 f6 25 tDh5 l::tf7 26 g4 �h8 27 tDf4 l::tg7 28 �h1 :ta8 29 .:th3�xf4 30 'ii'xf4 'ii'b6 31 tDe4a2 0-1 Game summary The Zaitsev Variation with 5 tDC3axb 5 6 e4 can be dangerous if Black plays incorrectly, but if he responds in the Black is clearly better, as White can right way the risks are reversed! In that hardly defend his existing weaknesses, case White's best is probably to defend nor safeguard his knight without the the e-pawn with 8 �d3 which leads to further weakening a2-a4. an unclear game. The old main line 14 b3 0-0 15 0-0 �d7 16 �g5 l1a5 with 8 i.f4 now seems to be too slow The knight on b5 is the obvious tar because of 8 ... g5!, when Black takes get and has no escape unless White over the initiative with the better posi plays a2-a4. tion. The white knight is quite vulner 17 a4 bxa3 18 tDc3tDg4 19 �d2?1 able on b5, while the trade of the e4 for 19 'iVc2fol lowed by tDh 5 might be a g5 has left the d5-pawn as a serious better chance of getting some play weakness. Hochstrasser was unable to 118 Other Options fo r White fi nd any real counterplay and lost a ttJf3 0-0 9 0-0 and White is slightly bet- rather one-sided game. ter. After4 ...'ii' as+ we have: Game 30 E.Bareev-V.Topalov Sa raje vo 2000 1 d4 ttJf62 C4 cS 3 ds bs 4 ttJf3 a) 5 ttJC3 ttJa6! 6 cxbS (now after 6 a3 b4 or 6 e4?! b4 Black is slightly bet ter) 6 ... ttJb4 7 'iVd2 jLb7 8 e4 (8 a3 ttJbxdS 9 ttJxdS 'ii'xd2+ 10 jLxd2 jLxds produces a balanced endgame) 8 ... e6 and Black is doing fine. Here 8 ... ttJxe4?! is tempting, but after 9 ttJxe4ttJC2+ 10 This is White's second-most popular c;to>dl'ii' xd2+ 11 jLxd2ttJxal 12 ttJC3th e move (after 4 cxbs). It's prime aim is to knight on al is lost and White's knights avoid certain lines by Black, such as an must be stronger than the rook, early ...e7-e6 in the bS-b6 variation. regardless of his wrecked pawn However, White does have some inde structure after... ttJb 3. pendent options as well, in particular b) 5 .1i.d2 (probably the only serious the idea of an early 'l'c2. attempt to get an edge) In fact White can play 4 'iVc2 right here. This looks a bit odd, but Black must still play precisely in order to equalize. In my opinion, the rarely played 4 ..:iVa s+! is the best reply, as it forces White to place his pieces in a slightly awkward way. Against other moves I think White has good chances of an edge; for instance, 4 ... ttJa6 5 a3! bXc4 6 e4 e6 7 jLxC4 exds 8 exds or 4 ... bxc4 5 e4 d6 5 jLxC4 g6 7 f4 jLg7 8 119 Play the Benko Gambit S ...b4 6 e4 d6 7 .td3 tDbd7 8 f4 (8 .txb2 19 d6! "iIIxa3? 20 'iWe4! and tDf3 g6 9 0-0 .tg710 h3 0-0 11 a3 .tb7 mates) 18 l:lael .td7 19 .te4 �f8 20 d6 is TOughly equal) 8 ... g6! (8 ...gS?! was 0-0-0 21 dxe7 lIxfl+ 22 ':'xfl .l:%.e8 with tried in S.Koutsin-D.Trifonov, Kiev 1999, an unclear position. but then 9 fxgs tDg4 10 tDf3 .tg711 0-0 4 ...g6 tDges 12 .te2 was good for White, and 9 eS dxes 10 fxg s might be even stronger) 9 tDf3 .tg7 and then: This is Black's most consistent and popular continuation, still intending to offer the gambit pawn after 5 cxbS a6. bl) 10 0-0 0-0 11 h3 (if 11 I:Iel "fIC7 From the other side, White must also 12 eS, then 12 ... dxes 13 fxes tD94) be prepared for 4 ...b4, 4 ....tb7, 4 ...bxc4, 11... 'ii' C7! (keeping an eye on es) 12 lIel 4 ...d6 and 4 ... e6, but we don't have to as 13 eS (or 13 a3 ':b8 14 .te3l:te8 15 worry about those. The only drawback tDbd2 eS 16 fxes tDxes 17 tDxes I1xes to the text is that, should White decide 18 axb4 axb4 19 .tf4 �e8, followed by to play an early e2-e3 or tDc3, Black is ...tDd7 with an equal game) 13... dxe s 14 now committed to a fi anchetto. We fxes tDh S! is okay for Black, since 15 g4 will have a quick look at these lines in is met by ls ... tDxes! (lS ... .txes 16 .tfl! the notes to White's 6th move below. is less clear) 16 tDxes .txes 17 gxh S scxbS .txb2! 18 "fIxb2 "fIg3+ and draws. 5 'iWc2is a positional way of playing b2) 10 eS dxes 11 fxes tDg412 e6! is and leads to a game with equal the direct try and gives White some chances. One example: S ...bxC4 initiative, but it is hard to foresee any (s ..."fIa s+ is possible here too) 6 e4 ..tg7 advantage without the queen side 7 .txC4 d6 8 tDC3 0-0 9 0-0 .ta6 pieces in play. For example, 12 ...tDde s (9 ...tDbd7 is a sound alternative) 10 13 tDxeS tDxeS 14 exf7+ tDxf7 15 0-0 tDd2 (or 10 tDbs "fId7 11 a4 ..tb7, in ..td4+ 16 �hl tDeS 17 tDa3 (or 17 .te4 tending ...tDa6 -b4) 10 ..:ltJfd7 11 .txa6 "fIa6!) 17 ... a6 (not 17 ... tDxd3?! 18 Vxd3 tDxa6 12 tDC4 tDb6 13 'i!Ve2 tDxC4 14 120 Other Options fo r White 'iVxC4 'iVb6 15 l:tbl l:tab8 16 J.gs lUe8 8 g3 d6 9 .i.g2an d 8 e4 .i.xfl 9 'ltixfl d6 17 :fel ttJb4 18 'ii'e2?! (18 .i.f4, plan reach the main lines of Chapters One ning e4-eS, is' still okay for White) and Two respectively; 6 b6 'ii'xb6 7 ttJc3 18 ... 'fWa6 (now Black's pieces are well is the main line of Chapter Five; while 6 placed and he is slightly better) 19 'ii'd2 e3 and 6 ttJC3 are examined in the next 'iVd3 20 l:tedl 'ii'xd2 21 l:txd2 .i.xc3!? game. (beginning a correct combination 6 •••axb5 based on White's weak back rank, Previously Topalov had played though the preliminary 21 ... h6! would 6 ... .i.g7 7 e4 axbs 8 .i.xbs 'ii'as+ 9 ttJc3 have made this even stronger) 22 bXc3 .i.a6 10 .i.xa6 ttJxa6 11 0-0, when Black ttJxdS 23 l:txb8 ':xb8 24 f3 f6 25 l:txdS did not have full compensation for the fxgs 26 eS l:tb6 27 exd6 exd6 28 ':xgs pawn in V.Korchnoi-V.Topalov, Ant c;t>f7 with a difficult rook endgame for werp 1997. This time he was better White, which he was unable to hold in prepared and showed that Black is fine E.Vorobiov-V.Zvjaginsev, Serpukhov after his strong 8 ..:ii' a4!. Incidentally, I 2008. searched for other ideas for Black, but I 5 ...a6 6 'fWc2 think it is difficult to equalize without this move. This is one of the additional options provided by White's 4 ttJf3 move order, 7 e4 and has been tried by such uncom Naturally 7 'ii'xcs is critical, since promising players as Korchnoi, Yer this was one of the ideas behind "C2, molinsky and Ivan Sokolov. It is a but after 7 ...e6 8 'ii'xbs (or 8 d6 ttJc69 "greedy" continuation where White is ttJbd2 "as 10 e3 .i.a6 and White can often out for the c-pawn, but Black can hardly hope to hang on to the d-pawn) expect good compensation if he plays 8 ...ttJxdS Black has good compensation accurately. thanks to his active piece play, aided by Otherwise 6 bxa6 .i.xa6 7 ttJC3 .i.g7 the exposed queen on bS. 121 Play the Benko Gambit 7 .•Ji'a 5+ 8 �d2 11... liJg4 12 �f4 Or 8 liJbd2liJa6 9 �e2 liJb410 'ifbl 12 liJbs 0-0 13 �C3 f6 14 e6 dxe6 15 �g7 11 0-0 0-0 and Black is fine; e.g. 12 �C4 �a6 is also unclear. a3 liJa613 b4 'iVb6 14 �b2 d6 with an unclear game. 8 ...'ii' a4! 12 ...f6 12 ...d6 !? is an interesting alterna tive, when 13 exd6 (or 13 liJbs 0-0 14 This was a novelty at the time. Now exd6 exd6 and Black is doing fine) on 9 'ifxcS, Black can gain time on the 13 ...a3 ! (the point) 14 'iii>d2 (other queen after 9 ...'ii' xe4+ 10 �e3 (or 10 moves give Black the edge) 14 ...axb 2 15 �e2 liJa6 11 'ii'xbs 'ili'xds 12 'ii'xds .l:i.el e616 dxe6 �xe6 17 �bS+ liJd7 18 liJxds 13 liJC3 liJac7 14 0-0 �g7 - liJgs 0-0 19 liJxe6fx e6 20 �xd7 .l:i.xf4 21 Bareev) 10 ...liJa6 11 'ii'xbs liJxdS, again �xe6+ 'iii>f8 is again unclear. with an unclear game. 13 exf6 liJxf6 14 a3 9 'ii'xa4 bxa4 The safest move, preventing ...a4-a3 Although this pawn is potentially and thus fixing the pawn on M. vulnerable and might be rounded up 14... 0-0 15 0-0-0 d6 by liJC3an d �bS, as we will see it is not Black can even improve here with so easy for White to arrange that. The ls... liJe4! 16 liJxe4.l:i.xf4 17 liJfd2 (or 17 pawn can even be an asset if Black gets liJxcSd6 18 liJd3.l:i.c4+ 19 'iii>bl �fS with to play ...M-a3 at some point, sup compensation - Rogozenko) 17 ... d6, as ported by the bishop on g7. in 5.Ernst-D.Rogozenko, Dieren 2001, 10 liJC3 �g711 e5 when Black had the initiative and a The immediate 11 �bS?! would be slight advantage. met simply by 11...a3, while 11 a3 d6 12 16 l:tel lita717 �b5 �a6 18 �xa4?! �d3 liJbd713 �bS!? 0-0 14 �xa4 �a6, 50 White finally wins the a-pawn, followed by ...J:tf b8, gives Black excel but to no avail. He should limit himself lent compensation for the pawn. to 18 �xa6 liJxa619 lle2 with equality. 122 Other Options fo r White 24... tLla s 2S lib6tLlb3+ 26 Came 31 A.Sorin-D.Serrnek Preventing ....l:tb B. White might be World Student Team said to have some advantage now, Championship, Paranana 1993 though it is hard to prove it, given Black's still active pieces. 1 d4 tLlf6 2 c4 cS 3 ds bs 4 tLlf3 g6 5 123 Play the Benko Gambit cxbS a6 6 e3 White has one more attempt at a tricky move order with 6 4:Jc3axb s 7 d6 (otherwise 7 4:JxbS ..ia6 8 4:JC3 returns to the main lines), but this is not a theoretical threat at all for Black. In fact after 7 .....ib7 it is White who has to think about equalizing: By holding back e2-e3 until this moment White is trying to transpose to the 4 cxbS a6 5 e3 variation with Black committed to playing ... g7-g6 (whereas S ... ..ib7 was my recommendation in Chapter Three). Although the natural 6 .....ig7 7 4:JC3 is probably okay for a) 8 e4?! 4:Jxe49 4:JxbS (or if 9 4:Jxe4 Black, there seems to be no reason to ..ixe4 10 ..id3, as in Y.Gozzoli-A.wirig, acquiesce to the opponent's wishes in Hyeres 2002, then simply 10... ..ic6 11 this way. Instead, Black can try to ex 0-0 e6 and Black is slightly better) ploit the omission of 4:Jc3 by playing 9 .. :iWas+ 10 ..id2 4:Jxd2 114:Jxd2?! (but 6 ...axb s 7 ..ixbs 'ili'as+ 8 4:JC3 ..ia6, if 11 4:JC7+ 'it>d8 12 4:Jxd2, then which, 12 ...exd6 ! 13 4:Jxa8 ..ih6 14 ..ie2 ..ixa8 with strong play for the exchange) 11...4:Ja612 'ifb3 ..ih6 13 4:JC3..ixd2+ 14 'it>xd2 ..ic6 was good for Black in G.Hertneck-M.5choeneberg, German League 1990. b) 8 4:JxbS 'iias+ 9 4:Jc3exd6 10 ..id2 was seen in T.Radjabov-S.Halkias, Greek Team Championship 2002, when 10 ...'ifb 6 11 e4 ..ig7 12 ..id3 0-0 fol lowed by ...4:Jc6 or ...d6-dS would have been finefor Black. as we will see in the game, offers c) 8 e3 exd6 9 4:JxbS (after 9 ..ixbS?! him good compensation. dS 10 0-0 ..ig7 11 ..ie20-0 Black was 124 Other Options fo r White slightly better in Se.lvanov-A.Khalif ently Sorin was worried about his oppo man, St Petersburg 1997, due to his nent's compensation since he opted for nice centre, more active pieces and a safe line with little chance of gaining space advantage) 9 ... d5 10 .i.e2 d6 11 an edge: 14 e5 dxe5 15 t"Llxe5 t"Llxe5 16 b3 .tg7 12 .i.b2 0-0 13 0-0 l:.e8 and l:txe5 'ilVd6 17 .i.f4 t"Llh5 18 l:txh5 'ii'xf4 19 Black's centre is hardly weak. ':h3 lIfd8 and Black soon regained his D.Blagojevic-I.Marinkovic, Niksic 1997, pawn with no further problems. continued 14 'ilVC2 t"Llc615 .to?! (15 a3 9 •..'ilxa6 10 We2 would be a better try) 15... d4! (opening In D.Blagojevic-S.Kasparov, Cutro up the game with a nice initiative) 16 2006, White tried the direct 10 e4 d6 11 exd4 t"Lld5 17 .i.b2?! (White had to play e5 dxe5 12 t"Llxe5 .i.g7 13 'iWd3 0-0 14 17 .tC4) 17 ...t"Llf4 and Black was clearly 0-0 'ili'xd3 15 t"Llxd3 t"Lla6, when the better. The game concluded 18 l:tfel? weak d5-pawn and open a- and b-files (18 .i.dl was necessary) 18... t"Llb4 19 gave Black his usual compensation. The 'iVdl':xe2 20 l:[xe2.i.xf3 21 �e8+ 'iVxe8 game continued 16 .te3 (if 16 l:tdl, 22 'ii'xf3 ':xa20-1. trying to protect the d5-pawn, then 6 ...axbs 7 .i.xbs'ilVa s+ 8 t"Llc3.i.a6 16 ... .l:tfd8 17 t"Llf4 t"Llb4 18 a3 t"LlC2 19 .l:tbl t"Llg4with compensation) 16 ... l:tfb8 17 �adl (after 17 t"Llxc5 ':xb2 Black has active play against the weak a- and d pawns) 17 ...c4? ! 18 t"Lle5':xb2 19 t"LlxC4 l:lc2 20 t"Llb6 and White was slightly better. Instead, Black should have played 17... t"Lld7 !, preventing t"Lle5,with ideas of ...c5 -c4 and ...t"Lla c5, or ...t"Llb6- c4, when he still has good play. 10 ... .i.g7 11 e4 d6 9.i.xa6 In later games Sorin switched to 9 .te2, but Black has few problems here either. For instance, 9 ....i.g 7 10 0-0 d6 11 e4 t"Llbd7 12 l:tel 0-013 .i.xa6'iVxa6 was A.5orin-R.Felgaer, Rosario 2000, reach ing a standard-looking Benko position where Black's typical ideas include ... l:tfb8, ... t"Llb6-c4/a4, ...t"Lle8-c7-b 5, or sometimes ... c5-c4 and ... t"LlC5-d3. Appar- 125 Play the Benko Gambit 12 lbd2 22 0-0 �d4+ 23 cithl lbe5 (winning a 12 lbb5 0-0 13 a4 lbbd7 14 �d2 piece) 24 lbg5 llxC4 25 litbl lbf6 26 d6 l:.fb8 15 0-0 lbe8 16 �C3 lbC7 17 �xg7 'ii'xd6 27 �f4 'ii'd5 28 l:tbdl l:1Ca4 29 was played in Z.Jasnikowski-U.Nielsen, l:!.xd4 lIxd4 30 'ii'xe5 ':dl 0-1 A.5orin Aarhus 1996, and here 17.. .'itxg7 G.Malbran, Buenos Aires 1997. should be fine for Black. For instance, 14... :fb8 15 l:[Cl lbb616 lbxb6 lixb6 18 I1a3 lbxb5 19 axb5 'irb6 20 l:.xa8 ':xa8 21 e5 ltb8!? (21 ...dx e5 22 lbxe5 lbf6 is more straightforward, and if 23 l:.dl,th en 23... .:ta4! 24 lbc6 l:te425 "ii'd3 c4 26 Wd2 "iixb5 27 f3 c3! 28 'ii'xC3 l1C4 with equality) 22 e6 (or 22 exd6 'iWxd6 23 l:tdl lbf6 and the d5-pawn will soon drop off) 22 ... lbf8 23 exf7 litb7 24 lbd2 "iixb5 25 lbC4 'ii'd7 (on the way to f5 and possibly d5 or f7) 26 llel 'iWf5 with an unclear game. Play might continue 27 lba5 l:tb6 28 b3 'ii'xd5 29 'ili'xe7 .:ta6 Black has typical compensation for 30 lbC4'ilxf7 31 We2 Wf6 with chances the pawn with his active pieces and for both sides. Black has a slightly ex play on the a- and b-files - and, as is posed king, but also nice central pawns usual for the Benko, this continues into and good squares for the knight on d4 the endgame. or f4. 17 b3 lbd7 18lba4 l:tb7 19 'ilxa6 ':xa6 12 •.. 0-0 13 lbC4 lbbd714 �d2 20 cite2f5 Another Sorin game went 14 f4?! (as Once more we see this standard we have seen before, this move is often thrust in the Benko endgame, seeking risky for White in the Benko) 14 ... litfb8 to undermine the white centre. (Black is already fine) 15 e5 dxe5 16 21 exf5 gxf5 22 l:lhdl citf7 23 .:tc2 lbe5 fxe5 lbe8 17 �f4?! (but 17 e6 fxe6 18 24 f4 lbg4 25 h3 lbf6 26 �el h5 27 citf3 'ii'xe6+ 'ii'xe6+ 19 dxe6 lbb6 is more :ta828 litC4 lbd729 �h4?! than comfortable for Black) 17 ... .:tb4! 29 litc2 lbf6 would repeat, though 18 b3 (forced) 18 .. Jbb3 19 l:tcl?! (19 Black might also try 29 ... �d4 with axb3 'ii'xal+ 20 lbdl would offer more compensation. chances of holding the game, though 29 ...lbb6 Black is at least slightly better) Now Black breaks in on the queen 19 .. J:tb4! (back again; now the white side. centre is collapsing) 20 lbe4? (20 lbdl 30 lbxb6 J::[xa2 31 I:r.el Itxb6 32 ':'xe7+ was necessary) 20 ... .i.xe5 21 .i.d2 l:ta4 citf8 33 b4? 126 Other Options fo r White 33 .l:te3 would still give White good ing the 5 e3 .i.b7 lines of Chapter hopes of making a draw. Three. In return Black can exploit this 33 ....i.d4 specific move order by playing 6 ...cxb S 7 .i.xbs 'iVas+ 8 liJC3 .i.a6. after which he obtains the usual active play for the pawn. As in many lines of the Benko. Black's compensation continued into the endgame and in the end it was White who went astray and lost. Occasionally 6 liJC3 axbs 7 d6 is tried as well. but if Black plays accu rately. starting with 7 ... .i.b7. it seems to be White who has difficulties in equal izing. Suddenly mate is a theme and White is about to lose the b4-pawn. After examining 4 liJf3 in the previ 34 lie6?! ous two games we have now pretty After this Black is just winning. The much dealt with all the critical varia last chance was 34 l:lh7 llba6 35 'it?g3 tions in the Benko. What remains are a .tf2+ 36 'it?h2 .i.xh4 37 lIxhS .i.e7 38 few sidelines which some players like bxcS dxcS 39 l:txfS+ 'it?g740 l:tes .i.f6 41 to try from time to time. None of these .l:te3 .i.d4 42 ':'g3+ 'it?f6. when Black is are particularly challenging from a much better but there is still some theoretical perspective. but of course work to do. it's as well to be prepared. 34 ...l:tba6 3S .i.e7+ 35 l:txd4 lit6a3+! 36 l:te3 cxd4 37 l:txa3 .l:1xa3+ would have been equally Game 32 hopeless in the long run. E.Grivas-A.Khaliflnan 3S ...'it?f7 36'it?g3 .i.f2+ 37 'it?h2:6a 3 38 Leningrad 1989 .l::[c1 .i.g3+39 'it?g1 l:ia140 1:1f1 l:lxf1+41 'it?xf1 :a1+ 42 'it?e2l:te1+ 43 'it?f3 l:txe6 1 d4 liJf6 2 c4 cS 3 ds bs 4.i.gS 44 dxe6+ 'it?xe7 4S bxcS h4 46 cxd6+ White is trying to play in a Trom 'it?xe647 d7 'it?xd7 48 'it?e3 'it?c6 49 'it?f3 powsky-inspired way against the 'it?dS SO 'it?e3 'it?c40-1 Benko. which can be dangerous against an unprepared opponent. Superficially Game summary it resembles the line 2 liJf3 cS 3 dS bS 4 With 6 e3 White attempts to manoeu .i.gS liJe4in Chapter Eight. but here the vre the game into his territory by avoid- weakness of the queenside dark 127 Play the Benko Gambit squares (due to White's C2-C4) enables a) 6 tiJd2?! bXc4 7 "'C2tiJxg s 8 hxgs Black to obtain immediate counterplay g6 9 "C3?! (but 9 e4 .i.g7 10 .i.xC4 d6 with ...tiJe4, .. :ii'a5+ and ...g7-g 5, and in also gives Black a slight advantage) fact has good chances to gain the ad 9 .....xc3 10 bxc3 .i.g7 11 :1cl d6 12 e4 vantage. tiJd7 13 tiJxC4 tiJb6 14 tiJf3 was played in N.Murshed-D.5ermek, Moscow Olympiad 1994, and was shortly drawn; but after 14 ...hS Black must be at least slightly better. b) 6 .i.d2 is hardly very ambitious, and Black can get an easy game quite comfortably, but it is objectively the best move here. For example, 6 ...tiJxd2 7 "xd2 "xd2+ 8 tiJxd2bxc4 9 tiJxC4d6 10 tiJf3 tiJd7 11 e4 g6 12 0-0-0 lLlf6 13 .i.d3lLlg4 with an equal game. 4 ...tiJe4 1 5 ...'ii' a5+ 6 lLld2g51 The most testing move. White must now be careful not to fall behind in development. 5.i.f4?1 Surprisingly, after this obvious move Black may already have the edge. 5 .i.h4? "a5+ 6 tiJd2 g5 is similar but even worse for White. Instead, there is 5 h4!? (a thematic idea from the Trom powsky) 5 .. :ii'a5+ and then: 7.i.e3?1 After this Black is already slightly better, but other tries are fail to give White anything either: a) 7 "c2 gxf4 8 "xe4 bxc4 9 lLlgf3 d6 10 'ii'xf4 .i.g7 and Black's position was preferable in E.Kahn-J.Brustkern, Budapest 2002. b) 7 b4!? 'ii'xb4 8 lIb1 'ii'C3 9 llb3 can be met by 9 ...tiJxf2 ! (9 ...'ii' f6 10 lLlxe4 128 Other Options fo r White 'iVxf4, as in A.Botsari-So.Polgar, Thessa 8 ...d6 (8 ...g4 9 "c2! fs 10 tLlh4 is also loniki Olympiad 1988, is less clear after unclear) 9 b4 cxb4 (or 9 ...'ifxb4 10 J:tbl) 11 'iVbl!) 10 ItxC3 (or 10 �xf2 'ii'f6) 10 tLlxe4dxe s 11 e3 tLld7 12 cxbs fs 13 10 ...tLlxdl 11 �xb8 tLlxC3 12 �es :g8 tLlg3 tLlb6 14 tLld2 (14 tLlxes?! b3+ 15 (or even 12 ...tLlxa2 !? 13 �xh8 b4) 13 "d2 "xd2+ 16 'iii>xd2 �g7 is good for �xC3 b4 14 �b2 �g7 15 �xg7 ::'xg7, Black) 14 ... g4 15 tLlb3 'ifa3 16 �e2 and when the queenside pawns give Black there is still everything to play for. the better chances. 7 . .. bxc4 8 'ifC2 tLlf6 c) 7 �es :g8 and now: Also possible was 8 .. .fs!? 9 f3 tLld6 10 �xgs �g7 11 �f4 tLla6 12 �xd6 exd6 and Black is doing fine with his lead in development and the bishop pair. Cl) 8 �C7?! 'ifxd2+ (the simplest and best solution) 9 "xd2 tLlxd2 10 'iti>xd2 bxc4 11 e3 �a6 12 tLle2 d6 and Black is doing fine. C2) 8 b4?! (fighting for the initiative, 9�xg5?! but objectively not the strongest) The fourth indifferent move with 8 ..:ii' xb4 9 tLlgf3 (9 l:.bl? "xd2+ is this piece finally gives Black a clear ad clearly better for Black) 9 .....as ! (plan vantage. Preferable was 9 'iixC4 �g7 10 ning ...gs -g4, ...d7-d6 and possibly %1Cl, although 10 ... tLla6 11 �xcs tLlxcs ... bsxC4; the immediate 9 ... g4?! allows 12 'ii'xcs 'ii'xd2+ 13 'iii>xd2 tLle4+, fol 10 llbl! tLlc3 11 l:txb4 tLlxdl 12 l:lxbs lowed by ...tLlxcs and ...�x b2, would gxf3 13 exf3 tLlxf2 14 �xf2 d6 15 �al still be somewhat favourable for Black �h 6 16 tLle4an d White had the edge in due to the bishop pair. N.Murshed-J.Hodgson, London Lloyds 9 ...tLlxd 5 10 "XC4 tLlb6 11 "c3 Bank 1992) 10 'iic2 fs 11 llbl a6 and Keeping the queens on with 11 "C2 White has very little for the pawn. �g7 12 e3 tLlc6 is no improvement for c3) 8 tLlgf3 ! (this development move, White. preparing b2-b4, seems the best here) 11•• :iIi'xc3 12 bXc3 �g7 129 Play the Benko Gambit it could hardly be long before Black won material somewhere. In despera tion White gives up two knights for a TOok, but it doesn't change the inevita ble result. 26 ':'c2 .u.xb1 27 tiJxb1 .i.xb1 28 .l:tc1 .i.d3 29 a4 tiJbC4 30 i.b3 tiJas 31 i.a2 c4 32 i.e3tiJec6 33 i.b1i.xb1 34 11xb1 tiJb3 3S l:tb2i.xC3 36 ':c2tiJcd4 0-1 Game summary Black is clearly better. The bishop on The bishop's excursion to g5 doesn't g7 and knight on b6 are working well achieve a great deal in the Benko. In together, while the b-file is always use fact, as we saw in the game, Black can ful for the TOok on a8. just use it as a target to generate coun 13 1:[c1tiJa4 14 tiJb1?! terplay by ... tiJe4 and ...g7 -g5. By the Very passive; it would have been time Grivas had made four moves with better just to give up the pawn, with 14 this bishop in the opening, Khalifman e4 for instance. had obtained a clear advantage as 14... tiJc6 15 e4 l:tb816 tiJf3d6 17 tiJfd2 Black and went on to win comfortably. i.e6 18 :c2 'iitd7 19 .i.e2litb6 Although White still has a few ways The immediate 19 ... lir.b2 20 ':xb2(or of achieving equality after 4 i.g5, I 20 l:tc1 .i.xa2) 20 ... tiJxb2 21 a3 l:tb8 don't think we will see much of this should also be decisive. line in the future. If White really wants 20 0-0 lithb8 21 l:lfc1 l:ib2 22 a3 .i.a223 to play in this style, it is better to play l:txb2 llxb224 i.d1tiJb6 25 f3 tiJes i.g5 on move two. Game 33 V.Kramnik-P.Leko Dortmund 1998 1 d4 tiJf6 2 c4 cS 3 ds bs 4 tiJd2 Not really a big threat to the Benko Gambit. White's idea is to neutralize his any hopes his opponent might have of an initiative and build- up slowly be- White is completely tied down and hind the centre; 1 3 0 Other Options fo r White for example, after 4 ...bxC4 5 e4 d6 6 6 bxc3 i..xC4 g6 7 b3 i..g7 8 i..b2 0-0 9 lLlgf3 {as The alternative 6 lLlC4!? tries to be Khalifm an had played against Leko the tricky, setting the trap 6 ...lLlxe 4?? 7 'iWe2 previous year}, while if Black delays lLlf6 8 lLld6 mate. Instead, Black can taking on c4 White might accept the either play 6 ... d6 with equality {and gambit pawn after all in more favour then 7 i..d3 g6 8 bxc3 transposes to the able circumstances. However, as we game}; or accept the pawn with will soon see, Black has a simple anti 6 ...cx b2, when 7 i..xb2 d6 8 i..d3 g6 9 dote which gives him an easy and com i..xf6 exf6 10 l:.bl i..e7 11 "ii'a4+ c;i;>f8 is fortable game. unclear. 4 ... bxC4 5 e4 6 ...g6 7 lLlc4 5 lLlxC4?! i..b7 is already worse for Other moves are examined in the White, who must either play the awk next game. ward 6 lLle3e6 7 dxe6 fxe6, or else try 6 7 •••d6 d6 - but then 6 ...exd6 {6 ... lLle4 and Leko is not tempted by 7 ...lLlxe4? ! 8 6 ... i..d5 are also good} 7 lLlxd6+i.. xd6 8 'ii'e2 i..g7 9 'iWxe4i.. xc3+ 10 i..d2 i..xal "ii'xd6 lLle4! 9 'ii'e5+ '&W8 10 lLlf3 {not 10 11 i..a5 "iVxa5+ 12 lLlxa5 i..C3+ 13 c;i;>dl f3 ? "ii'a5+ and wins} 10 ...d5 followed by i..xa5, when Black has rook, bishop and ...lLlc6 gives Black a very strong initia two pawns for the queen, but they are tive. mostly sitting at home. After 14 h4! 5 •••c31 White has a useful initiative according Interfering with White's desired set to Kramnik; for instance, 14... i.. C3 up. 5 ... C3! is Khalifman's move, which {14... d6 ?? drops the bishop to 15 "iVa4+} has become a standard equalizing idea 15 i..b5 i..b7 16 .lir.h3 and White is bet against 4 lLld2. Indeed, from a practical ter. perspective White must be careful not 8 i..d3 i..g7 9 lLlf3 0-0 10 0-0 lLlbd7 11 to end up in a worse position. i..d2 lLlb6 1 3 1 Play the Benko Gambit 1S... tLlxc 3 Stronger was 15 ...iLa6 ! 16 tLlxd6 'ii'xd6 17 exd5 tLlxd5 and Black is clearly better. 16 �xC3 'ii'xC4 17 �b2 �b7 18 �b3 'ii'a6 19 eS tLle4?1 Here 19 ...tLlg4 20 l::tbl tLlxe5 21 tLlxe5 dxe5 22 �xe5 �xe5 23 ':xe5 e6 24 �g4 .l:tab8woul d also be clearly bet ter for Black, while Leko's 20 exd6!? �xb2 21 dxe7 lUe8 22 �xf7+! c;.t;>xf7 23 12 tLlas? 'ii'b3+ can be met by 23... �d5 ! 24 tLlg5+ Unless tLlc6 is a serious threat the (or 24 'ii'xd5+ c;.t;>g7) 24... c;.t;>g7 25 'ii'xb2+ knight is badly placed here. White 'ii'f6, when White does not seem to should have preferred 12 tLlxb6axb6 13 have enough for the piece. a4 (Leko) or 12 tLle3c4 (or 12 ... e6 13 C4) 20 l::tb1 l::tac8 21 exd6 'ii'xd6 22 �xf7+?1 13 �c2 with a level game. But after 22 �xg7 c;.t;>xg7 23 'ii'xd6 12 ...c41 exd6 Black has a healthy extra pawn. Now Black gets the better chances. 22 ...l:txf7 23 �xg7 'ii'xd1· 24 :exd1 13 �C2 tLlxf2 2S c;.t;>xf2 �xf3 26 gxf3 ':c2+ 27 13 tLlxC4tLlxC4 14 �xc4 tLlxe4is very c;.t;>g3c;.t;>xg7 comfortable for Black, with a strong knight on e4 and bishop on g7; e.g. 15 tLld4 "ilC7 16 tLlc6e6 17 l:tel tLlxd2 18 'ii'xd2 �b7 19 �b5 exd5 with a slight advantage. 13... 'ii' c7 14 l:te1tLlbxd s is tLlxC4 This endgame should be lost for White. I won't really comment on the rest as this is not an endgame book. As it turned out Kramnik managed to draw afterLeko went wrong: 28 Itb3 .l:.xa2 29 ':d7 l:tas 30 h4 h6 31 1 3 2 Other Options fo r White lIC3 e6 32 l::td6 'iW6 33 lIe3 lle7 34 l:te4 �g7 9 l:tbl 0-0 10 �d3 lDbd7 11 0-0 g5 35 J:[c6 l:te5 36 .i:ta4 gxh4+ 37 ':'xh4 lDb6, which leads to similar play. h5 38 .l:!.a6 l:tg7+ 39 �f2 llb7 40 l:tha4 .l:r.b2+ 41 �g3 .l:tg5+ 42 �h3 l:tb3 43 �h4 1lf5 44 f4 l:.b7? According to Leko 44... .l:te 3! would have won; but not 44 ...l:tf3 ? due to 45 .l:r.xe6+!�xe6 46 l:ta6+and Black cannot escape the checks as the white king is stalemated. 45 lIe4l:te7 46 l:tea4l:tb5 47 ':xa7 l::txa7 48 1lxa7 llb449 �g3 :b3+ 50 �g2 �f5 51 1Ia5+ �xf4 52 .l:!.xh5 Yz-Yz The position is balanced and strate Game summary gically complex. I studied it in my After 4 lDd2 Black has very few prob preparation for the game with Petkov. lems in equalizing. Black gets a com a) 12 lDb3?! (now Black gets an ini fortable position with good chances of tiative on the light squares) 12... e6 13 an active game, and later on Leko even h3? (13 'ii'c2 exd5 14 exd5 �g4 15 reached a winning ending. This line lDbd2 l:te8is equal, but it doesn't seem can only be recommended as a surprise like a position worth aiming for with weapon for White, not as a serious the white pieces) 13 ... exd5 14 exd5 theoretical test of the Benko. �a6 15 lDa5 Game 34 V.Petkov-N.V. Pedersen Skanderborg 2010 1 d4 lDf6 2 c4 cS 3 d5 b5 4lDd2 Petkov used to decline the Benko with 5 b6. However, I noticed he had been venturing 4 lDd2 more recently and, as it is my conclusion that this line is rather unproblematic for Black, I was 15 ...lDbxd5 {15 ...'ii' C7!, threatening happy to play against it. ...lDb xd5, would be slightly better; e.g. 4 ...bxC4 5 e4 C3! 6 bxc3 g6 7 l:.b1 16 'iWd2 lUe8 17 J:[el llxel+ 18 lDxel Another option is 7 C4 d6 8 lDgf3 .l:.e8fol lowed by ...lDe4 an d Black is do- 1 33 Play the Benko Gambit ing fine) 16 liJc6 liJc3 17 'ii'el 'ii'd7 18 16 g3 liJd7 17 'iiC2 liJes 18 liJxes dxes liJe7+ c;t>h8 19 liJxg6+ fxg6 20 .xC3 19 liJfl 'iid4 20 liJe3liJf6 21 liJds litab8 ii.b7 21 liJel?! (quite passive; 21 liJgs, 22 l:.b3 �xds 23 exds e4 24 �xe4 (24 intending 21...h6?! 22 liJe6! �xg2 23 �e2 l:tb6 25 lIdl'ii' es 26 1hb6 axb6 27 �xh6! would still be unclear) 21... l:tae8 c;t>g2 l:td8 is also equal) 24... litxb 3 25 22 'iib3 �c6 23 �e3 liJh s (or 23 ... liJe4! axb3 l:te8 26 d6 l:[xe4 27 litdl'ii es 28 with a clear advantage) 24 �e2 liJf4 25 d7 l:tel+ 29 c;t>g2 ':'xdl 30.xdl liJxd7 �g4 .f7 26 �xf4 .xf4 27 �f3 �d7 28 31 'iixd7 'ii'e4+ with a level endgame. ii.dsii.fs 29 l:ldl?! (29 liJd3woul d be a 7 ...� g7 better try) 29... lite2 30 a4 �es 31 g3 Not yet 7 ... d6 in view of 8 �bs+. 'ii'h6 32 liJd3 ii.d4 33 liJf4? �C2 34 .f3 8 liJgf30-0 9 �d3 d6 10 h3 ii.xdl 0-1 D.5vetushkin-V.Baklan, Ru manian Team Championship 2009. A nice game by Baklan and a good illus tration of the dangers awaiting White ifhe makes a few inaccuracies. b) 12 l:tel seems to be the most pre cise move, reaching an equivalent posi tion to the main game (minus h2-h3 and ...'ii' C7). Petkov had played this before. White is delaying c3-c4 for a while, so I hold back on ...e7-es too, since this would give his knight a nice post on c4. 10 ... liJbd711 0-0 liJb612 litel'ii' C7 12 ... liJh s is a good alternative; e.g. 13 c4 'ii'C7 14 liJfl l:lb815 �d2 es with a roughly equal position. White might try 15 g4!?, which I suppose is one of V.Akopian-P.Tregubov, Mainz (ra the ideas of h2-h3, but it's very com pid) 2009, continued 12 ... es 13 dxe6 (if mittal and gives Black counter-chances. White doesn't take on e6 the position is For example, ls ... liJf6 16 ii.b2 es 17 balanced as well; typically Black will dxe6 (17 liJe3?! hs is better for Black) continue with ... liJh s) 13 ... �xe6 14 17 .. .fxe6 18 es liJe8 19 liJld2'ii' f7 is un ii.b2, and now 14 ... liJh s offers balanced clear; both sides have some serious chances; for example, 15 �xg7 c;t>xg7 pawn weaknesses. 134 Other Options fo r White 13 c4 eS 14 lbf1 lbhS is g4 lbf4 16 in trouble . ..Itxf4 exf4 17 'ii'd2 23 lbg4 lbd724 .i.C2 17 ....i.a6 24 gxf6 1:[ae8 25 .i.c2 hS 26 lbh2 More ambitious was 17 ... lbd7! (I lIxf6 27 lbf3 would be slightly better also considered 17 ... lba4, but thought for White, because of the weak pawns it was insufficient) 18 'iVxf4 lbes, when on eS, cS and g6. Black has decent compensation in view 24..• fx gs 25 .i.a4 lbb6 26 .i.c6lbxC4 27 of the strong dark-squared bishop and .i.xaSlixa S 2S lled1?! loose pawn structure in front of the Too straightforward. Instead, 28 d6! white king. For example, 19 lbxeS .i.xes 'iVxd6 29 1:[edl "fie6 30 1:[dSwoul d give 20 �e3 i..d4 21 1Ve2 "fie7 22 lbe3hS 23 White a dangerous initiative. gxhS .i.xh3 24 'ii'f3 'iih4 with interest 2S ...�d6 ing play for the pawn. 1S :bc1 The greedy 18 'ii'xf4 .i.xc4 19 lIxb6 .i.xd3 20 :c6 'iVas 21 l:txd6 C4 would give Black good play with the bishop pair and active pieces. 1S ....i.eS 19 lbxes dxes 20 lbh2 lbcS?! This casual move grants White some initiative based on the unex pected weakness of the es-pawn. I should have played the solid 20 ...f6 , hindering g4-gS, with equality. Now the position is unclear. 21 �c3 f6 22 gsl lbb6 29 l:.b1 135 Play the Benko Gambit Here Petrov had around eight min 44 .li!e3g3+?? utes to reach the time control at move But this should just have lost. The 40, while I had about three. correct course was to retreat with 36 l:ta3lDC5 44 ...�5 , keeping a slight advantage, The black pieces are now very well but then we wouldn't have seen the arranged. Petkov should probably tried forthcoming entertaining king march. to prevent this formation. 45 fxg3+ �xe3 46 lDg2+ �d3 47 .l:r.d1+ 37 "a5 .l:r.C73s lDf3g4 39 lDd2�g 5?1 �C2 4S lDe3+ �b2 The beginning of a risky winning at 49 'ii'a4?? tempt, in which the king makes a solo Here 49 'iVb5+ �c3 50 lDxC4 wins trek across White's heartland to the outright; the move played allows Black distant a3-square. Instead, 39 ... gxh3 back into the game. would be fine for Black, but I was in an 49 •.•"C 5 50 'ii'C2+ �a3 51 'ii'xe4 �f5 52 optimistic mood after the balance of 'ii'f3 c31 the position had changed in my favour, and was also in some time pressure. Of course my intuition told me there should be something for White, but it's not always so easy to find it over the board. Okay, maybe I wouldn't play this way against Ry bka or Fritz. 40 �h2 f3 41 .1:.g1 42••• lDxe41 43 lDe1?1�cs Blocking offth e pressure from the Now I was actually slightly better. queen on f3. 1 3 6 Other Options fo r White 53 d6 .l:[f7 54 l:td5 'iVb4?? A final exchange of errors. 54 ... .i.e6 Game 35 should draw, whereas the text loses to WI.Schmidt-G.Kasparov 55 'iVe2. Instead the mistakes con Dubai Olympiad 1986 tinue ... 55 ll:lc2+?.i.xc2 56 'iVxf7 'iVb2 1 d4 ll:lf6 2 C4 c5 3 d5 b5 Kasparov has played the Benko a few times, and it's an opening which suits his dynamic style. 4a4 57 'iVxa7+?? As Tartakower said: the winner of the game is the player who makes the next-to-last mistake. Instead, 57 'it>gl would keep it going. White wants to force his opponent 57... .i.a4+ 58 'it>gl'iVb1+ 59 'it>h2Wxa 2+ to make a decision about the b5-pawn. 60 'it>gl'iVxd 5 0-1 However, Black can reach an unclear game quite easily and in more than Game summary one fashion, so 4 a4 is not really the Th e game was equal from the opening way to test the Benko Gambit from a and I had a few ways to get a balanced theoretical viewpoint. Chris Ward is the position. Instead, I made a few inaccu only GM to play this line regularly as racies and Petrov gained the advan White. tage, that is until he let it slip. The final 4 ... b4 phase of the game was fun if not objec Both this move and 4 ... bxC4 (exam tively correct. Of course the computer ined in the next game) give Black a doesn't like it, but Petrov was unable to good game. On b4 the pawn controls find a refutation over the board. So I the c3-square, supporting the fightfor was quite content to play ...'it>g 5-f4-e3- the long diagonal after ...g7 -g6 and d3-c2-b2-a3 with that many pieces on ....i.g7. Th en Black can break in the cen the board and win. tre with ...e7-e6 as in the game, or in 1 37 Play the Benko Gambit some cases consider ...e7-es instead. 7 ...i.. g7 S g3?! S ttJd2 g6 6 e4 8 ..td3 is a more natural move, sup 6 b3 ..tg7 7 ..tb2, opposing the g7- porting the key squares e4 and c4. Then bishop, is a logical alternative but too play might continue 8 ... 0-0 9 0-0 e6 quiet to offerWhite any advantage. For (9 ... es is also possible) 10 h3 ttJbd7 11 example, 7 ... 0-0 8 e4 d6 9 g3?! (this ttJb3 itC7 12 ..te3l:.e8 13 'ii'c2 ..ta614 move is often dubious in these posi l::tfel l:[e7with balanced chances. tions, as we will see again in the main S •..e6 9 ..th3 game; 9 ttJgf3 would be roughly equal) Aiming for the exchange of light 9 ...e6! 10 ..tg2 exds 11 cxds lIe8 and squared bishops, which has some logic Black was already better in A.5urjadnji to it so that Black can't play .....ta6 hin R.Ponomariov, Swidnica (rapid) 1998. dering ttJC4. 6 ••.d6 The natural 9 ..tg2 exds 10 0-0 0-0 (not 10... dxe 4?! 11 ttJxe4! ttJxe4 12 'iWds winning material) 11 cxds ..ta6 12 l:[el ttJbd7 gives Black a comfortable posi tion and good chances of dynamic play in the middlegame, with most of his pieces active and well placed already. 9 ••• exds 10 i..xcs 'iVxcs 11 cxdS 7 ttJgf3 7 f4 ..tg78 ttJgf3 0-0 9 ..td3 is more ambitious, but Black has sufficient counterplay here too: 9 ...e6 10 0-0 exds 11 cxds l:te8 12 l:tel ..ta6 13 ji'c2 (or 13 ..txa6 ttJxa6 14 es dxes 15 fxes ttJxds 16 ttJC4 ttJac7 17 ..tgs f6 18 ..th4 Y2 -Vz C.Ward-lM.Degraeve, Gent 2005) Now we have a sort of improved 13 .....txd3 14ji'xd3 ttJbd7 15 as c4! 16 Benoni for Black. 'ii'xC4 l:tc8 17 'ii'xb4 ttJcs 18 l:ta3 was 11••• 0-0?! seen in G.Flear-C.Adrian, French Team 11 ... ji'a6! 12 'iVe2 0-0 was stronger. Championship 1997, when 18 ... ttJxds! Garry probably didn't like the exchange 19 exds l:txel+ 20 ttJxel ttJa6 followed of queens, but now White has prob by ... l:txclwoul d have equalized. lems defending the e4-pawn. For ex- 1 3 8 Other Options fo r White ample, 13 'ii'xa6 lLlxa6 14 0-0 lUe8 15 19 llel 'ii'd5 20 'iVdl lLlSd7 21 lta2 l:1el l:.ac8 (with the dangerous threat .i.xd4 of ...c4 -c3) 16 tt:)c4 lLlxe4 17 l:txe4 l:txe4 In his notes Kasparov prefers 18 lLlxd6.l:tee8 19 .l:tbl c4 20 lLlxe8l:lxe8 21.. .l:Ue8 22 .i.e3 lLlC4 with a decisive 21 .i.e3 .l:.d8 and Black is clearly better advantage, while 21...lLlcS!, threatening due to the advanced queen side pawns ....i.xd 4 and ...lLlbxa4, looks even and passive rook on bl. stronger. 12 0-0 e4 13 'ii'e2l! 22 I1d2 lLle5 13 "e2! keeps the balance: 13 ...c3 Declining the second pawn by 14 bxc3 J:te8 (14... bxC3 15 lLlC4'ii' a6 16 22 ... .i.xf2+, since White might obtain lLld4 lLlxdS 17 lLlbS lLlb4 18 lLlC7'ii' c6 19 some counterplay on the open files. lLlxa8 c2 is unclear) 15 cxb4 lLlxdS 16 23 lLlxd4 lLlbe4 24 :e2 .lilacS .l:ta3 lLlxb4 17 l:te3 lLl8c6 18 lLlC4 and White has reached a defendable posi tion. 13 ... e3 14 bxe3 bxe31 I will leave the rest without notes. Garry managed to convert the extra pawn to win this one-sided game. 25 ':'e3 lLlb626 .i.b2':'xe3 27 .i.xe3 liteS Now White has a very difficult de 2S .i.al l:te4 29 lLlb5 .l:txa4 30 'iVxd5 fensive task. The quiet 14 ... 'ivxc3 would lLlxd5 31 .i.xe5 dxe5 32 .lilxe5 lLlf6 33 lead to a lesser advantage. For instance, ':'e7 a5 34 lLld6 l1d4 35 lLlxf7 'it;JfS 36 15 l:ta2 lLlbd7 16 'ii'bl b3 17 .l:.a3 lLlcs .l:.a7.l:r.d7 37 l:txd7 lLlxd7 3s lLld6'it;Je7 39 18 lLlxb3 'i'b419 lLlfd2 lLlcxe4 20 lLlxe4 lLle4 a4 40 'it;Jfl 'it;Je6 41 'it;Je2 'it;Jd5 42 lLlxe421 lLld2'ii' xbl 22 lLlxbl l:tfb8 and lLle3+ 'it;Jd443 'it;Jd2lLle5 44 'it;Je2lLld 3 45 Black is only slightly better. lLldllLlel+ 46 'it;Jb2lLlf3 47 h4 lLle5 4S 15 lLlb3 'ilkg4 16 lLlfd4 'ii'xe4 17 'ii'xe3 'it;Ja3 'it;Je4 49 h5 gxh5 50 lLle3+cM3 51 lLlxd5 lS 'iVd2 lLlb6 lLld5 lLlg4 52 lLle7lLlxf2 53 lLlf5 lLle4 54 Black is a pawn up and has the 'it;Jxa4lLlxg3 55 lLlh4+'it;Je4 56 'it;Jb4lLlf5 more active pieces as well. 0-1 1 39 Play the Benko Gambit Game summary and 'iie2 or lDgs-e6 (after .. .fxe6). That With 4 ...b4, followed by ...g7 -g6, ....i.g7, Black cannot respond casually to this and later the ...e7-e6 break, it is quite plan can be seen by 7 ...lDbd7 ?! 8 lDf3 easy for Black to equalize against 4 a4. g6?! 9 es! dxes 10 fxes lDg4 11 e6! The game was similar to the Benoni fxe6? (but if 11... lDde s, then 12 exf7+ Defence, in a somewhat improved ver lDxf7 13 lDgs! lDxgs 14 'iixg4 lDf7 15 sion for Black, and illustrates well some .i.xc4 and White is clearly better) 12 of the dangers awaiting White if he lDgs lDgf6 13 lDxe6'i!i'b6 14 lDbs .i.xbs plays inaccurately or too passively 15 axbs :c8 16 .i.xc4 and White had a against the Benko. All in all, a convinc huge position in I.Kazantzidis ing win by Kasparov. N.Galopoulos, Greek Team Champion ship 2006. Instead, Black should play 7 ...g6, Game 36 leaving the d7-square free, and then: O.Dancevski-J.M.Degraeve European Championship, Ohrid 2001 1 d4 lDf6 2 c4 cS 3 ds bs 4 a4 bXc4 a) 8 es (continuing with the aggres sive approach) 8 ... lDfd7 9 lDf3 .i.g7 10 'iie2 was played in C.Ward A.5ummerscale, British League 2000, and now 10 ...'ii C7, attacking the pawn A different way of playing to the on es, would lead to an unclear game. 4 ...b4 of the previous game. For example, 11 e6 (ll lDbs?! .i.xbs 12 5 lDC3d6 6 e4 .ta6 axbs dxes 13 'ii'e4 'iid6 is slightly bet Making White work a bit to regain ter for Black) 11 .. .fxe6 12 1r'xe6 .i.xc3+ the pawn. 13 bxc3 .i.c814 'iie3 (14 .txc4!? lDes 15 7lDf3 .tbs+ 'iit>d8 16 lDgs is tempting, but The alternative is 7 f4, intending to after 16... 'ii as! 17 fxes .i.xe6 18 lDxe6+ attack in the centre with lDf3, e4-es-e6, 'iit>c8 19 .i.d2 a6 20 .i.d3 c4 21 .i.xC4 140 Other Options fo r White lLld7 22 exd6 exd6 23 0-0 lLle5 White Black has equalized without too doesn't have enough for the queen) much effort,just playing natural active 14 ...lLlb6 15 lLlgS �7 is roughly equal. moves. b) 8 lLlf3 �g7 9 �e2 is a quieter way 14 �d2 lLlg415 .l1bl'iVc8 16 0-0 lLlges of playing the position, when 9 ... 0-0 10 17 b3 lLlxf3+ 18 �xf3 �d4 19 �e2 lLles 0-0 lLlbd7 11 lLld2 lLle8 12 �xC4 �xC4 20 �h6 lite8 21 lLlxes �xes 22 .tC4 13 lLlxC4f5 was unclear in A.Yusupov .tbS 23 b4?1 V.Chekhov, Vilnius 1978. From a practi 23 'iVd3 'iVa6 24 �d2 l:rb7 25 Ilfcl cal point of view this continuation is l:teb8 26 b4 keeps the balance. attractive as it attacks the white cen 23 ..:ifa 61 tral pawns. 7 ...lLlbd7 8 'iWc2 8 �e2 g6 9 0-0 �g7 10 lLld2?! (10 'iWc2 was stronger) 10 ...lLle 5! 11 f4 lLld3 12 lLlxc4 �xC413 �xd3 �xd3 14 'iWxd3 0-0 gave Black a nice game in A.Moran Nuque-J.R.Koch, World Student Team Championship, Paranana 1993. The rook will be a factor on the b-file. 8 ...g6 9 lLlbSlLlb6 10 lLla3 Going round to regain the pawn. Another way was 10 �e2 �g7 11 lLld2 Black is slightly better: the pawns lLlfd7 12 lLlxC4 lLlxC413 �xC4 0-0 14 0-0 on b4 and as are weak, and in some lLlb6 15 �e2 'ii'd7 16 .:tbl c4 with an lines the rook on bl is hanging as well. unclear game. White faces a difficult defensive task. 10... �g7 11 as lLlbd7 12 lLlXC4 0-0 13 24 :fcl? .te2 l:lb8 After this Black wins a pawn for nothing. White had to play 24 �xb5 :xb5 25 bxc5 l::txcS 26 'iVa4, when 26 ... :ec8! (26 ..J:td 8?! 27 .l1fcl 'iWxa5 28 'ii'xa5 l:.xas 29 ':C7 �f6 30 g4 would give White compensation for the pawn) 27 'iVd7 'iVxas 28 'iVxe7 :5C7 29 'iYh4 'iVa4 gives Black the better chances, due to his active pieces and potentially dangerous passed a-pawn. 24 ...�xc4 25 'iVxC4 'iVxC4 26 l:[xC4 .l:tbS 27 .td2 l:[eb828 a6 fS 141 Play the Benko Gambit Game summary 4 a4 is hardly a serious try for an ad vantage. Both 4 ...b4 and 4 ...bxC4 lead to unclear positions and it seems a matter of taste which one Black should play. In this game Degraeve chose the latter and then defended the pawn by 6 ...i.a6, forci ng White to waste time in winning it back. Later on, standard counterplay on the b-file enabled him to win another pawn, after which the Undermining the centre, a common passed a-pawn marched through to theme in Benko endgames. Whether victory. White plays 29 f3 fxe4 30 fxe4 or takes In the opening White might prefer 7 on fS, the advanced centre becomes f4 as a more aggressive approach, but isolated and thus more vulnerable to as long as Black is careful not to play attack. ...liJbd7 too soon, he shouldn't have 29 exfs gxfs 30 .l:.b3 l:tSb6 31 g4 ':'xa6 anything to worry about here either. 32 gxfS cxb4 33 l:tcxb4 l:lxb4 34 l:lxb4 'ittg7 35 'ittg2 !!a3 36 ':g4+ 'iW6 37 .l:.h4 �xfS 38 l:txh7'iW6 39 i.e3as Game 37 JJirka-B.Vuck ovic Sozina 2004 1 d4 liJf6 2 c4 cS 3 dS bS 4 f3 The a-pawn decides the game. 40 �f3 a4 41 l:thS�g6 42 lIh6+ �g7 43 .l:r.h3.l:r.x e3+! 44 fxe3 a3 45 :g3+ �f7 46 .l:r.g2i.b2 47 e4 a2 48 llxb2a1 "iV 49 l:.f2 'iVd4 50 h4 �g6 51 ':'g2+ �hS 52 �f4 We are now really down to the odds "ilfes+ 0-1 and sods. 4 f3 was once played by An- 1 4 2 Other Options fo r White and in a rapid game, which is perhaps was just two pawns up. the best thing that can be said about it. 7 ....i g7 8 0-0 0-0 9 ltJec3 4 ... bxC4 Intending to deploy both knights on Here 4 ... g6 5 e4 d6 6 cxbs a6 would the queen side in order to combat transpose to the 5 f3 variation, except Black's counterplay there. that I am recommending s ... e6! in 9 ••.ltJbd 7 Chapter Four (rather than s ... g6 6 e4 d6, which I think is good for White). 5 e4 d6 6 .ixC4g6 10 .ie2 Prophylaxis against ...ltJb 6 or ...ltJes, which might now be met by a2-a4 or 7ltJe2 f2-f4 respectively, but Black can now 7 ltJC3 .ig78 .ie30-0 9 ltJge2 ltJbd7 refrain from either for the time being. 10 0-0 ltJes is quite comfortable for Instead, Y.Duhayon-N.V.Pedersen, Black. In B.Lajthajm-V.Malakhatko, Krakow 2008, saw 10 ltJa3ltJb 6 11 .ie2 Yugoslav Team Championship 2001, e6 12 dxe6 .ixe6 13 .ie3 We7 (I had White tried the optimistic 9 h4!? .ia6 this position in my preparations and 10 .ixa6 ltJxa6 11 Wd2 Was 12 ltJge2 saw that a subsequent ...d6-ds would l::tfb8 (aiming for counterplay on the b be finefor Black) 14 1:[cl 1:[fd8 15 Wc2 file and the dark squares) 13 hs?! (a ds 16 ltJxds?! (but 16 .igs h6 is very thematic pawn sacrifice to open the comfortable for Black) 16 ...ltJf xds 17 file for the rook, but here it fails tacti exds ltJxds and I was clearly better, cally; instead 13 0-0 ltJC7 would still be since the black pieces are very active, unclear) 13 ...ltJxh s 14 g4? ltJf6 (the un- while the white knight on a3 is out of expected 14 ...ltJg 3! was more accurate, play. The finish came rather quickly: 18 with the same idea of 15 ltJxg3 ':xb2! .ixcs Wgs 19 .l:.fdl?? ltJC3! winning etc) 15 ltJf4? ltJxg4! (spotting the tactic material and the game. this time; i.e. 16 fxg4 l:txb2! and wins) 10.•• 1:[b 8 11 ltJa3 lle8?1 16 ltJdlWxd2+ 17 .ixd2ltJe s and Black I would prefer 11... ltJe8 with an un- 1 43 Play the Benko Gambit clear game. Black has similar ideas as 24... tDe5 was still clearly better for in the b5-b6 line, such as ... tDC7 (and Black. After the text White might play possibly ...tDb 5-d4), ...e7-e6, or ...f7-f5 . 25 g3 with good chances of holding the 12 'iVC2 tDb6 13 i.e3 e6 14 dxe6 i..xe6 position, but instead he blunders. 15 l1ad1'ii' e7 16 tDcbS? 2S b3?? 16 i.g5 would keep control of the position with a slight advantage to White. The text provides Black with a tactical solution to his problems. 16..• tDf dsl 17 exds?1 But if 17 i.f2, then 17... tDb4 18 'ii'd2 d5 with the initiative. 17 ...i.. fs 18 i..d3 'iVxe3+ 19 l:If2 i.xd3 20 I:[xd3 'ii'e7 2S•.. tDhs Now there is no defence. 25.. .l:ie l! 26 l:.dltDh 5! was even stronger. 26 tDe4fS 27 g3 'iVh3 28'iVg2 'i\Vxg 2+ 29 Game summary Black has a strategically won posi Against 4 f3 Black can equalize easily tion, due to his strong bishop and con with 4 ... bxC4, followed by ... g7-g6, trol of the e-file, versus the weak white ...i.g7, an d usually a rook on the b-file, pawns and passive rooks and knights. with further ideas of ... tDbd7-e5, or 21 l:tf1tDd 7?1 ... tDb6 and ... e7-e6, or sometimes The knight was better where it was. ...tDe8-C7. Black could have forced his way into In the game White played 10 i.e2in enemy camp by 21 ... a6 22 tDC3 i.d4+ time - after 10 tDa3 there is 10 ... tDb6 23 1 44 Other Options fo r White istic for this line that, even from a fa While 4 e4 might be said to be too vourable position, White can quickly active, White's final option, 4 b3?!, suf land in trouble with inaccurate play, as fers from being too passive. Black can occurred after 16 ltJcbS? ltJfdS!. Vukovic get a comfortable game after either went on to win, though not without 4 ... bxC4 (or 4 ... e6 5 dxe6 fxe6 6 e3, as in further mistakes from both sides. L.M.Rodrigues-S.Kasparov, Figueira da Earlier on, the attacking idea of ltJc3 Foz 2009, when 6 ... bxC4 7 bXc4 ltJc6 is and h2-h4, as in Lajthajm-Ma1akhatko, fine for Black with ideas of ... iLe7, ...0-0, is not dangerous for Black. In fact it is ... !Ib8 and ... d7-dS) 5 bXc4 d6 6 ltJc3 g6 White who is more likely to get into 7 iLb2?! (the bishop is vulnerable here difficulties from the opening, given with the b-fi1e open; 7 e4 iLg7 8 llbl Black's activity and counterp1ay on the would make more sense, though Black queenside. is certainly not worse) 7 ...iLg 7 8 f3?! ltJbd7 9 e4 llb8 (White is already in trouble) 10 'it'C2'iWas 11 'ifilf2? (allowing Game 38 a combination; White had to try 11 1:[cl D.Shchukin-A.Khalifman and 12 iLal,albeit with an unenviable St Petersburg 1998 position) 11... ltJxe4+ ! 12 fxe4 ':'xb2! 13 'ii'xb2 iLxC3 14 'ii'cl ltJf6 15 ltJf3 ltJxe4+ 1 d4 ltJf6 2 c4 cS 3 d5 b5 4 e4?! 16 'ifile3 iLxal 17 'iYxal ltJf6 18 iLd30-0 19 h3 iLa6 20 I:[bl e5 21 dxe6 fxe6 22 1:tb3 "fiC7 23 1:ta3 iLc8 24 'iWb2 d5 25 'it'e5? ltJg4+! 0-1 U.Masera-P.Benko, Reggio Emilia 1970/71. An easy and instructive game from Pal Benko. 4 ...ltJxe4 5 'iYf3 The point of the previous move, hoping for 5 ... ltJf6? 6 d6! ltJc6 7 cxb5 ltJd4 8 'it'xa8 ltJc2+ 9 'ifildl ltJxal 10 'iVxa7 with a winning position (Mirk ovic). 5 cxb5 would hardly be consis Here White tries to seize the initia tent, and indeed s ...'iVa5+ 6 ltJd2 (6 tive by offering the e-pawn, but it's not iLd2?! ltJxd2 7 ltJxd2 g6 is worse for really a very good strategy to give away White) 6 ... iLb7 gives Black a nice game, a central pawn like this. Especially since as the trade of the e4- and b5-pawns is the loss of the e4-pawn often leads to naturally favourable for him. the ds-pawn becoming a weakness, as 5 ...'iYa 5+1 we have seen before in this book. A familiar idea, combining with the 1 45 Play the Benko Gambit knight on e4 to disrupt White's play, as are also quite unclear) 10 .1i.d3 .i.g7 11 we have seen before in Game 32. 0-0 b2 12 .l:tael 0-0 13 l:lxe7d6 14 lLlgs lLld7 ls lLlxh7! "d8 (ls ...'iit> xh7? leads to mate: 16 "iVhS+ 'iit>g8 17 .i.xg6 et c) 16 .i.gs f6 17 lLlxf8 'ii'xf8 18 'iWe2 lLles 19 .l:[xes dxes 20 .i.e3 and White went on to win in D.5hchukin-K.Andreev, St Pe tersburg 2010. Rather than go into these complica tions it is simpler to play 7 ... 'iib6! 8 bxc3 (or 8 .i.xC3 b4 9 .i.d2f6 'ii' and Black is just a pawn up) 8 ...bxC4 9 .1i.xc4 (or 9 lLlh3 'iWf6) 9 ....i.a6 10 .i.xa6 'iWxa6 6lLld2?! (Black has no problems after 10 ...lLlxa6 White doesn't want to give up the 11 lLle2 "iVf6 either) 11 lLle2 d6 12 0-0 bishop pair by 6 .i.d2?! lLlxd2 7 lLlxd2 lLld7 13 c4 lLles 14 'iib3, as in I.Nemet d6 when Black is better, but after the A. Robert, Bern 1990, when 14 ... lLlxC4! text the knight is somewhat badly would have been good for Black, since placed. 15 'ii'd3 can be met by ls ... 'iWa4! 16 lLlc3 Shchukin later improved with 6 lLles. lLlC3!? lLlxC3 7 .i.d2! (after 7 bXc3 d6 8 6 ...lLld6 7 cxbS lLlxbS cxbS lLld7 9 .1i.d2 a6 10 c4 'ii'a4 Black has a nice position) Black is already slightly better: his knight has a great square on d4 and 7 ...b4 8 bxc3 b3 9 lLlh3!, when White the white king is rather exposed. has definite compensation for the White's strategy has failed at quite an pawns. For example, 9 ... g6?! (but 9 ... b2 early stage! 10 .l:[bl'ii' xa2 11 "iVfS ! and 9 ... h6 10 .1i.d2 8 d6? 1 4 6 Other Options fo r White A blank shot. On the other hand, 8 tDxe2 24 �xe2 �e6 25 fxe5 fxe5 26 .lixb5 'iixb5 9 tDe2 i.a6 10 tDe4 d6 11 �3 :'e6 27 �e4 i.xd6 28 l:td5 i.e729 tD2c3'iVd3 is hardly pleasant for White tDxe5 :d6 30 ]:txe5 i.b6 31 :b5 i.xe3 either. The endgame after 12 'iVxd3 32 �xe3 lld5 33 tDf3 l1xb5 34 tDd4+ .lixd3 is quite favourable for Black, due �d5 35 tDxb5J:b8 0-1 to the bishop pair and good, long-term activity. Game summary 8 ... tDe69 i.xb5'iVx b5 10 tDe2 e6 Khalifman won convincingly against 4 10 ...i.b 7 and 10 ... i.a6 look even e4?!. Indeed, after 4 ...tDxe4 5 'iVf3 stronger. 'iVa5+! White may already be worse. 11 0-0 i.a612 :tel'iVd 3 13 'ii'xd3 i.xd3 The best try is probably 5hchukin's 14 tDf4 tDb4 later idea of 6 tDC3!? tDxC3 7 .lid2,in 14 ...i.f5 would allow White to de tending 7 ...b4 8 bxc3 b3 9 tDh3! with fend the pawn with 15 tDc4, when dangerous play for the pawn, but Black 15 ...tDb4 16 i.d2!? tDc2 17 tDd5 creates can avoid this with the simple 7 ...'iVb 6!, some mischief. when it is harder for White to prove 15 a3?! sufficient compensation. Conversely, After 15 tDxd3 tDxd3 16 :dl tDxcl the passive 4 b3?!, as seen in the noted 17 J:1axcl i.xd6 18 tDb3 i.e7 19 tDxC5 game Masera-Benko, gives Black no d5 White is a pawn down for nothing. problems at all. 15... tDe2 16 tDxd3 tDxal Summary In this chapter we examined various sidelines for White on the fourth and fifth moves. The Zaitsev Variation, S tDc3 axb5 6 e4, aims for a quick attack with moves like .lif4, i.C4an d e4-e5, but after6 ... b4 7 tDb5 d6 8 .lif4 g5! Black is already slightly better. 50 White should try an other option such as 8 i.d3.White has no advantage here either, but at least The knight cannot be prevented he is not worse. from emerging at c2 or b3, so Black is Another idea is 4 tDf3, which is of the exchange and a pawn up and wins ten just a transpositional tool to other easily once he gets his pieces out. lines, in particular 4 ... g6 5 cxb5 a6 6 b6 17 tDe4 tDb3 18 i.e3 :e8 19 tDf4 f6 20 as we saw in Chapter Five. Instead, in .l:tdl �f7 21 tDe2 e5 22 f4 tDd4 23 �f2 Game 30 White played 6 'iVc2, a choice 147 Play the Benko Gambit of Korchnoi and Ivan Sokolov, based on dlegame. With a creative effort involv the greedy idea of 'iixC5; but this was ing an unusual king march I managed neutralized by Topalov's strong novelty to win the game all the same. 8 ...'iIi'a4 and it is not so popular any 4 a4 gives the Benko player a com more. Game 31 considers 6 e3, by fortable choice between taking the which move order White avoids the 5 pawn or pushing past. 4 ...b4 was cho e3 iLb7 of Chapter Three. Here 6 ...cxb 5 sen by Garry Kasparov in Game 35 and 7 iLxb5 'ili'a5+ 8 lLlC3 .ia6 looks to he quickly seized the initiative in a sort guarantee Black sufficient play for the of improved Benoni position. Degraeve pawn in a typical Benko position. preferred 4 ...bxc4 5 lLlc3 d6 6 e4 iLa6in 4 iLg5 is inspired by the Trom Game 36 and also equalized quite eas powsky (2 iLg5), but it doesn't really ily. work against the Benko. After 4 ... lLle4!, Finally, the idea of 4 f3 intending with ideas of ...'ii a5+, ...g7 -g5 and e2-e4 (as in Game 37) is a lot less chal ...iLg7, it is White who must take care lenging than the 4 cxb5 a6 5 f3 of not to come out worse from the open Chapter Four; while the immediate 4 ing. In Game 32 Grivas played weakly e4 is somewhat suspect, and 5hchukin with 5 iLf4?!, 7 iLe3?! and 9 .ixg5?!, soon got the worse position as White in after which Khalifman was clearly bet Game 38, even if he later found a bet ter and won quite easily. ter way of playing (6 lLlc3!?) to reach a A sounder, if unthreatening, idea is more equal game. 4 lLld2, often leading to an unclear po To sum up, none of these fourth sitional game. This was played by and fifth move deviations should trou Kramnik against Leko in Game 33, ble Black at all as long as he knows where Kramnik was lucky to escape what he's doing - in which case some with a draw after a mistake in the of them are even a bit risky for White. opening. Vladimir Petkov tried the And with that we complete our cover same line against me in Game 34, and age of the Benko proper and now turn even gained the advantage when I our attention to the anti-Benko varia played inaccurately in a balanced mid- tions: 2 C4 cS 3 lLlf3 and 2 lLlf3 cS. 148 Chapter Seven The Kasparov Gambit 1 d4 li:Jf6 2 c4 cS 3 li:Jf3 pawn on dS. Perhaps unsure whether, after 2 ...cS To name this gambit after Kasparov 3 dS, his opponent intends the Benko is a little unfair, given that Vaganian or some variety of Benoni, White de and others had already played it sev cides to play it safe. Or so he thinks - eral times in the late 1970'S. However, since Black has another rather danger it was Kasparov's high-profile win in ous response: The Kasparov Gambit! the first game below which led to it 3 ... cxd4 4 li:Jxd4 eS 5 li:Jbs dS 6 cxdS becoming popular for Black, and his i.cs name has stuck. Game 39 A.Mikhalchishin-G.Kasparov USSR Championship, Frunze 1981 1 d4 li:Jf6 2 c4 cS 3 li:Jf3 This position can also be reached from fl ank openings, such as lli:Jf3 li:Jf6 2 c4 cS 3 d4 or 1 c4 cS 2 li:Jf3 li:Jf6 3 d4. Black can gain good activity with 3 ... cxd4 4 li:Jxd4eS s li:Jbs this line, and I think it is sound as well. Other moves for White, including 5 Black already has a lead in develop li:JC2, are examined in the last game of ment, while the white knights on bS this chapter. and bl are a bit clumsy, since both of S ... dS 6 cxdS .lies them would like to go to the c3-square. Obviously not 6 ... li:JxdS?? 7 'ii'xdS! Often a theme is simply to regain the and White wins a piece. 149 Play the Benko Gambit 7lLlsC3 involves .. :fle7 and .. ,IUd8, followed Answering the immediate threat of either by ...lLlb d7-b6 (or ... lLles) or just 7 ...'iVb 6, with a double attack on bS and ... lLlc6 if White allows the ds-pawn to f2. be captured (as in the game). The im 7 e3 0-0 8 lLlSC3com es to the same portant thing is for Black to develop thing. White occasionally plays 8 �e2 rapidly and play aggressively. lLlxdS9 0-0 with a reversed Sicilian, but 9 �e2 has no chance of an advantage here; Alternatively: e.g. 9 ... lLlc6 10 a3 �e6 11 b4 �b6 12 a) 9 �C4?! secures the ds-pawn but lLllc3 lLlxc3 13 lLlxc3 f6 and Black was exposes the bishop to attack; for in fine in B.Kurajica-B.Vuckovic, Bosnian stance, 9 ...lLlbd7 10 'ilc2 lLles 11 lLlxe4 Team Championship 2009. �fS 12 lLlxf6+ 'ilxf6 13 e4, as in Finally, the tempting but incorrect 7 R.Parker-C.Matamoros Franco, London d6? is refuted in Game 42. 1987, when 13 .. JUe8! would have been 7 ...0-0 8 e3 very strong, as 14 0-0 runs into The first of three alternatives for 14 ... lLlf3+! (another reason the bishop White; 8 h3 and 8 g3 (the most critical should be on e2) 15 �hl (or 15 gxf3 line) are seen in the next two games. �h3 and mates) lS .. :irh4 16 gxf3 i..d6 8 •.•e4 and wins. b) 9 a3 'ile7 10 lLld2 J:!d811 b4 (here 11 �e2 returns to the main game) 11...�d6 12 lLlc4 �c7 13 lIa2 (the idea is to switching the rook to d2 to defend the d-pawn, but this is easily neutral ized) 13... �g4 14 �e2 �xe2 15 l:[xe2 lLlc6 16 �b2 (or 16 l:.d2 lLles 17 d6?! �xd6! 18 lLlxd6 lLld3+! - Krasenkov) 16 ...lLles 17 lLlxes �xes 18 ':'d2 .:tac8 19 'iWb3 lLlg4! (seeking more than 19 ...�xc3 20 �xC3 lLlxdS with equality) The advance ...eS- e4 is the standard 20 h3 'iVh4 and Black had a strong ini move here and it the best. The pawn tiative in M.Krasenkow-I.Smirin, Vilnius takes control of d3- and f3-squares and 1988. rules out any later idea of the white e 9 ...'iWe7 pawn supporting its colleague on dS. Further protecting the e4-pawn and My evaluation of this position is making room for ... lId8, putting pres that Black has sufficientcompen sation sure on the ds-pawn - cl central theme for the pawn. The usual plan of action in the Kasparov Gambit, since regain- 150 The Kasparov Gambit ing the pawn will give Black a comfort 14 'ii'Xe4 'ili'xe4 15 lLldxe4 lLla6 was able game. equal in I.Morovic Fernandez-J.Liew (hee Meng, Dubai Olympiad 1986, as Black will soon regain his pawn. 11... lLlxd SI 10 lLld2 10 a3 l:td8 11 lLld2 transposes be low, while 11 b4 �d6 12 �b2 as (12 ...�es is also good) 13 bxas �es 14 Black can simply take the ds-pawn, 1ia4lLla6 15 lLld2 lLlcS 16 'iWb4�g4 17 since it is rather risky for White to take lLlC4 �xc3+ 18 'ivxc3 �xe2 19 �xe2 on e4 (see the next note). 'iWd7 20 J:thdl .:tac8 gave Black great 12 lLlxdS play for the pawns in lTimman After 12 lLlcxe4 Black has excellent V.Topalov, Wijk aan Zee 1996. compensation for the pawn, and the 10 ...l:.d8 sacrifice on e3 is a serious threat. For instance, 12 ... �b6 (even 12 ... lLlxe3!? 13 fxe3 �xe3 14 'ili'b3 �xd2+ 15 lLlxd2 lLlc6 16 lLlf3 1:te8 gave Black a strong initiative in E.Barkovsky-Ma.Tseitlin, Leningrad 1981) 13 lLlc3 (after 13 'ifb3 lLlc6 14 0-0 lLlas 15 'ili'a4 �d7 16 'ili'dl �e6 or 13 lLlg3 lLlc614 0-0 �e6 Black is doing fine with his far more active pieces) 13 ... lLlxe3!? (a very tempting sacrifice, though the simple 13 ...lLlxc3 14 bXc3 �fS 15 0-0 lLla6is also good, as ll a3 White is quite passive and has weak 11 'ili'c2 �fS 12 a3 lLlxdS 13 lLlxdS squares and weak pawns) 14 fxe3 l:[xdS is another transposition below, 'ii'xe3 15 l:tfl lLlc6. whereas 12 g4!? lLlxg4 13 �xg4 �xg4 This position is very tough for White 151 Play the Benko Gambit to defend: his king is stuck in the cen- concessions at all. tre and it is difficult for him to extri cate his pieces, whereas Black has a great many active possibilities. 16 ... 'ii'gS The most active move in the posi tion and the most uncompromising; The game E.Postny-I,Smirin, Israeli Garry is ready to give up the e4-pawn Team Championship 1999, continued to accelerate his initiative. 16 'ifa4?! (16 'ifb3 i..e6 17 lDC4 was a Otherwise Black has several ways to better try, though Black still has great maintain at least equality, such as play after 17... 'ifC 5, threatening ...lDd4) 16 ...lDe5 17 i..xe5 lIxe5 18 lDC4:c8 19 16 ...lDe5 17 ltdl i..e6 18 Ilel? (18 ltc2 llfdli.. C7 20 lDxe5 i..xe5 21 'ifbl i..xal was perhaps the last chance, when 22 'ii'xal h6 23 'ii'd4 .l:tc2 or 16 ... i..C7 17 18 ... l:tac8 19 lDC4 'ifxe2+! 20 lDxe2 l::tfdl lIc8 18 lDC4 l:txdl+ 19 l::txdl a6, i..xc4! 21 ltbl i..xe2 22 l:[el i..d3+ 23 while 16.. Jk8 17 lDC4! i..C7 18 .l:.fdl lta2 l:td4 24 b4 .l:.d5 25 i..f4 i..C4+ 26 .l:.cd8 19 l:txd5 .l:lxd5 20 l:.dl ':'xdl+ 21 ltbl i..b5 regains the queen with a 'ii'xdl a6 22 lDd2 lDe5 23 g3 lDd3 24 clear advantage) 18 ...I:td4 (now the i..d4 h6 is unclear; White has a good white king cannot escape and the at bishop on d4 and Black a good knight tack is decisive) 19 i..f3 'ii'f4 20 'ii'c2 on d3. lDxf3 21 l::te4 'ii'f6 22 :xe6 fxe6 23 gxf3 17 lth1nd6?! 'ifxf3+ 24 ltel 'iff2+ 0-1. A nice game Perhaps a bit too direct. 17 ...1:.c8 by Smirin, illustrating the dangers for looks stronger, and if 18 :adl then White if he underestimates Black's ac 18 ....l:tdd8 (Kasparov) or just 18 ...lDxb4, tivity in the Kasparov Gambit. while 18 lDC4 i..C7 19 ltfdl 1:.xdl+ 20 12... l:txd S 13 'ii'C2 i..fs 14 b4 i..b6 lS l:txdl b5 21 lDd2 i..xh2! 22 ltxh2 lDe5 i..b2 lDc616 0-0 23 'ii'xc8+ i..xc8 24 lDxe4'ii' h4+ 25 <;itgl Black has no problems here, having lDC4 remains unclear. regained his pawn without making any lS lDxe4i.. xe4?1 152 The Kasparov Gambit And here Black might have pre 23••• .i.a5 24 llcS? ferred 18 ..:.g 6 19 f3 l:.ad8 20 'ii'C3 Both the obvious move and the cru i.xe4 21 fxe4 ':'d2 22 1::tf2 (or 22 :ae1 cial mistake. After 24 a4! White might i.C7 with compensation) ':xb2 23 still hope to draw; for instance, 24... l:tb4 'Wxb2 i.xe3 24 i.C4 (if the rook moves, 2S ':'c4l:.xc4 26 'ii'xc4 'ii'e7 27 'ili'c6 :d8 Black has 24 ... l:I.d2 or 24 ... .i.d4) 28 :c1 i.C7! 29 g3 i.b8 and although 24 ... i.xf2 2S 'it'xf2 ttJes 26 i.dS b6 27 Black has finally has the passed pawn 'Wf4 l:Ie8 28 h3 h6, intending ...l:.e7 an d under control it is not clear that ...'ii' gs with an unclear game. whether can make any progress. 19 1i'xe4 .l:td2 24... 'it'b5! 20 b5?! Now Mikha1chishin has nothing Now it is White's turn to go wrong. better than to go into an endgame 20 i.d3? fails to 20.. :� g6!, so Mikhal with a queen for two rooks, which is chishin decides to sacrifice a piece to favourable for Black. obtain a pawn on the seventh, which 25 1:.fc1 'ii'xb7 26 'ii'eS 'ii'xcs 27 'ili'xcs may well be sufficient to draw. How i.d2 2S h3 h6 29 'ili'C4 i.xc1 30 'ili'xc1 ever, he had a stronger, if unexpected, Iixf2 continuation in 20 i.a6! bxa6 21 'ili'xc6 The rest is quite simple. (Kasparov), when 21 ... l:tad8 22 i.c3 is 31 'Wc7 a6 32 'ili'a7 :f6 33 a4 :dS 34 as actually good for White. .l:1d1+ 35 153 Play the Benko Gambit keep the pawn, Black obtains easy be quite dangerous if Black is not pre compensation with his lead in devel pared to a certain level, so he has to opment and active pieces. In the game play with some precision to achieve an Kasparov regained the pawn but, char unclear game. The drawback to White's acteristically, was not satisfied with plan is that it creates weaknesses in his equality as Black and sacrificedth e e4- kingside. pawn in order to sharpen the play. Ul 8 ....i.fS ! timately his uncompromising strategy Getting the bishop out before White was successful, although Mikha1chishin prevents it with g2-g4, while planning should probably have drawn, and to meet that move with ....i.e4. Then might even have been better at one Black will continue to develop rapidly, point. with moves like ...ttJbd7- b6(e5), ...l1a c8, ...ttJc4, and possibly ....i.d4 (should White play e3-e4). If White pushes on Game 40 the king side with g2-g4 and h3-h4, A.Raetsicy-D.Tyomkin Black can consider ...h7 -h6 or ...h7- h5 or Biel 1999 sometimes just ignore it. The advance ... e5-e4, on the other 1 d4 ttJf6 2 c4 cS 3 ttJf3cxd4 4 ttJxd4eS hand, is more risky for Black in this 5 ttJbsds 6 cxds i..cS 7 ttJSC30-0 8 h3 variation and, if played at all, needs to be well timed. For instance, the imme diate 8 ...e4? ! would met met by 9 g4, when the disruptive pawn sacrifice 9 ... e3 isn't really very good after 10 .i.xe3 .i.xe311 fxe3 'iid6 12 'iid4! 'iig3+ 13 �d2 intending 'iif4. In the spirit of the gambit, Black has to play at a good pace with constant active and aggres sive moves, but it's possible to take that too far. 9 a3?! A more ambitious approach than This move, preventing ...ttJb4 or the immediate 8 e3 of the previous ....i.b4 and preparing b2-b4, is often game. With h2-h3 White is planning to useful in the Kasparov Gambit, but just undermine the black e-pawn, if it ad here it seems like a luxury. Instead, vances to e4, with g2-g4 and possibly White has mostly preferred: g4-g5, and at the same time intends to a) 9 g4 is answered by 9 . . . .i.e4! and develop the bishop to g2. This line can then: 154 The Kasparov Gambit bly CS) 13 ttJd2 (preparing e2-e4) 13 ... ttJd6 14 hs i.fs 15 e4 i.d7 16 ttJC4 fs 17 h6 g6 18 a3 ttJxe4 19 ttJxes i.d6 20 'iVd4 with a very unclear position. White's pieces are quite unorthodoxly placed, but he does have a mating theme on g7 and a potential check on c4 or ds (if ds-d6 is allowed). For in stance, 20... ttJxC3 (or 20.. :iIi'e8 21 f4 ttJxgs 22 �dl) 21 i.e3! 'iVe7! 22 ttJxg6! hxg6 23 h7+ �f7 24 h8'iV l::txh8 25 l::txh8 al) 10 f3 ? is quite bad for White, be l:1xh8 26 'ifxh8 f4 27 'iVh7+ �f8 with a cause of 10 ... ttJxds! 11 ttJxe4?! (11 h4 is draw. required, even though 11... ttJxC3 12 b) 9 e3! is the best continuation "it'xd8 ':xd8 13 ttJxC3 i.c6 is very com here, and then: fortable for Black) 11 .. JWh4+ 12 �d2 ':d8 (Palliser), when 13 ttJxcs ttJe3+ 14 ttJd3 ttJxdl wins the queen and proba bly the game, since most of White's pieces haven't moved at all. a2) 10 ttJxe4 ttJxe4 is also finefor Black; for instance, 11 e3 i.b4+ 12 ttJd2 "it'f6 13 'i!Ve2 .l:[c8 14 a3 i.xd2+ 15 i.xd2 ttJa6 16 i.g2 ttJacs 17 i.xe4 ttJxe4with more than enough compensation, thanks to White's numerous weak nesses. hi) 9 ...ttJa 6?! gave Black a convinc a3) 10 l::th2 seems to be the best ing win in the game quoted below. I move when the position is unclear, al will give a bit of coverage on this, since though I would prefer Black in practice, it is useful for the understanding of the given White's vulnerable king on el variation, though I think other moves and displaced rooK. One possible con are stronger. tinuation might be 10 ...i.g 6 11 gs (see b11) 10 .1i.xa6?! bxa6 11 'iVf3 .1i.g612 ing as White can't castle short he might e4 ttJe8 13 g4?! (a risky move, as it cre as well use the rook actively and try to ates weaknesses around the white do something on the kingside) 11 ... ttJe8 king; stronger was 13 i.e3! i.d414 0-0, (the knight will be well placed on d6) afterwh ich Black has insufficient com 12 h4 ttJa6! (on the way to b4 or possi- pensation) 13... ttJd6 14 0-0 i.d4 15 ttJd2 1 55 Play the Benko Gambit 17.. .fS 18 hS .i.e819 g6 h6 (or 19 ...hxg6 20 h6!) 20 'Wb3 and White still has the extra pawn. b2) 9 ...e4 !? is quite challenging, given that h2-h3 specifically prepared to meet this move; i.e. 10 g4 .i.g6 11 .i.g2, but then 11... 4:Ja6 (or 11... 4:Jbd7, as in note 'b33' below) 12 gS?! (12 a3 4:JC7 13 h4 looks a better try) 12 ... tiJb4! 13 gxf6 4:Jd3+ 14 �d2 (or 14 �fl 'ikxf6 15 'ii'd2 'ikh4) 14 ...'ii'xf6 gave Black ex lS.. .fS ?! (not really sound but it cellent compensation for the piece in worked in practice; lS ... ':c8 was more K.Aseev-I.Smirin, Klaipeda 1988. The sensible) 16 exfs 4:JxfS 17 gxfS ':'xfS 18 game continued 15 l::tfl (not 15 f4? 'We2? (after the awkward-looking 18 .i.xe3+! and wins) ls .. .1Ue8 16 a3 (16 �1! Black has no obvious way forward �C2 might offer more hope) 16 ...bS 17 and White is clearly better) 18 ...'ili'h4 19 :ta2 hS 18 4:JxbS 'Wgs 19 .i.hl? 4:Jxf2! 4:Jde4?? (one mistake follows another; 20 'iVe2 4:Jxhl 21 l:txhl 'WxdS+ 22 �C2 White had to accept the inferior end 'Wxa2an d Black went on towin. game after 19 �h2 'iVf4+ 20 �hl l:th S b3) 9 ...4:Jbd7 ! looks like the strong 21 'Wf3 'Wxf3+ 22 4:Jxf3 .i.xc3 23 bxc3 est reply. Black is happy to allow e3-e4 .i.e4 24 �h2 .i.xf3 25 d6 with reason as his pieces can work around the cen able drawing chances) 19 .. :ii'xh3 (now tre: there is no defence) 20 4:Jgs l:[xgS+ 21 .i.xgs .i.hs 22 'We4 .i.f3 23 'Wxf3 'Wxf3 24 .i.e3 l:tf8 0-1 J.Werle-V.Mikhalevski, Hoogeveen 2000. A nice attack, again illustrating the dangers for White in the Kasparov Gambit, even if it wasn't entirely correct. b12) 10 a3! (preventing ...4:Jb4) 10... 4:JC7 11 g4 (this seems like the right time for this move; after 11 b4?! .i.e7 Black has compensation for the pawn) 11... .i.g6 12 .i.g2 'ike713 gs 4:Jfe8 14 h4 b31) 10 a3 transposes to the main e4 (if 14 ...1:tc 8, then 15 e4 and h4-hS is a game. strong threat) 15 4:Jd2 4:Jd6 16 b4 .i.b6 b32) 10 4:Jd2?! .i.b4 (a positional 17 .i.b2 with a slight advantage; e.g. move, aiming to undermine the dS- 156 The Kasparov Gambit pawn and/or weaken the white struc vantage for Black) 18 ...:c 8! (18... bxc3 ture) 11 a3 Jtxc3 12 bXc3 4JC5 13 c4 19 bXc3 Jtxf2+ 20 'ii'xf2 4Jfxe4 21 Jtxe4 'ii'a5 14 Jte2 4Jfe4 15 Jtb2 b5! with a 4Jxe4 22 4Jxe4..txe4 22 d6 is equal) 19 strong attack and the advantage. 4Jb5 4Jxd5 and Black regains the pawn Black's pieces will soon be very active, with an unclear game. while the white king is stuck in the cen 9 ...4Jbd7 10e3 e4 tre. Other moves were good too: b33) 10 g4 Jtg6 11 Jtg2 e4!? 12 g5 a) 10 ...4Jb 6 !? 11 b4 Jtd6 12 'i'f3 4Je8 13 4Jxe4 Jtxe4 14 Jtxe4 'ii'xg5 15 'ii'c8! 13 Jtb2 (not 13 e4? Jtxe4!) 13... e4 'iVg4 'ilie7 16 4JC3 4Jef6 17 'ii'f5 g6 18 14 'ii'dl l:td8 with compensation 'iVg 5 4Jxe4 19 ii'xe7 Jtxe7 20 4Jxe4 f5 (Tyomkin). 21 4Jc3 4Jb6 with good compensation. b) 10 ...':c8 11 g4 Jtg6 12 Jtg2 4Jb6 Black might well regain the d-pawn or 13 h4 (after 13 b4 Jtd6 14 e4':'C4 Black in some cases play more actively with has good play with his lead in devel ...Jtf6, ...l:tac8 and ...4JC4. opment and White's weaknesses on e4 b34) 10 'ii'f3 Jtg6 11 e4 Jtd4 12 Jtd3 and on the kingside) 13 ...h6 14 e4 4JC4 ttJC5 15 g5 hxg5 16 hxg5 4Jh7 17 "g4 Jtd4 18 0-0 4Jd6with compensation. 11 b4?! This creates weaknesses on the dark squares and looks mistimed. A less committal move such as 11 4Jd2mi ght have been better. 11 ... ..td6 12 Jtb2 13 JtC2 fter(a 13 O-O?! 4Jxd3 14 'i'xd34Jxd5 15 'i'e24Jxc3 16 4JxC3'iVh 4 Black had the more active position and the bishop pair and went on to win R.Palus-A.Vaulin, Polanica Zdroj 1998) 13 ...b5 14 a3 as 15 0-0 b4 16 axb4 axb4 17 l:r.xa8 'ii'xa8 18 ttJd2! (White can leave the knight on c3 because the 12 ... 4Je5 bishop on d4 is trapped after b2xC3; 12 ... 4Jb6, intending to target the d other moves lead to at least a clear ad- pawn after ... Jte5, was also fine for 157 Play the Benko Gambit Black, since 13 g4 i.g6 14h4 is safely protected d-pawn is a constant danger met by 14... hS. for White. 13 tl)d2l:tc8 14 i.e2 24 rJilf2?! 14 g4 i.g6 15 gs is risky here, be 24 rJild2 i.bs 25 :cl l:ld8 26 'ifl'el! cause of lS ...l:tx C3! (a nice and very tl)C4 27 i.d2 was a better defensive logical sacrifice: Black gets the dS- and configuration. gs-pawns for the exchange) 16 i.xC3 24 ...i.a4 25 l:tC1i.c2 26 i.d4 tl)C4 (16 gxf6 tl)d3+ 17 i.xd3 l:lxd3 18 fxg7 Now all White's pieces look more or l:[e8, followed by ...i.es, is good for less displaced and he was also in severe Black) 16 ...tl)xds 17 i.d4 "ii'xgs, when time trouble. White faces long-term problems re 27 i.C5b6 28 i.e3f6 garding the safety of his king. 14... tl)d3+ 15 i.xd3 exd3 16 11c1?! White might still have maintained equality with 16 'ii'f3 'iid7 17 1:cl i.es 18 0-0 l:tfe8 19 l:tfel i.xc3 20 l:txc3 l:[xc3 21 i.xC3 tl)xds 22 i.d4 i.g6, when a draw is the most likely result. 16••• i.e5 17 'ii'b3 i.xC318 1ixC3 IixC319 i.xC3'ii' xd5 29 tl)d2?! Giving up the a-pawn in the wrong way. Instead, 29 a4 tl)b2?! 30 rJileltl)xa4 31 rJild2 tl)b2 32 tl)a3 tl)c4+ 33 tl)xC4 liIxC4 34 l:tal (Tyomkin) or 29 ... i.xa4?! 30 tl)d2 i.bs (or 30... bS 31 :al) 31 tl)xC4 i.xC4 32 rJilel would offer White reasonable drawing chances due to the opposite-coloured bishops, although The endgame is slightly better for 29 ...rJilf7 30 tl)d2 tl)b2 still leaves Black Black, who has a dangerous d-pawn, with a favourable endgame. while the white pieces are rather pas 29.•. tl)xa 3 30 lla1tl)b5 sive. Black has won a pawn while keep 20 'iixd5 tl)xd5 21 tl)b1tl)b6 22 f3 :c8 ing the knights on and stands clearly 23 e4 i.d7 better, especially as White had no time On its way to c2, after which the on the clock to try and find a defence. 158 The Kasparov Gambit 31 h4 hS 32 g4 hxg4 33 fxg4 1;f7 34 hS .l:!.C3 3S 1;f3 .l:.a336 .l:[h1 .l:.a437 h6 gxh6 Game 41 38 ':'xh6 lbb4 39 1;f4 ttJC3 40 l:th7+ M.Matlakov-Alexa.Evdokimov 1;g6 41 ':xa7 Russian Team Championship 2010 1 d4 ttJf6 2 c4 cS 3 ttJf3 cxd44 ttJxd4es 5 ttJbsds 6 cxds i.cs7 ttJsc30-0 8 g3 41 ...ttJxe4 ! 42 ttJxe4 �xe4+ 0-1 A nice concluding tactic. If 43 1;xe4 then 43 ... d2+ and the pawn promotes, or if 43 1;f3 then simply 43 ... bS is win ning; e.g. 44 l:1b7b4 4S ltb6b3 46 gs S g3 is the theoretically most testing l:!.xe3+! 47 '\tlxe3 1;xgs (Tyomkin) and line in the Kasparov Gambit. It was the one of the pawns will go through. choice of Aronian in a game against Sutovsky (see below). Given one mo Game summary ment and White will have played i.g2, Although the move S h3 can be dan 0-0 and e2-e4 with a clear advantage, gerous against an unprepared oppo so Black needs to disrupt this plan right nent, after the accurate S ... i.fS! Black now. shouldn't have any problems. Even 8 ...'iW b6!? against White's best reply of 9 e3, ei It is not so easy get an unclear game ther 9 ... e4!? or 9 ... ttJbd7! offers Black in another way here. I spent some time excellent counterplay for at least on other continuations, but I won't go equality. In the game, 9 a3?! and 11 into detail since I could not find a clear b4?! were not the most precise moves, path to equality. and the further error 16 l:tc1?! allowed ge3 Black to reach a favourable endgame. Obviously the only move. Aided by his opponent's subsequent 9 ... i.g4! time trouble, Tyomkin went on to win Instead, 9 ... l:tdS has some logic to it convincingly. - attacking the ds-pawn - but then 1 59 Play the Benko Gambit White gets to play 10 �g2, supporting Black is about to regain the pawn with the pawn, and stands slightly better. the superior structure. And blocking For instance, 10... �g4 11 'ii'c2 tLlbd7 12 the bishop by 10 f3 weakens the white a3 ':'ac8 13 0-0 'ii'a6 14 tLld2 �e7 15 kingside, so that after 10... �d7 Black tLlb3 tLlb6 16 e4 tLla4 17 l:teltLlxC3 18 has good compensation for the pawn bxc3 'ii'a4 19 lla2wh en Black has some with his lead in development. For ex compensation, but probably not ample: enough. a) 11 'ii'e2 e4 and then: 10 �e2 al) 12 tLlxe4tLlxd5 13 tLlxC5'ii' xC5 14 The best reply, even though it 'ii'C4 'ii'd6 15 tLlC3 tLlxC3 16 'ili'xC3 l:tc8 17 means Black has achieved his first aim 'i!Vd3 'i!Ve7 18 "ilr'a3 'i!Ve8 19 e4 tLlc6 (with of preventing the fianchetlo. Moving ideas of ...tLld4 or ...tLle 5) 20 i.e3 as 21 the queen would allow Black to play �f2 tLlb4(ta king another route) 22 :Ldl ...�f3 with an easy game; e.g. 10 'ii'h3?! f5 23 exf5 �xf5 24 �f4 �e6 25 l:td2 'ii'xb3 11 axb3 �f3 'ii'f7 and the white king is still exposed with compensation, while Black is ready to regain the pawn on a2. a2) 12 fxe4!? was tried in V.Tverdohlebov-I.Foigel, Ukrainian Team Championship 1980, when 12 ...tLlg4 13 tLla3 (or 13 tLld2 i.xe3 14 tLlC4 �f2+ 15 �d2 'ii'c5) 13 ... �xa3 14 bxa3 ':'c8 15 �d2 tLle5 16 'i!Vf2 tLla6 gave Black good play for the pawns. He might also have considered 12 ... �g4!? 13 'ii'f2 (or 13 'ii'b5 �f3 14 'ii'xb6 axb6 12 l:tgl l:r.d8 13 �C4 �b4, when 15 ':'gl tLlbd7 16 tLld2 tLle5 17 i.e2 160 The Kasparov Gambit i.xe3 18 .l:tfl i.g2 19 :fS l:tfe8) worth investigating. The only drawback 13 ...ltJbd7 14 h3 i.hS 15 i.g2 ltJes 16 is that White doesn't need to allow it. 0-0 ltJd317 ltJa4 ltJxf2 18 ltJxb6 ltJxh3+ Therefore I would prefer the popular 19 i.xh3 i.xb6 again with compensa 11 ...e4 !, after which Black can obtain tion, since the pawns in the centre are sufficient compensation for the pawn weak and Black is still very active. with correct play: b) 11 'ii'd2 keeps an eye on the dS pawn, but then White's pieces are even more clumsily placed, and Black can gain active counterplay with moves like ...ltJa6-b4, ...l:.ac8, and ...i.b4. Play might continue 11... ltJa6 (not now 11... e4? ! 12 ltJxe4! and White is better) 12 ltJa3 lUe8 13 e4 .:tac8 14 ltJc4 'iVd8 15 a4 h6 16 b3 ltJh7 17 i.a3 i.d4 18 l:[cl 'ii'f6 19 'ii'dl ltJcS 20i.g2 a6 21 as i.bs 22 11c2 i.xC4 23 bXc4 hS 24 h4 ltJb3 25 'iVd3 'iVd8 with ongoing com a) 12 'iVhS?!(th is unambitious move pensation. gives Black an easy game) 12 .. :iVxbS 13 10 ...i.xe2 ltJxbs ltJxds 14 ltJlc3ltJxc3 15 ltJxC3fS 10 ... i.h3?! 11 g4!, threatening l:tgl 16 �e2 ltJc6 17 i.d2 ltJes18 ltJdsllad8 g3, was good for White in V.Korchnoi V2 -V2 P.H.Nielsen-B.Gelfand, Monte D.Gurevich, Las Vegas 2007. Carlo (blindfold rapid) 2006, and Black 11 'iVxe2 certainly had the edge final position. b) 12 ltJd2 (not the most precise; 12 a3 offers White more chances of gain- ing an advantage) 12... i.b4! (now the ds-pawn is weak; 12 ...l:te8 ?! would al- low White to rectify his mistake with 13 a3!, transposing to line 'd' below) 13 0-0 :te8 14 lIdlltJbd7 15 a3 (or if 15 l:tbl, as in C.Lingnau-A.Kadatsky, Par dubice 2001, then lS ... l:.ac8 16 a3 i.f8 17 ltJC4 'iVh3!? 18 ltJas 'ilb6 19 ltJC4 with a repetition) lS ... i.xc3 16 bxc3 11 ...ltJb d71? ltJxds 17 ltJxe4'ii' c6 18 'ii'd3 ltJ7b6 (or This prepares a very interesting idea 18... ltJsb6 !? 19 ltJd6 ltJes 20 'iVhs'ii' xbs on the next move, which is certainly 21 ltJxbs l:ted8 22 ltJd4 ltJa4 23 i.d2 161 Play the Benko Gambit J:tacS24 l:tabl b6 25 l:tb4 1DxC326 i.xC3 In his annotations Aronian sug l1xC3 with a level position} 19 litbl? (19 gested 12 .. ..l:teS! as a possible im c4 is equal) 19 ...l:tadS 20 c4 was provement, and in my opinion Black F .Kristensen-Ca.Hansen, correspon has sufficient compensation after this dence 1994, and now Black could just move. Now 13 0-0 lDbd7 would trans have taken 20... lDxC4! with a clear ad pose to line 'c' above, while 13 b4 i.d6! vantage. The threat is 21...f5 and is also fine for Black. So that leaves 13 22... lDdxe 3, while 21 lDg5? loses to lDd2an d then: 21...lDf4!. c} 12 0-0 lDbd7 (12 ...i.b 4!? is possi ble here too, when 13 a3 i.xc314 1Dxc3 lDbd7tran sposes to 13 0-0 in the main game) 13 a3 lUeS (13 ...l:tacS 14 lDd2 lUeS 15 b4 is slightly better for White - see 12 O-O! in the main game) 14 b4 (or if 14 lDd2, then 14 ... 'iIi'c7!) 14 ... i.d6! (possible because White doesn't have lDC4 yet) 15 lDd2 i.e5 16 lDC4 fidS 17 i.b2 i.xc3 is i.xc3 lDxd5 19 i.b2 was seen in the freestyle game "King dl} 13... a5 14 'ili'C4?! (better was 14 Crusher" IRybka-"Equidistance", Inter b3! lDa6 15 i.b2 l:tadS 16 1Da4 'iVd6 17 net 2006. Black should be able to with lDc4'iVxd5 is i.xf6 gxf6 19 0-0 with an stand White's slight pressure here; for edge for White) 14 ...lDbd7 15 lDdxe4 instance, 19 ...fie7 20 %1adlWe6 21 l:td2 lDxe4 Y2 -Y2 A.Maly-O.Berezin, Kharkov l:tacS 22 ,Ufdl 1D7b6 23 lDxb6 lDxb6 24 2000. Black is okay here and could have i.alh6 25 'Wh 5 1Dc4 26 .l:.d7b6 27 'itg2 played on. For instance, 16 lDxe4 lDe5 a6 2S J:.ld5b5 29 i.d4 l:taS 30 g4 %:tacS 17 'iVa4 lDd3+ is 'itfl i.b4!? 19 axb4 31 h4 l:te7 32 l:txe7'iVxe7 33 g5 :c6 34 l:[xe4 20'itg 2 l:txb4 21 'iVc2'iVg 6 22 'iVe2 'ili'g4 fie6 35 .l:r.dS+ 'ith7 remains very 'iWe4+ 23 Wf3 f5! is good for Black, due unclear. to the strong knight on d3 and White's d} 12 a3! is critical. This was played weaknesses at f2, d5 and b2. in the aforementioned game, d2} 13... 'iVC7 (Ar onian's move, which L.Aronian-E.5utovsky, European Team hinders b2-b4 for the time being) 14 Championship, Heraklio 2007, in which g4!? (a drastic response thrown up by White gained the advantage after the computer to undermine the e4- 12"'lDbd7 13 lDd2 l:tfeS 14 b4 i.fS 15 pawn; otherwise 14 0-0 a6 15 ,Udl i.d6 0-0 'fiC7 16 i.b2 lDb6 17 lDb5! and offers reasonable compensation, with went on to win convincingly. ideas of ...lDbd7, ...l:tacS and ...lDc 5) 162 The Kasparov Gambit 14 .. .'ii'es 15 ttJC4 'ii'gs leads to a very Again 13 O-O! would have been unclear position. Play might continue more testing, although Black gets 16 d6 (or 16' b4 .i.fS) 16 ....i.b6 17 b4 enough compensation now: ttJbd7 lS .i.b2 .l:r.acS 19 h3 l:tc6 20 bS .l:lxC4!? 21 'ii'xc4 ttJes 22 'ii'e2 ttJd3+ 23 'iitfl lIe6with compensation due to the strong knight on d3 and the vulnerable king on fl. 12 a3 Seemingly following Aronian's plan, but the slight difference in the position gives Black an unexpected resource. More accurate was 12 O-O! lilacS 13 a3! e4 14 ttJd2lUeS 15 b4 i..xe3 16 ttJC4 and White regains the piece with ad 13... .i.xc3 (13 .. .l:ifeS?! is a bit too vantage: 16 ... 'iYC7 17 .i.xe3 (or 17 ttJxe3 fancy: after14 ttJa2! ':'acS 15 exd4 exd4 'iVxc3 lS i..b2) 17 ...'ii' xC4 lS 'ii'xC4 1IxC4 16 'iYdl ttJxdS 17 ttJb4 ttJxb4 lS axb4 19 ttJbS lIaS Vz -Vz S.5wapnil-E.Ghaem ttJes 19 ttJd2 a6 20 ttJe4 'ii'xb4 21 f4 Maghami, Chennai 2010. Presumably White is clearly better) 14 ttJxc3 e4 15 Black offered a draw before his disap b4 ttJes! 16 .i.b2'iVd6 17 lIadl(17 ttJbS pointing tournament got any worse. 'iYxdS! lS ttJC7 ttJf3+! 19 'ii'xf3 exf3 20 12 ... .i.d41? ttJxdS ttJxdSis roughly equal) 17... a6 lS 'iitg2 lilfeS 19 l:ld4bS 20 lilfdlttJC4 21 a4 'iVd7 and now White might try 22 l:txc4!? (after 22 axbs axbs 23 Ital hS 24 h3 %itxal 25 .i.xal ttJd6 Black has sufficient compensation due to his control of the light squares) 22 ...bxc4 23 'ii'xC4 'ii'fs 24 d6 hS 25 h4 ttJg4 26 'iWe2 ttJes 27 ttJdSttJd3 2S ttJf4 (or 2S ttJe7+lIxe7 29 dxe7 ];teS30 i..d4 ':'xe7 31l:tbl l:tb7 32 bS axbs 33 as ':'c734 a6 b4 35 a7 ];tcS) 2S... ttJxf4+ 29 exf4 l:tabS This is only a temporary piece sacri 30 l:td4 l:tbdS 31 .i.a3 (if 31 'iVxa6e3! 32 fi ce. Otherwise 12 ...e4 or 12 ....l:tfeS 13 'iYe2 exf2 33 'ii'xf2 lIe6 or 31 as lIe6) ttJd2 e4 would return to Aronian 31 ...'iVcS with an unclear game. Sutovsky. 13..• exd4 13 exd4 The threats of ...1IfeS an d ... dxc3 en- 163 Play the Benko Gambit sure that Black regains the piece. 21 c;t>g2 tiJxd4 22 tiJd6 l:ted8 23 tiJxb7 14 0-0 lUe8 ':'xdS) 19 ... tiJxC3 20 l:txb2 tiJxdl 21 .l:1xdltiJf3 + 22 �g2 tiJxd4 23 .l:txd4l:te7 24 d6 l:td7 25 .:tc2c;t>f8 is a drawn end game. b2) 17 c;t>g2 %1ad8 18 .i.e3(or 18 .i.gs "ii'xb2 19 .i.xf6 gxf6 20 "ii'cl "ii'xcl 21 litaxcl l:td7 with equality) 18... 'ilxb 2 19 .i.d4 tiJxdS! 20 tiJxds 'iibs 21 tiJe3 (not 21 tiJC3?! "ii'C4 22 tiJe2tiJc6, wh en Black emerges a pawn up) 21 ...tiJc6 22 l:tbl 'ii'gs 23 tiJC2bS! 24 .i.e3!? (otherwise 24 c;t>gl tiJxd4 25 tiJxd4 'iif6 26 l:tb4 as! 15 'iVC2 regains the piece) 24 ...l:lxe 3 25 tiJxe3 Alternatively: ':'xdl 26 l:.fxdl a6 27 a4 (or 27 lld6 a) 15 'ifbs dxc3 16 tiJxC3 'ilxbs 17 'iies and the knight is indirectly cov tiJxbs tiJxds is equal, since 18 tiJd6 ered by ...'ii e4+) 27 ...'ii es (Black has to l:teb8 is only a temporary inconven play actively here) 28 axbs axbs 29 ience. c;t>gl g6 30 l:tdS"ii' e4 31 litbxb5 tiJd4an d b) 15 'ii'dl dxC3 16 tiJxC3is more in Black is doing fine. The knight and teresting, when the queen defends the queen combine well together against d-pawn, but after 16... tiJes Black has the weak light squares around the compensation due to the weakness of white king - one of the points behind the ds-pawn and the light squares ...'ii b6 and ....i.g4 back in the opening! around the white king. For instance: 15•.• :tac8 16 .l:!d1 bl) 17 .i.e3 'ilxb2 18 .i.d4tiJe4! 19 16 ...tiJ e5 llbl (or 19 tiJxe4'ii xd4! 20 'iixd4 tiJf3+ Another option was 16 ...tiJcS !? 17 164 The Kasparov Gambit �f4 liJce4 18 d6 l:te6, followed by offer White some chances of gaining ...d4 xC3 and liJxd6with equality. an edge. My conclusion is that Black 17 'ii'fs should be doing okay after 8 g3 if he Or 17 �g2 liJc4 18 a4 dxc3 19 liJxc3 plays precisely. hS with compensation. 17 ...g6 18 'ii'f4 dxe3 19 liJxe3l:.e4 Game 42 A.Shvedchlkov-Ma. Tseitlin USSR Army Championship, lvov 1977 1e4eS The game starts as a symmetrical English, but it soon transposes. 2 liJf3liJf6 3 d4 exd4 4 liJxd4eS S liJbS dS 6 exds .ies7 d6? Black is very active here, but White can keep the balance. 20 �e3 "xb2 21 l:.ab1 'iie2 22 libe1 l::txf4 23 ':'xe2 :e424 .:tee1.l:.ee8 2s liJbs YI-YI Game summary 8 g3 is currently a popular line in the Kasparov Gambit and I think it is also the most critical. In response 8 ... 'ii'b6 9 e3 �g4! prevents the white bishop go It is unlikely that any strong players ing to g2 in an efficient way. After the will try this move nowadays, since it is further 10 �e2 .ixe2 11 'ii'xe2 my rec well known that after 7 ...0-0 Black is ommendation is 11 ... e4, and if 12 a3 fine. The greedy 8 liJC7 is met by then Aronian's idea of 12 ....:.e8 13 liJd2 8 ...liJe4! with ideas of ....ix f2+ or "C7, temporarily preventing b2-b4, ... .ib4+ and ...'ii' xd6, when White can't seems to provide Black with sufficient take the rook in a good way. The calm compensation. In the game Evdokimov computer already prefers Black. found interesting counterplay with 7 ••• 0-01 11 ...liJbd7 12 a3 .id4!? and made a This move can almost be regarded as draw, though in this line 12 0-0 would a refutation of 7 d6, since it gives Black a 165 Play the Benko Gambit definite advantage. 7 ...tDe4? !, on the the knight and, behind it, the b2-pawn) other hand, only leads to an unclear 12 ...tDe4 and Black has a strong attack. game, and so must be regarded as ob d) 8 'it'ds!? �xf2+ 9 c;t>dl tDf6 10 jectively worse, which is why I have ap 'ii'xes+ (worse is 10 tDC7+?! c;t>f8 11 'iVd3 pended the dubious mark. Nevertheless, e4! 12 'ilfa3 tDa6 13 tDxa8 �cs and in the book Dangerous Weapons, Flank Black has a dangerous initiative) Openings, Palliser suggests that it might 10 ... c;t>f8 11 e3 (or 11 �gs tDbd7 12 iff4 be ventured by players who favour "a 'iVb6 with compensation - Ca.Hansen) quick knock-out", and it is certainly in 11...a6 12 tDsC3�h 4 13 'iVcs tDbd7with teresting and dangerous for White: a very unclear position in which Black should not be worse. S J..e3 If White doesn't do something dras tic there is simply ... tDc6 and ...a7-a6 with a fine game for Black. However, the intended 8 tDC7run s into 8 ...tDe4! a) 8 tDC7+?? (your opponent must be very cooperative to play like this) 8 .....xC] ! 9 'iVa4+ (9 dXc7 �xf2 is mate!) 9 ...'iVc6 0-1 was the rapid end to R.Hoque-lHodgson, London 1987 (along with two other games in the database). 9 e3 (other moves just lose; e.g. 9 b) 8 e3 O-O! 9 'ii'ds (9 tDC7tran sposes �e3? �xe3 10 fxe3 'ii'f6, coming in on to 8 tDC7in the main game) 9 ...�b4+ 10 f2, or 9 'iVds �xf2+ 10 c;t>dl tDxd6 11 �d2 tDxd2 11 tDxd2 a6 12 tDC3 (or 12 tDxa8tDc6 and the white king is a sit tDC7 l:ta7) 12... tDc6 with excellent com ting duck) 9 ... �b4+! (stronger than pensation for the pawn, supposing 9 ... �xd6?! 10 tDxa8 J..b4+ 11 tDd2 1r'h4 White even manages to hang on to it. 12 g3 'iVf6 13 f3 tDxd2 14 �xd2 l':td8 15 c) 8 �e3 �xe3 9 fxe3 0-0 10 'ili'ds J..xb4 ':'xd1+ 16 l:rxdl tDc6 17 �cs tDf6 11 'fixes?! (but after 11 'iVC4 tDc6 'iVxf3 18 :gl �g4 19 l:td2tDd4 20 �xd4 Black is clearly better in any case) exd4 21 l:td3 and White managed to 11...tDc6 12 'iVc3 (or 12 'iVf4 'iVb6 hitting hang on in P.Harikrishna-E.5utovsky, 166 The Kasparov Gambit Hoogeveen 2005) 10 .i.d2 (there is cult defence; e.g. 11 'iid2 .i.d7 12 a4 nothing better: 10 lLld2'ii' xd6 11 lLlxa8 .i.xbs 13 axbs lLlcs,threa tening ...lLlb 3 l:td8 is clearly worse; 10 �e2? loses af or .. JUd8xd6. ter 10 ...'ii' h4 11 g3 lLlxg3+! 12 hxg3 11 'ii'd2?! �xhl; while 10 lLlc3 lLlxc3 11 bxc3 No better is 11 'iih3?! lLlcs 12 'iih4 �xC3+ 12 .i.d2.i.xal 13 'ii'xal 'ii'xd6 14 'ii'b6 13 a4 a6 14 as 'ii'c6 15 lLlC7'ii' xd6 lLlxa8 lLlc6 followed by ... .i.e6 and (threatening ...lLld3+ winning the ...l::t xa8 leaves Black a pawn up) queen) 16 'iib6 .i.c6! 17 l:tdl (or 17 10 ... lLlxd2 11 lLlxd2 'ii'xd6! 12 lLlbS (af lLlxa8 lLlfe4 18 lLlxe4 lLlxe419 l:tdl'ii' f6) ter 12 lLlxa8? l::td8 Black has a close to 17 ...'ii' e7 18 lLlxa8l::t xa8 leaves White in decisive advantage) 12 ...'ii g6 and Black serious trouble. The threats are ...':'c8 is very active without even being a and ...lLlf d7, or ...lLlfe4 and ...'ii h4+; for pawn down, while White will have instance, 19 e4 lLlfxe4 20 lLlxe4 lLlxe4 great difficultydevel oping his king side. 21 g3 lLlf6 22 :gl lLldS 23 'iWb3 'ii'cs 8 ....i.xe3 9 fxe3 lLla6! and Black has a big advantage. Here 11 a4! seems to be the best de fence, though after 11 ...'iVas White would still have his work cut out to try and save the game. 11.•. .i.xb sI 12 lLlxbs lLle4 Again Black has a clear advantage: the weaknesses on e3, d6 and espe cially el give him a very nice game. 9 ...lLle8 and 10 ... a6 was also good, but Tseitlin's move puts more pressure on his opponent. And now Black is winning, since 10 lLllC3 .. :ii'as+ or ...'ifh 4+ cannot be prevented. There is no time to fianchetto with 13 'iids 10 g3 because of 10 ... .i.d7an d 11 ... .i.c6. 13 'ii'c2 'ii'h4+ 14 g3 lLlxg3 15 hxg3 10 ....i.d7 'ii'xhl is also a disaster for White. Another option was 10 ...'ii'b 6!?, 13 ..•'ii' as+ 14 lLlc3 lLlacslS l::tcl l:tfd8 16 when White would again face a diffi- g3 l:txd6 17 'ii'c4 167 Play the Benko Gambit 32 'iti>d3 tbf6 33 .i.C4 'iti>d6 34 :f2 :g6 35 liIc2 l1g4 36 .i.b5 l:tb4 37 .i.C4 l:tb1 38 'iti>e3e4 39 .i.e2 'iti>e5 tbd40 'iti>f2 5 41 :c6 l::tb2 42 h3 g5 43 a6 tbb4 44 :C7 tbd3+ 45 'iti>f1 :b1+ 46 'iti>g2 .l:tb2 47 'iti>f1 'iti>d4 48 .i.xd3 'iti>xd3 49 ':xa7 .l::tb1+ Vz-Vz Game summary Although 7 d6, threatening tbc7,might look tempting for White, it is well known that Black has no problems 17 •••:1c8 11 here. In fact after 7 ...0-0 Black already 17.. Jlb6 would have been decisive. has the advantage. In the game, The threats are 18.. J:tb4 19 'il'dS .l:td8 Tseitlin, one of the originators of the winning the queen, or just .. Jbb2. The gambit, played strongly to reach a text gives White an unlikely reprieve winning position, until he blundered which he is very happy to take. on move 17 and let his no doubt grate 18 b41 tbd3+ 19 exd3 l:ixc4 20 bxa5 ful opponent get away with a draw. l:[xc3 21 llxc3 tbxC3 22 'iti>d2 tbxa2 23 .i.g2 Game 43 F.Santos Garcia LAlvarez Naves Valladolid 1975 1 d4 tbf6 2 c4 cS 3 tbf3 cxd4 4 tbxd4e5 5 tbc2 Somehow White has escaped into a slightly worse endgame, which he eventually managed to draw. I will not comment on the rest. 23 •••tbb4 24 l:1b1 tbxd3 25 .i.xb7 f5 26 'iti>C3 'iti>f7 27 e4 fxe4 28 .i.xe4 tbC5 29 :f1+ 'iti>e7 30 .txh7 .l:th6 31 .i.g8tbe4+ 168 The Kasparov Gambit I wouldn't recommend this for (rapid) 2000. White. Avoiding the gambit with S liJc2 S ..• dS 6 cxdS 'ii'xds 7 'ii'xdS liJxdS (or 5 liJb3) simply means that Black can play ... d7-dS without having to sacrifice a pawn, and then it is White who must react precisely to get an equal game. True, exchanging queens sidesteps many of the complications of the Kas parov Gambit, but I don't think the re sulting endgames (or queenless mid dlegames) are especially drawish, if it is that White is aiming for. Other knight moves are: a) 5 liJb3 is certainly inferior, since 8e4 the knight is badly placed on b3 after After a g3?! tLlc6 9 �g2 �e6 10 0-0 S ... dS 6 cxdS 'ii'xds 7 liJc3?! (7 'ii'xds 0-0-0 Black is far more active, and 11 liJxds 8 e4 liJb4 9 liJa3 it.e6 is also in it.d2c;i;>ba 12 tLlc3liJdb4 13 tLlxb4�xb4 Black's favour; White should probably 14 it.gs f6 15 it.xc6 fxgs 16 �g2 l:td2 opt for 7 it.d2 and 8 liJc3, when Black gave him a very nice endgame in has only a slight edge) 7 ...'ii' xdl+ 8 U.Andersson-B.Gelfand, Tilburg 1990. <;i;>xdl liJc6 9 liJbS?! (9 e4 it.e6 10 f3 8 ...tLl b4 9 tLlxb4 .l:.d8+ 11 'ittel a6 12 it.e3 liJb4 13 'ittf2 9 �bS+ liJac6 10 liJxb4 it.xb4+ 11 would keep Black's advantage within it.d2com es to the same thing. limits) 9 ....l:.b8 10 liJc7+ 'ittd8 11 liJbs 9 ...it. xb4+ 10 �d2 liJc6 tLlg4 12 'ittel liJb4 13 liJa3 it.e6 14 f3 tLlxa2! 15 ':xa2 it.xb3 16 l::t.al it.b4+ 17 it.d2 it.xd2+ 18 'ittxd2 liJf6 and Black was a clear pawn up in R.Cifuentes Parada-I.Morovic Fernandez, Santiago 1989, and went on to win. b) 5 tLlf3 tLlc6 6 tLlc3it.b4 7 it.d2 0-0 8 e3 (here 8 a3! it.xc3 9 it.xc3 tLle4 10 'iVc2 tLlxc3 11 'ii'xc3 dS 12 cxdS 'ii'xds is roughly equal) 8 ... e4 9 tLld4 (or 9 tLlgs l:[e8)9 ...tLlxd4 10 exd4 dS! 11 'ii'b3 it.xc3 12 bxc3 it.e6 13 cS?! 'ii'C7 14 litbl litab8 In symmetrical positions such as 15 it.e2 b6 was good for Black in this, the player "with the move" has J.C.Dominguez-V.Topalov, Villarrobledo whatever chances are going, and here 169 Play the Benko Gambit that player is Black who has one more 18 �fl White should make a draw. The piece out. White, on the other hand, text can be met by 14 �xc6 �xc6 15 after just ten moves is now playing for lLlc4 f6 16.l:tfdl �e7 17 f3 with equal a draw at most. ity, but instead White starts to make 11 �b5 more trouble for himself. Instead, 11 �C4 �e6 12 �xe6 fxe6 14 �C4 0-0 15 lLlf3?! 13 lLlC3 was seen in A.Pomar Sala 15 lLlb3 was better. manca-Bellon Lopez, Las Palmas 1977, 15 ... �g4 16 lLlh4?!lladS?! when 13 ...0-0 14 ':Cll:.ac8 woul d give 16... lLld4 ! would have caused the Black some pressure in the endgame. white knight some problems in return The doubled e-pawns provide very ing to the game. For instance, 17 �d5 good central control, the knight (or l::tab8 18 h3 �e6 19 lLlf3 (or 19 �xe6 sometimes the bishop) has a nice fxe6!) 19 ... lLlxf3+ 20 gxf3 �xd5 21 exd5 square on d4, while the rooks can go to 1:tfd8 22 lUdl J:[bc8 23 Jitacl �8 and c- and d-files with activity on the Black has good winning chances due to queenside. the weak white pawns, especially the d 11... �d7 12 0-0 pawn. Another of Ulf Andersson's game 17 f3?! went 12 �xc6 �xd2+ 13 �xd2 �xc6 14 17 h3 would have freed the knight lLlc30-0-0+ 15 �e2 l:.d4 16 f3 .l:lhd8 17 again with equality, whereas the text lithdl and this time he managed to takes away its main retreat square. hold this slightly worse position in 17... �e6! U.Andersson-F.Langheinrich, German League 2003. 12... �xd 2 13 lLlxd2 a6 is b3?! 18 J:[acl�xc4 19 ':xC4':d2 20 l:tf2 lUd8 21 ':c2 .l:tdl+ 22 .l:[fl g6 is still After 13... lLld4 14 �xd7+ �xd7 15 slightly better for Black due to his more J::tfdl �e6 16 lLlb3! lLlxb3 17 axb3 11hd8 active pieces. 1 70 The Kasparov Gambit 18 ...J:td 2 ct;xb4hS 46 a4 h4 47 as h3 48 a6 h2 49 More accurate was 18 ...ttJd4 ! fol a7 h1'i10-1 lowed by 19 ... 95 with a clear advan tage. The main idea is 20 .ltxe6 fxe6!, Game summary threatening ...g7 -g5, when White The alternatives to 5 ttJb5 are all sig would have to give up the f-pawn in nificantly inferior. Indeed, White has to order to save his knight. work to prove equality in these lines, 19 l:tfd1?! and the exchange of queens surrenders 19 f4 would offer more drawing the initiative completely to Black, even chances, when at least the knight gets if White might expect to make a draw. back in to the game. In the game, Santos played a succes 19... l:tfd8 20 iLdS? l:txd1+ 21 ':'xd1 sion of inferior moves, which inevitably ttJd4?! led to a lost ending. Although Black Here 21... g5 22 ttJf5 iLxf5! 23 exf5 conducted the technical phase uncon �g7 threatening ...ttJb4 or ...ttJe7 woul d vincingly, I think the game is quite in have won much more easily. structive up until the final diagram. 22 g3 iLxdS 23 exds :xdS Summary The Kasparov Gambit is a sound pawn sacrifice for which Black gets active counterplay. In response White has three main moves: The original main line with 8 e3 is quite logical in that closes offth e c5- bishop, and this move continues to be popular, even though Kasparov himself demonstrated back in 1981 (see Game 39) that 8 ...e4, followed by ...We 7 and I won't comment on the rest. Black .. .lId8, gives Black a perfectly good is a pawn up and eventually won in a game. Indeed, White has yet to show rather unconvincing way. any substantial improvements here. 24 1 71 Play the Benko Gambit possibly ...tba6 -b4, again with good 0-0, so I would prefer 11 ...e4! as given play for the pawn. in the notes. Here Black might improve The third line, involving 8 g3, is cur on Sutovsky's play by 12 ...:te8 13 tbd2 rently regarded as the most critical, 'ilC7, deterring a quick b2-b4, and with Aronian-Sutovsky being a key en reaching an unclear game. counter. The primary idea is to rein Attempts by White to avoid the force the ds-pawn by �g2 and, ulti main line or the gambit altogether are mately, e2-e4. Obviously Black cannot markedly inferior. Game 42's 7 d6? is allow this plan to succeed, and 8 ...'ifb 6 simply a mistake, since 7 ...0-0! already 9 e3 �g4! is an effective hindrance, so gives Black favourable chances. The that after10 �e2 �xe2 11 'iWxe2 White various knight retreats on move 5 (seen is left with weaknesses on the light in Game 43) allow Black to free his po squares. Nevertheless, he still has his sition without sacrificing, when it is extra pawn on dS, so Black must play White who is put on the defensive. precisely to obtain sufficient compen In conclusion, the Kasparov Gambit sation. appears to be fully viable. Some new Game 41 features an interesting ideas and novelties in this chapter, es idea with 11... tbbd7 12 a3 �d4!? with pecially in the line with 8 g3, show that which Evdokimov achieved equality. Black can prove definite compensation However, White does better with 12 for the pawn. 1 72 Chapter Eight The Topalov Variation 1 d4 ttJf6 2 ttJf3 cS 3 ds bs 4 i..gs ttJe4 first three games below. The variation bears close resem blance to the Trompowsky 1 d4 tDf6 2 i.g5 ttJe4 with the further 3 ... c5 4 d5. From Black's point of view the move ...'ilVb 6 is no longer an effective equal izer since the b-pawn blocks the queen's path to b2. On the other hand, White is committed to d4-d5, when he might have preferred to play d4xc5 or f2-f3, while his own developed knight prevents the latter move altogether. This position is the starting point Other fourth moves are examined for the Topalov Variation, which will be in Game 47; while White's third move our main focus in this final chapter. As options, 3 e3 and 3 C3, simply defend you will quickly notice, the main differ ing the d-pawn, are covered in Games ence from the variations in the rest of 48 and 49 respectively. the book is that the white c-pawn is still on c2, and yet Black is attempting to play in Benko-style regardless. Game 44 Of course, if White were to oblige O.Romanishin-R. Wojtaszek with 4 c4 we would reach the Benko European Championship, after all, specifically the 4 tDf3 line dis Dresden 2007 cussed in Chapter Six (Games 30 and 31). Here we will look at different ideas 1 d4 ttJf62 tDf3 cS for White, of which 4 i.g5 ttJe4 is the An ambitious continuation, played most important and the subject of the by Topalov and Sergey Kasparov among 1 73 Play the Benko Gambit others. Now White has to decide what other options, such as 4 ...ii.b 7, 4 ...d6 or to do with the nice central d4-pawn. 4 .. :iVb6, but I think these offer White 3 dS more chance of an advantage. With One of the two main tests of 2 ...cS. 4 ... tDe4 Black aims to gain the bishop The other, 3 c4, transposes to the pre pair (by ...tDxg S) without making any vious chapter on the (Garry) Kasparov concessions to his king side. On the Gambit. downside, the advanced knight can 3 ... bS become a target, since it cannot be supported by ...d7-dS, so Black has to play with some precision in order to equalize. But White, too, can easily end up with the worse position, so the game is nicely in the balance. S ii.h4 With this move Black gains space on the queen side and aims to get either a Benko (after 4 c4) or an improved ver sion of the Schmid Benoni, 1 d4 cS 2 dS, where a quick tDb1-C3 usually prevents ...b7- bS. The white ds-pawn is now a target. Then again, so is the bs-pawn, The first of four possibilities for which can be attacked with moves like White. The others - 5 .1i.f4, 5 h4 and 5 e2-e3, 'ii'd3 or a2-a4. 'iVd3 - are examined in the next two 4 .1i.gS games. Of these, the two bishop re The principled and critical move in treats are the most testing. the position. White threatens to com S ...ii. b7 promise the black pawn structure. In With pawn already on bS, this is the stead, 4 c4 transposes to the Benko natural development, putting pressure Gambit proper (see Games 30 and 31), on the vulnerable ds-pawn. while 4 a4 and other moves are covered 6e3 in Game 47. A modest-looking move compared 4 ...tDe4 with the alternatives, but it is quite Championed by Topalov. There are reasonable. The other two main op- 1 74 The Topa/ov Variation tions are: (14 tiJd4 0-0 15 h4 tiJcs16 hxgs hxgs a) 6 'ii'd3 and now 6 .. .fs is the typi 17 f3 eS 18 dxe6 tiJxe6is more unclear) cal reaction and the most precise in my 14 ...tiJcS 15 axbs tiJe4 16 'ii'e2 'iWxbs 17 opinion. Black simply defends the .tes .txes 18 tiJxes .txdSan d Black is knight on e4, since if White takes on bS doing fine. then the ds-pawn is hanging. Black a2) 7 g4!? (an uncompromising at would be quite content to exchange his tempt to undermine the knight; it has b-pawn for the more valuable one in been a trend for some years to play g2- the centre. g4 in many different opening varia tions) al) 7 tiJbd2 (a position ally sound way of challenging the knight on e4) 7 ... c4 8 'ili'd4'iWa s+! 9 c3 (or 9 tiJbd2 7 ... c4 (a thematic way of eliminating e6! 10 dxe6 dxe6 11 gxfs tiJxd2 12 the threat to bS) 8 'ii'd4 tiJa6 (threaten fixd2fix d2+ 13 'iti>xd2ex fs with equal ing ...tiJb4) 9 c3 (9 tiJxe4fxe4 10 'ili'xe4 ity) 9 ...b4! (Black equalizes by precise tiJb4 is unclear, as 11 l::tdl is met by play) 10 cxb4 (10 gxfS?! .txdS 11 cxb4 11 ... tiJxdS! 12 .li:txdS 'ii'as+! and ...'ii' xa2) 'ii'xb4+ 12 tiJc3 e6 13 fxe6 dxe6 14 0-0-0 9 ...tiJxd2 10 'ii'xd2 (R.Hiibner-V.Topalov, tiJxC3 15 'ii'xC3 tiJc6is slightly better for Istanbul Olympiad 2000, saw 10 tiJxd2 Black since the white king is quite vul 'i!i'b6 11 'ii'es, when 11 ... tiJC7 12 'ii'xfs nerable) 10 ...'ii' xds 11 'ii'xds .txdS 12 tiJxds would keep the tension in an :gl (not 12 gxfS? gS! 13 fxg6 .tg7,fol- unclear position) 10... 'ii' b6 11 l::.dl (in lowed by ....tx b2, and Black is clearly stead 11 0-0-0 g6 12 'iVd4?! 1:[g8 gave better) 12 ... e6 with an unclear game. Black an active position in V.Kramnik a3) 7 e3 (this is too quiet to give V.Topalov, Monte Carlo rapid 2000, a Black any problems) 7 ...c4 8 'ii'd4 'ii'as+ game which Topalov won) 11... h6 12 e3 9 tiJbd2?! (this puts the white queen at gs 13 .tg3 .tg7 and Black has equal risk, but after 9 c3 e6 10 dxe6 dxe6 11 ized comfortably. For example, 14 a4?! 'ii'dl tiJd7 Black has an active position) 1 75 Play the Benko Gambit 9 ...e6 (now Black is slightly better) 10 for White to play as well. C3?? (giving up a pawn and the centre; b} 6 a4 is one of the critical tests of White had to try 10 dxe6 dxe6 11 b4!, Black's set up. This was Burmakin's im although 11... 'ii b6 12 lbxe4 fxe4 13 provement after his loss in the above lbd2 lbc6 14 'it'c3 as is still good for game. Black) 10 ....ixdS 11 lbxe4fx e4 12 lbgs lbc613 'it'd2 .ie7 bl} 6 ... a6?! 7 'it'd3(th e insertion of 6 a4 a6 to line 'a' improves White's and Black is winning: the knight on chances substantially) 7 ...fS 8 lbbd2 gs and bishop on h4 are awkwardly 'it'as+ (unfortunately ...lba6 is impossi placed, and Black plans ...h7 -h6, ...g7 -gS ble) 9 c3 e6 (9 ...c4 10 'ii'd4 doesn't help or ...bS- b4. here) 10 lbxe4 fxe4 11 'it'xe4.ixdS 12 V.Burmakin-A.5hariyazdanov, Biel 'it'f4 lbc613 e4 .ib3 14 .ie2bxa4 15 0-0 2000, continued 14 f3 h6?! (14... exf3 15 and White was clearly better in gxf3 0-0, intending 16 e4 lbes!, looks a V.Burmakin-B.Avrukh, Ubeda 2001. simpler way to win) 15 fxe4 .ixgs 16 b2} 6 ...b4 ?! 7 'ili'd3fS 8 lbbd2is simi .ixgs .ixe4 17 .if4 eS 18 .ig3 dS 19 larly good for White. Notice that Black .ie2 0-0 20 .if3?! (20 ltf1! might offer is unable to play ...cS -c4 and ...e7-e6 more chances to hold) 20 ....ixf3 21 due to his b-pawn advance, while gxf3 d4 22 0-0 l:tad8 23 :adl 'it'b6 24 8 ....ia6 9 'ili'e3 lbxd210 lbxd2d6 11 g4! l:.del dxc3 25 'it'xc3 l:td326 'it'cl 'it'cs 27 fxg4 12 h3 g3 13 .ixg3 .ib714 lbe4led l:te2'it'd s 28 l:tef2 'iVcs 29 l:te2lbb4 30 to a crushing win for White in a3 lbdS 31 l:tdl'ii' b6 32 .ixes'ii' g6+ 33 F.5chmitt-L.Guidarelli, French Team .tg3 l:txf3 34 l:tdel 'it'e4 35 .ib8I1f xe3 Championship 2007. 36 l:txe3 lbxe3 0-1. Quite a convincing b3} 6 ..:iV as+?! has been the usual win by Shariyazdanov against a solid choice for Black. Then the obvious 7 grandmaster, which again illustrates c3?! is met by 7 ...b4! 8 1rd3 bXc3 9 bxc3 that the position is sharp and difficult (not 9 lbxC3?! lbxC3 10 bxc3 .txdS! win- 176 The Top a/ov Variation winning a pawn) 9 .. .f5 10 g4 (or 10 e3 numerous times since G.Kasparov .ta611 'iVc2 �xfl 12 l:txfl c4 13 tLJbd2 A.Miles, Basel (3rd matchgame) 1986. 'ilVxd5 14 tLJxe4- 'iVxe4 15 'iVxe4 fxe4 16 Although Kasparov won that game, its tLJd2 tLJa6 17 tLJxe4 d5 18 tLJd2 e6 and theoretical status is less clear. I would Black has the better chances in the prefer the simple 9 e3, when White is endgame) 10 ...c4! 11 'ii'xc4 was played again at least slightly better. in J.Maherramzade-S.Citak, European b4) 6 .. :�b6! (according to my data Championship, Batumi 2002, when base this is a novelty, and I think it's 11 ... g5 12 �xg5 �g7 would have given quite an interesting one given that Black a nice initiative. other moves fail to equalize): However, 7 tLJbd2! is a lot stronger, transposing to the old line 5 .. :�a5+?! 6 tLJbd2 �b7 7 a4!, long known to be good for White. Then 7 ... �xd5 (or 7 ...tLJxd2 8 tLJxd2! �xd5 9 axb5) 8 axb5 b31) 7 tLJbd2 (following line 'b3') 7 ...i.xd5 8 axb5 'iVb7! (the difference) 9 tLJxe4 �xe4 10 e3 a6 11 .:ta5 (or 11 c3 g6 12 'iVa4�g7) 11.. .f6 (this might look a bit artificial, but it shuts out the 8 ...'ifb 4? 9 c4 �b7 10 'ilc2 a6 11 %:ta4 bishop on h4 and prepares to play ...e7- won quickly in J.Nogueiras Santiago e5) 12 �d3 �xd3 13 'iVxd3e5 14 'iWb3 R.Vaganian, Mexico 1980. Of course the (not 14 O-O?? e4, winning a piece) queen has other squares than b4, but 14 ... �d6 15 0-0 i.C7 16 l:ta4 axb5 17 White gains the advantage in any case. l1xa8 'it'xa8 18 'iVxb5 'iVc6 and Black is For instance, 8 ... 'iVb6 9 e3 tLJxd2 10 fine. 'iVxd2'iVb 7 11 c4 �e4 12 �d3 d6 13 0-0 b32) 7 'iVd3 bxa4! 8 tLJbd2 (not 8 tLJd7 14 �xe4 'iVxe4 was played in 'iVxe4?? 'iVxb2) 8 .. .fS 9 tLJC4 'iVh6 10 g4!? E.Cosma-B.Chatalbashev, Metz 2002, tLJa6 11 c3 (11 gxfS tLJb4 12 'ilxe4?? and now 15 tLJg5 'iVb716 f4 would have loses to 12... �xdS) 11.. .f xg4 12 'iVxe4 been good for White. Finally, 8 ...'iVC7 9 gxf3 with an unclear position. l:ta4!? 'ii'b7 10 c4 tLJxd2 has been seen b33) 7 tLJfd2 'iVh6! 8 tLJxe4 'ilxh4 9 1 77 Play the Benko Gambit 'iWd3 lDa610 'i!Vxbs �xds l1 lDec3lDb4 clear game with chances for both sides. 12 lDxds lDxds 13 lDd2 e6 is also okay 7 ...Wa s?! for Black, since 14 'iWb7 is safely an In my opinion 7 ...'iWb 6! is more pre swered by 14 ...'i!Vd8 !. cise, when Black should equalize: b34) 7 axbs a6 8 lDc3 (otherwise 8 bxa6?! Wxb2 is good for Black; 8 Wd3 Wxbs 9 lDa3'iWb 4+! 10 c3 'iWxb2 l1 l:tbl 'i!Vxa3 12 'ii'xe4 lIa7 is unclear; while 8 lDfd2 .l1.xds 9 lDxe4 �xe4 10 lDc3 �b7 11 e4 axbs 12 l:1xa8 �xa8 13 �xbs .l1.xe4 14 0-0 �c6 15 Wd3 gives White just enough play for the pawn) 8 ... lDxc3 9 bXc3 g6 10 c4 (or 10 bxa6 �g7!) 10... �g 7 11 :bl axbs 12 :xbs WC7 with compensation. 6 ...g6 7 c3 a) 8 'ii'd3 fs 9 lDbd2c4 10 "iid4 'ii'xd4 11 lDxd4 lDxd2 12 'ittxd2 �xds 13 lDxbs lDa6 14 b3 �g7 15 bxc4 .l1.c6 was fine for Black in O.Romanishin-D.5ermek, Solin (rapid) 2002. b) 8 a4 b4 9 as (or 9 C4 .l1.g7with an unclear game) 9 ...Wd6 10 c4 e6 11 a6 �xa6 12 lDbd2 lDxd2 13 lDxd2 �g7 14 lDe4 'ii'es 15 �d3 was E.Gausel P.5vedenborg, Norwegian Team Cham pionship 2000, and now Black should have played ls... .l1.xc4! 16 �g3 'iWxb2 Blocking the long diagonal on which 17 lIbl i.c3+! 18 'ittfl �xd3+ 19 Wxd3 Black's dark-squared bishop is intend 'ilfa2 20 �xb8 exds with four pawns for ing to take up residence. Timman won the piece. a game against Topalov with this c) 8 lDbd2 (a thematic move in these move, though I don't think Black is positions) 8 ... �xds 9 lDxe4�xe4 10 a4 really worse. b4 11 lDgs bXc3! 12 bXc3 (not 12 Alternatively, 7 a4 �g7 (the most lDxe4?! cxb2 13 l:tbl "iVb4+ and Black natural move; 7 ...b4 would cede the regains a piece on the fourth rank with initiative) 8 axbs �xb2 9 lla4 lDc3 10 a clear advantage) 12 ... �c6 is unclear lDxc3�x c3+ l1 lDd2a6 12 l:ta3�b4 13 once more. d6 f6 14 dxe7 'ii'xe7 reaches a very un- S 'ii'd3 178 The Top a/ov Variation In this position Timman played 8 bxc3 I:tf7 16 'iVd2 would have given tLlbd2!? tLlxd2 (after 8 ...�xd5 ?! 9 tLlb3 White a slight but safe advantage. In �xb3 10 axb3- 'iVb6 11 1i'd3 tLld6 12 stead, the players now trade mistakes. �e2 �g7 13 0-0 White has more than enough compensation for the pawn, with the bishop pair and lead in devel opment) 9 'it'xd2 (9 tLlxd2!? is still pos sible; e.g. 9 ...�xd5 10 tLlb3 �xb3 11 axb3 'iWb612 1i'd3 a6 13 �e2 ttJc6 14 0-0 �g7 15 l:tfdl l:ta7 16 f4 0-0 17 �f2 d6 18 e4 with compensation) 9 ...�g 7 10 e4 d6 11 �d3 tLld7 12 0-0 and White was slightly better in J.Timman V.Topalov, Wijk aan Zee 2001, due to his useful space advantage. 13 •..e6?? 8 ...fS 9 tLlbd2 �g71 Afterth is Romanishin could have won with 14 dxe6! d5 (or 14... dxe6 15 �xe6+ ..ti>h8 16 0-0) 15 e7 J:.e816 1i'e4!dxc4 (or 16... ..ti>h 8 17 �xd5) 17 'ii'e6+ ..ti>h8 18 ttJe5 etc. Conversely, Wojtaszek should have played 13 ...bx C3! 14 b3 e6, when 15 dxe6 d5 16 e7 1:[e8 17 'iVe4?! is thwarted by 17 ... c2+ 18 ..ti>e2 1i'a6! 19 �xa6 �xa6+ 20 'iVd3 �xd3+ 21 ..ti>xd3 �xal 22 ]::txal ttJc6 and Black comes out on top. 14 .l:tdl?bxC3 lS 0-0 exds 10 tLlxe4 10 1i'xb5 'it'xb5 11 �xb5 tLlxd2 12 tLlxd2 �xd5 is nothing to worry about. The endgame is balanced. 10... fxe4 ll 1i'xe40-0 12 'iVC2 Not 12 �xe7? J:te8 13 ttJd2 �f8 and Black wins material, while 12 tLld2 e6 13 �e7 �xd5 14 1i'h4.uf7 15 �xc5 tLla6 16 �d6 b4! is unclear. 12 ... b4 13 �C4?! Here 13 ':cl�xd5 14 �xe7 bXc3 15 1 79 Play the Benko Gambit 16 .i.xdS+?? White's own sore spots at dS and on A further error. 16 Wh3! dXc4 17 the queen side. The trade of the bS 'iWxb7 'ili'a6 18 'ifdS+ ii'e6! 19 ii'xe6+ pawn and ds-pawns is usually in (after 19 'ili'xa8?! cxb2 20 'ifxa7 c3 21 Black's favour. Sometimes the b-pawn ii'a4 dS the black pawns are very dan can be advanced further (as in the gerous) 19 ...dxe6 20 bxc3 would have game) to soften up the white queen kept a slight advantage for White. side. 16 ....i.xd S 17 :xdS 'ii'xa2 The conclusion is that Black is okay Hitting the rook on dS as well as the after 5 .i.h4 as long he plays with preci pawn on b2. White can hardly let the sion. For the three main lines above I rook go, but after 18 ...cxb2 the passed am recommending 6 'ii'd3 fS !, 6 a4 queen side pawns will be decisive. 'ifb6! and 6 e3 g6 7 c3 ii'b6! as the best 18 'ii'e4 cxb2 19 ttJd2asl ways for Black. Game 45 E.Relange-Ch.Bauer French Championship. Marseilles 2001 1 d4 ttJf6 2 ttJf3 cS 3 ds bs 4 .i.gsttJe4 s .i.f4 .i.b7 20 .i.g3 .t:ta621 'uxcS.l:tc6 22 %:txc6 ttJxc6 23 ii'd3 ii'e6 24 ttJC4 Ilc8 2S ttJd6 ttJb4 0-1 Game summary This was a very sharp game with both grandmasters making serious mis takes, which only reflects that the posi tions are difficult for both sides. As we saw, White's play in this line generally Again this natural developing move focuses on the vulnerable pawn on bS seems best. So we have the same posi and knight on e4. In turn Black must tion as in the previous game, except play precisely to minimize these tacti that the white bishop is on a different cal weaknesses, while targeting square. White also has the same op- 180 The Topalov Variation tions as before, though there are natu 0-0-0 (11 c3 lDxdS or 11 l:tdl lDxdS! is rally some differences in the play, the fine for Black) 11...e6!? (otherwise main one being that the black knight is 11...lDxa2+! 12 �bl 'iias 13 lDes lDb4 free to return to f6, attacking dS, since 14 'iifs 0-0-0 gets at least a draw; e.g. the bishop is no longer able to take it. 15 lDf7 'iia2+ 16 �Cl C3! 17 bxc3 e6 18 6a4 dxe6 'iia3+ 19 �bl 'iia2+ etc) 12 'iies The critical move here. �xdS 13 e4 lDc614 'ifhS+ g6 15 'ifh 3?! a) 6 e3 is now well met by 6 .. :ifas+ 7 (15 'iVgs would be unclear) ls... �xe4 c3 lDf6 !, when 8 d6 (8 a4?! lDxdS 9 16 lDgS?! 'iif6! and Black was clearly �xb8 ':xb8 10 �xbS 'iiC7 11 0-0 e6 12 better in Butnorius-R.Felgaer, Gibraltar lDbd2 �e7 13 'iiC2 0-0 was good for 2008. Black in S.Yuferov-D.5ermek, Bled 1994) b2) 9 c3 lDacs?! was played in 8 ... 'iYb6 9 dxe7 �xe7 10 �e2 0-0 11 G.Mittelman-B.Avrukh, Israeli Team 'iVd3 a6 12 lDbd2 dS 13 0-0 lDbd7 gave Championship, and now 10 g4 fxg4 11 Black a comfortable game in E.5uppa lDxe4 lDxe4 12 'iVxe4 gxf3 13 'ii'xf3 V.Lazarev, Porto San Giorgio 2000. would be slightly better for White. b) 6 'iid3 is more frequently played, Avrukh concluded back that 8 ... lDa6 but after 6 ...fs Black is doing fine. was dubious, but I don't think so. It was Again Black would be more than con his next move which was not the best. tent to see 7 'iixbs �xdS, while 7 lDC3 Instead, Black should play 9 ...lDc7 ! c4 8 'iid4 'iias (8 ...lDa 6!? 9 lDxe4fxe4 10 lDxe4fx e4 11 �xC7 (11 'iixe4 lDxdS 10 'iixe4 lDb4 is also promising) 9 lDd2 would be unclear) 11...'ivxc7 12 lDgs, as lDxd2 10 �xd2 b4 11 lDdl 'iixds was in M.Hebden-lHodgson, British Cham fi ne for Black in S.Boroday-A.Zubarev, pionship, Millfield 2000, and now Poltava 2009. So play usually continues 12... 'iYb6 would lead to an unclear 7 lDbd2c4 8 'iVd4lDa6 andth en: game. bl) 9 lDxe4?! fxe4 10 'iixe4 lDb4 11 For instance, 13 'iixe4 g6 (Black 181 Play the Benko Gambit plans ...�g7, ...0-0 and ...e7-e6) 14 11dl the move 10 ...fS would lead to bal (or 14 h4 �g7 15 e3 0-0 16 hS h6 17 anced chances. liJf3 gs with compensation) 14... �g 7 15 e3 0-0 16 'iixe7 (or 16 �e2 e6 17 0-0 l::tae8 18 'ii'h4 h6) 16 ... l:tae8 17 'iVxd7 �c8! 18 'ii'c6 l:Ixe3+! 19 fxe3 'fi'xe3+ 20 �e2 'ii'f2+ 21 'ii;Jd2 'ii'f4+ 22 'ii;Jel 'ii'f2+ with perpetual check. 6 ... b4! For instance, 11 lta2 (not 11 �es?! �xes 12 liJxes 'ii'C7 and Black has a slight advantage due to the weak d pawn; while after 11 liJbd2 liJxd2 12 liJxd2 �xb2 13 l:tbl �g7 14 0-0 d6 White lacks sufficient compensation for the pawn: 15 liJf3 liJd7 16 liJgs 0-0 Tsesarsky recommends the imme 17 liJe6can be met by 17 ... 'ii'as 18 liJxf8 diate 6 ...liJf6 in his MegaBase annota ':'xf8 and ...liJb 6, when Black has a very tions, intending 7 axbs liJxdS. How dynamic position) 11...0-0 12 h4 (or 12 ever, in that case White can retreat the as �a6) 12 ...liJd6 13 �xd6 exd6 14 as bishop, chase the knight from dS and liJa6 15 'ii'd3 'ii'c8 (or lS... b3!? 16 'ii'xb3 gain a slight advantage. l:tb8 17 c3 liJC7) 16 liJbd2 l::tb8 with an 7 'ii'd3 unclear position. Black is planning Other moves: ...liJC7 and and possibly ...'ii' e8-f7 trying a) 7 c4 e6 8 'ii'd3 liJf6 9 e4 trans to win the ds-pawn, or if 17 �b3 then poses to the game. 17 ... c4! 18 liJxC4 liJcs with compensa a) 7 e3 g6! (7 ... e6?! 8 dxe6 fxe6 9 tion. liJbd2 was slightly better for White in 7 ...liJf6 8 e4 A.Rotstein-1M.Degraeve, German In this position White has more League 2001, due to his slight lead in space in the centre, but Black can neu development and the weaknesses on tralize that by a plan of ... e7-e6, ... e4xds the light squares) 8 �es f6 9 iLf4 �g7 and ... �e7, followed possibly by ... liJh s 10 �c4 was played in T.Minogina and .. .f7-fS. E.Fatalibekova, Moscow 1981, and now 8 ...e6 182 The Topa/ov Variation 12... 0-0 (12... lLlbd7 !? might be more accurate, when White must also con sider the possibility of the black king going long) 13 'iVc2 iLc8 (13 ...b3!) 14 iLd3b3 ! 15 lLlxb3 and now, rather than ls ... iLd7?? 16 lLlxcs! and White won in D.Komarov-1M.Degraeve, French Team Championship 2002, simply ls ...lLla6 followed by ... l::tb8 and ...iLd7 offers Black good compensation for the pawn. Of course White could prevent the 9 (4 pawn sacrifice with 13 b3, but then 9 dxe6 is tame and allows Black to Black has more time to organize his equalize rather easily: 9 .. .fxe6 10 lLlbd2 forces with, say, ...lLlh s/lLlg4, ...f 7-fS, iLe7 11 eS lLlhS! (the knight doesn't ...g7 -g6, ...lLld7 and ...lLle s. For example, block the bishop here and will soon re 13 ...lLlbd7 (or immediately 13 ... lLlh S 14 turn to the centre; 11 ... lLldS would be iLe3fS 15 g3 iLf6) 14 'iVc2 lLlhs 15 iLe3 less precise due to 12 iLg3 0-0 13 'iVe4!) fS 16 g3 g6 17 iLd3 iLf6 with an un 12 iLe30-0, followed by ...lLlf4 is unclear. clear game; or 14 h3 lle8 15 'ii'C2 iLf8 13 'ii'C4 fails to prevent this due to 16 iLd3 g6, and if 17 g4?! then 13 ... lLlf4 14 iLxf4 iLdS!, while 14 iLxcS 17... lLlx dS! 18 cxdS 'iVf6 19 iLg3 "ii'al+ .i.dS! 15 iLxe7 (or 15 'iVxb4 lLlc6) 20 "i!Vbl (20 lLlbl?! iLxdS) 20... 'iVC 3+ 21 ls ...'iVxe7 16 'iVd4 lLlc617 'iVe3 'iVf7 gives 'iVc2 (21 iLc2?! iLxdS) 21... 'iWal+ with Black very good play for the pawn. perpetual check. 9 ...exd s 10 exds d6 11 iLe2 11••• iL e7 12 0-0 0-0 13 lLlbd2Yz -Yz More challenging is 11 lLlbd2 iLe7 12 0-0-0!? Here the game ended in a draw. 13 ...lLlh S 14 iLe3 fS 15 lLlb3 g6 16 iLh6 183 Play the Benko Gambit l:[eS 17 as ..tf6 could be a possible con but this is not really all that threaten tinuation with a roughly equal posi ing, as long as Black develops rapidly tion. and fights for the initiative. The final option is S 'iVd3, hitting Game summary the knight on e4 and pawn on bS. The S .i.f4 is one of the critical responses to drawback to this move is that the ex Black's ambitious play with 4 ...liJe4. Th e change on gs leaves the knight is one drawback of this bishop move is rather misplaced. White can try to jus that it allows the black knight to return tify it with ideas of 'iVf3 or liJxh7, or to f6 unmolested. After ...liJf6 and ...e7- sometimes 'iVC3 (after ... g7-g6), but e6, followed by ...e6xds, ...d7-d6, .....te7 Black can equalize without too many and ...0-0, with the further idea of problems. ...liJh S and ...f 7-fS, Bauer was able to After S ...liJxg s 6 liJxgs g6!? (the equalize the game and a draw was most ambitious move; 6 ... e6 is a safer agreed. option, since the tactics don't work for White; e.g. 7 liJxf7?! 'iVf6 ! S liJxhS 'iVxb2 9 'iWC3?? 'ii'Cl mate, or similarly 7 Game 46 liJxh7?! c4 S 'iVf3 llxh7 9 dxe6 'ii'f6 ! 10 Z.Kozul-V.Topalov exf7+ 'ifi>dS 11 'iVxaS?? 'iVxb2), there are Istanbul Olympiad 2000 several interesting lines to consider: 1 d4 liJf6 2 liJf3 cS 3 ds bs 4 .i.gs liJe4 s h4 a) 7 'iWxbS?! .i.g7 gives Black more than enough compensation, since it is quite difficult for White to defend b2. This continuation hasn't found For example, S c3 (or S liJd2 liJa69. c3 many followers, since Black can equal :bS 10 'iVd3 l:txb2 11 liJC4 :bS and ize rather easily. The idea is to open the Black was fine in J.Bonin:M.Molner, US h-file the rook after the likely ...liJxg s, Chess League 2007) S ...liJa6 9 'iVa4 l:tbS 184 The Topa/ov Variation 10 'ii'c2 'ii'b6 11 b3 ttJb4 12 'iVd2 fS with when Black's superior structure and excellent play for the pawn. bishop pair provide full compensation b) 7 'fic3 !?- (Vaganian) 7 .. .f6 8 ttJf3 for the pawn . .1g7 9 e4 (or 9 'fixcS?! fS 10 c3 ttJa6 11 d) 7 e4?! was played in V.Golod 'ii'e3 0-0 with compensation) 9 ...�6 10 V.Topalov, European Cup, Bugojno a4 bxa4 with an unclear game. 1999, which continued 7 ....i.g 7! 8 c3 (8 c) 7 �f3 !? should be met by fS! (the 'iVxbS?! e6 9 ttJf3 exds 10 exds 0-0 idea was 7 .. .f6 8 ttJe6! - Avrukh) and would allow Black a very strong initia then: tive) 9 ... 0-0 (White is already somewhat worse) 9 .i.e2 e6 10 ttJf3 exds 11 'iVxds (or 11 exds .l:le812 0-0 d6 with an edge for Black) 11 ... 'fib6 12 ttJbd2 (not 12 Wxa8?? .i.b7an d the queen is trapped) 12 ... .i.b7 13 'irb3 c4 14 'iVdld6 15 a4 a6 16 axbs axbs 17 llxa8 .i.xa8 18 0-0 ttJd7 Cl) 8 'ii'C3 ':'g8 9 'iVxcs ttJa6! (9 ...e6 10 "iie3 is good for White) 10 �e3 (or 10 "ii'xbs l1b8 11 'iVd3 l:txb2 12 c3 ttJcS 13 "iie3 'fib6 14 ttJd2 h6 15 ttJgf3 .1g7) 10 ...h6 (or 10 ...'ifb 6!? 11 'iVxb6 axb6 intending ...ttJb4 or ....i.g7) 11 ttJf3 .i.b7 12 d6 e6 13 ttJC3 'iVas and Black has excellent counterplay with moves like and Topalov had a clear advantage, ...bS -b4, ...ttJb4, ....i.g7 an d ...':c8. with the bishop pair, more space, and (2) 8 d6!? ttJc6 9 'iVds e6 10 ttJxe6! targets on e4, b2 and C3. dxe6 11 'iVxc6+ .i.d7 12 'iVxcs 'ifb6 The game went on: 19 l:.el ttJcS 20 (12 ....i.g 7!? 13 c3 :c8 also gives Black a .1fl 11e8 21 'iVc2h6 22 g3 .1b723 'ifbl surprising amount of play for the hS 24 h4 'iVc6 25 .i.g2 ttJd3 26 lIfl?! pawns) 13 'iVc3 (13 'iVxb6 axb6 hardly (the rook is passive here; but 26 l:te3 lessens Black's initiative at all) 13 ...:g8 .i.h6 and 26 l:te2 b4 are also very good 14 e3 .i.xd6 15 ttJd2 :c8 16 'ifb3 �e7 for Black) 26... dS (26 ...b4! looks even 17 0-0-0 .i.c6 18 �bl .1dS 19 'iVxbs stronger) 27 b3?! (27 ttJel ttJcS 28exds "iixbs 20 .i.xbs .i.xg2 21 .:thgl .i.ds, 'ikd7 29 d6 .i.xg2 30 ttJxg2 would have 185 Play the Benko Gambit been more tenacious) 27... Wc S 28 exds Alternatively, 9 iLxbS l:tb8 10 a4 iLxdS 29 bxc4 bxc4 30 lLlgs .i.xg2 31 .i.xb2 11 .l:tbl (after 11 lLlC4 .i.xal 12 �xg2 WdS+ 32 lLlgf3 iLxC3 33 'iVC2 'ikxal 0-0 13 'iVes d6 14 'iVh 2 hS! 15 .i.xd2 34 'iWxd2l:les (the knight on f3 is gxh6 lLlC7 or13 d6 exd6 14 lLlxd6 .l:.b6 lost) 35 l:.al l:.fS 36 'ili'e3 lLles 0-1. A 15 'ii'es 'ii'C7 16 'ii'h2 hS White lacks convincing win by Topalov. compensation for the exchange) s ...g6 11... .i.g7 12 0-0 (or 12 d6 0-0 and Black This time S ... .i.b7 is not quite as ef is fine) 12 ... 0-0 13 l:tel d6 is roughly fective, as after 6 .. .fS 7 lLlbd2 c4 (or if equal. Black might follow with ... lLlC7or 7 .. :ifas 8 c3 e6, as in H.Obermeier ...lLlb4. H.Teske, Bad Wiessee 2009, then 9 9 ... .i.xb2 10 axbs lLlC7 11 .l:.a2 .i.g7 12 lLlxe4fx e4 10 'iVxe4.i.xdS 11 'iVf4! with c4 d6 13 'ii'b3 an edge for White) 8 Wd4 lLla6 9 lLlxe4 After 13 .i.d3 e6 14 0-0 Black takes fxe4 10 'ii'xe4 lLlb4 11 l:tdl, the usual control of the h-file by 14 ... h6! 15 gxh6 11 ...lLlx dS? fails to 12 l:.xdS 'ikas+ 13 .l:.xh6 15 .l:lel�f8, as in the game. In �dl! e6 14 'iVf4! iLxdS 15 lLles and stead, Tsesarsky suggests 13 est?, when wins; while 6 ...lLlxg s 7 hxgs c4 8 Wd4 13 ...dx es (or 13 ....i.fs 14 'iWa4) 14 lLle4 e6 9 e4 also offers White a slight ad .i.fs 15 lLlxcs 'iid6 16 lLla6 would still vantage, as the rook on h8 and bishop be unclear. on f8 are not that easy to activate. 13 ... e6 14 .i.d3 �f81 6lLlbd2 Of course the Trompowsky idea of 'ii'd3-C3 is not possible here, since 6 'ii'd3 lLlxgs 7 'ii'C3?? just loses a piece to 7 ...lLlxf3 +. 6 ...lLlxgs 7 hxgs .i.g7 8 e4 lLla6 Topalov decides not to castle, since the king on f8 is less badly placed than the king on el; while if Kozul castles short, then ... h7-h6 becomes possible again, with chances for an initiative on the h-file. lS :a3 exds 16 exds a6 17 0-0 186 The Topa/ov Variation Rather obliging, but White had 22... �f 6?! nothing better at this point. 17 b6?! A mistake in turn, which allows would be met by 17... 11b 8 and 18... ttJa8 White a surprising tactic (see the next winning the pawn for nothing. note). Instead, 22 ...�x f3 23 ttJxf3 axb5 17... !:tb8 18 l:tel h6! 19 gxh6 :xh6 20 24 cxb5 �f6 would have been good for ttJe4 11b621 'ii'a2!? �g4 Black. After 21 ...axb 5 22 lIa7 White gets 23 'iVd2 l:thS some counterplay; for instance 22 ... b4? loses to 23 'ii'a5! lIb7 24 ]::ta8, threaten ing ttJxd6. 22 ttJegs?! The exchange sacrifice 22 l:txa6! ttJxa6 23 bxa6 was stronger, when the passed pawn gives White sufficient compensation. For example: 24 'iWas?! Here White could have played 24 ttJe6+! fxe6 25 �xg6 lir.h8 (or 25 ...'iti>g 7 26 it.xh5 �xh5 27 bxa6 exd5 28 ttJh2! and the a3-rook swings across to the kingside) 26 'ikf4 with an unclear game. The alternative 24... ttJxe6 25 dxe6'iti>g7 26 exf7 'ii'h8 27 'iti>f1 is actually better a) 23 .. .lIb8 24 a7 lla825 'ii'a6 �xf3 for White, as the black king is more vul 26 'iVxd6+! 'ii'xd6 27 ttJxd6 (threatening nerable; for example, 27... :h 1+ 28 'iti>e2 lIe8+!) 27 ... g5 28 ttJb5 �xd5 29 cxd5 l:txe1+ 29 'iti>xe1 axb5 30 cxb5 d5 31 J:r.a6 30 ttJC7l::t 6xa7 31 ttJxa8':'xa8 with f8'ii'+! 'ii'xf8 32 lta7+'iti>g8 33 �xg6. equality. 24... ttJa8 2S ttJe4? b) 23 ...�xf3 24 a7! �xe4 (or A third mistake. The only chance 24 .. :iVa8 25 'ii'a5! l:tb2 26 gxf3 �d4 27 was 25 l:.e4! and if 25 ...�f5 then 26 g4! J:r.e2':'xe2 28 �xe2 'iti>g7 29 �f1 l:lh8 30 �xe4 27 gxh5 �xd3 28 ttJh7+ 'iti>g7 29 'iWC7) 25 a8 Q l:tb8 26 'ii'xb8 'ii'xb8 27 ttJxf6 'ii'xf6 30 l:txd3 (Tsesarsky) re it.xe4 �d4 28 g3 'iti>g7 when Black is mains unclear. perhaps slightly better, but a draw is 2S ...�xf3 26 gxf3 �d4 the most likely result. Now Topalov gets a decisive attack. 187 Play the Benko Gambit 5 ...liJxg 5 6 liJxg5 and then either 6 ...e6 or, more ambitiously, 6 ... g6!?, when Black has a comfortable game, espe cially if White doesn't reply with 7 'ii'C3 or 7 'ii'f3. Game 47 R.Vag anian-V. Topalov Istanbul Olympiad 2000 27 'itfl?1 1 d4 liJf6 2liJf3 cS 3 ds bs 4 a4 Making it easy. 27 'itg2 'lid728 Game summary With 5 h4 White aims to get an unbal With this move White seeks to clar anced position, giving up the bishop ify the situation on the queen side be pair to open the filefor the rook on hl fore deciding what to do next. Apart and keep its counterpart on h8 out of from 4 ..tg5 and 4 c4, White has a cou play. Ironically, in the game it was ple of further tries, but neither of them Black who attacked down the h-file af are very ambitious and Black equalizes ter Kozul later castled short, while To rather easily: palov played ...'it>f8 and ... h7-h6. Per a) 4 g3 i..b7 5 c4 e6! (5 ... g6 6 i..g2 haps deterred by the result, few other bXc4 7 liJC3..tg 7 8 0-0 0-0 9 liJe5 d610 players have tried 5 h4, but the game liJxC4 liJbd7 11 .l:[el ..ta6 12 'ii'a4 'ii'c8 was actually unclear for a long time 13 liJa5 liJb6 14 'iVh4 lle8 15 ..tg5 fiC7 and Topalov only won after Kozul went 16 liJc6 i..b7 17 e4 liJbd7 18 f4! was wrong in the middlegame. somewhat better for White in The alternative variation seen in the S.Johannessen-R.J.Fischer; Havana notes, 5 'ii'd3, can be answered by Olympiad 1966) 6 dxe6 (or 6 cxb5 188 The Topa/ov Variation tbxds) 6 .. .fxe6 7 cxbs i.e7(after 7 ...ds?! lead in development. 8 ..th3! 'iVb6 9 tbgs is a bit annoying) 8 c) 5 tbbd2 is the sharpest move and .i.g2 0-0 9 O-O'was played in G.Marras requires a precise response: s ...tbxds 6 R.Bernardi, Palau 2009, and now 9 ... a6 a4 c4! 7 c3 e6 8 e4 tbf6 9 tbd4 ds is (or 9 .. :ifb6 10 tbc3 a6) 10 bxa6 tbxa6or comfortable for Black. 10 ... l:Ixa6 would give Black standard d) 5 i.d2(th e most common reply in and good Benko compensation for the practice) s ... b4 and now: pawn. b) 4 e3 can also be met by 4 ...i.b7, but I like 4 ..:ii' as+!? dl) 6 c4 e6 (this break is both the matic and logical here; 6 ... g6 7 i.d3 i.g78 a3 0-0 9 0-0 tba6 10e4 d6 is also This check can often be annoying fine for Black) 7 dxe6?! (7 'iVc2 would be for White when the ds-pawn is vulner more consistent) 7 ...dxe6 8 i.d3 i.b7 9 able, since all responses mean placing 0-0 tbbd7 10 llel i.d6 11 e4 tbes 12 the pieces in a less than optimal way. i.gs ':'d8 was good for Black in a) 5 tbC3 is the natural square for R.5eifarth-T.Zoltek, Pardubice 1992. the knight, but it can be chased away; d2) 6 i.d3 i.a6?! (6 ...tbx ds! would i.e. s ...b4 6 tbe2 i.b7 7 tbf4 e6 8 i.C4 give White sufficientcompen sation, but fiC7inten ding ... i.d6 is unclear. no more than that) 7 'ife2?! (7 c4! would b) 5 'iVd2 (a rather passive move) be slightly better for White, since s ...b4 6 c4 (or 6 i.d3 i.b7 7 e4 c4 8 7 ...tbxds is now met by 8 i.e4! tbb6 9 i.xc4 tbxe49 'iVd4tbf6 10 tbbd2'iVb6 is i.xa8tbxa8 10 b3 g6 11 a3 i.g712 l:[a2 also unclear) 6 ...e6 7 b3 i.b7 8 i.b2 i.b7 13 0-0 tba6 14 i.cl and Black exds 9 i.xf6 gxf6 10 cxds i.g711 a3 fs doesn't have enough for the exchange) 12 lla2 i.O 13 tbxC3 bxc3 14 WCl 7 ...tbxds 8 e4 tbC79 0-0 g6 10 tba3 i.g7 i.xds 15 :C2 i.xb3 16 l:txc3i.e6, wh en 11 tbC4?! i.xc4 12 i.xC4 i.xb2 13 l:tabl Black's extra pawn is balanced by his i.g7 was also good for Black in slightly worse structure and White's E.5osulin-V.Krasnov, Donskoj 2002. 189 Play the Benko Gambit 4 ... b4! ship, Chartres 2005, saw 6 b3 i.g7 7 �b2 0-0 8 g3?! (it wasn't yet too late for 8lDbd2) 8 ...e6! This is the best move in this position and leads to an unclear game. Second best, in my opinion, is 4 ...'ii a5+ with (now this break is quite effective chances for both sides, though I'm not since White is unable to support the saying any more about that. After 4... b4 centre with e2-e4) 9 �g2?! (here 9 dxe6 Black gains space, reinforces his control probably offers more chances of keep of the long dark-squared diagonal in ing the balance) 9 ... exd5 10 cxd5 !le8 co-operation with his (soon to be) fi (now Black has a kind of improved anchettoed bishop, and takes the c3- Modern Benoni and a slight advantage) and a3-squares away from the bl 11 0-0 �a6 (the bishop is well placed knight. The one drawback is that White on a6, exerting pressure on the light now has access to c4, which might be a squares) 12 J:1el lDe4! (12 ... d6 13 lDfd2 nice square for a knight. lDg414 i.xg7 c;txg7 15 lIa2 "iff6 16 lDf3 5�g5 lDd7 17 lDbd2 'ili'e7 18 lDfl f6 19 .l:!.c2 Threatening to compromise Black's lDge5 20 lDe3fS 21 h4 was less clear in pawn structure with �xf6. This follows A.Bagheri-J.Lautier, French Champion the example of the previous games in ship, Chartres 2005) 13 �xg7 �xg7 14 this chapter, and we will continue in lita2 'ii'f6 15 l:tc2 d6 16 e3 lDd7 17 lDh4 the same fashion with 5 . ..t2Je4. c4! 18 bXC4?! (but if 18 �xe4 .l:!.xe4 19 White's other main option is 5 c4, lDd2, then 19 ...lIx h4! 20 gxh4 c3 21 bringing about a familiar Benko struc lDe4 "ifxh4 and Black is clearly better) ture from Game 35. Indeed, 5 ... g6 (but 18 ...b3 19 lIcl 'ii'xf2+ 20 c;thl lDdc5 not 5 ... bxC3?! 6 lDxC3 and White is gave Black a great position in slightly better) 6 lDbd2 �g7 7 e4 d6 S.suvrajit-J.Deepan Chakl 1 9 0 The Topa/ov Variation 6g3?! Cl) 7 'iVd3 liJf6 8 liJbd2 .i.a69 liJC4 Not really a very good move. In d6 10 e4 .i.g711 .i.e2 0-0 12 as liJg4 13 stead, White should investigate the 'i¥b3 liJd7with an unclear game. various continuations from Games 44- C2) 7 .i.es f6 (7... liJf6 , allowing 8 46 and see how the inclusion of the .i.xf6, would be inconsistent) 8 .i.f4 two extra pawn moves affects things. .i.g79 liJbd2 fs!? (9 ...liJxd2 10 'iixd2 fs In those games Black did not always is also fine) 10 liJxe4 fxe4 11 ..tes 0-0 answer a2-a4 with ...bs -b4, whereas 12 .i.xg7 xg7 13 liJd2 e3! 14 fxe3 e6 here he is already committed, but apart and Black has good play for the pawn. from that Black has more leeway, since c3) 7 liJbd2 liJxd2 8 'iWxd2 ..tg7 9 the b-pawn is no longer vulnerable. In l1bl (or 9 c3 bxc3 10 bxc3 'iVas) 9 ...0-0 particular he can forego ....i.b7 in fa 10 .i.h6 d6 11 h4 (11 ..txg7 xg7 12 e4 vour of ...g7-g6. Then: .i.g4 is equal) 11 ....i.xh 6 12 'iVxh6 f6 13 a) 6 h4 g6 7 liJbd2 liJxgs 8 hxgs .i.g7 e4 'iVe8 and Black is fine. 9 liJC4d6 is fine for Black. d) 6 ..th4 g6! (6 .....tb 7?! is wrong b) 6 'ii'd3 liJxgs 7 liJxgs g6 8 'iVf3 fs! here - see the notes to Game 44) and: (8 .. .f6, as in M.Hell-P.5ahm, Volklingen 2003, could still be met by 9 liJe6!) 9 d6 liJc6 10 'iVds e6 11 liJxe6 dxe6 12 'i¥xc6+ .i.d7 13 'iixcs .i.g7 (or 13 .. :iVb6!?) 14 c3 as and with a lead in development and ideas of ....:.c8 and ...0-0, or 15 e4 'iWgs!, Black has very good play for the pawns. c) 6 .i.f4 g6! (otherwise 6 ....i.b7 transposes to Game 45) and now we have: 1 9 1 Play the Benko Gambit dl) 7 'iVd3liJd6 (now the main ideas have anything on the light squares to are ... liJfs, ... .lta6 and ... .ltg7; Black is compensate for the bishop pair. For comfortable) 8 c4?! (but after 8 liJbd2 example, 8 .ltg2 .ltg7 (8 ... d6?! allows 9 .ltg7 9 Itbl .lta6 10 'iVe3 liJfs 11 'ii'xcs liJe6!) 9 d6 (or 9 c3 as 10 d6 lita6) liJxh4 12 liJxh4 "iVb6 13 'iVxb6 axb6 14 9 ... liJc6an d Black is doing fi ne. b3 bS 15 axbs .ltxbs Black has active 7 .ltg2d6 8 .ltcl?! play for the pawn) 8 ....ltg 7 9 .liIa2 b3! 10 'iVxb3 liJa6 11 liJbd2 .l:tb8 was good for Black in M.Tratar-A.Escobar Foreno, World Student Team Championship, Paranana 1993. d2) 7 liJbd2 liJxd2 8 'iVxd2 .ltg7! (8 ...d6 9 e4 .ltg7 10 .ltbS+ liJd7 11 c3 a6 12 .lte20-0 13 0-0 as 14 'iVc2 liJb6 15 liJd2 bxc3 16 bxc3 'iVC7 17 litfbl was good for White in A.5ultanov V.Grechihin, Russian Championship, Samara 2000) 9 c3 (9 d6?! is nothing to Vaganian decides to keep the fear: 9 ... .ltxb2 10 .ltxe7 "iVb6 11 'iVf4 bishop after all, but he will have severe .ltC3+12 �dl fS 13 litbl liJc6 and Black trouble getting it out again later. is better) 9 ...bxC3 10 bXc3 'iVas 11 .l:ta3 8 .••.lt g7 9 liJfd2 liJxd2 10 liJxd2 liJd7 11 .lta6 12 e4 .ltxfl 13 �xfl d6 14 �gl 0-0 0-0 12 lite 1 as liJd7 15 h3 c4 16 �h2 liJcs with a com Consolidating the queen side and al fortable games for Black. lowing the knight to go to b6 unmo lested. 13 e4 13 liJC4 would just be met by 13 ....lta6. 13 ...liJ b6 14 .l:i.a2 Preparing to let the dark-squared bishop out again, so Black must strike quickly. 14.•. e6! An often occurring move with the knight on b6. Here it is even more ef 6 ...g6 fective, since ....ltxe6 threa tens the rook 6 ...liJxg S! 7 liJxgs g6 seems more on a2; for instance, 15 dxe-6 .ltxe6 16 b3 precise to me. White doesn't really .ltC317 .ltb2? runs into 17 ...liJXa 4!. 1 9 2 The Topa/ov Va riation A desperate attempt to free his po sition. White might have tried prepar ing this with 26 ttJdl, when 26 ..J:te8 27 c4 .lir.el+ (or 27... bxc3 28 'iVxc3) 28 .tfl .ta6 29 .:te2! l:txdl 30 J:.e6 'ii'al 31 'ii'h3! would be unclear, but after 26... .tx dS! 27 'iVxdS+ ttJxdS 28 .txdS+ 'iti>g7 29 .txa8 'iVd4! the queen triumphs over the uncoordinated white forces; e.g. 30 ttJe3 (or 30 .tf3 'iVxf4 31 'iti>g2 'iVcl and Black wins by advancing the pawns 15 b3?! exd5 16 exd5 .tb7 17 ttJfl .:te8 since White has no moves) 30 ...'ii' xf4 31 18 Iixe8+ 'iVxe819 h4 'iti>fl 'iWd4 32 .tf3 f4 33 ttJC4dS 34 ttJxas Unfortunately, White can't play 19 'ii'd2, when ...'ifc l-bl wins the rook. .tb2 because of 19 ....txb2 20 'iVxb2 26 ... bxC3 27 l:le2 'iVes, winning the ds-pawn. The text 27 1:[c2lIe8 is little better. prepares the development .tf4 and 27 ... .ta628 ttJC4ttJXC4 29 bXC4 .l:lb8 ttJe3 by preventing ...g6 -gs, but also weakens the king side. 19 ...f5 20 .tf4 .te521 ttJe3.txf4! Topalov is only relinquishing control of the long diagonal temporarily, as his queen will soon take up residence there instead. 22 gxf4 'iVe7 23 h5 'iVh4 24 hxg6 hxg6 25 'iVf3 'iVf6! Black has a decisive advantage. The c-pawn is very dangerous and the white king is vulnerable. Vaganian's next move allows Topalov to end things quickly. 30 lite6?'iVx e61 Only a temporary queen sacrifice as Black will soon get it back. 31 dxe6 c2 32 'ii'e3 l:tbl+ 33 'iti>h2 Cl'iV 26 c4 34 'iVg3 'iti>g7 35 'iVh4'iVgl+ 36 'iti>h3 0-1 1 93 Play the Benko Gambit Game summary and .i.b2 is a reasonable try for an ad Black got the upper hand from the vantage. opening phase afterVag anian wasted 2 e 3 liJf6 3 liJf3 g6 time with his dark-squared bishop. To palov's realization of his advantage in the middlegame was very instructive and he won convincingly. White would do better to retreat the bishop on move 6, either to h4 or f4, following the example of Games 44 and 4S. However, Black can get a com fortable game by discarding an early ....fi.b7 in favour of ...g7 -g6 as examined in the notes. In short, Black has nothing to fear after4 a4 b4!. 4.i.e2 White has also tried: a) 4 .i.d3 dS S c3 (very passive: the Game 48 Colle set-up is not very effective D.Brumen-I.Smirin against ...g7 -g6, so Black should be Croatian Cup, Rabac 2004 happy if anyone plays like this) S ...liJbd7 6 liJbd2 .i.g7 7 0-0 0-0 8 l:tel ld4c5 (not S liJes? liJxes 9 dxes liJg4 and Smirin probably played this way to Black won a pawn in F.Wegerer-M.Roiz, divert Brumen from his usual Trom Oberwart 2004) 8 ... b6 (or S ... lIeS 9 e4 powsky (1 ...liJf6 2 .fi.gS). The recom cxd4 10 cxd4 dxe4 11 liJxe4 liJxe4 12 mended move order in this book is 1 d4 .fi.xe4 liJf6 with an equal game) 9 e4 liJf6 2 liJf3 cS, when 3 e3 transposes to cxd4 10 cxd4 (not 10 liJxd4?! liJcs 11 the game. If somebody plays 3 e3 one .i.c2?! dxe4 12 liJxe4?? liJfXe4 13 .i.xe4 can probably assume they are rather liJxe4 14 l:txe4 eS 0-1 A.lbanez Lozano unambitious or else want to avoid a R.Cifuentes Parada, Dos Hermanas debate on opening theory. 2004) 10... dxe4 11 liJxe4 .i.b7 12 liJc3 Similarly, 3 b3 is not the most ag liJdS 13 ii'e2 f6liJ7 14 liJxdS liJxdS and gressive continuation. Black replies Black was slightly better in P.short 3 ...g6 and often ...cs xd4 and ...d7-dS (cf S.Conquest, Irish Championship, Dublin 4 b3 below). The queenside fianchetto 2007, as the isolated queen's pawn is a makes more sense to me when Black long-term weakness. has played ...e7-e6, aiming for a Nimzo b) 4 b3 .i.g7 S .i.b2 0-0 6 liJbd2 (6 Indian or Queen's Indian, when b2-b3 .i.d3 allows 6 ...cxd4 7 exd4 liJdS! 8 0-0 1 94 The Top alov Variation ttJf4 9 l:telttJxd3 10 'ilt'xd3 d6 11 C4 ttJc6 ble, trying to open up the game with a 12 h3 d5 and Black was better in well-timed ...d6-d5 or ...a7-a6 and ...b6- R.Wiemer-F.Gheorghiu, Remscheid b5) 17 ttJela6 18 �f3 ':'c7 19 h3 l:[dc8 1986) 6 ...cxd4 7 exd4 ttJc6 8 a3 d5 9 20 .i.xb7 'ii'xb7 21 1:.c2 ttJe8 22 �xg7 .i.d3 ttJh5 10 g3 was seen in ttJxg7 23 l:tdcl ttJe6 24 'ilt'a2 ttJe5 25 V.Vainshtein-Z.Gofshtein, Israeli 'ilt'b3 ttJd7 26 'ilt'd3 ttJe5 27 1i'b3 b5! 28 Championship, Ramat Gan 1992, and cxb5 axb5 (this structure is fine for now 10 .. :iWb6! 11 c3 l:te8 would have Black) 29 l:txC7 l:txC7 30 ':xC7 ttJxC7 31 given Black the edge. ttJef3ttJxf3 + 32 ttJxf3 'ilt'c6 c) 4 dxc5!? (White can't keep this pawn, but he does gain a bit of time to seize some space on the queen side af ter Black recaptures) 4 ...'iIt'a 5+ 5 ttJbd2 'i!i'xC5 6 a3 .i.g7 7 b4 'fIC78 �b2 0-0 9 c4 b6 10 �e2 .i.b7 11 ':cl d6 12 0-0 ttJbd7 13 11fh3 11ac8. (the endgame is favourable for Black, who has the potential to create a passed pawn in the centre, whereas it is difficult for White to advance on the queenside) 33 ttJd4 'ili'cl+ 34 rJi>h2 e5 35 'iVc2?! (exchanging queens only makes things worse for White) 35 ... 'iVxc2 36 This is actually a standard position ttJxC2f5 (the threat is now clear: ...<3;f7- for the 4 dXc5 variation. White has a bit e6-d5-c4-b3) 37 rJi>g3 ttJd5 38 f4 r3;f7 39 more space, as intended, while Black r3;f3 ttJc3 (preventing both e3-e4 and has an extra centre pawn and well a3-a4) 40 fxe5 dxe5 41 ttJal <3;e6 42 placed pieces; chances are TOughly ttJb3 <3;d5 43 ttJC5?! (43 ttJd2 was a bet equal. We'll have a look at one game to ter try, intending 43 ... ttJe4 44 ttJxe4 see how play might continue: 14 ':'fdl fxe4 45 'it>g4 <3;C4 46 <3;g5 <3;d3 47 a4!, (or 14 ttJd4 a6 15 .i.f3 ttJe5 16 .i.xb7 though the resulting queen endgame 'ii'xb7 with an equal game) 14 ... 1Ifh815 would still be lost) 43 ...e4+ 44 <3;f2 <3;C4 'i!i'a2 'fIa8 16 'ilt'al l:tfd8 (Black has ar 45 g4 ttJbl 46 gxf5 gxf5 47 <3;g3 ttJxa3 ranged his pieces as actively as possi- 48 <3;f4 <3;xb4(now the b-pawn decides 1 95 Play the Benko Gambit the game) 49 iDe6 4 •...i.g7 S 0-0 0-0 6 c4 cxd4 7 exd4 d S because the d-pawn would drop off. Black has equalized. In fact, it is White who has to play accurately to maintain the balance. 10 �e3 b6 I like this idea ...b7-b6 and ...iDxC 3, trying to play against the weak c3- and a2-pawns. 11 cxb6 iDxc3 12 bxc3 axb6 13 'ii'd2 .i.g414 l:.fbl iDas Now White has a reversed Queen's Gambit Tarrasch with an extra tempo, but it is not really of much help to him since the play is primarily strategical in nature. 8cs Alternatively, 8 iDc3 iDc6 9 iDes?! (but 9 h3 dxc4 10 .i.xc4 iDa5 11 .i.d3 .i.e6 12 'ii'e2 :'c8, as in J.Levitt C.Beaumont, British League 1999, is ls iDes?! fine for Black with ideas like ...�c4, It is a strategic mistake to exchange ...iDd5 and ... iDc6) 9 ... dxC4 (now it is the light-squared bishop, as it only White who has to prove that he can makes the c4-square even more attrac equalize) 10 iDxc6 bxc6 11 .i.xC4 'ii'd6 tive for the black knight. 12 .i.e2?! (12 h3 was a preferable way lS ....i.xe2 16 'ii'xe2 f6 17 iDf3 'ii'd6 18 to prevent ...iDg4 , though Black still has "iWbS?! a comfortable game with long-term Ceding the c4-square straightaway. 1 9 6 The Topa/ov Variation 18 tDd2 �fc8 19 'iVd3 :c6 20 %:tb2woul d ter of a reversed Queen's Gambit Tar have been more tenacious. rasch, where White's extra tempo was 18 ...tDC4 19 �C1 'iWe6 20 �b3 :as 21 not enough to outweigh the long-term i.f4 'iWfs strategic weaknesses in his position: the queenside pawns. White might hold the balance with active play, but he cannot really hope for more than that. As it turned out Smirin scored a convincing win in a rather one-sided game. The various alternatives for White given in the notes are hardly any more promising either. Game 49 22 i.e3?1 P.Velikov-M.Tal This just creates another target on Berlin 1986 e3. 22 i.g3 was correct, followed by l:[dl and tDd2, when White still has 1 d4 tDf6 2 tDf3 cS 3 c3 some drawing chances. This rather diffident move is often 22 ..•:a3 23 �b4 :fa8 24 1If1 played when White is aiming for a solid Hoping for 24 .. J:lxa2? 25 ':'xa2 ':'xa2 queen's pawn opening, such as the 26 'iVxe7, but that is easily dealt with. Torre, Colle, or London systems. 24.. :iVd 71 3 •••e6 Now White can't help losing a pawn and the result is virtually assured. 2S tDd2 tDxe3 26 fxe3 .l:r.xa2 27 ];txa2 l:txa2 28 c4 i.h6 29 :f3 'iVe6 30 'i!i'C3 dXC4 31 l::th3? Hastening the end. 31.•. ..t>g7 32 tDxC4'i!i'g4 0-1 Game summary Brumen played unambiti ously in the opening with e2-e3 and i.e2, when Black equalized rather easily by fi The most popular reply: White is anchettoing his king's bishop. The building a rock-like foundation on the game subsequently took on the charac- dark squares as a prelude to later ac- 1 97 Play the Benko Gambit tion, so Black adopts a counter-strategy tiJxds was good for Black in G.Kamsky on the light squares, intending to fol M.Carlsen, Moscow (blitz) 2008) ls ...h6 low with ...b7-b6, .....tb7 an d .....te7. 16 tiJge4tiJxe4 17 ..txe4..txe4 18 tiJxe4 4..tgS "iVc6 19 tiJd2 ds 20 ..tf4 bs with easy This takes the game into Torre terri equality in K.Arke11-Hmms, Southend tory. Black can play in the same way 2009. against other systems: 4 .....t e7 5 e3 a) 4 e3 (Co11e) 4 ...b6 5 ..td3 ..tb7 6 Here 5 ..txf6 ..txf6 6 dxcs!? tiJa6(or tiJbd2 ..te7 7 0-0 0-0 8 'ilke2 (or 8 e4 6 ...b6 !? 7 cxb6 axb6 8 e3 0-0 9 ..td3 ds cxd4 9 cxd4 ..ta6 with equality) 8 ...tiJc6 10 0-0 tiJd7 with compensation arising (alternatively, 8 ... cxd4 9 exd4 d6 10 .l:[el from Black's bishop pair and more tiJbd7 and 8 ...ds 9 tiJes tiJbd7 10 f4 space) 7 e4 (7 b4? just a110ws 7 ...tiJxb4 !) tiJe4! are also perfectly sound) 9 e4 tiJxcs8 es ..te79 b4 tiJe410 ..td3ds 11 cxd4 10 tiJxd4 (10 cxd4 a110ws 'ilfe2 as 12 bS tiJxc3!? (otherwise 10 ... tiJb4!) 10 ... tiJes 11 ..tc2 'ili'c8 12 f4 12 ... tiJcS is equal) 13 tiJxc3 ..tb4 14 l:1cl ..ta6 13 'ilfdl tiJc6 14 l:If3 g6 with an d4 15 0-0 dxc3 16 llfdl ..td7 led to unclear position in E.Co11e great complications in I.Miladinovic lR.Capablanca, Carlsbad 1929. P.Kiriakov, Halkida 1996. b) 4 ..tf4 (London) 4 ...b6 5 e3 ..tb76 S ..• b6 6 tiJbd2 ..tb77 ..td3 0-0 8 0-0 tiJc6 tiJbd2 ..te7 (threatening ...tiJh s) 7 h3 0-0 8 ..td3cxd 4 9 exd4 d6 (blunting the bishop on f4) 10 0-0 tiJbd7 11 l::tel l:te8 12 ..th2 a6 13 a4 'ilfC7 14 'ilfb3 ..tf8 A standard and balanced position for this opening. g 'ili'e2 Another Tal game saw 9 l:.el l:tc8!? (a typical reorganization of Black's (9... cxd4 10 exd4 tiJds is more usual) 10 pieces) 15 tiJgs (or 15 l:te2 ..tc6 16 c4 a3 tiJds 11 ..txe7 'ilfxe7 12 'ilfc2 cxd4 13 "ilfb7 17 lIael?! ds! 18 cxds ..txds 19 ..txh7+? (13 exd4 fs 14 c4 tiJf4 15 ..tfl ..tC4 llac8 20 ..txds 'ii'xds 21 "iVxds 'iWf6 or 15 ds tiJxd3 16 'it'xd3 'it'd6 would 1 9 8 The Topa/ov Variation be unclear) 13 10 . . . lDdS! dxe6! gxf3 22 exd7 l:.ff8 23 'ii'd6 lDe2+ The thematic equalizing move. 24 lilxe2 fxe2 25 'iVe6+ ltf7 26 'ii'e8+ 11 .i.xe7lDcxe 7 with a draw) 21 lDxg5 :g6 22 'it'h 5 lDf6 This set-up leads to a roughly equal 23 'iVf3 ! would be a simpler way to position, though I would prefer to be keep the balance. Black, as he has greater prospects for 20••• g4 21 lDel:cS 22 'ii'd2 l:.g6 dynamic play. Although the position is objectively 1 99 Play the Benko Gambit still about equal, Black has more prac which makes the position very difficult tical chances. Indeed, White now be for White. Although Tal did not con gins to go astray, wasting time with his duct the endgame with complete accu rook. racy, he managed to win in the end. 23 :e5?1 d6 24 ':e3h5 25 :c1 b5?1 38 ... 27..• bxC4 28 l:lec3'ii xd4 29 'iixd4 tDe2+ The set-up with ...e7-e6, ...b7- b6, ...i.b7 30 200 The Topalov Va riation options which can be rather too i..b7 6 e3 g6 7 c3 iVb6! and 6 a4 'iVb6! drawish for an ambitious Black player. (seen in the notes to Game 44) and 5 In the middlegame Velikov had diffi .i.f4 .i.b7 6 a4 b4! followed by ...ltJf6 culty finding a constructive plan, made and ...e7-e6 (as in Game 45). If White a few inaccuracies and ceded the ini opts for 5 h4 or 5 'iVd3 Black equalizes tiative, which Tal later converted to a more easily by taking the bishop and superior endgame. playing ...g7 -g6. Other continuations are less threat Summary ening. 4 a4 is well met by 4 ... b4, as the In this chapter we have seen that 2 ...c5 inclusion of these moves in previous is also feasible against 1 d4 ltJf6 2 ltJf3 variations (i.e. after 5 c4 g6 or 5 i..g5 without c2-c4. If White advances with 3 ltJe4) does not help White at all. De d5, Black can play 3 ...b5 in any case, fending the d4-pawn with 3 e3 or 3 c3 trying to gain space and undermine is of course a valid option, but these the d5-pawn. The most testing reply to moves are hardly challenging from a this plan is probably 4 .i.g5, and here I theoretical perspective. Black obtained recommend 4 ...ltJe4, as championed by good play in both the reversed Tarrasch Topalov. of Game 48 and the Torre Attack of Then the critical moves are 5 .i.h4 Game 49. and 5 .i.f4, retreating the bishop. Black To conclude, the Topalov Variation should be okay if he responds with is a verysh arp and double-edged open 5 ....i.b7, but he needs to proceed with ing for Black, but at the moment it some accuracy. Key lines are 5 i..h4 seems to be holding up pretty well. 201 Index of Variations 1 d4 tiJf6 2 c4 cS 3 d5 b5 Chapter 1: The Fianchetto Variation 4 cxb5 a6 5 bxa6 g6 6 tiJC3 693 d6 7 b3 - 24 6 ...�xa6 7 g3 �g7 8 �g2 8 tiJf3d6 9 �h 3 - 26 8 . . . d6 9 tiJf3 9 tiJh3 - 29 9 . . . tiJfd7 10 0-0 10 'ili'c2tiJb 6 11 h4 - 16 10 �d2, 10 �f4 - 20 10 ...tiJb6 11 Vc2 tiJ8d712 l:tdl0-0 13 libl 13 e4 -13 13 b3 -16 13 ... tiJC4 14 b3 14 tiJd2 -18 14 .. :i!i'a5 - 20 202 Index of Variations Chapter 2: The Classical Variation 4 cxb5 a6 5 bxa6 g6 6lDc3 .i.xa67 e4 7f4 - 67 7 ....i.xfl 8 'itxfld6 9g3 9 lDf3 .i.g7 10 h3 lDbd711 'itgl 0-0 12 'ith2 Was 13 l:tell:lfb8 14 l:te2 -56 14 Wc2 - 60 9lDge2 - 61 9 g4 hS 10 gs lDfd7 - 64 9 ....i.g7 10 �g2 0-0l1 lDf3 lDbd712 h3 12 :el lDg4 13 We2 -34 13 lDd2- 40 13 others - 34 12 ..:ii'b6 12 ...l:ta6 13 l:telWas 14 WC2 - 50 14 J:.e2-5 3 13 l:tel l:tfb8 14 l:te2-4 3 14 eS -47 Chapter 3: The 'Quiet' 5 e3 4 cxb5 a6 5 e3 .i.b76 lDC3 6 bxa6 - 73 6 .••Wa5 7 .i.d2 7 bxa6 - 80 203 Play the Benko Gambit 7 ...axb5 8 �xb5 'ifb6 9 'ilb3 - 73 9�C4-76 Chapter 4: The 5 f3 Variation 4 cxb5 a6 5 f3 e6 6 e4 exd5 7 e5 'iVe78 'iVe2 liJg89 liJC3�b7 10 liJh3 c4 11 �e3 ll liJf4 - 95 11 ... axb5 12 0-0-0 - 86 12 liJxbs - 92 Chapter 5: White plays b5-b6 4 cxb5 a6 5 b6 'iVxb66 liJC3g6 7 liJf3 7 e4 d6 8 a4 �g7 9 as 'ilb4- 100 7 ...�g7 8 e4 d6 9 liJd2 0-0 10 �e2 10 a4 liJbd7 11 as 'iVd8 12 �e2 .l:tb8 13 0-0 - 10 �e2 1o... liJbd7 11 0-0 204 Index of Variations 11 4.:JC4'fiC7 - 110 11..• l:tb8 12 a4 12 ...'iVa7 - 103 12 ...'iVd8 - 108 Chapter 6: Other Options for White 44.:Jf3 4 cxb5 a6 5 4.:JC3 - 114 4 'ili'c2 - 119 4 �g5 4.:Je4- 127 4 4.:Jd2bXc4 5 e4 c3 6 bxc3 g6 7 4.:JC4- 130 7 l:.bl - 133 4a4b4-137 4 ...bxC4 -140 4 f3 - 142 4e4 - 145 4 b3 - 145 4 •••g6 5 cxb5 a6 6 'fic2- 119 6 e3 - 123 6 4.:JC3-1 23 6 b6 - Chapter 5 6 bxa6 �xa6 - Chapters 1 and 2 Chapter 7: The Kasparov Gambit 1 d4 4.:Jf62 c4 cS 3 4.:Jf3 cxd4 4 4.:Jxd4e5 54.:Jb5 205 Play the Benko Gambit 5 tDc2(an d other moves) - 168 S •••dS 6 cxdS i..cS 7 tDSC3 7 d6 -165 7 •••0-0 8 e3 -149 8 h3 - 154 8 g3 -159 Chapter 8: The Topaloy Variation 1 d4 tDf62 tDf3 CS 3 dS 3 e3 -194 3 c3 -197 3 c4 - Chapter 7 3 ••• bS 4 i..gs 4 a4 (and other moves) - 188 4 •••tDe4 5 i..h4 -173 5 i..f4 - 180 5 h4 - 184 206 Index of Com plete Ciames Aagaard.J-Nisipeanu.L.D, Berlin 1997 ...... 50 Adianto.U-Polgar.J, Jakarta (rapid match, game 2) 1996 ...... 95 Aseev.K-Ponomariov.R, European Championship, Ohrid 2001...... 29 Bareev.E-Topalov.V, Sarajevo 2000 ...... 119 Blees.A-Greenfeld.A, Tel Aviv 1988 ...... 56 Brumen.D-Smirin.l, Croatian Cup, Rabac 2004 ...... 194 Burmakin.V-Kasparov.S, Deizisau 2003 ...... 13 Campos Moreno.J-Topalov.V, Aviles 1992 ...... 100 Comas Fabrego.L-Topalov.V, Pamplona 1994...... 53 Cossin.S-Vachier Lagrave.M, Chalons (rapid) 2008 ...... 34 Dancevski.O-Degraeve.J.M, European Championship, Ohrid 2001 ...... 140 Furman.S-Geller.E, USSR Team Championship 1975 ...... 26 Gabriel.C-Sermek.D, Pula 2000 ...... 60 Gallardo Garcia.A-Ortega Hermida,D, Mondariz 2007 ...... 20 Goganov,A-Kasparov.S, Peterhof 2008 ...... 16 Grivas.E-Khalifman.A, Leningrad 1989 ...... 127 Gurevich.D-Benjamin.J, Chicago 1986 ...... 76 Hellsten.J-Baklan.V, German League 1997 ...... 108 Hochstrasser.M-Pikula.D, Pontresina 2000 ...... 114 Ivanov.Se-Khalifman.A, St Petersburg 1997 ...... 92 Jirka.J-Vuckovic.B, Sozina 2004 ...... 142 Kachar.V-Arbakov.V, Russian Cup, Moscow 1999 ...... 4 7 Karpov.A-Miles.A, Tilburg 1986 ...... 73 Kozul.Z-Topalov.V, Istanbul Olympiad 2000 ...... 184 Kramnik.V-Leko.P, Dortmund 1998 ...... 130 Lalic.B-Khalifman.A, Linares Open 1997 ...... 86 Lugovoi.A-Khalifman.A, St Petersburg 1996 ...... 40 Mancini.M-Degraeve.J.M, French Team Championship 2004 ...... 67 Matlakov.M-Evdokimov.Alexa, Russian Team Championship 2010 ...... 159 207 Play the Benko Gambit Mikhalchishin.A-Kasparov.G, USSR Championship, Frunze 1981 ...... 149 Milov.V-Topalov.V, Prague (rapid) 2002 ...... 103 Mohr.R-Polgar.J, Amsterdam 1989 ...... 80 Neverov.V-Pedersen.N.V, Dieren 2010 ...... 18 Nielsen.P.H-Pedersen.N.V, Danish Team Championship 2006 ...... 110 Pavlovic.S-Giorgadze.T, Lugano 1985 ...... 24 Petkov.V-Pedersen.N.V, Skanderborg 2010 ...... 133 Piket.J-Topalov.V, Monte Carlo (rapid) 1997 ...... 43 Raetsky.A-Tyomkin.D, Biel 1999 ...... 154 Relange.E-Bauer.Ch, French Championship, Marseilles 2001 ...... 180 Romanishin.O-Wojtaszek.R, European Championship, Dresden 2007 ...... 1 73 Sakovich.I-Lanka.Z, Riga 1980 ...... 64 Santos Garcia.F-Alvarez Naves.L, Valladolid 1975 ...... 168 Schmidt.WI-Kasparov.G, Dubai Olympiad 1986 ...... 137 Shchukin.D-Khalifman.A, St Petersburg 1998 ...... 145 Shulman.Y-Khalifman.A, FIDE World Cup (playoff), Khanty Mansiysk 2005 ..... 61 Shvedchikov.A-Tseitlin.Ma, USSR Army Championship, Lvov 1977 ...... 165 Sorin.A-Sermek.D, World Student Team Championship, Paranana 1993 ...... 123 Vaganian.R-Topalov.V, Istanbul Olympiad 2000...... 188 Velikov.P-Tal.M, Berlin 1986 ...... 197 208