Transpo Tricks in Chess Finesse your Chess Moves and Win
Andrew Soltis
BATSFORD
First published in the United Kingd...
1003 downloads
4460 Views
9MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Transpo Tricks in Chess Finesse your Chess Moves and Win
Andrew Soltis
BATSFORD
First published in the United Kingdom in 2007 by Batsford 10 Southcombe Street London Wl4 ORA An imprint of Anova Books Company Ltd Copyright © Batsford 2007 Text copyright © Andrew Soltis The moral right of the author has been asserted. All rights reserved. No part ofthis publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the copyright owner. ISBN 9780713490510 A CIP catalogue record for this book is available from the British Library. 15 1413 1211 10 09 08 07 10987654321 Reproduction by Spectrum Colour Ltd, Ipswich Printed and bound by Creative Print & Design, Ebbw Vale, Wales This book can be ordered direct from the publisher at the website www.anovabooks.com. or try your local bookshop. Distributed in the United States and Canada by Sterling Publishing Co., 387 Park Avenue South, New York, NY 10016, USA
Contents
Page
Introduction
5
Chapter One:
Double KP Openings
14
Chapter Two:
Ruy Lopez
34
Chapter Three: Sicilian Defense
54
Chapter Four: Semi-Open Games
94
Chapter Five:
Double QP Openings
120
Chapter Six:
Indians
153
Chapter Seven: Dutch Defense and Flank Openings
191
Index of Opening Variations
213
3
4
Introduction
Openings have become so complex and convoluted that we've forgotten our basic goal in them. The reason we pick, say, 1 e4 over 1 d4 - or 11 h3 rather than 11 .te3, for that matter - is simply to reach a middlegame we want to play. Unfortunately, our opponents are making choices too, in order to reach the middlegame they want to play. It's rarely the same
In 99 of 100 games Black continues 5... ttJd6 or 5... .te7 and life goes on.
middlegame.
But lurking in the warren of footnotes of 'book' is 5... a6!? It's How do you get the one you usually followed by the comment want? Unfair as it may be, you can't 'If 6 .ta4, then 6... b5 7 .tb3 d5 rely on your legendary positional transposes to the Open Defense. ' skill, your better-than-Fritz calculating ability or your iridescent personal charm. What you can use is trickery - the trickery of transposition.
Since it's a footnote we take little notice. After all, 5 ... a6 'just transposes.' That tells us it doesn't really matter because it only leads to some other opening, something irrelevant on that page.
Consider this opening: 1 e4 e5 2 ttJO ttJc6 3 .tb5 ttJf6 4 0-0 ttJxe4 5 d4, the main line of the Ruy Lopez's Berlin Defense.
But 5... a6 has been a valuable weapon for players from Paul Morphy to Vasily Ivanchuk. Its power lies in how it gets Black 5
Introduction
where he wants to go - to the Open Defense - and avoids what he wants to avoid - the Exchange Variation (3 ... a6 4 .txc6).
We don't judge transpositions by the same standard as we do other moves. An original opening idea, a TN as they're called, is evaluated by the new position it creates. But a Finesses like that are rarely transposition by definition reaches appreciated except when they make an old position, as 5 a3! brilliantly new 'book'. That was the case in did. Instead, you should judge it by Botvinnik-Capablanca, A.Y.R.O. its effects, especially: 1938, an instantly famous game that How it degrades your opponent s began 1 d4 ttJf6 2 c4 e6 3 ttJc3 .tb4 choices or improves your own 4 e3 dS. A basic strategy in any opening is to increase the options at your disposal. Take this familiar position.
White's S a3 .txc3+ 6 bxc3 seems obvious today. But in 1938 it was a masterstroke. 'The idea of the move is typically modem - to transpose into a favorable variation which would not be reached in any normal manner,' Reuben Fine wrote.
This is a 'tabia,' that is, a standard starting point in a major opening. In this case it's a tabia that has served as the launching pad for thousands of Dragon Sicilians, which often continue 10 ... ttJe5 and .. J:tc8/ ... ttJc4.
In truth, there is a normal manner, the discredited 4 a3 .txc3+ But in a 1997 game, Anand5 bxc3 d5?!, which allows White to rid himself of a doubled pawn. Ki.Georgiev, Black tried 10... Botvinnik used 5 a3! to trick ttJaS!? White quickly appreciated Capablanca into that favorable the difference: After the natural 11 .tb3 Black could transpose into version of 4 a3. 6
Introduction
more familiar lines with 11..J1c8 slow moves often prove fatal, as in and 12 ...tbc4 13 i.xc4 l:hc4. But this case: 10 ... tba5 gives Black an extra option, ... tbxb3+, that he may employ depending on White's next few moves. In other words, Black gets to choose whether he wants to transpose with ...tbc4 or not. Instead of trying to figure out how dangerous ... tbxb3+ would be White made a practical choice, 11 i.e2!. This is a counter-finesse. It takes away Black's extra option 13 g4! bS 14 h4 e6 1S a3 hS and leaves the knight with nothing 16 i.gS hxg417 hS! gxhS 18l:1xhS better to do than go to c4. l:1cS 19 fxg4 l:1xc3 20 i.xf6 'iixf6 This had a bonus effect because a 21 'ifxc3 tbc4 22 i.xc4 bxc4 move order finesse can also be 23 'ifxc4 and White won. judged by: This shows how transpositions How it unnerves or confuses your play tricks not just with move order opponent but with your opponent's equanimObjectively, 11 i.e2 is no better ity. When he realizes he is being than 11 i.b3. But psychologically dragged into your middle-game, he it was a potent blow - and may lose the nonnal composure that transpositions typically have players enjoy in the opening, when greater psychological power than they confidently rattle off the first 15 moves. Players who lose their objective strength. confidence make mistakes. After 11 .. J1c8 12 ~b1 Black couldn't bring himself to play the Let's go back to 1 e4 eS 2 tbfJ best move, 12 ... tbc4, because it tbc63 i.bS tbf6 4 0-0 tbxe4 S d4 would create the middlegame White a6. Giovanni Vescovi was rated No. wanted to play. Black had more or 60 in the world when he first saw less decided, when he passed up that position from the White side, in 10 ... ~e5, that he didn't want that 2005. He decided not to be tricked middlegame. into the Open Defense. So he chose a very different policy, 12... a6? In the Dragon such
But that meant choosing 6 i.xc6?!, which turned out to be a 7
Introduction
prelude to a worse idea, 6... dxc6 7 'Yi'e2 .ltf5 8 g4? .ltg6 9 h4.
Anand began calculating furiously, trying to find out what was wrong with White's move. But there's nothing wrong with it. It's Black could have refuted that just unfamiliar. After spending two with 9...'Yi'd7! 10 ~xe5 'Yi'xd4, as of his precious five minutes, he Johannes Zukertort played way played 4... d6 5 ~f3 d5 and reached back at London 1883 (!). a book position. Vescovi spent 40 minutes to find This was a case of a move whose that dubious line. That leads to a major benefit was simply to give another criterion of a transposition. the other player something to think It can be measured by: about. Typically, these moves do not reduce his options. Rather they increase them, giving him more to consider.
How it gets your opponent to think The real battle of the opening begins when you can force your opponent out of his book knowledge. Only then does he risk making errors and spending costly minutes.
There really isn't much value, for example, to 1 d4 c6 and then 2 c4 d5 compared with the normal route, 1 d4 d5 2 c4 c6. But this order, employed by Anatoly Karpov among others, gets White debating with himself over whether he knows more about the Caro-Kann (2 e4) than the Slav. Or what he would do about 2 c4 b5!? A strange move order can do that.
That was case when Vishy Anand was Black in a speed playoff game at a big-bucks event in 1994. In the Petroff Defense, 1 e4 e5 2 ~f3 ~f6 3 d4 ~xe4, his opponent failed to play 4 .ltd3, the move considered virtually automatic. He played 4 ~xe5!? instead.
A 'something to think about' move can cost your opponent more 8
Introduction
And there's a fourth way of evaluating a crafty transposition, by:
than minutes. It can prompt a bad decision. The most drastic recent example of that befell Vladimir Kramnik in what was then the most important game of his life, the final game of a 1994 Candidates match.
How it preserves your mental health
We all have to deal with an everHis opponent, Boris Gelfand, expanding amount of book analysis. opened with 1 c4 and there Almost as bad as trying to followed 1..• c5 2 liJc3 liJf6 3 g3. memorize all that at home is trying This pOSItIOn had occured to remember it at the board. This gazillions of times before - but can be maddening. never to Kramnik. As routine as We'd love to cut down the 3 g3 was, it confused him. He amount of book we need to know replied 3... d5 4 cxd5 liJxd5 and and still reach the middlegames we then on 5 .i.g2 like. The best ways to save our midnight oil - and our sanity - is through transpositions. Consider the main line of the Winawer French, 1 e4 e6 2 d4 d5 3 liJc3 .i.b4 4 e5 .
... he played 5... e6?? 'almost without thinking,' according to his opponent. Kramnik was assuming he would transpose, after 6 liJf3 liJc6, to another tabia that is quite good for Black.
The familiar path is 4 ... c5 and then 5 a3 .i.xc3+ 6 bxc3 liJe7. But But 6 liJxd5! cxd5 7 'ii'b3 won a White has numerous sidelines such pawn (7 ... c4 8 'iib5+), the game and as 5 dxc5, 5 .i.d2, 5 'iVg4 and 5liJf3. the match. The confusing effect of Theory regards these as not quite as 3 g3 set back Kramnik's world good as 5 a3. But in practice they championship aspirations for are dangerous to an ill-prepared Black. several years. 9
Introduction
Rather than spend hours studying them. Black can play 4...liJe7!. This used to be purely a prelude to 5 a3 i.xc3+ 6 bxc3 b6. But Wolfgang Uhlmann showed that 6... c5! was a simple transposition to the 4 ... c5 main line. Black gets the benefits of the normal move order without having to agonize over the 4 ... c5 sidelines. It's the 'mental hygiene' move order. Every transpositional trick has to be judged by a balance sheet: What are the benefits? What are the drawbacks? Which order counts more? Some, like 4 ... liJe7, may have more plusses than minuses. Other's like Kramnik's 5... e6?? are disastrous.
This is another book POSItIon from the Keres Attack - but White has lost a move because he spent two tempi to get his h-pawn to h4. A lost tempo should make a huge difference. But what mattered most was Black's unfamiliarity with Kereslike positions. After 9...liJde5?! 10 i.e3 b5?! she was worse and after 11 liJxc6 liJxc6 12 'ifd2 i.b7 13 f4 ~c7 14 'ifflliJb8 15 a3liJd7 16 f5 liJe5? 17 fxe6 fxe6 18 i.h3 she was lost. White's objectively bad 8 g5? worked brilliantly as a transposition. It was really 8 g5!.
But the vast majority are somewhere in between. The bottom line is not whether they give you a superior position but whether it's a position you want to play. Ideally, it's also one your opponent wants to avoid.
'IT JUST TRANSPOSES'
Consider 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4 4 liJxd4 liJf6 5 liJc3 d6 and now 6 h3 e6 7 g4. This transposes into a conservative version of the Keres Attack, in which White relies on a slow buildup with i.g2, rather than g4-g5.
Some opening positions can come about from two, three or more logical orders. Books - and even experienced players - tend to dismiss an alternative route because 'it just transposes'. But each route is likely to have benefits and liabilities that a shrewd transposer knows to evaluate. Take this
However, Andreikin-Kosteniuk, Tomsk 2006 went 7... a6 8 g5liJd7 9 h4.
10
Introduction
position from the Taimanov Sicilian. It's White's move and when it was analyzed, by Mark Taimanov, in the 1972 edition of ECO he claimed Black wins.
realizes Black isn't going to allow 10 e5. So he'll be aware that Black is planning to transpose, with 9....tc5! to the diagram, and may look instead for alternatives such as 9 h3. The third route, S....te7!?, is the most deceptive. An experienced Sicilian player in White's chair would recognize that the thematic move is 9 f4. He would see that the naturaI9 ... d6 transposes into a book Scheveningen. So he'll play 9 f4 allowing Black to reply 9 ... .tc5 and get where he wanted to go. Of the three, 8... .te7 may be Black's best route to the diagram.
In fact, White has a good reply, 10 .!Df5!. The critical line is 10....!De7!, which forces 11 .!Dxg7+ ~f8 12 .txc5 'i'xc5+ 13 ~h1 ~xg714 e5.
THE LURE OF THE FAMILIAR
Whether this is sound is in There are players, like Uhlmann, dispute and that means both White Lajos Portisch and Mihai Suba, and Black may be interested in who used finesses of move order reaching the diagram. There are solely to reach the middlegame they three quite reasonable routes to it. want. There are others, like Bent One is 1 e4 c5 2 .!Df3 e6 3 d4 Larsen, who also use them as cxd4 4 .!Dxd4 .!Dc6 5 .!Dc3 'i'c7 'confuse-moves' to pose puzzles or 6 .te3 a6 7 .td3 .!Df6 S 0-0 and to get their opponents to think. And now S....tb4 9 f4 .tc5. But if Black then there are tricksters. seeks the diagram this order is The trickster looks for the crafty seriously flawed: White has a way to reach the middlegame he strong alternative in 9 .!Dxc6! and wants. He knows, for example, that 10.!Da4. when opponents are confronted with an unfamiliar move, they are strongly, even irrationally, tempted to look for a way to reach a recognizable position.
The second route is forcing. Black attacks the h-pawn with S....td6. White's instinct is to meet the threat with a threat, 9 f4. But he 11
Introduction
Laszlo Szabo fell victim to that temptation in the 1953 candidates tournament when his game with Paul Keres began 1 d4 d5 2 ttJf3 ttJf6 3 c4 dxc4 and then 4 ttJc3 and 4••• a6.
2 ttJf3. He knows that books recommend 2 d4!. But about half of all 1 e4 ttJc6 games continue 2 ttJf3 because White wants to transpose to the familiar (2 ... e5). Similarly, when a 1 d4 player faces 1...d5 2 ttJf3 ttJc6 3 c4 e5?!, he is likely to transpose into the Albin Counter Gambit, 4 dxe5?! even though he suspects 4 ttJxe5! is better. The trickster exploits that foible of human nature. The rest of us have to be willing to punish him for taking such liberties. For example, 1 e4 c5 2 ttJf3 ttJc6 3 d4 cxd4 4 ttJxd4 ttJf6 5 ttJc3 g6.
That was unfamiliar to him. But he remembered getting a very good game once with 4 'ifa4+ ttJbd7 5 ttJc3 a6. After studying the position for a while he played 5 'ifa4+, seeing that 5... ttJbd7 would transpose. A stunned Keres took 15 minutes to make sure he wasn't dreaming. Then he played 5... b5! and won easily (6 ttJxb5? Jid7). A much more common, and less painful, error occurs when a good player takes the easy way out of an unfamiliar position and transposes to a recognizable one. He may do this even if he suspects - or knows for certain - that he is playing an inferior move.
This order gives Black all the tactical benefits of an Accelerated Modem Dragon without allowing the Maroczy Bind as the Accelerated does. But there is a big minus 6 ttJxc6 sentences Black to either a dubious middlegame (6 ... bxc6 7 e5) or a no-fun endgame (6 ... dxc6 7 'ifxd8+).
You've seen this happen when a 1 e4 player meets 1...ttJc6 with 12
Introduction
Nevertheless the trickster will weigh the risk of being squeezed to death in a Maroczy Bind against the risk of White playing 6 tiJxc6!. The odds are on his side. A database check finds White played 6 tiJxc6! less than a third (327 of 1,000 games) of the time. Even superGMs like Bobby Fischer and Mikhail Tal played the inferior 6 i.e3 to return to a position they knew.
Examples like that should tell you to beware those words 'it just transposes.' There is usually a plus, a minus, an extra option to consider. If you don't, you are allowing your opponent to control the direction of the opening. And the price may be ending up in the middlegame he wants.
13
Chapter One: Double KP Openings Trickery begins with the oldest openings, like the King's Gambit Accepted. Today's theory says the KGA (1 e4 e5 2 f4 exf4 3ltJf3) is no longer dangerous because of 3... d5 and then 4 exd5 ltJf6 with equal chances.
Breyer Gambit, 1 e4 e5 2 f4 exf4 3 'fif3 d5 4 exd5. But the Breyer is harmless and Black can even try the sacrificial 4 ... ltJf6 5 J.b5+ c6!?, recommended by Yacov Estrin. This order can upset a Bishop's Gambiteer. There are few 1 e4 e5 2 f4 exf4 3 .tc4 players who are comfortable in KGA positions, which they would reach by playing 4 ltJf3 in the diagram.
If that makes Black happy, he should consider the 'something to think about' order of 1 e4 e5 2 f4 d5 3 exd5 exf4!?, as played by Anatoly Karpov and Artur Yusupov.
So they'll look at 4 J.c4 and realize it transposes into an ugly Bishop's Gambit - 1 e4 e5 2 f4 exf4 3 J.c4 and then 3... d5 4 exd5 rather than the recommended 4 J.xd5. That's a psychological trap because White doesn't really stand badly after 4 exd5!? ltJf6 5 lLlc3 and then 5... c6 6 d4. He has simply transposed into the main line of the normal Bishop's Gambit, that is 1 e4 e5 2 f4 exf4 3 J.c4 ltJf6 4 ltJc3 c6 5 d4 d5 6 exd5. But it takes a lot of clock time to figure that out over the board.
What is there to think about? Well, for starters, White wonders whether he should head into the book line with 4 ltJt3 ltJf6 or try 4 'ift3, which transposes into the
14
Double KP Openings
If Black doesn't like the ... d5 KGA or wants more than equality he can aim for this:
The natural 4... g5 falls into a trap because 5 h4 transposes into a better-than-usual version of Allgaier Gambit (5 ... g4 6 liJg5 h6 It's a tabia that's been studied 7 liJxf7 'l;xf7, rather than 6... f6 since the 1860s and today is 7 liJh3 gxh3 8 'ii'h5+). regarded as excellent for Black In contrast with the normal (8 g3 i.h3! 9 :f2 liJf6! or 9 gxf4 Allgaier, 1 e4 e5 2 f4 exf4 3 liJf3 g5 'i'd7! 10 :f2liJf6). 4 h4 g4 5 liJg5?! h6 6 liJxf7 'l;xf7 But Black can't force his way to 7 i.c4+ d5!, White has an extra it from the traditional 1 e4 e5 2 f4 move, liJc3, which is better than exf4 3 liJf3 order. He needs to use Black's extra ... d6?!. misdirection. Viktor Korchnoi claimed that 4... h6! favors Black in the diagram. It may be his best but after 5 d4 g5 6 g3! we've reached a position that often occurs in KGA lines and the evaluation is at best fuzzy.
Bobby Fischer's try was 3... d6 and then 4 i.c4 h6! 5 d4 g5 6 0-0 i.g7 7 c3 liJc6 reaches the tabia.
The problem with 3... d6 is that 4 d4 threatens 5 i.xf4 before Black For example, 6... fxg3 7 hxg3 can put his kings ide house in order i.g7 8 i.c4 and 8... i.g4 9:f1 'i'd7 with both ... g5 and ... h6. His best is 10 'iVd3, Gallagher-Jurgens, Bad 4... g5 5 h4 g4 after which 6 liJg5? Worishofen 1994. Remember the f6! is unsound but 6 liJgl leads to 6 g3 position. We'll see it again. double-edged play. Black can try to improve on the There is no objectively better Fischer defense with the neglected move than 4 d4. But there is a more 3.•. h6, again seeking that good-forconfusing one, 4liJc3!? Black tabia. This can confuse White 15
Double KP Openings
unless he knows that 4 d4 gS and S ttJc3! d6 6 g3! reaches the fuzzy line.
BISHOP'S GAMBIT If your opponent is a Bishop's Gambiteer, you have good reason to fear he's better prepared than you, a lot better. The careful transposer will answer 1 e4 e5 2 f4 exf4 3 .tc4 with the forgotten favorite of Mikhail Tchigorin, 3... ttJc6.
That leaves us with 3•.• g5. The Romantic-era continuation was 4 .tc4 g4 but 4••.ttJc6! clears a path to the tabia.
Black invites S 0-0 .tg7 6 d4 h6 and so on. Once White plays 4 .tc4 he can't insert h2-h4 effectively because S h4? g4! 6 ttJgS boomerangs badly after 6... ttJeS! 7 .tb3 h6 8 d4 hxgS 9 dxeS .tg7.
Black can then meet the natural 4 ttJf3 with 4•.. g5!, reaching that good form of the KGA that we considered in 3 ttJO gS 4 .tc4 ttJc6!. This is a psychological plus because, as noted before, Bishop's Gambiteers are often uneasy in a KGA.
The best answer to 3 ... gS is supposed to be 4 h4, which rules out the tabia after 4 ... g4 S ttJeS. However, the analysis of that position is immense and White may be tempted by the confuse-move 4 d4. Then 4 ... g4 S .txf4!? gxf3 6 'iVxf3 is an unusual Muzio-like line that served Alexander Morozevich well.
The bonus is that White has some natural but bad responses, such as 4 ttJc3? which allows a strong 4 ... 'iVh4+ S ~f1 .tcS. Whether White has an edge after 4 d4 ttJf6 S eS dS! is murky (6 .tb3 ttJe4). VIENNA GAME
The safe response is 4 d4 h6 since S h4 .tg7 6 g3 g4! is promising for Black. But S ttJc3! and S... d6 6 g3 puts us in the fuzzy position again.
There are two types of Vienna players. The first don't want to memorize volumes of Ruy Lopez theory. The second is booked up 16
Double KP Openings
and prepared to inflict volumes of Vienna theory on you, such as with 1 e4 eS 2 lbc3 lbf6 3 ii.c4 lbxe4 4 'iih5.
defend - any- cramped -posi tion reputation but it's based on White expanding with c2-c3 and d2/d4 or
'iib3.
Against either kind of opponent it may pay to be devious as early as 2... d6!?
Expansion like that isn't possible here and it's doubtful he has more than an optical edge after 5 0-0 lbf6 6 d4 ii.g4! or 5 d4 lbf6 6 h3 0-0 7 0-0 lbd7 8 d5 lbcb8 (MovsesianEfimov, Pula 1997).
This looks prohibitively passive. But White has to be careful since 3 f4 exf4 and then 4 lbf3 h6! S d4 g5 is that fuzzy KGA again.
There's much more experience with the other 'beginner's move,' 2 ••.ii.c5, which has been put to good use by Karpov, Bent Larsen and Vasily Smyslov. Books used to claim Black is already worse in view of 3 lbf3 lbc6 4 lbxe5! lbxe5 5 d4.
White's best is probably 4 d4 or 4 ii.c4. But few Vienna players are familiar or comfortable with KGA lines in which Black gets to play ... 'iih4+. Well, what about 3 ii.c4 ? Then 3 ... lbc6 leaves White to decide whether to allow 4 d3 lbaS!? or plunge into another unfamiliar KGA (4 f4 exf4 5 lbf3 h6 or 5 ... g5 6 h4 g4 7lbgS lbeS).
But Black has a good alternative in 3 ..• d6! since 4 d4 exd4 S lbxd4 lbe7 or 5 ... lbc6 heads toward a reasonable version of the Scotch Game after 6 ~e3, 6 lbb3 or 6lbxc6 'tWf6.
Instead, he may assume that simple development, 4 lbf3, will punish 2 ... d6. That transposes into a Hungarian Defense after 4...ii.e7!? The Hungarian has a stodgy, I-can-
For example, 5 ... lbc6 6 ~e3 ii.b6 7 lbdS lbf6 8 lbxb6 axb6 9 f3 0-0 10 ii.c4 d5 !. Or 10 c4 l:te8 lllbc2lbd7 12 'i'd2lbde5 13 ~e2 ~e6 14 b3 f5 with good play for 17
Double KP Openings
Black in Zarnicki-Bianchi, Buenos Aires 1989. White can stay in Vienna mode with 3 f4. Then 3... d6 is a King's Gambit Declined in which White has forsaken c2-c3, his most ambitious plan in the KGD, and books disagree about his chances for advantage. There is also 3 g3, which transposes into a main line of Louis Paulsen's variation. Black's only minus is relinquishing the bookrecommended defense (I e4 e5 2 ttJc3 ttJf6 3 g3 d5!). So 2... iLc5 has no major drawbacks or benefits and is more of a 'something to think about' move. CENTER GAME Most players who meet 1 e4 with 1... e5 know that 2 d4 exd4 3 'ir'xd4 ttJc6 4 "iVe3 is neutralized by 4•..ttJf6 5 ttJc3 ii.b4 6 iLd2 0-0. That's about all that anyone remembers. But White has a confuse-move in 5 iLd2.
This often has the effect of scaring Black into another defense, such as 5 ... iLe7 6 ttJc3 0-070-0-0that is, transposing into the less ambitious 5 ttJc3 iLe7 6 iLd2 0-0. But this is primarily a bluff because if Black meets 5 iLd2 with 5... iLb4! White has nothing better than transposing into the book line with 6 ttJc3!. The other finesses in the Center Game arise if Black is afraid of a line that hardly anyone knows about. This is 5 e5!?, rather than 5 ttJc3 or 5 iLd2. That seems to just lose a pawn but 5... ttJg4 6 'ife4 ttJgxe5? allows 7 f4. To play 4 ... ttJf6 with confidence Black should know a little about crazy lines like 6... d5 7 exd6+ iLe6 and then 8 iLa6 or 8 iLe2! ttJf6 9 dxc7 'ir'xc7. But he can save himself a lot of worry and/or study by picking another fourth move. One is 4 ..•iLb4+, with the idea of reaching the equalizing book line after 5 ttJc3 ttJf6. The drawback is Black may have to face White's extra options, such as 5 c3 iLa5 6 iLc4 or 6 'ifg3. Black's other alternative in 4 •••iLe7.
Then 5 ttJc3 ttJf6 transposes into that 4 ... ttJf6 5 ttJc3 iLe7 line, as Boris Spassky and Alexander Alekhine did.
18
Double KP Openings
This succeeds after 3... c5 4 c3 dxc3? 5 tbxc3 when he's transposed into 1 e4 e5 2 d4 exd4 3 c3 dxc3 4 tbxc3 c5? 5 tbf3, a ludicrous defense to the Goring Gambit. (But 3... c5 4 c3 d5 or 4 ... ife7 in the diagram are virtually untested.) Another version of the Goring arises after 3... i.b4+ and then 4 c3 dxc3 5 tbxc3. This forgotten variation is quite playable after 5...tbc6 6 i.c4 d6 7 tbg5 .Jl.xc3+ 8 bxc3 tbe5 and now 9 ..tb3 h6 10 f4! hxg5 11 fxe5 'iie7 12 ifd5 dxe5 13 ..txg5!.
But there's another wild card for Black to worry about, 5 'iVg3 and then 5...tbf6 6 e5, which Paul Keres said favors White, or 5... iu6 6 tbc3 tbge7 7 .td2 as in ZiemackiFatyga, correspondence 1992 (7 ... d6 8 0-0-0 .te6 9 f4).
The main benefits to 1 e4 e5 2 d4 So Black has to decide whether exd4 3 lbf3 are forcing Black to the drawbacks of 4 ... .tb4+ or 4 ... .te7 outweigh the merit of start thinking at move three and preventing him from playing a avoiding the rare 4 ... tbf6 5 e5!? Petroff, e.g. 3...lbc6 4 lbxd4 or 3... .tc5 4 tbxd4 tbc6. DANISH GAMBIT The Danish is a blood relative of the Goring and Scotch Gambits. The Danish has the poorer reputation of the three but White can try to improve with 1 e4 e5 2 d4 exd4 and now 3 tbf3!?, rather than the Danish 3 c3.
Black's chief alternative is 3 ...lbf6. Then 4 e5 is another Petroff line, I e4 e5 2 tbf3 lbf6 3 d4 exd4 4 e5, that Black has been avoiding for years (by playing 3... tbxe4!). It's not a bad line but few Black players will know the theory and that makes 3 tbf3 a worthy weapon against a Petrofessional. BISHOP'S OPENING Independent lines in the Bishop's Opening are somewhat rare today. Instead, 2 ..tc4 is used primarily as a route to the Vienna or Giuoco
19
Double KP Openings
Piano that allows White to seek or published theory. The primary avoid specific Vienna and Giuoco virtue of using 2 .i.c4 to reach those positions (2 ... 4Jf6 3 d3 4Jc6 44Jf3) positions. is that White avoids the Petroff. He benefits from 2 .i.c4 if he likes 2... 4Jc6 3 4Jc3 4Jf6 4 d3 or PHILIDOR'S DEFENSE 2... 4Jf63 d3 4Jc6 4 4Jc3 but doesn't want to try to get there via the Books used to ridicule the Vienna because of the chaotic 2 Philidor because of White's space 4Jc3 4Jf6 3 .i.c4 4Jxe4!? advantage. But young masters are Note that after 1 e4 e5 2 .i.c4 challenging that view. To them the key issue is how to reach this tabia. 4Jf6 3 d3 and the natural 3....i.c5:
Then 4Jc3 gets White into a Vienna without having to face 2 4Jc3 4Jf6 3 .i.c4 4Jc6 4 d3 .i.b4 or 4... 4Ja5. Black retains those options with the more precise 3•••4Jc6!. There's a reason they say 'Knights before bishops.'
The traditional order, 1 e4 e5 2 4Jf3 d6 3 d4 4Jd7, named after the U.S. Civil War major James Hanham, allows White to win the two bishops, 4 .i.c4 c6 5 0-0 .i.e7 6 dxe5! dxe5 7 4Jg5 .i.xg5 8 'ii'h5.
It has many of the benefits of a
If Black is concerned about
Aron Nimzovich helped popularThe other direction White can go ize another sequence, 1 e4 e5 after 2 .i.c4 is towards the Pseudo- 2 4Jf3 d6 3 d4 4Jf6. This is more Lopez. That is a Ruy Lopez-like likely to get Black to the tabia system that can begin 1 e4 e5 2 4Jf3 (4 4Jc3 4Jbd7 5 .i.c4 .i.e7) because 4Jc63 .i.c4 followed by 4 d3, 5 c3 few opponents are willing to wade and 4Jbd2, 0-0, .i.b3 and eventually into the 4 dxe5 4Jxe4 5 'ifd5 d3-d4. complications. regular Lopez, minus the 30 tons of 4 dxe5 he should consider offering 20
Double KP Openings
an endgame, because even fewer Whites will trade queens after 1 e4 d62 d4 e5!? Emanuel Lasker used this version in a world championship match against David ('I detest endgames') Janowsky. Today's players, like Michael Adams, are more confident on the Black side of 1 e4 d6 2 d4liJf6 3liJc3 e5!?
gives up the center, such as in 1 e4 e5 2 liJf3 d6 3 d4 exd4 and then 4liJxd4 g6 or 4 ... liJf6 5liJc3 iJ.e7. White can try to exploit this with Henry Bird's 4 iJ.c4!?
A key point is that 4 ... liJc6 5 liJxd4 is a fine Scotch Game. And 5 c3 dxc3 6 liJxc3 transposes into a good Scotch Gambit. It's as if Black met 1 e4 e5 2 liJf3 liJc6 3 d4 exd4 4 c3 dxc3 5 liJxc3 with 5 ... d6 6 iLc4.
Now 4 liJf3 liJbd7 transposes to Nimzovich's line and 4 dxe5 dxe5 5 'ii'xd8+ ~xd8 is a slightly worse endgame, e.g. Vaisser-Rontaine, French Championship 2006 went 6 iLc4 i.b4! 7 iJ.d2 ~e7 8 f3 c6.
The test of 4 iJ.c4 is 4...liJf6. In the first ECO Lev Polugayevsky gave 5 liJg5 d5 as favoring White (6 exd5 h6 7 liJf3 iJ.b4+ 8 c3 dxc3 9 'ii'a4+ and wins).
Black was soon equal (9 liJge2 liJbd7 10 liJc 1 b5! 11 iJ.e2 iJ.d6 12 a4 b4 13 liJdl a5), then better (14 liJe3 liJb6 15 liJd3 g6 16 b3 liJfd7 17 c3 bxc3 18 iLxc3 f6 19 liJb2 liJc5) and eventually won.
On the same page he considered 1 e4 e5 2 liJf3 d6 3 d4 liJf6 4 liJg5!? exd4 5 iJ.c4 d5 and called it equal. It's the same position.
Note that the 1...d6 orders have
Better, after 4 iLc4liJf6 5liJg5, is
5...iJ.e6. This dates back to Paul
the added benefit of cutting down White's 1 e4 e5 options, such as 2 f4, 2 iLc4, 2 liJc3 and so on.
Rudolf von Bilguer, who thought 6 liJxe6 fxe6 7 iJ.xe6 'ike7 (8 iLc8! 'ikxe4+ 9 'ike2) was even. To modem eyes this looks like a slight
There's a separate chapter in Philidor theory in which Black 21
Double KP Openings
White edge. Therefore 4 iL.c4!? is more than just a confuse-move.
brothers popularized 1 e4 eS 2 ttJf3 ttJf6 3 ttJc3 to reach a Four Knights Game, for example.
Observe that in traditional (1 e4 eS 2 ttJf3 d6) orders, White can try 3 iL.c4 with the idea of 3... ttJd7 4 0-0 iL.e7 S d4. In this way he avoids 3 d4 exd4 and traps Black in the inferior Hanham line.
Their 19th Century colleagues tried 1 e4 eS 2 ttJf3 ttJf6 3 d4 exd4 and 4 iLc4 in an effort to get into a Two Knights (4 ... ttJc6), which is terra incognita to a Petroff player. But the critical line is 4 ... ttJxe4!, the dubious Prince Yurosov Gambit.
Better is 3...iLe7! 4 0-0 ttJf6 which leaves White with a tiny edge after S 1::te I 0-0 6 c3 ttJbd7 7 d4. More ambitious is 3... ttJf6!? 4 ttJgS dS S exdS, which reaches a Two Knights Defense in which Black is a tempo behind.
These days 3 d4 ttJxe4 is more common, and theory says the chances are roughly balanced after 4 iL.d3 dS S ttJxeS. Yaacov Murey's discovery, 4 ... ttJc6, enables Black to avoid this. It's based on S .txe4 dS 6 .td3 e4, which appears to be sound. If White wants to avoid Murey and seek a main line he can do it with 4 ttJxeS!?
But the missing tempo is ... ttJc6, which may benefit Black since the knight is not attacked here. He can create an unclear position with S... h6 6 ttJf3 iL.d6 or 6... e4 7 'i'e2 ..tb4 8 c3?! O-O! (Miguel Najdorf). Then 4... dS S .td3 transposes into the main line as if 4 .td3 dS had been played.
PETROFF DEFENSE The Petroff is so annoyingly solid that a frustrated White will be sorely tempted to transpose into another opening. The Paulsen
This order gives Black an extra option, 4... d6!? S ttJf3 dS. He's transposed into the 'other' Petroff, 22
Double KP Openings
the 3 ltJxeS line, which White White reaches a Smyslov after presumably wanted to avoid when 7... 0-0 8 c3 dS 9 i.c2 lite8 10 d4. he chose 3 d4. This skirmishing - It's a more conservative form of it Murey's 4 ... ltJc6, the preemptive because White has given up on the 4 ltJxeS and the counter-finesse useful c2-c4. But it's also a much 4... d6 - has nothing to do with more modest position for Black finding the objectively best move than many Petroff players can but rather seeking the position in tolerate. which your opponent will be most The best challenge to S i.d3 is uncomfortable. 5... d5 and then 6 'iVe2 'ile7. A The 'other' Petroff is the most typical case was Romero Holmespopular today and it runs 3 ltJxe5 Garcia Padron, Las Palmas 1991: d6 4 ltJo ltJxe4 S d4 dS. The 7 0-0 ltJcs 8 lite 1 'ilVxe2 9 i.xe2 Smyslov alternative, S... i.e7 6 i.d3 ltJe6 10 d4 c6 11 c4! i.b4 12 ltJc3 ltJf6, has a stodgy reputation 0-0 13 a3! i.xc3 14 bxc3 b6 because 7 h3 leaves Black's QB IS ltJeS l:d8 16 a4! f6 17 ltJg4 and without a good sqaure. Then 7... 0-0 White won. 8 0-0 dS transposes to a favorable A later Petroff tabia begins with Exchange French in which h2-h3 is 3 ltJxe5 d6 4 ltJo ltJxe4 5 d4 d5 an extra move. 6 i.d3 followed by some Black White can try to trick his mixture of ... i.g4, ... i.e7 and opponent into the Smyslov line via ... ltJc6. But the order is in dispute. an order popularized by Alfonso In the definitive 19th Century Romero Holmes, starting with authority, the Handbuch, Emil 5 .id3!? Then S... ltJcs 6 i.e2 gives Schallop recommended 6 ... i.e7 him 7 d4 with tempo. So 5.•.ltJf6 7 0-0 ltJc6 followed by ... i.g4. But 60-0 .ie7 7 h3! is more common. others argued for 6... ltJc6 7 0-0 i.e7, or even 6... i.g4 7 0-0 ltJc6 and ... i.e7 as Carl Schlecter recommended in his version of the Handbuch. We can dispense with the last order, since 6... i.g4 allows 7 'iVe2! with an edge after 7 ... 'iVe7 8 0-0 ltJc6 9 i.bS! and i.xc6+, or 7... fS 8 h3 i.hS 9 g4!. 23
Double KP Openings
Garry Kasparov said 6.•. tbc6 7 0-0 i.e7 8 :tel i.g4 was fine for Black.
chooses between undermining the e4-knight with 8 c4 or attacking it with 8 .:tel. Isaac Kashdan had success in the 1930s with a line that runs 8 c4 tbb4! 9 cxd5 tbxd3 10 'iVxd3 'iVxd5 11 .:tel i.f5. The Kashdan line still holds up well today, and if Black likes it he can try to reach it even after 8 .:tel. He does that with 8...i.f5, rather than the popular 8... i.g4.
On 9 c4 Black has 9 ... tbf6 10 cxd5 i.xf3! with a good game. But he added that White gets an initiative if he undermines the knight a move earlier with 8 c4!. That's why Karpov preferred 6... tbc6 7 0-0 i.g4 in their 1985 match, so Black could meet 8 c4 Then 9 c4 tbb4 10 cxd5 tbxd3 with 8... tbf6!? Then 9 tbc3 is considered best (9 ... tbb4 10 i.g5! transposes to what Black wants or 9... dxc4 10 i.xc4 i.e7 11 d5!) which White probably doesn't as he and the results of 9 ... i.xf3 10 'iVxf3 could have sought it directly with tbxd4 11 'iih3 have been in White's 8 c4. favor lately. PONZIANI OPENING The bottom line is that no order The Ponziani (1 e4 e5 2 tbf3 tbc6 is universally endorsed. Jan 3 c3) gets no respect today. But if Timman and Alexander Belyavsky Black is caught off guard by 3 c3 he like 6...tbc617... i.e7 while Vishy can bail out with 3...tbf6 4 d4 d6. Anand, Vladimir Kramnik and There doesn't seem to be Alexey Shirov preferred 6... i.e7! anything in 5 i.c4 tbxe4!, so 7... tbc6 - and Yusupov plays both. 5 i.b5, transposing to the old Another tabia arises after, say, Steinitz Defense of the Ruy Lopez, 6...tbc6 7 0-0 i.e7. White usually is played most often. 24
Double KP Openings
best order because 4 ... i.h4?! is dubious. In contrast, in the traditional Scotch order, 1 e4 e5 2 ttJf3 ttJc6 3 d4 exd4 4 ttJxd4 Black can avoid this line via 4 ... i.c5, among others. There are several finesses in the Four Knights line after 5... i.b4 6 ttJxc6 bxc6 7 i.d3. The Steinitz is regarded as a clumsy antique but this is a Steinitz in which White has played conservatively (1 e4 e5 2 ttJf3 ttJc6 3 i.b5 d6 4 c3?! ttJf6 5 d4 instead of 4 d4!). This order was a favorite Ponziani evasion for GMs like Smyslov and Yefim Geller and it has no major drawbacks, e.g. 5... i.d7 6 0-0 i.e7 7 'iWe2 0-0 8 :dl 'i'e8 (threat of ... ttJxd4!) 9 d5 ttJd8 11 c4 c6 11 i.a4 b5! as in ManikMokry, Olomouc 1998.
Now 7... d5 8 exd5 cxd5 9 0-0 0-0 10 i.g5 is played with almost robotic uniformity and is very slightly in White's favor or even, depending on whom you read. But there are several diversions starting with 8...'iWe7+!?
But if, on the other hand, Black knows his way in a sharp line of the Scotch Gambit (1 e4 e5 2 ttJf3 ttJc6 3 d4 exd4 4 c3 ttJf6!? 5 e5 ttJe4) he can transpose into it from the Ponziani by means of 3... ttJf64 d4 exd4!.
Theory has gone back and forth about the value of 9 'iVe2. In any case, against an opponent who likes to keep queens on the board, 8... 'iWe7+ has obvious merits.
SCOTCH GAME
But if White is happy in an ending, he can try to force it by meeting 8... cxd5 with 9 ~e2+!? so that 9... 'i'e7 transposes. There are extra benefits when an endgamehating Black plays 9 ... i.e6
White can get into a major Scotch variation via the Four Knights (1 e4 e5 2 ttJf3 ttJc6 3 ttJc3 ttJf6 4 d4 exd4 5 ttJxd4). If that's a position he wants to play, this is the 25
Double KP Openings
(10 i.b5+! tbd7 11 i.c6 d4 12 'ife4) or 9 ... i.e7 (10 i.g5 0-0 11 O-O-O!? with better-than-usual chances).
These days Black is more concerned about the standard Scotch order, 1 e4 eS 2 tbo tbc6 3 d4 exd4 4 tbxd4 and then 4 ...tbf6 5 tbxc6 or 4 ... i.c5 5 tbxc6. For that reason, he can try to reach the Four Knights version with 4..•i.b4+!?
Black can anticipate this and rule out endgames through another route to the tabia devised by Georg Marco, 7..•0-0 and then 8 0-0 dS 9 exdS cxdS 10 i.gS. This appears to be the most exact - and the most deceptive since White will begin thinking about 8 i.g5 and the threat of 9 e5. Black is worse after 8... d6, so 8... dS:
Now 5 tbc3 tbf6 transposes to it, and considering the current view of theory, that would be a small victory for Black. The obvious objection to 4 ... i.h4+ is S c3. Then S...i.cs 6 i.e3 i.b6 reaches a book position - from 4 ... i.c5 5 i.e3 - but with an extra c2-c3 for White.
Black's point is 9 eS 'ife8! and White is the one surprised, e.g. 10 f4? tbg4 11 'ife2 f6 (12 exf6 gxf6 13 h3 £Xg5 14 bxg4 i.xg4! 15 iLxh7+ <Jiig7 and he resigned in Candela-Korneev, Ponferrada 1997).
Who benefits? White gets added protection of d4. But he loses one of the major assets of the Scotch, the knight hop from c3 to d5. Recent practice indicates Black is OK after 7 tbfS i.xe3 8 tbxe3 tbf6 because White cannot defend the e-pawn with tbc3 (and 9 tbd2 allows the equalizing 9... d5). Or 7 i.c4 d6 8 0-0 tbf6 9 tbd2 tbe5! 10 i.e2 0-0 11 h3 :e8 as in GoloshchapovBrodsky, Hoogoveen 2006.
If White spots that and heads instead towards the tabia, with 9 exdS, Black can use an extra option, 9...'ifxdS!. This hits at g5 and g2 and after 10 i.xf6 "iVe6+ and 11..:iVxf6 he stands well. 26
Double KP Openings
And what about the standard order, 4 llJxd4 .tcS S .te3, threatening llJxc6 ? Much midnight oil has been spent on 5.. :iVf6. The mental hygiene alternative is S.••.tb6!.
Now 6 llJc3 d6 and 7... llJf6 or 6... llJf6 transpose to a solid Four Knights/Scotch line. For example, Sveshnikov-Mi. Tseitlin, Budapest 1989 went 6 llJc3 d6 7 g3 llJf6 8 .tg2 0-0 9 llJxc6 bxc6 10 .txb6 axb6 with equality. Another trendy line is 4 ... llJf6 5llJxc6 dxc6 5 e5. Black can avoid that with 4••:i'f6. Then he is trying to reach, not avoid, the 5 .te3 .tc5 position. This order has the added benefit of dodging another difficult line, 4 ....tc5 5 llJb3. The best try at punishing 4 ... 'i'f6 is 5 llJb5 .tc5 6 'i'e2. But the verdict is out on 6....tb6 or 6... 'i'd8 (and ... a6).
the queen move is even playable. Its fate depends on how we evaluate this tabia.
The final word on 7 ... ..txc3+ 8 llJxc3 or 8 bxc3 is still to be written. But equally important is finding White's best way to the diagram. There are four routes. (a) S llJc3 ..tb4 6 llJdbS 'iYxe4+ 7 ..te2 gets there. But Black does much better with 6 .....ta5! and ... a6, e.g. 7 .te2 a6 8 llJd4 ..txc3+ 9 bxc3 llJf6 10 llJxc6 dxc6 11 e5 llJg4 (Renteria-Miktov, Minneapolis 2005). For that reason White should prefer: (b) SllJc3 ..tb4 6 ..te2! and then 6•.:i'xe4 7 llJdbS. If Black varies with 6 ... llJf6 7 0-0 .txc3 then 8 llJf5! is potent, e.g. 8... 'iYxe4 9 .td3 'iVg4 10 f3! ~a4 11 bxc3 0-0 12llJxg7! with a winning attack in Karjakin-Malinin, Sudak 2002. But White may opt for 5 llJb5 because it is more forcing and seeks something better than the diagram. For example:
At one time the fate of the Scotch seemed to hinge on the soundness of 4•..'i'h4. Today there's doubt that 27
Double KP Openings
(c) S tbbS 'fixe4+ 6 .te2 i.b4+ 7 tblc3 gets to the diagram. But White can improve with 7 .td2!, with elaborate lines following 7... .txd2+ 8 'iVxd2 'fie5 9 f4!, for instance.
•.• tbd8-e6/ ..•cS sets up a darksquare blockade. This is Black's best way to play for a win in the Four Knights, since Akiba Rubinstein's 4 ... tbd4 allows a draw-minded White 5 tbxd4 exd4 6 e5.
(d) Black can try to improve with S....tb4+ since 6 tblc3 'iVxe4+ 7 .te2 transposes. But again the bishop interpolation, 6 .td2!, is better and reaches ( c) after 6... 'iVxe4+ 7 i.e2.
There are ways for White to avoid the Metger - and still play to win - and other ways for Black to seek it. White's sidestep is 7 tbe2, which has been revived by Daniel Campora, who showed that White's edge in a !1ymmetrical position (7 ... tbe7) is something he can work with.
The main problem with 5 tbb5 is S.•..tcS and then 6 'iVe2 tbd4, when White doesn't reach the diagram or prove a clear edge. As it stands now, (b) may be the most accurate.
For example, 8 c3 .ta5 9 tbg3 tbg6 10 d4 c6 11 .td3 .l:te8 12 h3 .tb6 13 .l:tel h6 14 .tc2 'fic7 15 i.e3 i.e6 16 'fic 1! ':ad8 17 a4 a5 18 tbf5 d5 19 tbxg7!, Campora-
FOUR KNIGHTS GAME The main Four Knights line for most of the last century was Johannes Metger's 'unpin', after 1 e4 eS 2 tbf3 tbc6 3 tbc3 tbf6 4 .tbS .tb4 S 0-0 0-0 6 d3:
Baron Rodriguez, Lanzarote 2003. Black can avoid this by capturing on c3 earlier. Reuben Fine did it with 6..•'fie7 7 i.gS .txc3 8 bxc3. To avoid the drawish 8... d6 9 i.xc6 he played 8...tbd8!:
Black breaks the pin with 6... d6 7 .tgS i.xc3 8 bxc3 'iVe7 and
28
Double KP Openings
Now 9 d4 d6 is the Metger position. If White starts the knight maneuver with 7 lDe2, Black has 7... d5!?, e.g. 8 exd5 lDxd5 9 lDg3 lDf4 10 d4 lDg6 11 c3 .td6, Levenfish-Alatortsev, Moscow 1940. Nevertheless the more popular method is to delay .. .'ike7, as in 6....txc3 7 bxc3 d6 and 8 .tg5 Now 5... exd4 6 e5lands Black in 'iie7). This was a favorite order of the Lange he presumably wanted to avoid. If he still feels that way, Frank Marshall and Smyslov. 5....txd4! and 6 lDxd4lDxd4 7 ..tg5 d6 8 f4 'iie7 or 8... .te6 is the GIUOCO PIANO! antidote. But Yakov Estrin argued TWO KNIGHTS that White had reasonable chances. There's a recognized boundary What's certain is that this is much after 1 e4 e5 2 lDD lDc6 3 .tc4 that better than 4 d4 .txd4 5 lDxd4 separates the Giuoco, 3....tc5, from lDxd4 6 f4? (6 0-0 transposes) d5! the Two Knights, 3...lDf6. But the 7 exd5 'iih4+ 8 g3 'iih3, which border is often blurred by trans- favors Black. position.
Because 3... .tc5 makes no threat White has more freedom than in the Two Knights. He can head towards the Pseudo-Lopez with 4 d3, 4 c3 or 4 0-0 while retaining other options.
With 3... .tc5 Black is avoiding certain lines, such as 3 ... lDf6 4 lDg5, which have good reputations, or even the questionable Max Lange, 3...lDf6 4 d4 exd4 Theory says 4 c3 can be met by 50-0 .tc5 6 e5!? 4 ... lDf6 or 4...'iie7 5 d4 .tb6. But if holding e5 as a strong point is White can try to push Black into Black's intent, it's more precise to the Lange via 1 e4 e5 2 lDo lDc6 play 4....tb6. Then 5 d4 'iie7, as the 3 .tc4 i.c5 and now 4 0-0. Castling Berlin Pleiades played, transposes. was once ridiculed but today it's often seen as an alternate route to the Pseudo-Lopez (4 0-0 lDf6 5 d3). There is also 5 d4!?
The rationale for 4....tb6 is that .. .'ike7 is not as useful as ... ..tb6 in the Pseudo-Lopez. White can change his mind after 4...'iie7 and
29
Double KP Openings
switch to S d3 with idea of hitting the vulnerable queen later on with ttJe3-dS or ttJfS. The minus to this order is that after 4 ... ..tb6 S d3 Black will have lost a tempo if he feels ... a6 and ... ..ta7 are necessary later on.
Another move order issue in the Two Knights concerns the periodically revived 4 d4 exd4 S eS dS 6 ..tbS ttJe4 7 ttJxd4.
Note that 3 .te4 ..teS is a route to the Canal Variation (4 ttJe3 ttJf6 S d3 d6 6 ..tgS!?), one of the most promising Giuoco lines. White can't reach it via the Four Knights because of 1 e4 eS 2 ttJf3 ttJc6 3 ttJc3 ttJf6 4 .tc4? ttJxe4! and he needs Black's cooperation in the Vienna order. However, he can get to it in the Two Knights Defense, 1 e4 eS 2 ttJf3 ttJe6 3 .te4 ttJf6 ifhe avoids 4 ttJc3? ttJxe4! and plays 4 d3! and then 4.....teS S ttJe3.
It was once assumed Black must reply 7.....td7, Paul Morphy's move, after which 8 ..txe6 bxe6 9 0-0 ..teS and 10 ..te3 is fairly balanced. But recent experience suggests 10 f3! gives White the upper hand.
In a standard Two Knights order, 1 e4 eS 2 ttJf3 ttJe6 3 .te4 ttJf6 4 ttJgS, there are two offbeat but related lines, 4... dS S exdS ttJd4 and S... bS. They don't look at all alike. Yet they transpose after S... ttJd4 6 c3 bS! 7 .tfl! and S... bS 6 .tfl! ttJd4 7 c3.
To avoid that, Black can employ Adolf Anderssen's 7.....teS!? Then 8 .te3 .td7 9 ..txc6 bxc6 10 f3 would be met by 10 .. .'ifh4+ 11 g3 ttJxg3 !. Instead, White usually replies 10 0-0. The point is he transposed into the equal 7... ..td7 line in which 10 ..te3 is played and lost the promising 10 f3 option.
Black's choice should depend on which line is likely to provoke a bad response by White, such as S... bS 6 dxc6?! bxc4 7 ~e2 h6 8 'ilxeS+ JLe7 which favors Black. Hans Berliner revived S... bS in his world postal championship run, explaining that S... ttJd4 'is less clear' after 6 ttJc3.
The risk of this finesse is allowing 7 ... .tcS 8 ttJxc6!? and then 8 ... ..txf2+ 9 ~fl or 9 ~e2!? ..tg4+ 10 ~fl. White also gets an extra option after 8 ..te3 JLd7 9 ..txe6 bxe6 when he tries 30
Double KP Openings
10 liJd2!, which has been doing well lately, rather than transpose with 10 0-0.
The prime benefit is that 6 'tib3, which is a valid move against 5... i.a5, is poor here since 6 .. :i!le7 threatens 7... liJa5 !. But that benefit is trumped by 6 d4!. Black can't reach the safety of the Lasker Defense, 6... i.b6 (7 0-07 d6), in view of 7 dxe5!.
For example, Sveshnikov-Zaja, Bled 2001 went 10 ... liJxd2 11 'i'xd2 0-0 12 liJb3! i.b6 13 0-0-0 'i'e7 14 l::the1 a5 15 a3 a4 16 ..tg5! 'i'e6 17 liJd4 'i'g6 18 f4 and White eventually won.
Therefore he has to enter one of the main Evans lines, such as 6... exd4 7 O-O! d6 8 cxd4 i.b6, where White's claim to compensation has withstood the test of time a century and a half and counting.
On balance, 7 ... i.d7 appears better than 7... i.c5. And Morphy beats Anderssen once more.
Once Black decides on S...i.aS a key question is whether White EVANS GAMBIT should play the immediate 6 'ib3 or The first question to be asked in insert 6 d4 first. More attention is the Evans (1 e4 eS 2 liJf3 liJc6 paid to 6 d4 d6 7 'iVb3 but 7.. :i!ld7 3 i.c4 ..tcS 4 b4 i.xb4 S c3) is seems to be adequate. where Black should retreat the Compare that with 6 'ib3: bishop. There are two ways for it to reach b6 and set up the solid Lasker Now ... 'i'd7 is impossible. The Defense - 5... i.c5 6 0-0 d6 7 d4 other queen moves, 6 .. :iVf6 or ..tb6 and 5... i.a5 6 0-0 d6 7 d4 6.. :iVe7, seem to transpose into bad ..tb6. The order matters because versions of 6 d4 d6 7 'iVb3. For White has other options at move example, 6 'iVb3 'iff6 7 d4 d6? walks into 8 d5! liJ-moves 9 Vi'a4+. six. Consider S•.. i.cS: 31
Double KP Openings
Also on 6 'i!Vb3 'fie7 7 d4 exd4 8 0-0 .tb6 9 cxd4 or 6 ... 'fif6 7 d4 exd4 8 0-0 d6 9 e5! we're in variations that often prove fatal to Black before move 25.
He wants to save c2-c3 for later, such as with 5... d6? 6 d4 exd4 7 c3!?, as suggested by Savielly Tartakower. But Marshall showed that 5 0-0 can be safely answered by 5...tDf6! 6 d4 exd4 7 c3 dxc3 8 e5 d5!.
The real challenge to 6 'ifb3 is 6...'fif6 7 d4 tDxd4!. Then 8 tDxd4 exd4 9 0-0 has been tested a bit in correspondence (9 ... dxc3 10 e5 'fixe5! 11 .txf7+ ~f8) without a conclusion. If White likes the looks of this, 6 'tWb3 would be much better for him than 6 0-0 and perhaps superior to 6 d4, too.
SCOTCWGORING GAMBIT These two gambits are closely related and may reach Giuoco and Two Knights positions. The Scotch order (1 e4 e5 2 tDo tDc6 3 d4 exd4 4 .tc4) gives Black more choice.
Note that early in the Evans, each side can give his opponent a dubious 'something-to-think-about' move. Black's is 4 ... tDxb4?!, rather than 4 ... .txb4. But there's not that much for White to consider and after 5 c3 tDc6 all that's happened is Black transposed to the questionable 4 ....txb4 5 c3 .tc5. White's bid for cleverness worse. It is 5 0-0 after 4 ... .txb4.
The simplest reply is 4...tDf6 which transposes into the Two
1S
32
Double KP Openings
Knights (1 e4 e5 2 tDf3 tDc6 3 ~c4 tDf6 4 d4 exd4). This is a good choice for a Black who likes to face the Max Lange (5 0-0 ~c5) or the anti-Lange (5 0-0 tDxe4). But if Black normally meets 3 ~c4 with 3... ~c5 this is a bad order because he may land in unfamiliar territory, That's why he may prefer 4... ~b4+. His idea is that 5 c3 dxc3 6 tDxc3 transposes into what was long considered his best Goring defense (1 e4 e5 2 tDf3 tDc6 3 d4 exd4 4 c3 dxc3 5 tDxc3 ~b4! 6 ~c4).
His idea is not 6 c3 dxc3?! because that transposes into an unattractive Goring. Better is 6... ~g4 with unclear play. For example, 7 'ii'b3 'ii'd7 8 cxd4 ~xf3 9 dxc5 tDd4 10 'ike3 'ii'g4 11 'iVg5 with a complex endgame in Wist-Rausis, correspondence 1994.
That conclusion is in recent dispute. However White doesn't have to get involved in the debate since in the Scotch order 6 bxc3! favors him (6 ... ~e7? 7 'ii'd5).
One final point is that 4... ~c5 can also transpose, after 5 c3 tDf6 to a Giuoco Piano. That's a safer choice than 5... dxc3?! 6 tDxc3 d6, another unclear Goring line (7 ~g5 tDge7 8 tDd5 f6 9 ~xf6!?).
Well, then, what about the third option in the last diagram, 4...~c5 ? That is a not-so-veiled invitation to the Lange (5 0-0 tDf6 6 e5 d5). Compared with 4... tDf6, Black loses the anti-Lange possibility. But he benefits by not having to face the 5 e5!? line, which Yevgeny Sveshnikov has proven to be dangerous.
So the bottom line of the Scotch Gambit, 4 ~c4, is that 4... ~b4+?! is to be avoided, 4 ... tDf6 is the safe choice and 4 ... ~c5 is the most double-edged. And that's one of the more definitive conclusions we can make about those trap-happy 1 e4 e5 openings.
Another virtue is that after 4.•.~c5 5 0-0 Black has an extra 5... d6 option.
33
Chapter Two: RuyLopez There are far more tricks in the move order of the Ruy Lopez than in any other I e4 e5 opening and they often start as early as move four, as in the Berlin Defense. The bona fide Berliner is content to head into the endgame of 1 e4 eS 2 ttJf3 ttJc6 3 .i.bS ttJf6 4 0-0 ttJxe4 5 d4 ttJd6 6 ..txc6 dxc6 7 dxe5 ttJf5 8 'iixd8+. He might be upset by 4 'iWe2!?
Black should know the counterfinesse, 4 ... a6!. Then 5 .i.xc6 dxc6 6 ttJxe5 'iWd4 regains the pawn. White's best is 5 .i.a4!, but he has to know the WorralllWormald Variation, (usually reached by 3 ... a6 4 .i.a4 ttJf6 5 'iWe2), into which he's transposed. CLASSICAL DEFENSE In the Classical Defense (1 e4 eS 2 ttJf3 ttJc6 3 .i.bS .i.cS) there are two early tabias and various ways of reaching them. One tabia, with ... ttJge7, occurs after 4 c3 ttJge7 S 0-0 .i.b6 or 4 0-0 ttJge7 S c3 .i.b6.
He is unlikely to be happy defending the Steinitz-Defense-like 4....i.e75 c3 d6 6 d4. Nor will he be eager to gambit a pawn, 4 .....tcS!? 5 .i.xc6, e.g. 5... dxc6 6 ttJxe5 'iVd4 7 ttJd3!. That's probably not in his DNA.
There's a consensus that White is at least slightly better after 6 d4 and 34
RuyLopez
11 exd6 cxd6 12 l:.d1 as in Matulovic-Ilincic, Sombor 2004.
his main concern is what he wants to avoid: 4 0-0 will sidestep 4 c3 f5!?, for example. But 4 c3 will avoid 4 0-0 ttJd4.
There's psychology behind 4 0-0 and then 4.•.ttJf6 5 ttJc3!?
What would really turn matters in favor of 4 c3 is if someone found a way to get an advantage from White's extra option, 4... ttJge7 5 d4. But 5... exd4 6 cxd4 ..tb4+! has been equalizing for decades. If Black intends ... ttJf6 instead of ... ttJge 7, there are three orders worth considering. The first and most effective is 3.•.ttJf6! and then 4 0-0 ..tc5. This exploits White's preference for 4 0-0 when he expects Black to continue with a normal Berlin. In this way Black avoids an early c2-c3/d2-d4 in view of3 ... ttJf6 4 c3? ttJxe4!.
What's happened is we've landed in a Four Knights Game (1 e4 e5 2 ttJf3 ttJc6 3 ttJc3 ttJf6 4 ..tb5) in which Black has played a rare sideline, 4 ... ..tc5!? and White replied 5 0-0.
The other two orders start with 3••...tc5 - 4 c3 ttJf6 5 d4 ..tb6 6 0-0 0-0 and 4 0-0 ttJf6 5 c3 0-0 6 d4 ..tb6. Books give much more attention to the first order because it grants White an extra option, 6 'iWe2. But if he intends 6 0-0 anyway, and he doesn't remember the complex 4 c3 f5 theory, the second order is best.
That's fine for Black - if he knows the theory (5 ..• 0-0! 6 ttJxe5 ttJxe5 7 d4 ..td6!).
Alexander Be1yavsky was once surprised by Ljubomir Ljubojevic in this move order and was outplayed after 6.. J~e8?! 7 ttJf3 ttJxe4 8 d4 ttJxc3 9 bxc3 ..tf8 10 ttJg5! h6 11 ttJh3 d6 12 'iif3! ..td7 13 l:.b1 l:.b8 14 'iWg3 ttJe7 It has the added benefit of 15 ..txd7 'iixd7 16 ..txh6!. tempting Black into 4 ... ttJf6 5 c3 ttJxe4? That looks like the 3... ttJf6 FIANCHETTO/COZIO 4 c3? ttJxe4! mentioned above. But DEFENSES this one turns out to be bad after When Black develops his KB on 6 'iWe2!, e.g. 6... f5 7 d3 ttJf6 8 d4 ..te7 9 dxe5 ttJe4 10 ..tc4! d5 g7 he is playing a reversed Vienna, 35
RuyLopez
a mirror image of 1 e4 e5 2 ttJc3 version of the Giuoco Piano in ttJf6 3 g3 Jtb4. But White's extra which Black has an extra ... a6, e.g. tempo in the Lopez version (1 e4 e5 5... d6 6 d4 'ilie7 7 0-0 Jtg7. 2 ttJf3 ttJc6 3 Jtb5 g6) means he There's a back-door route to a can push his d-pawn faster. Black fianchetto and it avoids the He should do that quickly 3... g6 4 c3 issue as well as the more because after 4 0-0 Black has time troublesome 3 ... g6 4 d4 exd4 for 4... Jtg7 5 c3 ttJge7 6 d4 and 5 Jtg5!. This is the Cozio Defense then 6... exd4 7 cxd4 d5! with a (3 ...ttJge7), which transposes to a double-edged game. 3 ... g6 4 0-0 position after 3... ttJge7 4 0-0 g6 5 c3 Jtg7 6 d4 exd4 7 cxd4 More dangerous is 4 c3: d5. If White replies 4 d4 exd4 5 ttJxd4 we get 5... g6 6 ttJc3 !i..g7 7 Jte3 0-0:
Black is in trouble after 4 •.•d6 5 d4 Jtd7 6 'ifb3! (6 ... Jtg7 7 dxe5 dxe5 8 Jtc4 hits f7 and b7). Vassily Smyslov, the maestro of this variation, said 4... a6! is best. Black is pleased to transpose, after 5 i.a4, to a Modem Steinitz Defense (3 ... a6 4 i.a4 d6 and ... g6).
This becomes an old Steinitz Defense, in fact one of the better versions of the old Steinitz, after 8... d6. But Black may do better with Why would Black prefer 3 ... g6 to ... d5!? For example, 8 'ilid2 d5 the normal Modem Steinitz order? 9 exd5 ttJxd5 10 tLlxc6 bxc6 The answer is he avoids other 11 Jtxc6 ttJxe3! 12 'ilixe3 l:.b8 or Modem Steinitz lines, such as 12 !i..xa8 'iixd2+ l3 Wxd2 ttJc4+ 5 O-O! and 5 i.xc6+!? 14 WeI ttJxb2. White might prefer to meet 4 ... a6 with 5 .tc4. That's a pretty solid
There must be a drawback to the Cozio order or everybody would be 36
RuyLopez
playing it. It is 3...tbge7 4 tbc3! which gets very complex after 4..•g6 S d4 exd4 6 tbdS!. Whether White has an advantage after 6... i.g7 7 i.gS is in dispute. Yet few Lopez players are willing to try it and that makes the Cozio order a good practical weapon.
After S 0-0 the most popular move in recent years is S...i.cS.
BIRD'S DEFENSE Henry Bird's defense (1 e4 eS 2 tbO tbc6 3 i.bS tbd4 4 tbxd4 Now 6 d3 c6 7 i.c4 dS! heads exd4) looks absurd: Black moves his only developed piece a second towards the good S d3 line. More time in order to double his own accurate is 6 i.c4!, breaking the pawns. rule against moving a piece twice in But those are doubled center the opening. Black must deal with pawns and they do a good job of the threat of 7 i.xf7+ ~xf7 8 'iihS+ restricting White after S d3 c6 and doesn't have time for ... dS. 6 i.c4 dS!. For example, 7 exdS cxdS 8 i.bS+ i.d7 9 i.xd7+ 'ii'xd7 100-0 i.cS! 11 l:.e1+ tbe7 is quite nice for Black.
STEINITZIANA
Wilhelm Steinitz's solution to the Lopez problem was maintaining a White's error was committing himself too early (S d3?!). The best solid bridgehead on eS, after 1 e4 eS waiting move in many if not most 2 tbo tbc6 3 i.bS, by means of openings is castling. Here S O-O! 3 ... d6. His disciples tried 3... d6 allows him to decide between 6 d3, 4 d4 i.d7 S tbc3 tbf6 6 0-0 i.e7 or 6 c3, 6 i.c4 or even 6 'ii'hs after he its transpositional kin such as 3••• tbf6 4 0-0 i.e7 S tbc3 d6 6 d4 sees Black's fifth move. i.d7. He has excellent chances following S... c6 6 i.c4 dS 7 exdS A key position arises after White cxdS 8 i.bS+ i.d7 9 l:.e1+ tbe7 protects his e-pawn with 7 and 10 i.xd7+ 'ii'xd7 and now l1"iihS threatens to win Black's. That leads or 11 c4 (rather than transposing to a tabia, 7... exd4 8 tbxd4 0-0 with 11 d3?). 9 i.xc6 bxc6.
ne1
37
RuyLopez
White is better but Black has play along the b-file and from ... c5!/ ... iJ.c6, e.g. 10 iJ.f4 c5 11 ltJD iJ.c6 12 'it'd3 .!tJh5! 13 iJ.e3 l:te8 14 l:Iadl 'ifb8 15 iJ.c1 'iib7, Narciso DublanMizoev, Salou 2006.
And now 5... exd4! 6 ltJxd4 ltJf6 7 0-0 iJ.e7. Then 8 iJ.xc6 or 8 l:Ie 1 0-0 9 iJ.xc6 heads towards the double-edged ... bxc6 tabia that should make Black happy. Unfortunately for him there's a counter-counter-finesse. White can move up the capture on c6, as Jackson Showalter showed more than a century ago, 3... d6 4 d4 .td7 and then 5 iJ.xc6 iJ.xc6 6 .!tJc3. This transposes to the favorable line after 6 ...ltJf6 7 'it'd3 exd4 8.!tJxd4.
Experience shows that White should try to force Black to retake on c6 with the bishop, thereby eliminating ... c5! counterplay. He does that with 3... d6 4 d4 iJ.d7 5 .!tJc3 .!tJf6 6 iJ.xc6! since 6... bxc6? 7 dxe5 drops a pawn.
Of course, Black can accelerate his plan, too, with 4 d4 exd4, as Alexander Onischuk does. But there's a difference because White can retake with his queen, 5 'it'xd4 iJ.d7 6 iJ.xc6, which has been known to favor White since Morphy's day.
Instead 6... iJ.xc6 7 'it'd3 renews the dxe5 threat and favors White following 7 ... .!tJd7 8 iJ.e3 exd4 9 iJ.xd4! or 7... exd4 8 .!tJxd4 iJ.d7 9 iJ.g5 iJ.e7 10 0-0-0 and 8... g6 9 iJ.g5 iJ.g7 10 .!tJxc6 bxc6 11 e5 dxe5 12 'it'D 'it'd6 13 l:tdl (Lupulescu-Pessi, Bucharest 2004).
To resuscitate the Steinitz Black needs a new idea, perhaps the littleexplored 5....!tJge7!? in the last order, or something that confuses White about the timing of iJ.xc6. He might try the Berlin move order 3... .!tJf6 4 0-0 and then 4 ... d6 or
Black's counter-finesse lies in moving up ... exd4 before iJ.xc6 so that he can retake with the b-pawn. He accomplishes this after 3... d6 4 d4 iJ.d7 5 .!tJc3: 38
RuyLopez
4 ... i.e7 5 :tel d6. This has the benefit of denying White his most aggressive plans such as 'iVd3/0-0-0 and avoiding 4 d4. But White should be aware after 3•••lbf6 4 0-0 d6 5 d4 i.d7 that he has 6 i.xe6! i.xe6 7 lbe3.
The liquidation that follows 7••. lbxe4 8 lbxe4 i.xe4 9 :tel favors him (9 ... f5 10 lbg5 or 9 ... i.f5 10 dxe5 d5 11 lbd4). Instead, Black usually transposes with 7... exd4 8 lbxd4 into a ... i.xc6 line, rather than ... bxc6!.
most obvious improvement is that by inserting 3... a6 4 i.a4 before 4 ... d6, Black makes d2-d4 a trifle risky. After 1 e4 e5 2 lbf3 lbe6 3 i.b5 a6 4 i.a4 d6 5 d4 Black can reply 5...b5! 6 i.b3 lbxd4 7 lbxd4 exd4
Now 8 'ifxd4? is the Noah's Ark Trap (8 ... c5 9 'i'd5 i.e6 10 'iVc6+ i.d7 11 'i'd5 c4 and wins). But 8 e3! and then 8... dxc3 9lbxc3lbf6 10 0-0 is a worthwhile gambit.
White's most flexible fifth move is 5 O-O!, which gives him all sorts Bottom line: The Steinitz of finesses. He can, for example, deserves its reputation but only if meet 5... i.d7 with 6 c3. This White times i.xc6 correctly. If usually transposes into the old Black likes the ... bxc61ines he has a Modern Steinitz main line, 5 c3 much better chance of reaching i.d7 but without risking the Siesta them via a delayed Steinitz, such as Variation (5 c3 f5!?). 3... a6 4 i.a4lbf6 5 0-0 i.e7 6 :tel d6 when White's best is 7 i.xc6+ Another virtue concerns Oldrich bxc6 8 d4 exd4 9lbxd4 i.d7. Duras's idea, c2-c4, It's not considered dangerous because 5 e4 MODERN STEINITZ i.g4! exploits the d4-hole. But after DEFENSE 5 0-0 i.d7 White can seek a This is called the 'Improved superior Duras with 6 e4. Then Steinitz Defense' in Russia and the 6... i.g4 would cost Black a tempo 39
RuyLopez
and it's not clear if that or 6... tbf6 7 tbc3 tbd4 equalize.
7 .txc6 and get a favorable old Steinitz position. '
The most common continuation after 5 0-0 .td7 is 6 d4.
He means 7....txc6 8 %:tel exd4 9 tbxd4 because 8... tbxe4? 9 d5 and 8... .txe4 9 tbc3 favor White. As often happens with transpositions, both players may be right. White got the middlegame he wanted and Black avoided the middlegame he sought to avoid. Lajos Portisch added another finesse to the Modem Steinitz when he met 5 0-0 with 5....te7!?
This is an improved gambit in view of 6•.. b5 7 .tb3 tbxd4 8 tbxd4 exd4 9 c3 because ... .td7 is not as useful as ... .tb7. And 6 ... tbge7 7 c4! is an excellent Duras, e.g. 7...tbg6 8 tbc3 .te7 9 .te3 0-0 10 %:tel .tg4 11 .txc6 bxc6 12 h3 as in SmaginBabu, London 1990. This looks like an innocuous route to quiet lines, such as 6 c3 tbf6 7 !:tel 0-08 h3 .td7. But there are differences between 5... .te7 and 5....td7.
Theory prefers 6...tbf6!. Then both players can claim they won the battle of the opening. Black says, 'J tricked you into a Steinitz Defense Doubly Deferred without allowing you a chance to use your best weapon.'
One is that Black is better situated to play 6... .tg4 compared with the somewhat dubious 5 0-0 .tg4.
Translation: This position can be reached via 3... a6 4 .ta4 tbf6 5 0-0 d6 6 d4 .td7 but only if White passes up 6 .txc6+!.
For example, 6 c3 .tg4 7 d4 b5 8 .tb3 tbf6 intending ....txf3. Also 'No,' replies White. 'J got what J 7 d3 tbf6 8 tbbd2 tbd7 9 h3 .th5 wanted because J can now play 10 .tc2 tbf8!? was good for Black 40
RuyLopez
(11 g4 .tg6 12 d4 h5) in BrownePortisch, Milan 1975.
followed by ...ti:Je7. The forgotten 5....te7 deserves to be remembered.
Another difference is Black can try to upgrade the Siesta Variation by meeting 6 c3 with 6... f5!? 7 exf5 .txf5 8 d4 e4.
Black can also mix the Modern Steinitz with the Cozio if he plays 1 e4 e5 2lbf3lbc6 3 .tb5 a6 4 .ta4 and now 4...lbge7.
More of a test of 5... .te7 is 6 d4. Then 6...b5 7 .tb3 lbxd4 8 lbxd4 exd4 9 c3 is another center gambit. After 9 ... lbf6! 10 cxd4 i.b7 11 f3 0o we've transposed into a roughly even line (5 d4 b5 6 i.b3 lbxd4 7lbxd4 exd4 8 c3 .tb7 9 cxd4lbf6 10 f3 i.e7).
Alexey Suetin recommended '5 c3!' and cited one of his games that led to a 5...b5 6 .tb3 d5 7 'ife2! edge.
If there is a refutation of 5... .te7 it is 6 .txc6+ bxc6 7 d4. Then 7 ... f6?, which is thematic in comparable .txc6+ lines of the old Steinitz, makes the kingside pieces stupid. And 7... exd4 8 lbxd4 But Black improves with 5... d6!. Then 6 d4 .td7 transposes to the Modem Steinitz line that runs 5 c3 .td7 6 d4 lbge7. Theory likes White's position more today than it did a generation ago but what may be more important is Black avoided the non-5 c3 options of the Modem Steinitz. Moreover after 4 ... lbge7 5 0-0 Black can safely transpose into the fianchetto line (5 ... g6 6 c3 d6). White may have to challenge But 5... .te7 is playable if Black 4 ...lbge7 with 5 d4 exd4 6 lbxd4, delays ... c5 by means of 8....td7 which so far has remained unclear 9lbc3 lbf6. Or, better yet, 9....tf6! after 6... g6 or 6... b5!? 7 .tb3 lba5.
... allows White to set up a bind after 8... c5 9 lbc6! 'ifd7 10 lba5! (10 ... lbf6?! 11 c4).
41
RuyLopez
EXCHANGE VARIATIONS White can play Jlxc6 at move four, five or six in the main Lopez lines. Exchanging at move five makes sense ifhe wants to avoid the Open Defense and retain the 0-0-0 option. Eduard Gufeld beat Mikhail Tal in a game that went 1 e4 e5 2 ttJf3 ttJc6 3 Jlb5 a6 4 Jla4 ttJf6 5 Jlxc6 dxc6 6 d3 ttJd7?! 7 ttJbd2 Jle7 8 ttJc4 Jlf6? and then 9 'iVe2 c5 10 Jld2 0-0 11 g4 with a winning attack (ll...bS 12 ttJe3 g6?! 13 h4 ttJb8 14 0-0-0 ttJc6 IS :dgl Jle6? 16 ~bl ttJb4 17 Jlxb4 cxb4 18 gS! and ttJg4).
This can transpose to a Steinitz Defense Doubly Deferred line (7 :el d6) that Akiba Rubinstein liked to playas Black and is similar to the better forms of the old Steinitz after 8 d4 exd4 9 ttJxd4 Jld7. White might improve with 7 ttJxe5 ttJxe4 or 7 d4 ttJxe4 8 dxeS 0-0 (9 :e 1 dS 10 ttJd4 Jld7 11 f3 ttJgS as in I.Sokolov-S.Nikolic, Pancevo 1987) but there's too little experience to tell.
A major benefit of a delayed Jlxc6 is that Black finds it harder to defend his e-pawn. After 3 ... a6 4 Jla4 ttJf6 S Jlxc6 dxc6 he cannot play... f6 as he would after 4 Jlxc6 dxc6. However, 6... Jld6! is a good developing defense and Black stands well after 7 ttJbd2 Jle6.
OPEN DEFENSE Before getting to the main Open line, 1 e4 e5 2 ttJf3 ttJc6 3 Jlb5 a6 4 Jla4 ttJf6 5 0-0 ttJxe4 6 d4 b5 7 Jlb3 d5, let's consider 7...Jle7!?
The reason the doubly delayed exchange, 5 0-0 Jle7 6 Jlxc6 dxc6, has a better reputation is that 7 ... Jld6 would cost a tempo. Instead, Black obtains a solid game with ... ttJd7. If he wants a unbalanced middle game, with greater winning chances, he should consider the rare 6...bxc6!?: 42
RuyLopez
i..e3, lbbd2) and at Black's ( ... i.e7, ...lbc5, ... lba5) can be mixed in a variety of orders, making transpositions inevitable.
This tries to lure White into 8 :el and then 8... d5 9 dxe5 .te6, a dead even position that could come about after 7 ... d5 8 dxe5 .te6 if White plays the inept 9 l:.el?! .te7.
In the first edition of ECO, Viktor Korchnoi endorsed the conventional wisdom that 9 c3 .te7 10 .te3 and 9 c3 .te7 10 lbbd2 were so favorable that Black should avoid them with 9 ... .tc5!. But Andrei Sokolov showed how White can trick his opponent into a favorable line with 9 .te3!? and then 9 ....te7 10 c3!. This is very effective against booked-up 9 c3 .tc5 opponents.
The best try at punishing 7 ... .te7 is 8 dxe5. Experience with 8...lbc5 9 i.d5 .tb7 (or 9 ...lbe6) has been far from convincing, e.g. 9 ... .tb7 10 lbc3 0-0 11 .tf4 lbe6 12 .tg3 f5!. If Black likes these positions he can get to them even if White adopts a standard way of avoiding the Open Defense, 5 d4. This often favors a better-prepared White after 5... exd4 6 0-0 or 6 e5. But Black has 5.•.b5!? instead:
For instance, 10 ... 0-0 11 lbbd2 'ild7 12 :el :ad8 13 .tc2 lbxd2 14 'ifxd2 .tg4 15 'ifd3 g6 16 .th6 :fe8 17 :adl lba5 18 h3 .txf3 19 'ilxf3 c5 20 'ilg3 lbc4 21 b3 lba3 22 e6! 'ifc6 23 .tf5! and wins, Kamsky-Piket, Dos Hermanas 1995.
Then 6 dxe5! lbxe4! 7.tb3lbc5 8 .td5 i.e7 9 0-0 transposes to the previous line. The tabia of the Open Defense arises after 1 e4 e5 2 lbf3 lbc6 3 .tb5 a6 4 .ta4 lbf6 5 0-0 lbxe4 6 d4 b5 7 .tb3 d5 8 dxe5 i.e6. The moves at White's disposal (c2-c3,
There is no obvious antidote to 9 .te3 but attention has shifted to 9 lbbd2. It allows White to answer 9 ... .tc5 with 10 lbxe4 dxe4 43
RuyLopez
11 it.xe6 with at least a small White edge. That often prompts 9 ... it.e7, after which 10 c3! is another good transposition.
8 ... l:tb8 9 axbS axbS, he can reach a much improved Open Defense, 10 dxeS it.e6 11 c3. But 8 a4? was knocked into a footnote by the Lasker-Schlechter match of 1910 when Black equalized with 8... tiJxd4!'
The downside to 9 tiJbd2 is that White has diminished control of d4, compared with 9 it.e3 and 9 c3. That's significant after 9.•.tiJcS!.
The appeal of the tweaked move order, 8 dxeS it.e6 9 a4, is that 9 ... l:tb8 allows White to reach that ideal version of 8 a4 following 10 axbS axbS 11 c3. For instance, 11...it.e7 12 tiJd4! tiJxeS (12 .. :iVd7 13 it.c2!) 13 f3 tiJcs 14 it.c2 it.d7 IS b4 with advantage, AlekhineRohacek, Munich 1941.
Then White would be happy to see 10 c3 it.e7 11 it.c2 it.g4 because that transposes to yet another good 9 c3 line (9 c3 tiJcs 10 it.c2 it.g4 11 tiJbd2 it.e7). For example, 12 .:tel 'ifd7 13 tiJf1 lId8 14 tiJe3 it.hS IS tiJfS. But Black has been doing better, after 10 c3 with 10•.• d4. That has become a main line, with book analysis stretching past move 20.
So what's wrong with 9 a4 ? Vintage theory says 9•..tiJaS is the answer. But 10 axbS axbS 11 tiJd4! it.cs 12 c3 0-0 13 it.c2 is hardly convincing. More testing is needed (say of 9 •.• b4 10 as, threatening it.a4) before we can properly evaluate 9 a4!?
A trickster will prefer yet another ninth move, 9 a4!?
MOELLERIARKHANGEL DEFENSE
White is trying to slip into a forgotten line that begins 1 e4 eS 2 tiJf3 tiJc6 3 it.bS a6 4 it.a4 tiJf6 S 0-0 tiJxe4 6 d4 bS 7 it.b3 dS and now 8 a4, instead of 8 dxeS. If White gets control of an open a-file, 44
There's no better example of how transpositions can improve a group of related openings than the aggressive systems in which Black mixes ... a6/... tiJf6 with ... i.cs and possibly ... bS/ ... i.b7. Up until fairly recently they seemed like distant cousins unworthy of our attention: The Moeller Attack (1 e4 eS
RuyLopez
2 ~f3 ~c6 3 i..bS a6 4 i..a4 ~f6 S 0-0 i..cS) was considered too flaky because of 6 c3 i..a7 7 d4 or 6 ~xeS ~xeS 7 d4. The Arkhangel, S•.. bS 6 i..b3 i..b7, enjoyed 15 minutes of fame in the 1960s. That expired when White obtained an edge on both wings with 7 .l:tel i..c5 8 c3 d6 9 d4 i..b6 10 i..g5 and a2-a4. But then came the 'New ~ ~ Arkhangel,' S...bS 6 .... b3 .... cS.
commonly come from 5 'iVe2 b5 6 i..b3 i..c5 7 0-0 and lead to roughly equal chances. If White abandons c2-c3/d2-d4 and plays 7 d3 Black can safely revert to the old Arkhangel, 7... 0-0 8 a4 i..b7. That's an 'old' position that occurs after 5 0-0 b5 6 i..b3 i..b7 when White passes up the critical lines such as 7 .:tel. Again Black gets a free ride in the opening. A key test of the New Arkhangel is 7 c3 d6 8 d4 i..b6:
Black's cunning lies in holding back his QB until he knows whether it serves best on g4 or b7. For example, on 7 .l:tel he avoids 7... i..b7 - which transposes to the unfavorable old Arkhangel - and plays 7... ~g4! in the Moeller spirit. That's good for him, e.g. 8 d4? ~xd4 9 ~xd4 'ii'h4! or 8 .l:te2 ~d4 9 ~xd4 i..xd4, threatening 10 ... 'ii'h4.
The difference between new and old is revealed by 9 .:tel. Then 9 ... i..b7? transposes to an inferior 'old' position. But after 9...i..g4! we are 'new.' White can't support his center on the current squares and he has no edge after 10 d5 ~e7 11 h3 i..d7! and ... c6. What about a2-a4, the usual antidote to an early ... b5 ? That shouldn't trouble Black in the last diagram, e.g. 9 a4 i..g4! 10 axbS axbS 11 .l:txa8 'iVxa8 12 h3 i..hS
If White concludes that he should be playing 7 'i'e2 instead, he is merely transposing into one of the Worrall/Wormald tabias that 45
RuyLopez
has perfonned well, e.g. 13 "ifd3 exd4 14 ii.g5 ii.xf3 15 ii.xf6 ii.xg2!, Pavlovic-Carlsen, Reykjavik 2006.
face the more critical 8...l:tb8 and 8... ii.g4 lines. Despite that, the .. .ii.c5 family is respectable today - because so far Black has found more good transpositions than White.
Is there nothing but good news for Black in the New Arkhangel? Not quite. He has to know what to do after 7 ttJxeS. The amount of MARSHALL and theory on 7 ... ttJxe5 8 d4 ii.xd4 ANTI-MARSHALL 9 'ilfxd4 d6 has exploded in recent years and the debris hasn't settled In the 1930s Frank Marshall's yet. gambit was considered refuted. In Moreover, White can set his own the 1960s it was resuscitated by transpo traps. With 7 a4 he forces Boris Spassky - as a drawing Black to meet the threat ofaxb5. weapon. Today it is perhaps the most dangerous Ruy Lopez Then 7...ii.b7 defense. That adds huge significance to Black's seventh move: 1 e4 eS 2 ttJf3 ttJc6 3 .tbS a6 4 .ta4 ttJf6 S 0-0 .te7 6 l:tel bS 7 ii.b3.
8 d3 is another quiet line, an antiMarshall in which .. .ii.c5 has been played in place of ... ii.e7. That would seem to benefit Black but White can confuse matters with 8 c3, hoping for 8...d6 9 d4 ii.b6 10 ii.gS!, a favorable 'old' position.
With 7 ... d6 Black commits himself to a main-line variation (8 c3 0-0 9 h3 etc.). But 7... 0-0 allows him to threaten to play the Marshall Gambit, 8 c3 d5!?
What happened: We've transposed into 7 c3 d6 8 a4!? ii.b7 9 d4 ii.b6 10 ii.g5. This helped White because he didn't have to
It's just a threat because he can 46
RuyLopez
back out with 8... d6, tranposing into the main line. Yet the threat is so worrisome that today's strongest GMs skirt the issue with modest, if not cowardly anti-Marshall moves like 8 d3 or 8 h3. Why then should anyone play 7... d6 if it enables White to play more aggressively without risk? The only downside to 7... 0-0 is supposed to be that 8 a4 is stronger. It is - but not significantly, after 8 a4.i.b7. That's why top GMs have been hunting for an aggressive way to sidestep the Marshall. One idea is 8 d4.
Alexander Grischuk among others. And what about other antiMarshall lines? The simplest begins with 8 d3. Then 8... d5 is considered unsound since c3 is available for a knight (9 exd5 tiJxd5 10 tiJxe5 tiJxe5 11 :xe5 c6 12 l:te 1 .i.d6 13 tiJc3!). Black typically meets 8 d3 with 8...d6. Following 9 c3 White will eventually play d3-d4. This ensures he can execute the tiJf1 maneuver that is so important to the Lopez. He plans 10 tiJbd2, 11 tiJf1, 12 tiJg3 and eventually d3-d4. He doesn't get to do that in some 8 c3 lines, such as the Zaitsev Variation (8 ... d6 9 h3 .i.b7 10 d4 :e8 11 tiJbd2 .i.f8 12 tiJf1? exd4! and Black wins a pawn).
The downside to 8 d3 d6 9 c3 is White will be a tempo behind traditional 8 c3 d6 9 h3 lines when he eventually advances d3-d4. But that tempo may be a minor concern This works best when Black compared with dealing with the becomes concerned about the looks Marshall. of 8... tiJxd4 9 tiJxd4 exd4 10 e5 Note that one of the few active tiJe8 and bails out with 8... d6. responses Black has is 9 ... .i.e6. Then 9 c3! transposes into 8 c3 Black seems to be doing well when d6 9 d4 .i.g4. Books say that old you compare this with 8 c3 d6 9 h3 line is second-best to 8 c3 d6 9 h3. .i.e6 10 d4!, which theory says But second-best doesn't mean it's favors White after 10 ... .i.xb3 bad and the old line is used by Gata 11 axb3 :e8 12 d5! and 13 c4. Kamsky, Magnus Carlsen, and But is Black equal after 8 d3 d6
47
RuyLopez
9 c3 .te6 and then 10 d4 ?
The test of Geller's order is 9... tbxd4, since 10 .txt7+ l1xt7 He is a tempo ahead of the 11 tbxe511f8! 12 'iixd4 c5 is a good previous line. But all that means is gambit and 10 tbxd4 exd4 11 e5 that h2-h3 is missing. Black best tbe4 12 'iig4 d5 has performed well may be 10... i.g4!, transposing to for Black. Nevertheless, 9 d4 may 8 c3 d6 9 d4 .tg4, that old, be worthwhile because allowing the reasonable system for White. Marshall is worse - and because the Another anti-Marshall finesse is most natural response from Black is 8 h3. Then if Black continues fleeing with 9 ... d6 10 c3 into 8••• d6, as he often does after 8 d3, something more familiar to him. White can shoot back 9 c3!. He has transposed to 8 c3 d6 9 h3 while Because of the forcing nature of dodging the Marshall. the main Marshall lines, there is Books used to dispose of 8 h3 with the comment that 8 ....tb7 keeps the gambit alive. That is, 9 c3 d5! 10 exd5 tbxd5 11 tbxe5 tbxe5 12 l1xe5 tbf4! is a prohibitively strong attack.
little room for transposition following 8 c3 d5 9 exd5 tbxd5 10 tbxe5 tbxe5 11 l1xe5 c6. An exception occurs in Fischer's idea of 12 g3.
Yefim Geller pointed out the virtue of 9 d4!? instead. Then 9..• d6 10 c3! transposes to the Zaitsev Variation. If White prefers the Zaitsev to the Marshalland most Lopez players do - then he's scored a minor victory.
48
RuyLopez
This is designed to bypass the mega-theory of 12 d4 .td6 13 .:tel 'iVh4 14 g3 'iVh3.
Arthur Bisguier once explained in Chess Review why he used 5•.. b5 6 .tb3 d6!?
The simplest way of handling 12 g3 is 12 ....td6 13 .:tel: And now 13 ..:iVd7! and ... 'iWh3 transposes to the line White is trying to avoid. One of the few variables in the Marshall is the timing of .txd5, if that's White's wish. He can do it with 12 d4 .td6 13 .:tel 'iWh4 14 g3 'iVh3 15 i.xd5 exd5 or wait for First, he wanted to avoid his 15 .te3 .tg4 16 'iVd3 .:tae8 and opponent's favorite Delayed then play 17 .txd5. Exchange (5 ... .te7 6 .txc6) Or he can wait further, 17 ttJd2 .:te6 18 'iWfl 'ifh5 19 .txd5. He can
Variation. Second, White 'couldn't be sure I would transpose into the main line and he felt compelled to analyze the complexities of 7 ttJg5. ' Opinions vary about the soundness of7 ... d5 8 exd5 ttJd4.
even reach the same positions with 12 i.xd5 exd5 13 d4 i.d6 14 .:tel 'iVh4 15 g3 'ifh3. Which is best? Perhaps the last order because - aside from the extra 14 .:te3 option - it bypasses another Geller line, 12 d4 i.d6 13 .:te 1 'iVh4 14 g3 'iVh3 15 i.e3 i.g4 16 'iVd3 ttJxe3!? 17 .:txe3 c5.
White eventually selected 7 e3 and Black transposed after 7... i.e7 8 h3 ttJa5 9 i.e2 e5 10 d4. But by then he had gained more than an hour (!) on the clock.
ROAD MAPS TO THE RUYMAIN Black reaches the main Ruy tabia after 1 e4 e5 2 ttJf3 ttJe6 3 i.b5 a6 4 i.a4 ttJf6 5 0-0 i.e7 6 .:tel b5 7 .tb3 0-0 8 e3 d6 9 h3. Other orders give White more options but that also means more to think about. 49
RuyLopez
That's an extreme case of 'giving him something to think about. ' Like other alternative routes to the Ruy tabia, this one comes with a price. White has saved a tempo compared with the usual main lines. He doesn't have to play l:.el, since ~c2 protected the e-pawn. In other routes he can avoid h2-h3 because ... ~g4 is not a concern. Yet This position occurred several as we'll see White may have to play times in the 1950s as a means to a h2-h3 and l:tel after all and give Marshall-like gambit, 8... d5. back the tempi. If he does, Black can derive some short-term gains However, Black also has 8.••d6, through these alternative orders. with the hope of reaching an Another is 1 e4 eS 2 ltJf3 ltJe6 improved, or at least confusing, 3 ~bS a6 4 ~a4 ltJf6 S 0-0 bS version of the Ruy main line after 6 ~b3 and now 6... ~e7. Then 9 d4 ltJaS!? That leaves a pawn hanging but no one who has faced 7ltJg5? O-O! is pointless. this as White seems to trust 10 dxe5 This order has the usual ... b5 dxe5 llltJxe5 ltJxb3. drawback of allowing 7 a4. Then 7... b4 has a poor reputation because Instead 10 .te2 eS has been of 8 'iVe2 0-0 9 a5! and 10 c3. played. When Paul Keres reached Perhaps better is 7... ~b7 and this position he rejected 11 dxe5 8 axb5 axb5 9 l:txa8 and 10 ltJc3 dxe5 12 'i'xd8+ l:.xd8 13 ltJxe5 ~xe4! and played llltJbd2.
but there is little experience to prove it.
Another all-but-forgotten order What's happened is that we've (after 1 e4 e5 2 ltJf3 ltJc6 3 ~b5 a6 reached a main Lopez line but with 4 ~a4 ltJf6 5 0-0) is S•.• ~e7 6 l:.e1 the omission of h2-h3 and ... 0-0. bS 7 .ib3 ~b7. This also comes That should favor White, not Black. about via the Arkhangel if Black But it's hard to prove that if Black puts his KB on e7 rather than c5. continues to delay castling with White has nothing better than 8 e3: 1l ••• exd4 12 exd4 l:.e8.
50
Ruy Lopez
This is not just a trap, 13 dxe5 dxe5 14 liJxe5? 'iie7!. It deserves the serious testing it hasn't gotten.
Instead he usually plays 14 i.b1 and then 14... 0-0 15 liJf1. That transposes to a position that normally comes from 5 0-0 ~e7 6 :el b5 7 ~b3 0-0 8 c3 d6 9 h3 liJa5 10 .tc2 c5 11 d4 'iic7 12 liJbd2 cxd4 13 cxd4.td7 14 liJf1 :ac8 when White passes up 15 liJe3! and 15 .l:!e2! and settles for 15 ~b1.
OLD MAIN LINE The way the main Ruy tabia was reached for much of the 20th century was 5 0-0 ~e7 6 :e1 b5 7 ~b3 d6 S e3 and now S•••liJa5 9 .te2 e5. This was considered a finesse because it avoided 8... 0-0 9 d4, at a time when that line was regarded as potent. The 8... liJa5 order began to disappear in the 1950s but was revived by Bent Larsen and Oleg Romanishin.
In this order White didn't get that choice and the result is a roughly equal position (15 ... l:.fe8 16 liJg3 liJc6).
Why was it revived? One reason is that by delaying castling Black can meet 10 d4 'iie7 11 h3 with a quick ... l:tc8 that threatens the c2-bishop earlier than usual.
Another plus of the Old Main Line is that Black isn't castled after 8... liJa5 9 .tc2 c5 10 d4 'iic7 11 h3 liJe6. That means 12 d5? - which is
For example, 11 ...~d7 12 liJbd2 exd4 13 exd4 :eS. Since White hasn't had time for liJf1, he can't defend the bishop with :e2 or liJe3, as he does in the orthodox main line.
a good move when liJbd2 and ... 0-0 are added - turns out well for Black after 12 ...liJbS 13 liJbd2. Instead of 13 ... 0-0? 14 a4 ...
51
RuyLopez
use the 11.. ..id7 12 ltJbd2 cxd4 13 cxd4 :c8! 14.ib 1 0-0 trick. MAINLINES A modem main line runs 1 e4 e5 2 ltJf3 ltJe6 3 .ib5 a6 4 .ia4 tiJf6 5 0-0 .ie7 6 :e1 b5 7 .ib3 0-0 8 e3 d6 9 h3 and now 9...ltJa5 10 .ie2 e5 11 d4. Black typically defends the attacked e-pawn with ...'fic7, ... ltJd7 or ...ltJc6, with or without a trade of c-pawns.
... he can attack with 13 ... g5! (14 ltJxg5 :g8), with advantage according to Aleksandr Matanovic.
Often he adds ... :e8 so that after ... .if8 he smoothes out his development and puts pressure on e4, e.g. 11...'ifc7 12 ltJbd2 cxd4 13 cxd4 .ib7 14 ltJf1 :ac8 and 15 ...:fe8.
The drawback to the Old Main Line is that White saves a tempo by not playing h2-h3. But that may be temporary. For instance, 11 ltJbd2 0-0 12 ltJf1?, rather than 12 h3!, allows 12 ... cxd4 13 cxd4 .ig4! with excellent play.
More economical is Yaacov Murey's gambit idea, 11 ...:e8!? It's based on 12 dxe5 dxe5 13 ltJxe5 .ib7 with compensation for the sacrificed pawn.
Tal was plainly frustrated with the Old Main Line when Larsen adopted it against him in 1991. Tal knew that the books claimed 10 d4 'ifc7 11 a4 was the refutation. But as he studied the position he concluded 11 ... e4! (instead of the book 11...h4?) was not at all bad. He also noticed that after 11 ltJbd2 Black can reply 11 ....id7 12 d5 0-0 and when White plays ltJf1-e3, Black replies ... ltJg4!, which looked good. So after considerable thought Tal just played 11 h3 and transposed after all.
For example, 14 ltJd2 .id6 or 14 'iff3 .if8 15 .if4 :e6 16 ltJg4 ltJc4 17 ltJxf6+ :xf6 18 tiJd2 tiJe5, Chandler-Hebden, Millfield 2000.
This was the final game between these two great warriors and it ended in a draw after Black got to 52
RuyLopez
But the principal idea of 11 .. J::te8 What made the Zaitsev playable is to save a tempo by omitting was the realization that 9...:e8 ... 'iic7. For example 12 liJbd2.tts 10 liJg5: 13 liJf1 gives Black good counterplay after 13... cxd4 14 cxd4 exd4 15 liJxd4 i.b7. Compare this with 11...'iic7 12 liJbd2 :e8 13 liJfl .tfS when White gets the upper hand from 14 i.g5!. By avoiding ... 'iic7 Black may find a better use for the queen, e.g. Now 10...:ts! is perfectly safe 11...l:te8 12 liJbd2 .tfS 13 b3 liJc6 for Black. White has nothing better 14 i.b2 .td7 15 a3 'iib6! with good than 11 liJf3 !, repeating the play. In contrast, 11 ... 'iic7 12 liJbd2 position. In fact, grandmaster draws :e8 13 b3 i.fS 14 liJfl favors have ended here or after a few more White. repetitions. Igor Zaitsev's contribution to the Ruy occurs when Black meets the tabia's 9 h3 with 9...:e8 10 d4 i.b7. He prevents the liJd2-fl maneuver that is so valuable in the Ruy, as noted earlier.
Zaitsev's original move order is inexact if Black wants to avoid a draw as well as the complications of 9...:e8 10 a4 liJa5 11 .ta2. More precise is 9....tb7! first and then 10 d4 :e8, transposing.
53
Chapter Three: Sicilian Defense Transposition tricks punish routine moves more often in the Sicilian than in any other opening. That might be expected in sharp Sicilian lines but it also applies to the closed variations.
White has a strong sacrifice, 8 fS! gxfS 9 tbh4 fxe4 10 dxe4 followed by tbf5 or 'i'h5, e.g. 10 ... ~e6 11 tbf5 .ltxf5 12 exf5 f6 13 tbc3 0-0 14 tbd5, Korolyev-Varlamov, correspondence 1981.
For example, White can try to save a tempo compared with the traditional closed order, 1 e4 cS 2 tbc3 tbc6 3 g3, with 2 g3. The Russian postal champion Sergey Korolyev showed the benefits of this following 2...tbc6 3 .ltg2 g6 4 d3 .ltg7 S f4 d6 6 tbf3.
If Black chose 6... tbf6 instead, Korolyev was content to transpose to a normal closed with tbc3, since many players think ...tbf6 is one of the less effective closed defenses.
The drawback to 2 g3 is 2... dS!. White doesn't have time for 3 d3 and 4 tbd2 because of 3... dxe4! and Black has been playing auto-pilot he may not like 3 exd5 'iVxd5 4 tbf3 moves but if he continues 6..•eS .ltg4 5 .ltg2 'iVe6+!. 70-0 tbge7?!: Alexander Morozevich refined this order with 2 d3!? That enables White to transpose, after 2... g6 3 g3 .ltg7 4 .ltg2 tbc6 5 f4 d6 6 tbf3 e5 7 0-0 tbge7 8 f5!, to Korolyev's sacrifice. But it also allows him to safely meet a quick ... d5 with tbd2! and obtain a satisfactory King's Indian Reversed.
Thanks to his saved (tbc3) tempo 54
Sicilian Defense
CHAMELEON SICILIAN This elastic system is based on White's ability to delay a decision of whether to open the center with d2-d4. It works because the sophisticated Black moves of an open line, such as ... a6 and ...'iic7, will be out of place in a closed one, as Vladimir Simagin first showed in games like Simagin-Portisch, Plovdiv 1959, 1 e4 c5 2 liJf3 e6 3 liJc3!? a6 4 g3 b5.
liJc6 3 liJge2 e6 4 g3. Black indicates his willingness to play a Taimanov or Scheveningen variation and White is hinting at a closed line. But if Black naively responds 4... g6?, expecting 5 .tg2 .tg7
... he'll be surprised by 5 d4! cxd4 6 liJxd4. This exposes a nasty hole at d6 (6 ... .tg7? 7liJdb5!). Similarly 4... liJge7 5 .tg2 g6? 6 d4!. Black never has this problem in normal Closed Sicilian orders such as 1 e4 c5 2 liJc3 liJc6 3 g3 g6 4 .tg2 .tg7 because he has too much firepower focused on d4.
Black's setup would serve him well after 5 d4 cxd4 but not in the game's continuation, 5 .tg2 .tb7 6 d3! liJc6 7 0-0 d6 8liJg5. Despite his opening expertise in the open And suppose Black abandons Sicilian, Lajos Portisch appeared clueless after 8... h6 9 liJh3 liJf6 ... g6 and goes instead for a quick 10 f4 i.e7 11 f5! e5 12 liJf2 liJd4 ... d5, as in 1 e4 c5 2liJc3 e6 3 g3 d5 13 liJb I! liJd7 14 c3 liJc6 15 a4 4 exd5 exd5. That's a perfectly i.g5 16 .txg5 'iixg5 17 axb5 axb5 valid anti-closed weapon. 18 lha8+ i.xa8 19 liJa3 liJa7 But 2... e6 is a more committal 20 d4!. move than Black may suspect and On the other hand, if Black after 3 liJf3 he cannot stop 4 d4. If assumes he is headed into a closed Black is a Najdorf or Dragon Sicilian, White can switch to an player, he'll find himself tricked out open one. This is the case of 2 liJc3 of his repertoire. 55
Sicilian Defense
players are not familiar with. And 5... a6 is trickier than you might think - 6 'ii'd3 tbc6 7 ..tf4 'ii'c7 8 00-0 tbe5?? 9 tbxe5 dxe5 10 ..txe5! Resigns was Benjamin-Hrop, The first version is the most Parsippany 2003. deceptive. After 2 tbf3 a Taimanov The Dragon player may feel at Variation player might reply 2 ... e6 ease in this first Chameleon version because he wants to avoid 2 ...tbc6 because he can meet 2 tbf3 tbc6 3 ..tb5, the Rossolimo Variation. 3 tbc3 with 3... d6 4 d4 cxd4 But 3 tbc3 poses a problem. 5 tbxd4 g6. But this is a Modern Then 3... tbc6 allows one of the Dragon Variation with a slight better editions of the Rossolimo, difference: ...tbc6 has been played 4 ..tb5!. And 3 ...tbf6 transposes in place of ... tbf6. That means after 4 e5 tbd5 into the dubious 6 tbd5! is possible. Then 6.....td7 Nimzovich Variation (usually 7 ..tg5 is annoying and 6... e6 7 tbc3 reached via 2 tbf3 tbf6 3 e5 tbd5 exposes holes. 4 tbc3 e6).
There are three basic versions of the Chameleon: White can bring his knights to f3 and c3. Or he can play 2 tbe2 and 3 tbbc3. Or he can begin with 2 tbc3 and then 3 tbge2.
The most flexible response to 2 tbf3 e6 3 tbc3 is 3... d6. Then 4 d4 cxd4 5 tbxd4 opens a path to various mainstream Sicilians (5 ...tbf6, 5... tbc6, 5... a6). However, White can surprise him with 5 'i'xd4!?
If Black replies 6.....tg7 instead he is not lost - as Alexey Suetin claimed in the first edition of ECO. But he is worse after 7 tbb5! :b8 8 c4. The Dragon player might refme his order and meet 2 tbf3 with 2..• d6 3 tbc3 g6. But he lands in new and complicated territory after 4 d4 cxd4 5 'ii'xd4! (5 ...tbf6 6 e5 tbc6 7..tb5).
Then 5...tbc6 6 ..tb5 is a position (usually reached via 2 tbf3 d6 3 d4 cxd4 4 'iixd4) that most 2 ... e6 56
Sicilian Defense
But life isn't so simple for White But this is how Bobby Fischer in this Chameleon order. A major outsmarted himself in his 1992 drawback to 2 tDf3 tDc6 3 tDc3 or rematch with Boris Spassky. 2... d6 3 tDc3 is 3... e5!. Then d2-d4 Spassky switched to a closed is impossible and he is stuck in a Sicilian, 3 tDbc3 d6 4 g3! tDc6 closed Sicilian with his KN 5 i.g2 g6 6 0-0 i.g7 7 d3 0-0 8 h3 misplaced at D, where it blocks his and 9 f4 - an excellent weapon f-pawn. against Fischer, who had always preferred ... f5 or ... e6/ ... tDge7 For example 2... tDc6 3 tDD e5 systems, not those with ... tDf6. 4 i.c4 d6 5 0-0 i.e7 6 a3?! tDf6 Perhaps Black's best policy after 7 :b 1 0-0 8 b4 a6 9 d3 b5! as in J.Polgar-Kramnik, Paris 1994. 2 tDe2 is allowing an open Sicilian, White can improve on 6 a3?! but so say with 2...tDc6 3 tDbc3 d6 4 d4. He delays ... e5 until White has far 3... e5! has performed well. given up on tDd5/i.c4, such as 4 g3 Jose Capablanca and Paul Keres e5!. tried to solve the ... e5 problem with There are slight differences in the the second Chameleon version, 2 tDe2 and 3 tDbc3. The difference third Chameleon order, 2 tDc3 and is that 2... tDc6 3 tDbc3 e5 allows 3 tDge2. A Najdorf player can be 4 tDd5! followed by tDec3! and expected to reply 2..•d6 and 3... a6. i.c4, which coordinates the White Then after 4 d4 cxd4 5 tDxd4 he has a choice between transposing pieces well. into a normal Najdorf with 5... tDf6 The natural response to 2 tDe2 is and the extra option of 5... e6!? 2...tDf6. This attacks the e-pawn That avoids the 6 i.g5 Najdorf and invites the complications of 3 e5?! tDg4 or 3 ... tDd5 4 tDbc3 e6. and gets good versions of other Najdorfs, such as 6 f4 b5 7 i.d3 i.b7 8 0-0 tDd7 9 f5? 'iib6!. It also creates a reasonable Keres Attack after 6 g4, e.g. 6... tDe7 7 a3 tDbc6 8 tDb3 b5 as in Vallejo PonsTopalov, Leon 2006. The problem with the Najdorfwannabe order is Black has to justify ... a6 after 4 g3!.
57
Sicilian Defense
Then 4 g3 .ltg7 is an innocuous Closed Sicilian in which White's KN is once again misplaced. Better is 4 d4 cxd4 5 tDxd4 i.g7. If Black has any Dragon instincts he'd be happy here because he enjoys the benefits of the accelerated order without allowing White his strongest weapon, the Maroczy Bind.
Now 4 ... e5 is not nearly as effective when White's f-pawn is unblocked. He is virtually a move ahead of normal closed Sicilians after 5 .ltg2 tDc6 6 d3 g6 7 f4 .ltg7 8 0-0 tDge7 9 .lte3 0-0 10 'i'd2 tDd4 11 litf2. Note that Black also gets an extra option in both the 2 tDc3/3 tDge2 and 2 tDc3/3 tDf3 orders. After 2 ... tDc6 and 3 ... d6 4 d4 exd4 5 tDxd4 he can try 5... e5.
But White can also feel happy with this move order, because he discouraged Black from a slew of other openings he might have prepared, such as the Taimanov or Sveshnikov.
This has the benefit of reaching a Sveshnikov Variation (6 tDdb5 a6 7 tDa3 b5 8 tDd5) but with the possibility of 8 ...tDce7!?, which has tested well. However, White also has an extra option, 6 tDf5!?, and this entire 5 ... e5 line must be called unclear.
GRAND PRIX ATTACK The Grand Prix was once considered a patzer's opening. But no less a Sicilian authority than Yevgeny Sveshnikov said that allowing it by answering 2 f4 with 2 ... tDc6 was a serious mistake. 'Already Black is worse,' he wrote.
The biggest problem with all the Chameleon orders is the accelerated form of the Dragon.
White's basic plan in the attack is tDc3 and f2-f4 followed by tDf3 and posting his KB on c4 or b5. If Black
After 1 e4 c5 2 tDf3 tDc6 3 tDc3 g6!:
58
Sicilian Defense
could count on White not transposing, his best order would be 1 e4 c5 2 lLlc3lLlc6 3 f4 e6 4lLlnlLlge7:
What about a more flexible - and very common - order such as 1 e4 c5 2 lLlc3 lLlc6 3 f4 d6 4 l'bn e6 ? Then 5 i.b5 can be firmly answered by 5... l'bge7 and 6... a6!. But on 5 g3!:
In that way he can meet 5 i.b5 with 5... a6! and 5 i.c4? with 5... d5!. White gets a good Chameleon, e.g. 5...lLlge7 6 i.g2 and if 6... g6 then 7 d4! exposes the d6 hole again.
But the greatest strength of the Grand Prix today is its transpotential. White can switch to an open Sicilian, 5 d4 cxd4 6 lLlxd4. The result is perfectly good version of the Taimanov Variation. But there's quite a bit of theory for Black to know - and if he's not a Taimanov player, he's been tricked out of his repertoire.
Better is 4... a6. But 5 d4 cxd4 6 lLlxd4 lLlf6 would land Black in a strange Najdorf. That is, 1 e4 c5 2 l'bf3 d6 3 d4 cxd4 4 lLlxd4 l'bf6 5 l'bc3 a6 and 6 f4 l'bc6?!, e.g. 7 i.e3 i.d7 8 'iYn g6 9 l'bd5! ':c8 10 l'bxf6+ exf6 11 c3 is excellent for White (F 0 garasi -Ftacnik, Austria 2006).
This kind of switch happens because Black has to make pawn and piece commitments. That provides White with the information he needs to decide whether to transpose into another Sicilian line. In the last diagram for example, White could play 5 g3, creating a closed Sicilian in which Black has given up on the recommended ... lLlf6 and ... e5 defenses.
Most likely the best answer to 1 e4 c5 2 l'bc3 l'bc6 3 f4 d6 4 l'bf3 is neither 4... e6 nor 4... a6 but 4... g6. Black should be happy with 5 i.b5 i.d7, a familiar Grand Prix position. But he had better know a lot of Dragon theory as well because of 5 d4!? 59
Sicilian Defense
The Grand Prix is a particularly good weapon against Najdorf, Scheveningen and Modem Dragon players because they'll meet 1 e4 c5 2 liJc3 with 2•.. d6 in the hopes of transposing (3 liJf3 liJf6 4 d4) or punishing White's order (3 liJf3 e5!).
The trickster can try to exploit Kasparov's order with the paradoxical 5 .tb5+ .td7 6 .tc4!? His point is that Black's bishop gets in the way of supporting ... d5.
There are only a few examples of this so far (6 ... liJc6 7 0-0 e6 8 d3 liJge7 9 'i'el a6 10 f5!? gxf5 But there's a problem with 2... d6. 11 'i'h4 b5 12 .tb3 'iVb6, Some of the best forms of the Grand Ramon Perez - Movsziszian, Prix for White are those with ... d6. Balaguer 2006). But this may be a For example, 3 f4 liJc6 4 liJf3 g6 finesse with a future. 5 .tb5! is a pin that rules out And there is one more headache ... liJd4, a strong idea when the pawn for Black after 3... g6. White can is at d7. reply 4 d4 cxd4 5 'i'xd4. This Garry Kasparov's solution was allows him a promising pawn sack 1 e4 c5 2 liJc3 d6 3 f4 g6 4 liJf3 after 5... liJf6 6 e5 liJc6 7 .tb5 .tg7: (7 ... dxe5 8 'i'xd8+ ~xd8 9 fxe5! liJxe5 10 .tf4 liJed7 11 0-0-0).
ALAPIN VARIATION Simon Alapin - pronounced ah-LAH-peen, in case you were wondering - gets the credit for 2 c3. But the German player H. W.Popert played it, as well as the wily This is far superior to 4 ... liJc6 5 2 liJf3 I 3 c3, more than 50 years .tb5 .tg7?! 6 .txc6+!, which before Alapin. favors White. Black can meet The benefit of the second order is 5 .tb5+ with 5....td7!. that White gets to see Black's second move before he commits himself to c2-c3. That's significant because after 2 liJf3 liJc6 3 c3!? Black has lost some of his best 2 c3 options.
He also invites a Dragon, 5 d4 cxd4, and can meet 5 .tc4 with 5 ...liJc6 6 0-0 e6 followed by kicking the bishop with ... liJge7/ ... d5 60
Sicilian Defense
The difference is 2 ttJO e6 3 c3 ttJe7!? 4 d4 cxd4 5 cxd4 d5:
He can't play the best version oLb6/... i.b7 - which arises in 2 c3 ttJf6 3 e5 ttJd5 4 d4 cxd4 5 cxd4 e6 6 ttJf3 b6 - because the QN is in the way (3 ...ttJf6 4 e5 ttJd5 5 d4 cxd4 6 cxd4 b6 7 i.c4!).
And now 6 exd5 allows 6... ttJxd5!. What's happened is Black slipped into a harmless SemiTarrasch Defense position which Also 3... d5 4 exd5 'iixd5 leads to more commonly occurs in orders a 2 c3 d5 position in which Black's like 1 ttJf3 ttJf6 2 c4 e6 3 d4 c5 4 e3 QN is developed early. White soon cxd4 5 exd4 d5 6 cxd5?! ttJxd5. threatens to push his center pawns, White does better with 6 e5 even e.g. 5 d4 i.g4 6 i.e2 e6?! 7 h3 i.h5 8 c4 'i'd6 9 g4 i.g6 10 d5! though 6...ttJbc6 and 7...ttJf5 is quite a respectable version of the with advantage. Advance French for Black. It's In practice, the delayed form of respectable - but perhaps not the the Alapin appears infrequently kind of middle game that a after 2 ttJO d6 because of 3 c3 ttJf6! Sicilianista would like to defend. (4 d4? ttJxe4). It occurs most often That's why you'll see 1 e4 c5 after 2 ttJf3 a6, when White tries to 2 ttJO e6 3 c3 d5 4 e5 being played, prove that ... a6 is irrelevant. even though 4 exd5 is objectively It's also a good weapon against better. White is hoping a 1...c5 2... e6 but there is a counter-finesse player won't be comfortable in a in 3 c3 ttJe7!? and ... d5. This French. And that's also why Black favorite of Yevgeny Vasiukov's often avoids the objectively best avoids ... d5/exd5 lines in which 4... ttJc6 5 d4 and prefers a gambit Black has to retake with his queen with 4 ... d4!? 5 cxd4 cxd4 6 i.b5+ or e-pawn, that is 1 e4 c5 2 c3 d5 ttJc6 7 i.xc6+ bxc6 8 'ifa4 or 3 exd5 or 2... e6 3 d4 d5 4 exd5. 6 ... i.d7 7 ttJxd4!? 61
Sicilian Defense
MAIN ALAPIN LINES
unclear, especially after trendy continuations such as 7 .te2 'fIe7
After 2 e3 tLlf6 play usually continues 3 e5 tLld5 4 d4. But 8 'iVe2 g5!? Sveshnikov, the leading authority The main line of 2.•• d5 is 3 exd5 on 2 c3, prefers 4 tLlf3 and meets 'iVxd5 4 d4. Black should avoid 4 ... tLlc6 with 5 .i.c4 and then 5... e6 4... cxd4 because he's headed to one 6 0-0 d6 7 d4 cxd4 8 cxd4 to reach of the isolated pawn middlegames a typical Alapin position. we know to be inferior when they His point is that this order averts arise in the Queen's Gambit some troublesome lines such as Accepted and Semi-Tarrasch. For example, 4 ... cxd4 5 cxd4 tLlc6 4 d4 cxd4 5 cxd4 d6 6 tLlf3 tLlc6 6 tLlf3 e6 7 tLlc3 'fId8 8 i.c4 7 .Jtc4 dxe5 and 7... tLlb6 8 .i.b3 d5. transposes to a 1 d4 position with an Another virtue of 4 tLlf3 is that it extra tempo for White. gives Black a chance for 4..• e6 Better is 4... tLlf6 5 tLlf3. Both 5 ..te4 b6. He thinks he can transpose into the 4 d4 cxd4 5 cxd4 5... e6 and 5... .i.g4 are common. The e6 6 tLlf3 b6 line, which enjoys a sly alternative is 5•..tLle6!? good reputation.
This rules out popular attacking setups with .i.d3 because of 6 .id3 .i.g4! (7 .i.e3 cxd4 8 cxd4 i.xf3 9 gxf3 e6).
But White can cross him up with 6 .i.xd5! exd5 7 d3 and claim a slight edge. The more accurate route to the fianchetto line is 4 tLlf3 b6! 5 ..tc4 ..ib7.
Another benefit of 5... tLlc6 is that 6 .i.e2, which prepares c3-c4, can be handled by 6.•. exd4! 7 exd4 e6 with rough equality. White cannot take with his knight on d4 because g2 is hanging.
The drawback to 4 tLlf3 tLle6 5 .Jte4 is Gedeon Barcza's idea, 5... tLlb6! 6 ..ib3 e4. That's not necessarily bad for White. It's just 62
Sicilian Defense
In contrast, 6... e6 7 0-0 cxd4 8lbxd4! lbxd4 9 cxd4 followed by lbc3 or i.f3 favors White.
8 i.g5! and 0-0-0.
However, 7 ... h6!? leaves him without a plan. A cautionary tale, Lopez-Dobrov, Neiva 2005, went ROSSOLIMOIMOSCOW!'iIVxd4 8 i.e3 e5! 9 'ilVd3lbf6 10 0-0-0 i.e7 The two variations in which 11 h3? 'ilVd7 12 ~bl b5 13 lbd2? White places his bishop on b5 at 'ilVb7 14 f4 0-0 15 fxe5 dxe5 move three are naturally related. If 16 lbf3? b4 17 lbe2 i.b5 White Black is versatile enough to play resigns. As a result White has been open Sicilians with both 1 e4 c5 2lbn d6 and 2 ...lbc6 he can reduce anticipating 7 lbc3 h6 with the his investment in midnight oil by immediate 7 i.g5!. means of 2.•. d6 3 i.b5+ lbc6! and 2...lbc63 i.b5 d6!, transposing.
This almost always transposes (7 ... lbf6 8 lbc3) and so far no one has found a drawback to 7 i.g5!.
Another virtue is that 4 d4 cxd4 5 'ifxd4 i.d7 6 i.xc6 i.xc6 is a main line that can also occur after 2... d6 3 d4 cxd4 4 'ifxd4 and then 4... lbc6 5 i.b5 i.d7 6 i.xc6 i.xc6.
The easiest way to deal with 1 e4 c5 2lbn d6 3 d4 cxd4 4 'ilVxd4 with the prospect of c2-c4 is simply to avoid it by means of 3 ...lbf6! (4lbc3 cxd4!).
If Black spends his home time on that position, he can be prepared for all the major i.b5(+) and 'ilVxd4 variations.
Top GMs such as Vishy Anand have tried to find an edge for White in 4 dxc5 but failed after 4...lbxe4 5 cxd6 lbc6! 6 lbbd2 lbxd6.
Note that in the 6 i.xc6 i.xc6 position White's only bid for advantage or even for the initiative is supposed to be 7 lbc3 lbf6
There's a more direct, but little known, way for White to seek the 63
Sicilian Defense
Maroczy Bind, 1 e4 cS 2 tiJf3 d6 3 d4 cxd4 4 tiJxd4 tiJf6 S .td3!?
Objectively this isn't dangerous. Subjectively, it can be a major pain for Black because the reason he plays 4 ... tiJf6 is to prompt 5 tiJc3 and avoid c2-c4. Today we know that the pawn structures after 5 iLd3 e6 6 c4 or 5... g6 6 c4 aren't necessarily bad for Black. But they can be very difficult to handle if you're unfamiliar with them - as a Modern Dragon, Scheveningen or Najdorfplayer usually is.
For example, 8 ....tg7 9 .tb2 0-0 10 0-0 "ikc7 11 tiJc3 (FogarasiStarostits, Scanno 2005). So far a simple antidote to 5 .td3!? hasn't appeared. MODERN DRAGON VARIATION The Modern Dragon was born as a counter-finesse. When the Dragon began scoring victories in the 19th century it commonly came about via 1 e4 c5 2 tiJf3 tiJc6 3 d4 cxd4 4 tiJxd4 tiJf6 5 tiJc3 d6 6 .te2 g6. But White found an effective deterrent in 6 .tg5 and then 6... g6?! 7.txf6!.
Black has two forcing replies to . 5 .td3. One is 5... e5 when 6 tiJf3 Kurt Richter, who helped launch and 7 c4 reaches a good version of 6 .tg5, also popularized the refined the Kalashnikov Variation for order of 1 e4 cS 2 tiJf3 d6 3 d4 cxd4 White. 4 tiJxd4 tiJf6 S tiJc3 g6!. Then Perhaps better is S...tiJc6 and 6 .tg5 .tg7 is too late. then 6 tiJxc6 bxc6 and 7... g6. Unlike the 6 .td3 Najdorf we'll consider later, Black hasn't spent a tempo on ... a6. But experience indicates White stands well after 7 c4 g6 8 b3:
The price Black pays IS committing himself to a Dragon before .te2. This gives White more aggressive options, including the Yugoslav Attack - .te3/fl-f3J'iid2/ 0-0-0 - and some quiet ones like the 64
Sicilian Defense
fianchetto system, 6 g3. Even the latter contains pitfalls after 6 g3 .tg7 7.tg2:
More common is 6 .te3. It sets a small trap (6 ... lDg4?? 7 .tb5+) and typically leads (6 ....tg7 7 0) to another fork in the road. One path is 7... 0-0 8 'iVd2 lDc6 and another is 7... lDc6 8 'iVd2 0-0. Most players see no difference. But there is one because 7...lDc6 8 'iid2 allows Black to use 8....td7!? as a waiting move.
The natural 7...lDc6 walks into 8 lDxc6! bxc6 9 e5, a trap which has snared many a master. Black should either play 6... lDc6 (7 .tg2 lDxd4!) before ... .tg7 - or he should delay ... lDc6 in favor of 6... .tg7 7 .tg2 0-0. Then on 9 .tc4 he can transpose, with 9... 0-0, into main lines. That's useful during times, like these, when Black is faring better against .tc4 Yugoslavs than against non.tc4 ones. Thanks to this order he can temporarily keep his king in the center after 9 0-0-0 ~c8. Then 10 g4 lDe5 11 h4 h5! 12 g5 lDh7 In the Yugoslav Attack there is a makes it safer for Black to castle, 'something to think about' move, now that the g- or h-file can't be 6 D. Yefim Geller indicated in the forced open. 1975 ECO that '6 f3?!' was Another situation arises when punished by 6... 'i'b6 7 .te3 'iixb2. White plays .tc4 but delays 'iid2, But that allows 8 lDdb5! with as in 7..•lDc6 8 .tc4 in order to threats to trap the queen with accelerate the h2-h4-h5 attack. 9 ~b 1. This is strong enough to This order tries to tempt Black make 6... 'i'b6 the move deserving into 8... 'i'b6. That looks good in ofa '?!'.
This trap is important to recognize because it arises from other move orders, including 1 e4 c5 2 lDf3 lDc6 3 d4 cxd4 4 lDxd4 lDf6 5 lDc3 d6 6 g3 g6 when 7 .tg2 .tg7 8 lDxc6! transposes into what White wants and 7... lDxd4! 8 'iixd4 .tg7 is what Black wants.
65
Sicilian Defense
view of 9 ii.b3? liJxe4 or 9... liJg4. However 9 liJf5! 'ifxb2 10 liJxg7+ is such a promising sack that masters abandoned 8... 'ifb6?! in the 1960s. Black's best policy after 8 ii.c4 is 8... 0-0!, which usually transposes to a main line after 9 'ifd2. If instead 9 ii.b3 Black can get his queenside attack running before White inflicts kings ide damage, 9...ii.d7 10 h4 liJxd4 11 ii.xd4 bS 12 hS as.
For instance 13 hxg6 hxg6 14 liJdS liJxdS 1S ii.xdS l::te8 16 'ifd2 ':xe2! 17 'ifxe2 ii.xd4 18 0-0-0 ii.g7 19 'itb1 'ib6 20 g4 e6 as in Perez Candel-Guseinov, Kusadasi 2006.
Then 10 exd5 liJxd5 11 liJxc6 bxc6 12 ii.d4 is heavily analyzed and remains controversial and 12 liJxd5 cxd5 13 'ifxd5 is risky. White should avoid the alternative 10 liJxe6 bxe6 11 exdS.
This transposes to the pawn-grab line after 11... liJxd5 12 liJxd5 cxd5 13 'ifxd5. But there's no reason to grant a dangerous extra option, Vasily Panov's l1 ...'ifaS 12 dxe6 ii.e6!. A 1954 Soviet game went 13 a3 l::tfd8 14 'ife2 l::txd1+ 15 'ifxd1 l::td8 16 'i'e2 liJd5 17 liJxd5 'ifxd5 18 c7? ii.xb2+! 19 'itxb2 'i'a2+ and wins. If Black doesn't play 9... d5, he has two natural orders to reach a tabia. One is 9...liJxd4 10 ii.xd4 ii.e6 11 'itb1! 'i'e7 (not 11...'i'a5 12 liJd5! 'i'xd2 13 liJxe7+). Experience says White is better but not enough to discourage Black from seeking it.
DRAGON: 90-0-0
In the Yugoslav tabia (1 e4 eS 2 liJt3 d6 3 d4 exd4 4 liJxd4 liJf6 S liJe3 g6 6 ii.e3 ii.g7 7 t3 liJe6 8 'ifd2 0-0) White has three chief alternatives. One of them, 9 0-0-0, The other order is 9...ii.e6 allows only a little room for finesse followed by 1O ... liJxd4. Black may be skittish about it because ... after the forcing 9...dS. 66
Sicilian Defense
.. .instead of 10 'it'b 1 ttJxd4 11 i£.xd4 "ikc7, transposing, White can try 10 ttJxe6 fxe6 11 g3! and 12 i£.h3, which puts e6 under fire. But there's an upside to 9... i.e6. Black steers clear of 9 ... ttJxd4 10 i.xd4 i..e6 11 ttJd5!?, which many Dragonistas don't like to face. So it's up to him to choose which line he wants to avoid more, the 11 ttJd5!? one or the 9... i.e6 10 ttJxe6 one.
Not at all. A natural continuation is 12 ..."ika5 13 i.a4 l::f.fc8 14 i.b3. Then we've landed in a position that was reached in the 1950s via 9 i.c4 ttJxd4 10 i.xd4 i.e6 11 i.b3 'iia5 12 'it'bl l::f.fc8. It's been analyzed well into a king-and-pawn endgame (!) with the conclusion that White is much better, if not winning . This underlines a danger to both players in a Modem Dragon - the possibility of a 9 0-0-0 line suddenly turning into a risky 9 g4 position and vice versa. Or a 9 0-0-0 line transforming into a 9 i.c4 position. That's the case with 9 0-0-0 ttJxd4 10 i.xd4 "iia5 and 11...i.e6. This looks like just another routine Dragon. But Black's order is faulty because 11 i.c4! will transpose, after 11 •••i.e6 12 i.b3, into that bad 1950s line.
Whatever the route to that tabia White usually continues 12 g4 or 12 h4. A clever transposer may prefer 12 i.b5!? The bishop is headed for b3, its best square on the board.
DRAGON: 9g4 The 9 g4 thrust was long been considered a worthy alternative to 9 0-0-0. It's superior in some ways: It stops 9 ... d5 with 10 g5!. It discourages 9 ... ttJxd4 10 i.xd4 i.e6 because 11 ttJd5! is a better version of 9 0-0-0 ttJxd4 10 i.xd4 i.e6 11 ttJd5.
But doesn't that just lose time?
But 9 g4 was dealt a serious blow m the 1995 world championship
67
Sicilian Defense
match. What Team Kasparov came up with was 9.•. il..e6! 10 0-0-0 tbxd4 11 il..xd4 'ilVa5!.
DRAGON: 9 il..c4 Some Dragon experts, like Sergey Tiviakov, have circumvented this hyper-theoretical line by meeting 6 il..e3 il..g7 7 f3 tbc6 S'ilVd2 with S..•il..e6!?:
We are headed toward a standard position in the 9 0-0-0 tbxd4 line but with an extra tempo for Black. He saved it by getting his queen to a5 in one move, compared with 9 0-0-0 tbxd4 10 il..xd4 il..e6 11 ~bl! when l1..:iWc7 is played because 11 ... Via5? stumbles into 12 tbd5!.
Black is willing to transpose into non-il..c4 lines after 9 0-0-0 0-0 or 9 g4 0-0. The obvious drawback is 9 tbxe6 fxe6. But so far the pawn weakness hasn't proven significant in games that went 10 0-0-0 .a.c8 11 il..c4 'ilVd7 or 10 il..c4 ct;f7 11 a3 (else ... tba5) .a.c8 12 il..a2 tbe5 and ...tbc4.
An extra tempo is worth platinum in the Dragon and this was apparent in the first AnandKasparov game. After 12
Black can try to improve on this with S•..tbxd4 9 il..xd4 il..e6, when White is denied both tbxe6 and il..c4. But this fails to 10 il..b5+!. Then 1O... il..d7? 11 il..c4 gets Black into the position he was trying to avoid - and a move down to boot. There's been virtually no experience with 1O... ct;f8 (11 0-0-0 'ilVa5) so we'll have to let the tricksters test it for us.
There doesn't seem to be a downside to 9 g4 il..e6! as there is in the comparable 9 0-0-0 il..e6 position because 10 tbxe6 fxe6 11 g3 is, of course, illegal. 68
Sicilian Defense
In the main .tc4 line, 7...lDc6 8 'i'd2 0-0 9 i.c4, Black usually plays ....td7 and posts a rook at c8 while White plays 0-0-0, .tb3 and h2-h4. If Black is going to play a ... 'i'a5/ ...:fc8 system instead it's more accurate to do it with 9...'i'a5 and 1O ... .td7 rather than the reverse. Now 9 f4 allows 9•• :Wb6!, which threatens 1O ... ~xe4 or 1O ... ~g4/ 11.. ..txd4. No better is 8 f4 'i'b6!. This underlines White's difficulty in reaching Classical positions via the Levenfish (6 f4) Variation. The Levenfish sets a trap, 6 ....tg7 7 e5!?, that is best avoided by 6 ... ~c6. A devious White might play 6 f4 to see if Black falls into it, and then shift into the Classical line after 6 ... ~c6 7 .te2 .tg7 8 .te3 0-0.
The reason is that this order halts the quick attack of 10 h4? with 10 ••• 'i'b4! 11 i.b3 ~xd4 and 12 .txd4? ~xe4. White has to play the 12 'i'xd4 ending and, while it's not dreadful, it's certainly not what he had in mind.
But he's outsmarted himself because 9 O-O?! walks into 9•••'i'b6! again.
DRAGON: CLASSICAL
When White plays .te2 he usually continues fl-f4 and .te3. An early tabia is 1 e4 c5 2 ~f3 d6 3 d4 cxd4 4 ~xd4 ~f6 5 ~c3 g6 6 .te2 .tg7. There are few .te2 lines in which White castles long, so 7 0-0 would seem to best retain his options. However, the g I-b6 diagonal is a problem after 7... 0-0 8 .te3 ~c6:
69
The way to avoid that kind of surprise as well as other liberating ideas such as ... d5, ... ~xd4 and ... ~g4 is to retreat the knight from d4 to b3. Siegbert Tarrasch routinely gave ~b3 a question mark in the Sicilian. But a case can be made in the Modem Dragon for 6 .te2 .tg7 7 ~b3 rather than 7 0-0. The reason is the natural 7... 0-0 allows White a sharp extra option,
Sicilian Defense
8 g4!?
. White may be gun-shy about 8 f4 because of 8.. :ii'b6. But this time White has a strong 9 e5!, which Sammy Reshevsky once allowed (9 ... dxe5?? 10 tbxc6 wins).
Black's king posItion makes 8...tbc69 g5 tbd7 10 h4 or 10 ..te3 more dangerous than if he had played 7 ...tbc6. Some players don't like tbb3 because it allows Black to develop ... ..te6 without fear of tbxe6. But the bishop faces more jeopardy from pawns - 7 0-0 tbc6 8 tbb3 ..te6? favors White after 9 f4! and 10 f5!, e.g. 9 .. J:tc8 10 f5 ..td7 11 g4.
After 8... 0-0 instead, White reaches the Foltys Attack with 9 tbb3 and then 9... ..te6 10 g4. As for Black, his way of dodging it is 9... tba5. Then 10 0-0 ..te6! transposes into a position normally reached via 9... ..te6 10 0-0 tba5 that is considered safe enough.
In truth, tbb3 serves too many good purposes and the main This order allows Black to meet question is when to play it. It is essential to the most double-edged 10 g4 with 10... b6!? There is no Classical line, Jan Foltys' attack. bishop to kick with f4-f5, and Perhaps the most accurate order for following 11 g5 tbd7 and ... e5 he White is 6 ..te2 ..tg7 7 ..te3 tbc6 has counterplay, e.g. 12 'ild2 ..tb7 and then 8 f4 0-0 9 tbb3 ..te6 13 ..td4 e5 14 fxe5 tLJxe5 15 tLJxa5 10 g4!? (see next diagram) bxa5 16 0-0-0 'ilc7 as in Oll-Van der Wiel, Dutch Team ChampionThis order avoids 8 tbb3 ..te6 9 f4 'iYc8!, which stops g2-g4. Then ship 1996. White has nothing better than to White, too, gets an extra option transpose into the older lines with 10 0-0 0-0 that work well for Black from this order and it is 10 e5. 70
Sicilian Defense
Black can transpose into a Modem Dragon with 6... g6 and then 7 i.e3 or 7 ltJb3 is normal. But there is also 6...ltJxd4!? 7 'it'xd4 g6.
Then 1O... dxe5?? loses a piece (11 'it'xd8 and ltJxa5). But 10...ltJe8 needs testing to see if White is overextended. The most popular alternative to Then on 8 i.e3 i.g7 White has 9... ltJa5 is 9... a5. Then on the semi- no edge because his queen is automatic 10 a4 Black can safely vulnerable (9 'ifd2 0-0 10 0-0-0 play 10... i.e6 since 11 g4 d5! is i.e6). This would come about from much stronger when Black's QN the Modem Dragon's 5... g6 6 i.e2 cannot be driven offb4 (12 f5 i.c8! ltJc6 7 i.e3 and 7 ... ltJxd4 8 'it'xd4? 13 exd5ltJb4 14 i.D gxf5). If White is going to punish this White usually plays 11 0-0 and order the best bet is 8 i.g5 i.g7 9 transposes into a position normally ltJd5. The critical position is likely reached via 9... i.e6 10 0-0 but with 9... 0-0 10 i.xf6 gxf6. But Black's the addition of the a-pawn moves. bishops seem to offset the bad There are plusses and minuses but pawns, e.g. 11 0-0-0 f5 12 'ilfd3 the addition tends to help White in fxe4 13 'it'xe4 :e8 14 'it'D :e5 lines such as 11 ...'it'c8 12 h3 ltJb4 15 i.c4 i.e6 as in Michiels-Van der 13 ltJd4 i.c4 14ltJdb5. Weide, Amsterdam 2000. Finally, there's a move order that If that's true, 6 ... ltJxd4!? IS a doesn't fit into any traditional Dragon category. It arises from 1 e4 significant finesse - and a reward c5 2ltJf3 d6 3 d4 cxd4 4ltJxd4ltJf6 by the Gods of Theory to Black for 5 ltJc3 ltJc6 when White responds reviving the old Dragon move order after all these years. with the vanilla move 6 i.e2. 71
Sicilian Defense
ACCELERATED DRAGONS
lines with ~hllf2-f4 in which .i.e3
The Accelerated Dragon, 1 e4 cS 2 liJf3 liJc6 3 d4 cxd4 4 liJxd4 g6, becomes a Modem Dragon after ... d6. The benefits of the Accelerated order lie in limiting White options and enabling Black to shoot for ... d5!.
is delayed. Some books claim 6 liJb3 is refuted by 6 ... .i.xc3+!? (7 bxc3liJf6 8 .i.d3 d5). But many GMs don't believe it: Anatoly Karpov has played 6 liJb3 anyway and Gata Kamsky and Bent Larsen passed up 6 ... .i.xc3+ when they had the chance as Black.
White's knight is immediately attacked after S liJc3 .i.g7. If he wants to fianchetto, he has to try something like 6 liJde2liJf6 7 g3.
Nevertheless if White wants to play .i.e2 but not .i.e3, a more precise route is 5 .i.e2 or 5 liJb3 (after 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4 4 liJxd6 g6). For example, 5 .i.e2 .i.g7 6 liJb3 and 6 ... d6 7 0-0 liJf6 8liJc3. If you look up SliJc3 .i.g7 6 .i.e3 in books you will only find 6...liJf6. The reason is this is how Black sidesteps some Yugoslav lines. For example, 7 f3?! 0-08 'iid2 and now not 8 ... d6 but 8 ... d5! with excellent chances. The same goes for 8 g4 'iib6!.
tl>
Now 7 ... 0-0 8 ... g2 d6 9 0-0 gets him to a respectable Modem Dragon (1 e4 c5 2liJf3 d6 3 d4 cxd4 4 liJxd4 liJf6 5 liJc3 g6 6 g3 .i.g7 7 .i.g2 0-0 8 0-0 liJc6 9liJde2). But Black improves with 7 ...bS! (8 liJxb5 liJxe4). In contrast he usually has to spend a tempo on ... l:.b8 or ... a6 to engineer ... b5 in the Modem Dragon. More often you'll see 6 liJb3. White is seeking a Classical Dragon (6 ... liJf6 7 .i.e2 0-0) in which he retains the .i.g5 option as well as
But there is a rarely mentioned downside to 6 ... liJf6. It grants 72
Sicilian Defense
White an extra option, 7 ltJxc6!? SCHEVENINGEN VARIATION bxc6 8 e5, which leads to a space The basic elements of the edge following 8 ... ltJg8 or into Scheveningen are Black pawns at razor-sharp tactics (8 ...ltJdS) that e6 and d6 and a ltJ on f6 facing have been analyzed past move 20. White knights at c3 and d4 and a Do you really want to subject pawn at e4. This can come about yourself to memorizing lines such from a variety of orders but many of as 8... ltJdS 9 ltJxdS cxdS 10 'fixdS them prematurely commit Black to l:.b8 11 ii.xa7 ltxb2 12 ii.d4 ltxc2 moves like ... 'fic7. 13 ii.d3 e6! 14 ~a8 ltc6 IS ii.bS lta6 ? Why is it premature? After all, Black is always playing ... ~c7 in If the answer is no, the mental the Sicilian, so 'it just transposes,' hygiene move 6... d6 is preferable. You may end up in the Yugoslav right? this way but it may help you to sleep at night.
No, not when White develops ii.g2, for example. One of the best features of the modem Scheveningen order, 1 e4 c5 2 ltJf3 d6 3 d4 cxd4 4 ltJxd4 ltJf6 5 ltJc3 e6, is that after 6 g3:
Because of the ways that the Accelerated move order degrades White's options it would be a perfect defense to 1 e4 - if it weren't for the Maroczy Bind. To avoid it Black has tried orders such as 1 e4 c5 2ltJf3 g6. Then if White tries to transpose into the Bind with 3 c4 ii.g7 4 d4 or 3 d4 ii.g7 4 c4 (and then 4 ... cxd4 SltJxd4ltJc6) Black has a variety of independent options, including 4 ... ~aS+ or 4 ... d6. Those extra options have to be weighed against the downside. To play 1 e4 cS 2 ltJf3 g6 3 d4 ii.g7 Black has to feel comfortable in other openings, including a Schmid Benoni (4 dS) and a Panov CaroKann after (4 c3 cxd4 S cxd4 dS! 6 exdS!) as well as 4 dxcS!?
Black can dispense with ... 'iic7 and obtain equal play from 6...ltJc6 7 ii.g2 ii.d7!, e.g. 8 0-0 a6 9 a4 ii.e7 10 ii.e3 0-0 11 ~hl ltJeS 12 f4 ltJc4, Malakhov-Vogt, Feugen 2006. No better is 8 ltJdbS ~8 9 ii.f4 ltJeS and ... a6. 73
Sicilian Defense
The Scheveningen order is also effective against White players who try to escape to a more exciting Sicilian with 6 ~c4 or 6 ~g5. They expect 6 ... tiJc6 or 6 ... a6, reaching familiar territory.
5 ... a6 6 ~c4 e6 7 ~e3?! ~e7. For example, 8 'iWe2 b5 9 ~b3 b4 or 9 ~d3 .tb7 10 a3 tiJbd7 11 0-0 tiJc5 12 f3 0-0 and Black stood splendidly in Ulko-Ulybin, Moscow 2005.
But Black has a better move, 6 ... ~e7!. Then both sides risk being trapped in an inferior Najdorf or other Sicilian.
KERES AND ENGLISH ATTACKS Two major alternatives to quiet Scheveningens involve g2-g4. There is the Keres Attack, with an immediate 6 g4, and the English Attack, with ~e3, f2-f3 and 'iWd2 added.
For example, 6 ~c4 ~e7! 7 ~b3 and 7 ... a6? 8 f4 favors White, e.g. 8 ... b5 9 e5! with advantage or 8... 0-0 9 'iWf3. (In the Najdorf move order this happens with 5 ... a6 6 ~c4 e6 7 .tb3 when Black plays 7 ... ~e7?! instead of7 ... b5!.)
For decades it was thought the Keres Attack could only come about from the traditional order. There's no point to g2-g4-g5 if Black hasn't put a knight on f6, the thinking went. As a result a 1967 Shakhmatny Bulletin article recommended 1 e4 c5 2 tiJf3 d6 3 d4 cxd4 4 tiJxd4 tiJc6 5 tiJc3 e6.
Black can get better counterplay from ... tiJa6-c5 rather than ... a6, e.g. 7... 0-0 8 ~e3 tiJa6! 9 f4 tiJc5. If White is alert he may try 7 ~e3 instead of 7 ~b3. He wouldn't mind reaching a book Sozin Variation (7 ... tiJc6). But 7... a6! is much better.
Then 6 ~e2 tiJf6 gets Black into This is a superior Najdorf the Scheveningen while avoiding position that could come about via the Keres.
74
Sicilian Defense
But White doesn't need to with a2-a3 and has already played attack a knight. In the 1985 world g4-g5. That's enough to ensure an championship match Karpov advantage, 8... tiJb6 9 :gl tiJ8d7 improved with the simple, but at the 11 f4 and 12 f5. time, stunning 6 g4!. Then 6 ... ttJf6 7 Another example, is 6.....te7 7 g5 g5 is an excellent Keres. Avoiding tiJfd7 8 h4 tiJc6 9 ..te3 when ... tiJf6 is better but not necessarily 10 ..tc4! gives White a favorable equal (6 ... a6 7 ..te3 tiJge7 8 tiJb3). Velimirovic Attack. The sad truth is there is no From the Sozin order (1 e4 c5 simple anti-Keres route to the Scheveningen. Kasparov even 2 ttJf3 ttJc6 3 d4 cxd4 4 tiJxd4 ttJf6 resorted to the Najdorf (5 ... a6 6 5 tiJc3 d6 6 ..tc4 e6 7 ..te3) he ..te2 e6) to do it, although that could obtain something like that allows White a wealth of sixth only if Black was very cooperative - 7... a6 8 'i'e2 ..te7 9 g4? 'i'c7? move alternatives. (rather than 9... tiJxd4 10 ..txd4 e5! In the normal Keres order, 1 e4 and ... ..txg4) 10 g5 tiJd7 11 h4. c5 2 tiJf3 d6 3 d4 cxd4 4 ttJxd4 tiJf6 Another dangerous attacking 5 tiJc3 e6 6 g4, a consensus has emerged that 6... h6 is the only move setup is begun by 1 e4 c5 2 tiJf3 d6 that offers Black a shot at equality. 3 d4 cxd4 4 ttJxd4 tiJf6 5 ttJc3 e6 Other moves allow White to reach 6 ..te3 or 6 f4. In both cases White favorable versions of the English can carry out a Mikhail Tal plan of Attack, the Velimirovic Attack or ..te3, £2-f4, 'i'D and 0-0-0. somesuch. For example, 6... a6 7 g5 The pawn move seemed superior tiJfd7 8 ..te3 b5 9 a3: because 6... ..te7 7 ..tb5+ and 8 e5 poses problems for Black. The bishop move, on the other hand, allows White an extra option, after 6... tiJc6, of transposing to a Sozin (7 ..tc4), without having to face Pal Benko's 6 ... 'i'b6 in the normal Sozin order. But today 6 ..te3 is preferred for another reason. It is the main route into the English Attack.
This is an English Attack in which White has replaced £2-D
After 6... a6: 75
Sicilian Defense
10 g4! dxe4 11 'iVf2 'iVc7 12 g5 with advantage. CLASSICAL VARIATION
If White adopts the f2-f4 plan Black obtains quick counterplay (7 f4 b5! 8 'iVf3 .tb7 9 .td3 ttJbd7 10 a3 ':c8). But 7 g4 is a good response. Then the natural 7... h6 transposes into a kind of Keres in which White has added the useful move .te3 while Black has added the slow ... a6. As a result White has good chances whether he continues a la Keres (8 h4), in English Attack style (8 f3), or in a TaVTopalov hybrid with 8 f4. There is a drawback to 7 g4 inasmuch as 7••. e5 8 ttJf5 g6 forces White to sack material (9 ttJg3? .txg4). But 9 g5! gxf5 10 exf5, the Bela Perenyi line, has been prohibitively strong.
The wonderfully flexible system that comes about via 1 e4 c5 2 ttJf3 d6 3 d4 cxd4 4 ttJxd4 ttJf6 5 ttJc3 ttJc6 should be named after Louis and Wilfried Paulsen, who explored it in the 1880s. But there already is a Paulsen Variation so 'Classical' may have to do. Black has a choice of two orders, 2 ... d6 and 5 ... ttJc6 versus 2... ttJc6 and 5 ... d6. Today his decision usually hinges on how he feels about 2 ... ttJc6 3 .tb5 as opposed to 2 ... d6 3 .tb5+, as well as how he would deal with 3 c3 or 3 ttJc3 . The Paulsens used the Classical to reach the Dragon, Scheveningen and other lines that came to prominence long after they were gone, including Isaac Boleslavsky's 6 .te2 e5. Today it's also used as a waiting policy. After 6 .te2, for example, Black can play a Scheveningen or a Dragon more safely (or 6 ... ttJxd4!? 7 'iVxd4 g6 as noted earlier).
But there is no absolutely correct In the majority of games White English move order after 6 .te3 a6, since 7 f3 and 8 'iVd2 or 7 'iVd2 and plays 6 .tg5 and 6 .tc4 and so 8 f3 also have merit. In the 1990s Black spends his home preparation White showed that his better mainly on the Rauzer and Sozin development trumped ... d5 even Variations. This makes him more when g2-g4 is delayed, as in 7 'iVd2 vulnerable to transpositional traps, .te7 8 f3 ttJc6 9 0-0-0 d5 and now beginning with 6 f4. 76
Sicilian Defense
This move can be a problem for a Black who uses the Classical to dodge the Keres Attack and reach the Scheveningen (6 i.e2 e6).
l:.e8 14 i.b3 a6 15 llJd4 llJa5 16 llJf5 llJxb3 17 axb3 llJxe4 18 llJh6+ gxh6 19 l:txe4 i.xe4 20 llJxe4 i.f8? 21 i.g5 with mixed chances. The safest response to 6 f4 may be 6 ... g6, transposing into the Levenfish Dragon. But while a Dragonista will have memorized those complex 7 llJxc6 bxc6 8 e5 and 7 i.b5 i.d7 8 i.xc6 lines, it's unlikely a Classical player has. That's another point in favor of 6 f4!?
If he replies 6... e6 White has an ambitious plan in 7 i.e3 and 'if01 0-0-0. This transposes into 1 e4 c5 2 llJf3 d6 3 d4 cxd4 4 llJxd4 llJf6 5 llJc3 e6 6 f4 llJc6 7 i.e3 when Black has lost the chance for the sharper 6 ... a6 and ... b5/ ... i.b7.
Other tricky Classical options are 6 i.e3 and 6 0, which are both aggressive and waiting moves. They can transpose into a Dragon (6 ... g6) or an English Attack (6 ... e6) - but if Black really wanted to play a Dragon he would have chosen 5 ... g6 and if he wanted an English he would have played 5... e6.
A Black who prefers to meet 6 i.e2 with the Boles1avsky Variation may also have a problem with 6 f4. Books recommended 6 ... e5 because 7 llJxc6 bxc6 8 fxe5 llJg4! gives him excellent darksquare play whether White accepts the pawn (9 exd6 i.xd6) or permits 9 ... llJxe5.
He loses some options this way. Theory prefers ... a6 to ... llJc6 in the English but after 6 i.e3 e6 Black is already committed to ... llJc6. Also, 6 ... e6 7 i.c4! will be annoying to a Black who usually meets 6 i.c4 with Benko's 6 ... 'iib6. There are two clouds on 6 i.e3 's horizon. One is 6 ... e5, which transposes into a line of the Boleslavsky that is considered even (7 llJb3 i.e7 8 i.e2) or into a Sveshnikov Variation that's rated good for Black (7 llJdb5 a6 8 llJa3 b5).
However, 6 ... e5 is tested by 7 llJO! followed by i.c4. An example of a Boleslavsky player finding himself in unfamiliar surroundings is Salmensuu-Atalik, Groningen 1999: 7 ... i.e7 8 i.c4 0-0 9 0-0 exf4 10 i.xf4 i.g4 11 'iid2 i.h5 12 l:.ae1 i.g6 13 ~h1 77
Sicilian Defense
The other cloud is 6 ... lbg4! which theory says is at least equal. That's why 6 f3! is more accurate.
It normally occurs via 6 ... e6 7 'it'd2 .te7 8 0-0-0 0-0 9 lbb3 'it'b6. But White can keep it from happening by playing 9 f4 or 9 f3. Both moves enjoy good reputations.
White can still reach Yugoslav Dragon and English Attack positions (6 ... g6 7 .te3 or 6 ... e6 7 .te3) but he avoids 6 .te3 lbg4. Experience with 6 ... e5 7 lbb3 .te7 8 .te3 .te6 9 lbd5 and 7 ... .te6 8 lbd5 is mixed.
That's why Black may prefer 6...'it'b6 and then 7 lbb3 e6 8 'it'd2 .te7 9 0-0-0 0-0, transposing. A third route to the tabia is 6... e6 7 'it'd2 'it'b6 8 lbb3 .te7 9 0-0-0 0-0.
In the Classical order, 6 .tg5 discourages a switch to the Dragon (6 ... g6 7 .txf6) or the Boleslavsky
Which order is best? The latter two threaten ... 'it'xd4 as well as ... 'it'xb2. But they share the demerit of allowing White to double pawns with .txf6.
(6 ... e5 7 .txf6 'it'xf6? 8 lbd5). Black usually replies 6 ... e6 and experience has shown that White's only way to seek an edge is 7 'it'd2 and 8 0-0-0.
In addition, White can answer 6.•. e6 7 'it'd2 'it'b6 with 8 0-0-0 (and transpose after 8 ....te7 9 0, for example). The only way to punish this is 8 ... lbxd4 9 'it'xd4 'it'xd4
RAUZER VARIATION
Black makes major decisions 10 Ihd4. Unlike other early about the timing or need for Sicilian endgames, this one favors ... lbxd4, ... h6, ... a6 and ... 'it'b6. If White, according to Peter Svidler. White meets ... 'it'b6 with lbb3 he might reach this position, one of the best Rauzer tabias for Black.
He preferred 6•..'it'b6. But that may have its own problem, 7.te3!? 78
Sicilian Defense
It's considered double-edged enough to be attractive to both White and Black and can come about from 6 i.g5 e6 7 .d2 i.e7 80-0-00-09 f4liJxd4 10 .xd4. If Black would rather defend this position than the one that arises after 9 liJb3, he can get where he wants to go with 8•••liJxd4 9 .xd4 0-010 f4.
Declining the pawn is bad (7 ...•a5? 8 liJb3 .c7 9 liJd5! or 8...•d8? 9 i.e2 e6 10 0-0, a Scheveningen a tempo or two down).
The problem with the second order is White doesn't have to cooperate by playing 10 f4. He also The same goes for 7 ... liJg4 has 10 e5, which Jan Timman said 8 liJd5! liJxe3 9 fxe3 in view of is the only way to punish Black's 9...•a5+ 10 b4! liJxb4 11 liJb3! sequence, e.g. 10... dxe5 11 .xe5 liJd3+ 12 ~e2, winning a piece. In i.d7 12 i.e2 :c8 13 i.e3 a6 14 g4 contrast, 6... e6 7 .d2 'iWb6 8 i.e3 was unclear in Akopian-Jobava, liJg4! is fine for Black. Beer Sheva 2005. So the answer to which order is White also has a promising best may depend on whether Black can play 6... 'iWb6 7 i.e3 .xb2 alternative in 10 f3. That works 8 liJdb5. If not, 6... e6 7 .d2 i.e7 better here than with knights on the and 8 0-0-0 0-0 9 liJb3 'iWb6 is board (6 i.g5 e6 7 'iid2 i.e7 better. 8 0-0-0 0-0 and then 9 f3 a6 10 g4 d5!). Another tabia is: That's the tradeoff: Black has to decide which middlegame to shun the one with 8... 0-0 9 liJb3 or the 8... liJxd4 9 .xd4 0-0 10 e5 and 10 f3 ones. A Black trickster may like the rare 7 'iid2 i.e7 8 0-0-0 0-0 9 f4 i.d7!? 79
Sicilian Defense
His idea is to answer the natural 10 Jic4 with 1O ... liJxd4 11 'iixd4 Jic6 and reach a promising position (12 .l:thel 'iia5). That's a line that commonly comes about via 9 ... liJxd4 10 'iixd4 'iia5 11 Jic4 Jid7 when White passes up 12 e5! for the less ambitious 12 ::thel.
This order is significant because on 9 f4 Black transposes not into the bad line (9 ... 0-0?) but into a very good one, 9... Jid7!. This comes about via 7 'iid2 a6 8 0-0-0 Jid7 9 f4 Jie7 when White has bypassed 9 D.
If he insists on 9 f3 Black responds 9...liJxd4 to 'iixd4 b5! and posts his bishop more usefully The apparent refutation of on b7 than on d7, e.g. 11 h4 Jib7 9 ... Jid7 is to liJdb5, threatening 12 ~bl 'iic7 13 'iid2 .l:tc8 14 Jid3 liJxd6. But to ... d5!? offers good h6 15 Jie3 b4 16liJa4 d5!, Sax-Acs, play for a pawn, e.g. 11 exd5 liJxd5 Hungarian Championship 2003. 12 liJxd5 exd5 13 'iixd5 Jig4! 14 'iixd8 Jixd8 15 ::td2 Jib6. Or And since 9 liJb3 allows 9... b5! 11 e5 liJe8 12 h4 a6 13 liJd4 f6, under better than usual circumSutovsky-Zviagintsev, Essen 2000. stances, the real test of Black's Another deceptive order is move order is 9 Jixf6. Then 7 'iid2 Jie7 8 0-0-0 a6 (or 7... a6 9 ... Jixf6? drops a pawn to 8 0-0-0 Jie7). This bears a 10 liJxc6. But experience with resemblance to 7 'iid2 Jie7 80-0-0 9... gxf6 10 f4 Jid7 and 10 Jie2 0-0 9 f4 a6? which Paul Keres 'iib6 has been inconclusive so this refuted with 10 e5! dxe5 11 liJxc6 order may be a real contribution to the Rauzer. bxc6 12 fxe5.
80
Sicilian Defense
SOZIN VARIATION
Pal Benko's great contribution to 6 Jtc4 theory was realizing that 6...'iib6 7 lbb3 e6 transposes to a fairly good Scheveningen for Black, even though he'll probably retreat his queen to c7 at the cost of a tempo. One of the secrets of the Sozin is that getting into the right line is often worth a tempo.
Joseph Blackburne used 6 Jtc4 against the Classical Variation so that 6 ... g6 could be punished by 7 lbxc6 bxc6 8 e5!. Later Os sip Bernstein demonstrated the finesse of6... Jtd7.
This is evident when White stays within Sozin borders by answering 6.. :ii'b6 with 7 lbdb5!? and then 7... a6 8 Jte3 'i'a5 9 lbd4 e6 (not 9 ... lbxe4? 10 'iif3):
Books recommend 7 0-0 because 7 ... g6 8lbxc6 slightly favors White. But 7... e6!? is an 'old' Sozin with 0-0 that isn't considered as dangerous as 0-0-0 lines, e.g. 8 Jtb3 Jte7 9 Jte3 0-0 10 'i'e2 lbxd4 11 .ixd4 .ic6. And there's been virtually no GM experience with 8lbdb5 'iib8 9 Jtf4 or 9 Jtg5.
White is a tempo down in an 'old' Sozin. (It would occur after 6 Jtc4 e6 7 Jte3 a6 if Black had the extra tempo .. :~i'a5.)
The subtle reply is 7 Jtb3, which waits for Black to choose between 7... e6 and 7 ... g6. But after 7 ... g6 8 .ie3 Black has 8 ... lbg4! and against 8 1'3 there is 8 ... lbxd4! 9 'i'xd4 Jtg7 with excellent play, e.g. 10 Jte3 0-0 11 'i'd2 b5 12 Jth6 .ixh6 13 'i'xh6 b4 14 lbd5 e6! 15 lbe3 a5 16 h4 a4 17 Jtc4 d5!, Shivaji-Motylev, Minneapolis 2005.
But even a move down, White has good chances following 10 0-0 Jte7 11 f4, theory says. In fact, GMs often gives the tempo back, say with 11...'i'c7 12 .tb3 0-0 to reach a book line, 13 'i'f3 lbxd4 14 Jtxd4 b5, with equal chances.
81
Sicilian Defense
In the orthodox 6 .tc4 e6 White chooses between the old 0-0 and f2-f4-fS plan and Dragolyub Velimirovic's attack with .te3, 'i'e2, 0-0-0 and g2-g4-gS. Black, meanwhile, has to pick either quick development ( ... .te7, ... 0-0) or more ambitious but slowdeveloping counterplay (... a6, ... bS, .. .'iVc7).
.td7 10 .tb3 ""8 11 g4! :c8 12 gS ttJe8 13 h4 ttJaS 14 g6! as in Velimirovic-Milic, Belgrade 1965, one of the earliest successes of White's strategy. Instead, Black does better against the Velimirovic when he begins queenside operations, e.g. 7 i.e3 a6 8 'ife2 "iic7 9 0-0-0 ttJaS or 9... i.e7 10 l:thgl ttJaS 11 i.d3 bS.
Matching the right defense with each attacking plan and vice versa is crucial. Queenside play didn't fare well against the old Sozin when 6 .tc4 e6 7 0-0 was young. After 7... a6 8 .te3 'i'c7 9 .tb3 ttJa5 10 f4 b5:
What this means is each side should be flexible enough to exploit the plan his opponent chooses. White's best waiting move is 7.tb3. Then on 7....te7 8 .te3 0-0
Black often got crushed by 11 fS ttJxb3 12 cxb3! (l2 ....te7 13 'iVd7 14 'i'f3 and IS eS or IS fxe6 fxe6 16 'iVh3).
... he can head toward the Velimirovic with 9 "iie2!, having the benefit of knowing where Black's king lives.
:c1
An antidote was found in quick development, 7....te7 8 .te3 0-0 9 f4? d5!.
The best waiting reply is 7... a6. If White makes a commitment with 8 f4, then 8... .te7 9 .te3 0-0 heads into an old Sozin tabia that offers chances for both sides (10 0-0 ttJxd4! 11 .txd4 bS!).
But that policy doesn't fare well against the Velimirovic, e.g. 7 .te3 .te7 8 'iVe2 0-0 and now 9 0-0-0 82
Sicilian Defense
The waiting game continues if 7 i.b3 a6 is followed by 8 i.e3. Strong players, such as Vishy Anand, have played 8.. :iic7 to keep the queenside counterplay option alive. However switching to the old Sozin with 9 f4! exploits Black's tardiness and he is worse after 9...b5 10 ttJxc6 'iixc6 11 f5. Or The novelty pays off when White 9... i.e7 10 0-0 ttJxd4 11 i.xd4 b5 plays 6 i.c4 and discovers after 12 e5!, which won quickly in 6 ... a6 7 i.b3 e6 that he's been Golubev-Lerner, Odessa/Istanbul pulled into a popular Najdorf line 2006 after 12 ... dxe5 13 fxe5 ttJd7 that may be good for Black. 14 :xf7! ~xf7 15 'iif3+.
In the normal Najdorf order, 5 ... a6 6 i.c4 e6, White can Both sides are finally out of circumvent that position (7 i.b3 passes after 8 i.e3 and 8...i.e7!. ttJbd7) by means of 7 O-O!, as Then 9 0-0 0-0 is a good Sozin for played by that devious transposer Black and 9 'iie2 0-0 10 0-0-0 is a Deep Fritz against Vladimir Velimirovic tabia that is still being Kramnik in 2006. evaluated. But there are no other 5... ttJbd7 tricks and plenty of drawbacks: On RARE FIFTH MOVES 6 i.g5 Black's best is probably ... a6, reaching a Najdorf (5 ... a6 6 We know a lot about 1 e4 c5 2 ttJO d6 3 d4 cxd4 4 ttJxd4 ttJf6 6 i.g5 ttJbd7) that hasn't had 5 ttJc3 when Black replies 5... g6, success since the 1960s. 5... e6, 5... a6 and 5... ttJc6 but very Also, White's d4-knight cannot little about two alternatives. The be exchanged off by ... ttJc6xd4 so it first is 5...ttJbd7, which has been dominates the center. He gets a tried by Tigran Petrosian, Mark superior English Attack after Taimanov, and Bent Larsen. In the 6 i.e3, e.g. 6... a6 7 g4 h6 8 f3 e6 Havana 1966 tournament book 9 'iid2 'iic7 10 0-0-0 and 11 h4, or Petrosian wrote, 'The only 8...b6 9 'iid2 i.b7 100-0-0 e6 11 h4 advantage of this move... consists d5 12 i.h3! with a strong initiative, Balogh-Bilek, Budapest 2002. of its relative novelty.' 83
Sicilian Defense
The more successful alternative is 5...i.d7, a specialty of Viktor Kupreichik and Aloyzas K veinys.
the unclear 7 ltJdb5 i.c6 8 i.xf6. Bottom line: 5... i.d7 has a few advantages over the Classical order but Black has to risk the worst-case scenario, a somewhat questionable English Attack or Rauzer. NAJDORF VARIATION The essential Najdorf move, ... a6, is useful in most Sicilian variations and that gives Black freedom to transpose into them. Whether that makes sense depends on which of the nearly a dozen reasonable alternatives White picks at move six. The move orders of some of them, such as 6 i.e2, 6 i.g5 and 6 f4, are too complex to consider here. Among the others:
This works as a waiting move in cases such as 6 i.e2 g6, when Black has dodged the Yugoslav Attack, or 6 ... e6, when he got a Scheveningen without encountering the Keres or English attacks. The principle drawback to 5 ... i.d7 is the bishop occupies the KN's best retreat square and that improves the impact of g2-g4-g5 in an English Attack (6 i.e3 or 6 f3).
(a) 6 i.d3
But in practice 6 i.g5 is White's favorite. Then 6 ... ltJc6 transposes into a 6 ... .td7 Rauzer. The only apparent benefit to Black for doing it that way, rather than 5 ... ltJc6 6 i.g5 i.d7, is to avoid the Sozin and other Classical alternatives like 6 f4.
This virtually unbooked move has great transpo value. After 6... e6 7 f4 or 7 0-0 White gets quite reasonable Scheveningen positions in which i.d3 replaces i.e2, e.g. 7 f4 ltJc6 8 ltJf3 i.e7 9 0-0 and "iVel-g3. If Black never plays
For the sake of originality, GMs prefer meeting 6 i.g5 with 6 ... e6. That offers White a choice of a very rare Najdorf line (7 f4 a6), another Rauzer (7 "iVd2 ltJc6 8 f4 h6) and 84
Sicilian Defense
Scheveningen positions, because he meets 6 ~e2 with 6 ... e5, then 6 ~d3 could be a problem.
2004. That can't be the best line for Black after 6 ~d3. But what is? (b) 6 a4
And if Black answers 6 ~d3 with 6... e5 White has a good retreat square at e2, e.g. 7 ttJde2 ~e7 8 0-0 0-0 9 f4 ttJbd7 10 ttJg3 (10 ... exf4 11 ttJf5! ttJe5 12 ttJxe7+ 'ilixe7 13 ~xf4 h6 14 ~g3! and .th4, Gallagher-Karjakin, Panomo 2002).
This is White's best waiting move in the Najdorf. Against 6... e5, he can continue 7 ttJf3 so that 7....te7 8 ~c4 0-0 9 ~g5 or 7... h6 7 ~c4 will put him a tempo ahead of 6 ~e3 e5 7 ttJf3 and 6 ~e2 e5 7 ttJf3 lines.
The drawback to 6 ~d3 would seem to be 6•.. ttJc6, since the attacked knight cannot be maintained on d4 (7 ~e3 ttJg4!). However after 7 ttJxc6 bxc6 8 0-0:
Instead, Black may transpose to a Scheveningen, Boleslavsky or Dragon after 6 ... ttJc6. Evidence indicates the a-pawn advances help White in the Dragon. For example, 6•..ttJc6 7 .te2 g6 8 0-0 ~g7 9 ~e3 0-0:
White can continue ttJa4/c2-c4! as he does in the Sozin line that runs 1 e4 c5 2 ttJf3 ttJc6 3 d4 cxd4 4 ttJxd4 ttJf6 5 ttJc3 d6 6 .tc4 'ilib6 7 ttJxc6 bxc6 8 0-0 and ttJa4 with a bright outlook.
The hole at b6 gives White chances for a queenside bind here after 10 'ilid2 ttJg4 11 ~xg4 ~xg4 12 ttJd5! lic8 13 a5. Also good is 10 f4 since 1O ... 'ilib6, which instantly equalizes when the a-pawns are on their original squares, can be met by the strong 11 a5! ttJxe5 12 e5. If Black has to settle for 10 ... ~d7 then
In the diagram he can answer 8... e5 with 9 b3 .te7 10 .tb2 0-0 11 ttJa4. Then 11...~e6 12 'ilie2 'ilic7 13 c4 ttJd7 14 c5! favored him in Dgebaudze-Wemmers, Belgium 85
Sicilian Defense
11 ttJb3 and ttJd5 guarantees a positional edge.
So Black turned to 7...b5. Then White found 8 0-0 was good since 8 ... iLb7 9 :tel! ttJbd7 10 .tg5 ttJc5? 11 iLd5! is a sound sack, as Fischer showed.
Another idea after 6...ttJc6 is 7 iLe2 e5. Books recommend 8 ttJxc6 and show how it only equalizes. Better is 8 ttJb3! iLe7 90-00-0 and now 10 iLg5!:
To preempt that Black began to play 7...ttJbd7 and ... ttJc5, much in the spirit of the Scheveningen (1 e4 c5 2 ttJf3 d6 3 d4 cxd4 4 ttJxd4 ttJf6 5 ttJc3 e6 6 iLc4 i.e7 7 .tb3 0-0 8 iLe3 ttJa6!?). After 8 0-0 ttJc5 Black is faster than in the Fischer line. The counter-finesse is 7 0-0, which had been discarded in the 1950s. Then 7 ... b5 8 .tb3 transposes to the 7 i.b3 b5 8 0-0 position that White wants and it avoids the 7 i.b3 ttJbd7 one that Black wants.
This favors White after 10...iLe6 11 iLxf6 and ttJd5. This is different from the normal Boleslavsky (5 ... ttJc6 6 i.e2 e5 7 ttJb3 i.e7) when Black equalizes immediately after 80-00-09 i.g5 ttJxe4!.
If Black plays 7...ttJbd7 anyway then 8 i.g5!:
The addition of a2-a4 and ... a6 means that 10 i.g5 ttJxe4? dooms the knight, 11 iLxe7 ttJxc3 12 iLxd8 ttJxdl 13 i.c7 ttJxb2 14 :tfbl. (c) 6 iLc4 There's general agreement that Black's best is 6... e6 but the consensus about White's reply is breaking down. The old favorite, This transposes to a 6 i.g5 7 i.b3, was considered best because it preserves the 0-0-0 Najdorf position that has been option, e.g. 7 ...iLe7?! 8 f4 0-0 9 scoring well for White since the 'it'f3 and then 9... ttJbd7? 10 g4 or 1960s. Strong players such as Leonid Yudasin have tried 9.. :ilc7 10 f5 ttJc6 11 i.e3. 86
Sicilian Defense
8... 'ii'c7?!. But 9 .txe6! fxe6 10 liJxe6 has been crushing Black
11 liJd5, as in Topalov-van Wely, Wijk aan Zee 2007.
since a Keres brilliancy (l0 .. .'iic4 11 liJd5 rJif7 12 .i.xf6 rJixe6 ~ 13 .&.c3! and wins). To play this line Black may have to risk 8... h6 9 .th4 g5 10 .i.g3 liJe5.
The crafty response to 6 .te3 is 6...liJc6.
He can avoid all this by remembering why 7 0-0 was abandoned half a century ago 7....i.e7! 8 .i.b3 0-0. Then 9 f4liJc6 reaches a fairly balanced main line of the old Sozin and 9 .te3 b5 is an equally double-edged line of the Najdorf.
This avoids 6... e6 7 g4 and forces an English Attacker to prepare g2-g4. On 7 'iVd2 Black has an active reply in 7... liJxd4 8 .txd4 e5 and .•..te6. And if White offers a Scheveningen, with 7 .te2, Black can force his way to a Classical Dragon, 7... g6, in which White has been deprived of the Yugoslav Attack and .tg5 lines. The price he pays is being committed to ... a6!?
(d) 6.te3 White is ready for a quiet Scheveningen or an English Attack. Some GMs, like Michael Adams, would rather play against 6 .te3 e5 than 6 .te2 e5 so they begin with 6 .i.e3. If 6... e6, then 7 .i.e2 allows them to transpose to 6 .te2 e6 7.i.e3. Some English Attackers prefer 6 f3 and 6... e6 7 .i.e3. In this way they avoid 6 .i.e3 liJg4. But they also surrender options such as 6 .i.e3 e6 7 g4 and 6 .i.e3 e5 7 liJb3 .i.e7 8 'ii'd2!?
TAIMANOVIKAN VARIATION The standard starting position of the Taimanov Variation can be reached in two different ways, and Mark Taimanov used both of them, 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4 4 liJxd4 e6 and 2... e6 3 d4 cxd4 4 liJxd4 liJc6.
In the last line White delays £2-f3 so that he rules out ... d5 and keeps the £2-f4 option, as in 8 'ii'd2 .i.e6 9 0-0-0 liJbd7 10 f4. He can also trick Black into a bad line after 8... 0-0 9 0-0-0 b5? 10 f3! .te6
The first order discourages the King's Indian Reversed (3 d3), which is promising against 2 ... e6. 87
Sicilian Defense
The second order eliminates the Rossolimo, in view of 2 ... e6 3 .tb5? a6!. Black also may choose an order based on how he intends to meet 3 lLlc3 or 3 c3. For example, 1 e4 c5 2 lLlf3 e6 3 lLlc3 is a good Chameleon for White but 2 ... lLlc6 3 lLlc3 e5! isn't. Both orders, as well as a third used by Salo Flohr, 1 e4 cS 2 lLlo lLlc6 3 d4 cxd4 4 lLlxd4 'iVc7!?, have the drawback of allowing a Maroczy Bind. For instance, 4... e6 S lLlbS threatens lLld6+ and usually buys time for 5... d6 6 c4!.
The timing of the three key Taimanov moves, ... a6, ... lLlc6 and ... 'iVc7, has turned out to be more complicated than it seemed a decade or so ago. Only recently was 1 e4 cS 2lLlo lLlc6 3 d4 cxd4 4 lLlxd4 e6 S lLlc3 a6 tested by the English Attack, for example. This has proven to be dangerous after 6 .te3 lLlf6 7 g4 or 6... 'iVc7 7 'iVd2 lLlf6 8 0-0-0. This provides an argument in favor of S•.• 'iVc7.
Black can bailout of the bind with S.•.lLlf6!?
Now on 6 ..te3 Black can respond more vigorously with 6•••lLlf6 7 'iVd2 ..tb4! and then 8 f3 lLlxd4 9 .txd4 e5 or 9 'iVxd4 ..txc3+ 10 bxc3 e5.
If White defends the attacked epawn with 6 lLllc3 then 6 ... d6 7 .tf4 e5 8 .tg5 is the Sveshnikov Variation. So is 6 .tf4 e5 7 .tg5 d6 8lLllc3.
The drawback to this order may be 7 lLldbS 'iVb8 8 f4 although Black gets a tempo back after 8... d6 and ... a6.
Of course, Black has a simpler route to the Sveshnikov if that's his wish, and 5... lLlf6 makes most sense when Black seeks a Taimanov and is upset to see 5 lLlb5.
Fischer, who liked lLlb5/c2-c4 lines as White, reached the Taimanov as Black through another order, Ilya Kan's 1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 lLlxd4 a6. After SlLlc3 88
Sicilian Defense
he transposed (S ... ~c6) into the S... a6 Taimanov.
and now S c4 ~f6 6 ~c3 .i.b4!) but not versus the Kan. Another minus Other Kan players delay ... ~c6 is that White can play S i.d3 further, by means of 5...fie7 in because his knight is not hanging on order to disturb White's d4 as it is in the Taimanov. That development with ... b4 or ... .i.b4. enables him to create a delayed For example, 6 i.e2 ~f6 7 i.e3? Maroczy Bind with c2-c4. allows 7...i.b4!, e.g. 8 fid3 ~c6 Black will have ample threatening 9... ~eS. opportunity in these orders to Or 8 i.d2 i.xc3 9 i.xc3 ~xe4 transpose into a Scheveningen with 10 i.b4 'iVb6 11 i.a3 ~c6 with ... d6. But that makes the most sense little compensation for the lost when White has played a move that pawn in Bonte-Cabrilo, Timisoara is useful in the Taimanov but isn't 2006. in a Scheveningen. An illustration is Another plus of the Kan order appears when White fianchettos, 5 ~c3 fie7 6 g3 and then 6....i.b4!:
a2-a3 to rule out ... i..b4 and to anticipate ... b5-b4. A popular order in the 1960s was 1 e4 e5 2 ~O ~e6 3 d4 exd4 4 ~xd4 e6 5 ~e3 a6 6 .te2 fie7 7 a3:
White cannot ignore ... i.xc3+. But neither can he claim an advantage after 7 ~e2 ~f6 or Here 7... ~f6 8 0-0 .i.e7 9 .i.e3 7 i.d2 ~f6 (8 i.g2 ~c6 9 ~b3 d6! is a good decision. Compare it i.e7! and ... d6 with an excellent with 1 e4 cS 2 ~f3 d6 3 d4 cxd4 Scheveningen). 4 ~xd4 ~f6 S ~c3 e6 6 .i.e2 a6 The drawbacks to the Kan order 70-0 fic7 8 f4 ~c6 9 .i.e3 i.e7 and begin with S c4. This works badly now instead of normal lines such as against the Taimanov (1 e4 cS 10 a4 or 10 fie1 White would 2 ~f3 ~c6 3 d4 cxd4 4 ~xd4 e6 transpose with 10 a3?!. 89
Sicilian Defense
But it's considered unwise for Mortensen). Black is ready to playa Black to transpose into a normal Scheveningen after 9... lLlf6 Scheveningen after g2-g3 when and 10 ... 0-0. he's already played ...'fic7. After The only way to prevent that is 1 e4 c5 2 lLlf3 lLlc6 3 d4 cxd4 9 l:.el, with the idea of 9... lLlf6? 4 lLlxd4 e6 5 lLlc3 'fic7 6 g3 Black 10 lLlxc6!, reaching the 1960s more or less has to play 6... a6 to position. But the difference is avoid 7 lLldbS 'iVb8 8 ..tf4. 9•••.i.f6!? Then 6... a6 7 ..tg2 d6 8 0-0 lLlf6 Now 10 lLlb3, 10 lLlde2 and 9 liel! allows White to carry out his 10 .i.e3 aren't much after strategic plan of lLlxc6 and e4-eS 10 ... lLlge7. The crucial lines are which ... ~c7 is intended to prevent 10 lLlxc6 bxc6 11 .tf4 eS and - by tactical means. That is, 9••. ..te7 10 lLlfS exfS 11 lLldS and unless 10 lLlxc6! bxc6 11 e5! dxe5 they prove more successful than 12 J::txe5 and 12 .. :ii'xe5 13 .txc6+. they have been, Mortensen's order There have been no major is a significant strengthening of the improvements in that line since the Taimanov. 1960s. For this reason many players feel S... 'fic7 is a slight error that is punished by 6 g3! and that S... a6 is a superior waiting move (6 g3 d6!).
THE ...~6 VARIATIONS There is a family of lines built around ... 'iib6 and ... e6, often with ... ..tcS. They all use pressure on the b6-f2 diagonal, at least at the start, but often are used to reach the Scheveningen. The family members are:
But Black has another finesse even if he's committed to ...'fic7 and ... a6.
(a) 1 e4 c5 2 lLlf3 lLlc6 3 d4 cxd4 4 lLlxd4 ~6, (b) 1 e4 c5 2 lLlf3 lLlc6 3 d4 cxd4 4 lLlxd4 lLlf6 5 lLlc3 ~6, (c) 1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 lLlxd4 ..tc5, and (d) 1 e4 c5 2 lLlf3 e6 3 d4 cxd4 4 lLlxd4 ~6.
This can come about after 5•• :ii'c7 6 g3 a6 7 .i.g2 and then 7•.. d6 8 0-0 ..te7!? (Erling
The last two, popularized by Michael Basman and Kveinys 90
Sicilian Defense
respectively, can transpose into one another, e.g. 1 e4 cS 2 tLln e6 3 d4 cxd4 4 tLlxd4 iVb6 and now S tLlc3 i.cs 6 i.e3 tLlc6.
7 i.d3 i.e7 8 0-0 0-0 9 i.e3 "fic7 10 f4 d6 11 iif3 a6 12 a4 or 12 g4. Theory says White is favored a bit but Black's position is solid. If that's what Black wants, which is the best route to it? In order (a) White has an extra option of S tLlbS. But Black has a surprising reply in S... a6 6 i.e3 'iVd8!.
Not 6.. :iVxb2? 7 tLldb5, which was refuted in a Morphy-Paulsen (!) game. Black has been getting quite good positions after 6... tLlc6, e.g. 7 liJa4 'i'aS+ 8 c3 i.xd4! 9 i.xd4 tLlxd4 10 'i'xd4 liJf6 11 i.e2?! eS! 12 'i'c4 d6. Black's most accurate route to the diagram is Kveinys' . One reason is that 4... i.c5 has not done well when met by the surprising 5 .te3 'i'b6 6 c3!. For example, 6... 'i'xb2 7 'i'b3 'i'xal? 8 tLlc2 or 7 ... 'i'xb3 8 axb3. Or 6 ... tLlc6 7 tLld2! tLlxd4 8 tLlc4.
His point is that 7 tLld4 tLlf6 8 tLle3 e6 is a Taimanov Variation with i.e3. That has a much better reputation today thanks to 'iVd2 and 0-0-0 but it may not be something White feels comfortable with, especially ifhe normally plays i.g2 against the Taimanov. This order may also confuse White since 7 tLlSc3 and then 7... e6 8 tLld2 bS! is fine for Black. If Black has any doubts about that, he should prefer order (b) because 6 tLldbS a6 7 i.e3? is a mistake (7 .. :iVa5 wins material) and 7 tLla3 e6 8 tLle4 "fie7 9 i.e3 bS has never done well for White.
The three other orders often reach Scheveningens after White retreats his knight to b3. The benefit to Black is avoiding the English and Keres Attacks. For example, in the Kveinys order, play may continue 5 tLlb3 tLlc6 6 tLlc3 liJf6 and then 91
Sicilian Defense
FOUR KNIGHTS VARIATION For decades the variation that begins 1 e4 c5 2 liJo liJc6 3 d4 cxd4 4 liJxd4 liJf6 5 liJc3 e6 was in the Sicilian's 'dubious' basket because the main line, 6 liJdb5 .tb4 7 a3, handed White the two bishops and better prospects. After 6 ... d6 Black says, 'I outsmarted you. I reached a Scheveningen in which your a2-a3 is a wasted tempo.'
This changed in the 1970s when Black adopted the Four Knights as a way of transposing, after 6 ... d6 7 .tf4 e5 8 .tg5 a6, to what was dubbed the Sveshnikov Variation.
White replies, 'Not at all. My goal was to avoid both the Sveshnikov and 6 ... ..tb4. The added
This denies White options of the normal Sveshnikov, 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4 4 liJxd4 liJf6 5 liJc3
benefit is that I'll play 7 ..tc4! and get a Sozin in which you've denied yourself the Benko line.'
e5 6 liJdb5 d6 7 .tg5 a6, such as 6 liJf5, 6 liJdb5 d6 7 liJd5 and 7 a4. Of course, there's a lot of theory to know in the Four Knights as well, if White decides not to join him on the road to the Sveshnikov. But Black can reduce that significantly by shifting to a Scheveningen with 6 ... d6.
Or Black can play 6 ... e5 and say, 'I'm the one who got what I wanted, a Sveshnikov. In this version you can't play 7 liJdb5 d6 8 .tg5 because your knight has no retreat aft er 8... a6".. To which White replies, 'Nonsense, I'm the one who came out ahead because after 7 liJb3 it's a Sveshnikov with an extra tempo for t me. That tempo sops ... A~b4'.. '
That is usually second-best. But it avoids heavy analysis such as 6 g3 d5 7 .tg2 ..tb4 8 exd5 liJxd5 or 6 .te2 .tb4 7 0-0 .txc3 8 bxc3 liJxe4 9 .td3 d5 10 ..ta3. Instead, Black is roughly equal after 6 g3 d6!? or 6 .te2 d6!?
There is no consensus on the best reply to 6 a3 but 6 ... d6 can't be bad, especially since 7 ..tc4?! really isn't a good Sozin after 7 .. :ikc7! 8 ..ta2 a6 9 0-0 liJxd4 10 'iVxd4 liJg4.
There's a who-tricked-whom quality to 6 a3.
92
Sicilian Defense
The Four Knights variation-into Zaitsev said he played this for the Sveshnikov order was bound to first time in a five-minute game spawn a counter-finesse. The ever- and later in a simul that went creative Igor Zaitsev found one in 8...~xd5 9 exd5 exf410 dxc6 bxc6 11 'ii'f3. 6 ~db5 d6 7 i.f4 e5 8 ~d5!?: Experience with 11...d5 12 0-0-0 remains mixed (12 ... .lid6? 13 'ii'c3! cxb5 14 'iVc6+ .lid7 15 'ii'xd6 favored White in a ZaitsevSveshnikov game). But so do the main lines of the Sveshnikov Variation. The Sicilian is never easy for Black - or for White.
93
Chapter Four: Semi-Open Games Black reveals something about himself when he answers 1 e4 with a move other than 1...e5 or 1...c5. An Alekhine's specialist, for example, almost certainly likes positional imbalances and quick counterplay. He may have a problem with 1 e4 liJf6 2 liJc3!? because the Vienna Game (2 ... e5) is not likely to give him what he wants.
can lead to a Steinitz French (4 f4 e6 5 d4). But the alternatives are somewhat questionable, such as 3... liJe4 or 3... d4 4liJce2liJg4. If White is going for a bigger edge in the Alekhine's, the traditional choice is the Four Pawns Attack, 1 e4 liJf6 2 e5 liJd5 3 c4 liJb6 4 d4 d6 5 f4. The standard response is 5... dxe5 6 fxe5liJc6. It sets some traps based on a liJf3/ ... .tg4 pin. For example, 7liJf3 hopes for 7... .if5? 8 d5! liJb4 9liJd4 with advantage but 7... .tg4! is excellent for Black.
Moreover, few Alekhine's players are prepared for a French (2 ... e6 3 d4 d5) or a Pirc (2 ... d6 3 d4), and a Sicilian (2 ... c5?! 3 e5) is dubious. On 2... d5 3 e5:
The same goes for 7 .te3 .if5 8 liJc3 e6 9 liJf3 and 9... .ig4!, despite the loss of time, e.g. 10.te2 .txf3 11 gxf3 'iih4+ 12 .tn 'l'f4. White avoids this by simply reversing the order, 9 .te2! and 10 liJf3. A more deceptive Black will delay both ... dxe5 and ... liJc6 as Viktor Korchnoi did with 5....tf5 6liJc3 e6.
Black may be reluctant to play the best move, 3... liJfd7! because it 94
Semi-Open Games
12 i.e3ltJc6 gets Black to where he wants to go. When this order was introduced in a 1960 GellerKorchnoi game White became confused and was worse after 11 ~f4? ltJc6 12 exf6 ..txf6. ALEKHINE'S: 4ltJf3 When White meets 1 e4 ltJf6 This pays dividends after 7 ~d3? 2 e5 ltJd5 with the more modest i.xd3 8 'i'xd3 c5 (9 d5? dxe5 3 d4 d6 4 ltJf3 the timing of pawn 10 fxe5 'i'h4+ and ... 'i'xc4). Better exchanges on d6 and e5 become is 7 ltJD, after which 7 ... dxe5 pivotal. After 4...i.g4 for example 8 fxe5 ltJc6 9 ~d3?! i.g4! is White looks for the right moment for exd6, since ... cxd6 allows him to another trap. grab space with d4-d5!. But after 9 i.e3! ~e7 White has But on the immediate 5 c4 ltJb6 reasonable chances in either the 6 exd6 Black has 6... exd6!. That's well-trod 10 i.e2 0-0 11 0-0 f6 or the enterprising 10 d5!? (1O ... ltJb4 a satisfactory version of the Alekhine's Exchange Variation, 11 ltJd4 i.g6 12 a3). more commonly seen after 4 c4 Since 10 d5 demands exact ltJb6 5 exd6 exd6. More accurate is preparation by Black, he may prefer 5 i.e2 e6 and then 6 c4ltJb6 7 exd6 to meet 7 ltJf3 dxe5 8 fxe5 with or 7 0-0 ~e7 8 exd6. 8•..i.e7!. This allows him to reach a Another question concerns Four Pawns tabia smoothly after h2-h3. In most cases Black will 9 i.e2 0-0: simply retreat to h5. But 5 ..te2 e6 6 h3 is mistimed in view of 6... ~xf3! 7 ~xf3 c6.
Now 10 d5 i.b4! collapses the center and 10 0-0 f6 11 exf6 ~xf6 95
Semi-Open Games
Black wants to set up a target on e5 (8 0-0 dxe5 9 dxe5 lbd7 10 'iVe2 'tic7). He benefits from White's sloppiness because 8 c4 lbb6 9 exd6 allows the favorable 9••• lbxc4!.
If the bishop were still on g4 Black would stand well with ....tf6. But here it's en prise and that gives White time for 14... i.g6 15lbd4!. His pieces dominated after 15 ... ..tf6 16 lbxc6 bxc6 17 'i'xc6, as in A.Sokolov-Arzhenkov, Bor 2000 (17 ... i.xb2 18 ::tadl 'i'a5 19 'tixd6 'iVxa2 20 ..tf3 ::tac8 21 c5! a5 22 c6).
If White intends to put the question to the bishop, a better order is 6 O-O! i.e7 7 h3. Then 7.....txf3 8 i.xf3 c6 9 c4 lbb6 loses material to 10 exd6 and 11 c5. The reason Black needs 8 ... c6 is that 8... dxe5 fails to 9 c4! lb-moves 10 i.xb7.
A separate branch of the 4 lbf3 tree is 4... g6. Today 5 .tc4 and 5... lbb6 are considered best. But the old 5... c6!? has a new use thanks to
Moreover 7...i.f5 permits White to carry out the space-grabbing plan
transposition.
under ideal conditions. For example 8 c4 lbb6 9 lbc3 0-0 10 i.e3 lbc6 and now 11 exd6 cxd6 12 d5! exd5 13 cxd5! lbe5 14 lbxe5! dxe5 15 "iVb3 with advantage.
After 6 0-0 White seems to be getting a good Exchange Variation in view of 6.....tg7 7 exd6! exd6 8 ::te1+.
The main benefit of h2-h3 lies in 7... i.h5 8 c4 lbb6 9 lbc3 lbc6 10 exd6 and then 10... cxd6 11 i.e3 0-0 12 d5!. This is quite good for White, particularly after the natural 12 ...exd5 13 lbxd5 lbxd5 14 'tixd5:
However, 6... dxe5! 7 lbxe5 ..tg7 is the order of a slyboots. What has happened is that Black transposed into a line attributed to Edwins Kengis. It usually comes about via 4 lbf3 dxe5 5 lbxe5 g6 6 ..tc4 c6 70-0 i.g7. 96
Semi-Open Games
11...0-0 12 b3 .tf6 13 .tb2 as 14 0-0 with a space edge, EmmsBaburin, Isle of Man 1997.
Black has been equalizing, e.g. 8 c3 lbd7 9 lbxd7 ..txd7 10 lbd2 0-0 11 lbf3 ..te6 12 'iie2 lbc7! 13 iH4 ..txc4 14 'iixc4 lbe6 IS ..teS 'iidS, Sarbok-Carlsen, Gausdal 2003.
This possibility wouldn't arise in a normal Exchange (4 c4 lbb6 S exd6 exd6 6 lbf3) if Black is careful, e.g. 6.....te7 7 lbc3 0-0 The point of the 4... g6 S ..tc4 c6 order is that Black avoids some of 8 ..te2 and 8 ...lbc6 9 .te3 .tg4!. the options in Kengis's order, such as (4lbf3 dxeS S lbxeS g6) 6 c4 and CENTER COUNTER DEFENSE 6 'iif3. If Black likes the Kengis Once upon a time a player who line, 4... g6 S ..tc4 c6 may be the answered 1 e4 with 1...d5 2 exd5 most precise way to reach it. 'iixd5 just wanted to get out of Finally, there is 4 lbf3 lbc6. The book and was hoping White didn't reputation of this line rests on the know that 3 lbc3 'iia5 4 d4 lbf6 latest analysis of a gambit, S c4 5 lbf3 .tg4 6 h3! favors him lbb6 6 e6!? fxe6. If White isn't significantly (6 ... ..thS 7 g4! .tg6 8 confident about what to do then, he lbeS). should bail out with 6lbc3. Today Black usually plays 1 e4 dS with ... ..tfS to obtain a CaroKann pawn structure with more active play. But he watches for a chance to transpose into a good version of ... ..tg4. For example 4 lbf3 lbf6 5 ..tc4 c6.
Then 6.....tg4 (or 6... dxeS 7 dS) 7 exd6 exd6 transposes to an Exchange Variation in which Black committed his queenside pieces too early. Now 6 0-0 ..tg4! has the benefits of a superior Caro-Kann (7 d4 e6 8 .:tel lbbd7).
For example, 8 ..te2 ..te7 9 d5! and 9... ..txf3 10 ..txf3 lbeS 11 ..te2 lbexc4? 12 ..txc4lbxc4 13 'iia4+ or 97
Semi-Open Games
But Black is the one most likely to be outfoxed in the Center Counter even when he thinks he is the fox. After 4 tiJf3 he might be tempted to seize the center with 4 ... e5. But then 5 d4! traps him in a discredited old line that used to arise after 4 d4 e5?! 5 tiJf3!.
The same idea succeeds, however, after 9 0-0-0 tiJd5 in view of 10 ~xd5 cxd5 11 'iib5+ tiJd7! (12 'iVxb7 l:.b8 13 'iVc6 ~a3). White also derives some extra options from 9 tiJe5 !, such as 9 ..•tiJbd7 10 tiJxd7 tiJxd7 11 g4 ~g6 12 h4, which has been dangerous the few times it's occurred.
The major finessing of the Center Counter begins after 4 d4 tiJf6 5 tiJf3 (or 4 tiJf3 tiJf6 5 d4). The waiting move 5 ... c6 anticipates White most dangerous strategic idea, d4-d5, and avoids queen traps (6 tiJe5 tiJbd7 7 tiJc4 'iWc7!).
The most dangerous situation for Black is when both ~c4 and 'iie2 are played early on so that the strategic d4-d5! break is in the air. For example, 1 e4 d5 2 exd5 'iVxd5 3 tiJc3 'iVa5 4 d4 tiJf6 5 ~c4 c6 6 ~d2 ~f5 and now 7 'iVe2! allows White's bishops to take command after 7 ... e6 8 d5! cxd5 9 tiJxd5 'iVd8 10 tiJxf6+.
The more common reply to 5... c6 is 6 ~c4. This reaches a tabia after 6... ~f5 7 iLd2 e6 8 'ife2 ~b4 when White chooses between 9 tiJe5 and castling. The two often transpose 9 tiJe5 tiJbd7 10 0-0-0 and 9 0-0-0 tiJbd7 10 tiJe5.
The best try to exploit White's order is 5 ..• ~g4!?, rather than 5 ... c6.
But the more accurate knight move cuts down Black's options. The simplifying 9 tiJe5! tiJd5? is dreadful (10 iLxd5! cxd5 11 'iWb5+! 'iVxb5 12 tiJxb5 iLxd2+ 16 <;i;xd2 tiJa6 17 tiJd6+ and wins).
Then 6 tiJf3? tiJc6! and 7•.. 0-0-0 gives Black the most aggressive setup he can get in the Center Counter, e.g. 7 ~b5 0-0-0 8 ~xc6 bxc6 9 0-0 e5 10 'ifd3 exd4 11 tiJe2
98
Semi-Open Games
.txf3 12 'iixf3 'iid5, DavidKosteniuk, Dresden 1999.
A more subtle order is 3 d4 tbxd5 4 .te2!?
And on 6 f3 .tf5 7 tbge2 e6 Black is secure, e.g. 8 .td2? 'iib6 9 g4 .tg6 10 h4 h6 11 .tb3 tbc6! 12 .te3 0-0-0 13 'iid2 .tb4 14 0-0-0 tbd5 15 .tf2 tba5 with advantage in Movsesian-A.Kogan, Nova Gorica 2000. CENTER COUNTER: 2...tbf6 If you're going to seek a big edge against 1 e4 d5 2 exd5 tbf6 you probably have to master the complex 3 .tb5+ lines. But if you just want a reasonable middle game you can save midnight oil with 3 c4. Black's best is 3... c6 4 d4 cxd5, transposing into the Panov Variation of the Caro-Kann.
This rules out 4 ... .tg4, and transposes into a main Whitefriendly line after 4 ... g6 5 tbf3. Black might prefer 4....tf5 on the grounds that he tricked White into one of the quietest 3 .tb5+ lines (3 ... .td7 4 .te2!? tbxd5 5 d4 .tf5 ).
But even if he's been tricked, And if you're looking for White still enjoys good prospects something in between, there is 3 d4 after 4...i.f5 5 tbf3 e6 6 0-0 i.e7 and 3 tbf3. After 3... tbxd5 Black 7 c4. e.g. 7...tbb4 9 tba3 tb8c6 will likely choose between two 10 .te3 0-0 11 ~d2 ~d7 12 .l:.fc1 basic plans of development - ... g6/ :fd8 13 tbb5! tba6 14 a3 ~e8 ... .tg7 and ... .tg4/ ... e6. Against 15 :d1, Van der Weide-Buschke, either plan White will operate with Senden 2006 . .te2, 0-0 and c2-c4. MODERN DEFENSE The virtue of 3 tbf3 and then 3...tbxd5 4 d4 lies in avoiding the sharpest versions of the Portuguese Gambit (3 d4 .tg4!?). If Black insists on meeting 3 tbf3 with 3....tg4, White seems to get the upper hand with 4 .tb5+! in view of 4... c6 5 dxc6 tbxc6 6 h3 .th5 7 .txc6+ bxc6 8 d3 or 4... tbbd7 5 h3 .th5 6 tbc3.
Black often uses 1.. .g6 and 2....tg7 as a form of reconnaissance. He waits for White to show his cards before he decides whether to transpose into a Pirc Defense with ... tbf6. This makes sense when, for example, White has given up his 99
Semi-Open Games
most dangerous anti-Pirc weapons 2 d4 ttJf6 3 ttJc3 g64 g3) but can be such as a quick .te3-h6xg7. It's interrupted in the Modem, 1 e4 g6 more a matter of taste after 1 e4 g6 2 d4 .tg7 3 ttJc3 d6 4 g3 ttJc6!. 2 d4 .tg7 3 ttJc3 d6 4 f4, when Then 5 ttJge2? .tg4!, 5 d5 ttJd4 Black can choose between and 5 .te3 e5, are not the kinds of transposing into the Austrian Attack position White wants when he plays of the Pirc (4 ... ttJf6) or staying in g2-g3. the Modem with 4 ... c6, 4 ... ttJc6 or 4 ... a6. But the Pirc option soon expires. After 4 f4 c6 5 ttJf3 ttJf6?:
This is a bad Pirc because ... c6 doesn't fit in with 6 .td3 0-0 7 0-0 b5 and then 8 e5! dxe5 9 fxe5 ttJd5 10 ttJxd5 'iVxd5 11 a4.
For example, 5 d5 ttJd4 6 ttJbl!?, to trap the knight, leads to doubleedged play like 6... c6 7 c3 ttJb5 8 .tg2 cxd5 9 exd5 ttJc7 10 ttJe2 ttJf6 11 0-00-0 12 c4 b5!, RaetskySakaev, St. Petersburg 1999. A more precise White will prefer
Black was clearly worse after 4 ttJge2!. Then 4•••ttJf6 5 g3 gets 11...c5 12 'iVe2! .tb7 13 c4 bxc4 him to the desired Pirc position. If 14 .txc4 'ii'd8 15 e6, as in Ghinda- Black insists on 4 ... ttJc6, White Bogdan, Rumanian Championship gains space with 5 d5! ttJe5 6 f4 and 1997, a case of Modem player .te3/ttJd4 gives him a much better finding himself in an unfamiliar position than he normally gets in a g2-g3 variation. Pirc. White has to make a key decision at move four in the Modem. Suppose he wants to devleop his KB at g2. The fianchetto is carried out smoothly in the Pirc (1 e4 d6
The .tc4 lines tell a similar story. In the Pirc (1 e4 d6 2 d4 ttJf6 3 ttJc3 g6) White has had some success with 4 .tc4 .tg7 5 'iVe2. But in the Modem he faces 4 .tc4 ttJc6!?
100
Semi-Open Games
Then 5 d5 is inconsistent and on other moves Black can shift into a pleasant Pirc, such as 5 ttJf3 ttJf6 6 0-0 i.g4 or 5 i.e3 ttJf6 6 f3 0-0 7 ttJge2 e5. The moral is that White needs ttJc3 in the Pirc order to defend e4 but he should delay it in the Modern if he wants the option of meeting ... ttJc6 and/or ... i.g4 with c2-c3!. The favored order is 3 ttJf3 and then 3... d6 4 i.c4 ttJf6 5 'iie2.
Aside from Modem-into-Pirc, Black can also try Modern-intoDragon by inserting ... c5. For instance, I e4 g6 2 d4 i.g7 3 ttJc3 and 3 ... c5 seeks an Accelerated Dragon (4 ttJf3 cxd4 5 ttJxd4 ttJc6) without having to deal with the Maroczy Bind. In practice he succeeds about one third of the time because I e4 players generally avoid the Schmid Benoni, 4 d5, and they have doubts about 4 dxc5!?
It also pays to play 3 ttJf3 d6
4 i.e2! before ttJc3 when White seeks the Classical setup, since 4••.ttJf6 5 ttJc3 transposes to a heavily trafficked Pirc.
This avoids some offbeat lines such as 4 ttJc3 i.g4 or 4 ... a6. In contrast, 4 i.e2 i.g4 5 h3 i.xf3 6 i.xf3 ttJc6 7 c3 e5 8 dxe5 favors White's bishops. By delaying ttJc3 White also discourages ... a6/ ...b5 because there is no knight for ... M to attack.
MODERN: 3... c6 When Black plays ... c6 before ... d6 he creates problems for a White who meets 1 e4 g6 2 d4 JLg7 3 ttJc3 d6 with a move of the QB or a KB fianchetto. For example 4 i.g5 is reasonable after 3... d6. But after 3 ... c6 it's a dubious gambit, 4 i.g5 'ifb6!. And 3•.. c6 4 g3 invites 4••. d5!, when White's bishop bites on proverbial granite, e.g. 5 exd5 cxd5 6 JLg2 ttJf6 7 ttJge2 0-0 8 0-0 ttJc6 9 h3 b6 10 i.g5 e6 and ... iLa6, in Ambrus-Eliseyev, St. Petersburg 2002. But ... d5 and ... 'ifb6 aren't the only ideas behind 3... c6. White has to be wary of being trapped in an unfamiliar Pirc if Black follows with ... d6 and ... ttJf6, e.g. 4 ttJf3:
101
Semi-Open Games
Note that White can discourage the Gurgenidze after 1 e4 g6 2 d4 .tg7 by means of 3 f4.
Black should transpose with 4 •.. d6! S .te2 ttJf6 into a main Pirc line. This usually arises from 1 e4 d6 2 d4 ttJf6 3 ttJc3 g6 4 ttJf3 .tg7 5 .te2 0-0 6 0-0 c6 and is considered fairly even. Thanks to the 3 ... c6 order Black has dodged the sharper Pirc lines, based on f2-f4 and/or .tg5. The sternest test of 3 ... c6 is 4 f4. Then 4 ... d6 5 ttJf3 ttJf6? is a bad Pirc, as noted above, and 5 ... .tg4 is a somewhat dubious Modern. Black often prefers 4 ... d5, seeking a Gurgenidze Variation. There are several routes to a Gurgendize tabia, including 1 e4 g6 2 d4 .tg7 3 ttJc3 c6 4 f4 ~6!? S ttJf3 dS and then 6 eSt, since 6 exd5 .tg4! would grant Black plenty of compensation for a pawn. Which route to the tabia is best? The only thing that seems certain is Black should avoid 1 e4 c6 2 d4 d5 3 ttJc3 g6 if he really wants a Gurgenidze. White has too many excellent alternatives to 4 e5 and 5 f4, such as 3 cxd5 and 3 e5.
And then 3 ••• c6 4 ttJf3 dS S eS. Unlike the normal Gurgenidze, when his knight is misplaced at c3, here he can support his center with c2-c3. And on 3 ... d6 4 ttJf3 White has a promising Pirc. 'I felt like a broken man after 3 f4,' Julian Hodgson once said when it was played against him. Nevertheless Black has a good response in 3 ... cS!, which is good here because 4 c3? does a poor job of supporting the center (4 ... cxd4 5 cxd4 'iib6!).
PIRC DEFENSE The standard Pirc order is 1 e4 d6 2 d4 ttJf6. White's e-pawn is attacked and he is more or less limited to 3 .td3, 3 ttJd2, 3 f3 and the main line, 3 ttJc3. The first two moves are attempts to reach Yefim Geller's quiet system in the Modern Defense (1 e4
102
Semi-Open Games
g6 2 d4 i.g7 3 c3 d6 4 lbf3 and i.d3/lbbd2). If he succeeds, it's a small victory since Pirc players are often uncomfortable in Modem lines. On 3 i.d3 Black can avoid the Geller system with 3... eS. Then 4 c3 dS!? is the active way of disturbing White's plan. The more solid way is 4...lbc6 Slbf3:
(12 dxe5 lbxe5 13 lbxe5 'i¥xe5 14 lbf3 ~5 15 e5 i.f5) in Rublevsky- Azmai parashvili, internet 2004. With 3 f3 instead, White is signalling he's willing to play a Samisch King's Indian. An example of what can happen to a strong GM who likes the Pirc as Black but not the KID was A.Rodriguez-Kuzmin, Minsk 1982: 3... g6 4 c4! i.g7 5 lbc3 lbc6 6lbge2 0-0 7 i.e3 a6 8 lbc 1 e5 9 d5 lbd4 10 lbb3 c5?! 11 dxc6 bxc6 12 lbxd4 exd4 13 .i.xd4 l:i.b8 14 'i¥d2 'iia5? 15 l:i.cl .l::i.d8 16 i.e2 i.e6 17 0-0 'iib4 18 b3 and White won. Black can try to avoid that fate with 3 ... eS. Then he's offering White a tiny-edge endgame, 4 dxe5 dxe5 5 'iixd8+. But 4 lbe2! followed by 5 c4 and 6 lbbc3 heads to the Samisch.
Who is more uncomfortable here? A Black who likes the sharp tactics of a Pirc may not like defending a Pseudo-Lopez after If White plays the conventional S...i.e7 6 h3 0-0 7 lbbd2. But 3 lbc3 after 1 e4 d6 2 d4lbf6 Black White can find himself, after has an alternative to 3... g6 in 3... c6, S...i.g4 6 dS lbe7 7 c4, in an Old a system worked out by the Omsk Indian - a favorable one but one master Anatoly Ufimtsev. that may not suit a 1 e4 player. Black can also tum 3 i.d3 or 3 lbd2 into a Philidor by means of ... lbbd7, as in 3 lbd2 eS 4 c3 i.e7 S lbgf3 lbbd7. White's setup is so passive that his edge usually disappears by move 12. For example, 6 i.e2 0-0 7 0-0 c6 8 l:i.el 'ikc7 9 i.f1 l:i.e8 10 'ikc2 i.f8 11 b4 d5! and Black was soon better 103
Semi-Open Games
Black waits: After 4 tiJt3 or 4 g3, After 6 0-0 cxd4 7 tiJxd4 0-0, he can bring about a tame version of he's transposed into a Classical Dragon. On 6 d5 0-0 it's a Schmid the Pirc with 4 ... g6. Benoni but a Schmid in which But if 4 ..tg5, then 4... tiJbd7 and White was denied the useful S... eS/... ..te7! is a good Philidor ..tbS+!. because White's QB is misplaced. The tactical justification is And on S f4 'iVaS 6 'iVd2 bS! 6 dxc5 'ifa5! with good play. This Black obtains faster counterplay explains why S... cS is better than than in a normal Pirc because he S... O-O 6 0-0 cS 7 dxcS dxcS, which hasn't spent tempi on ... g6/.....tg7, is regarded as favoring White after e.g. 7 ..td3 tiJb6 8 tiJf3 b4 9 tiJdl 8 'ifxd8 and 9 ..te3. dS! 10 eS tiJe4 as in WellsThe most dangerous Pirc line is Hodgson, Edinburgh 1989. 4 f4 and the usual continuation is 4.....tg7 5 tiJt3 0-0 6 ..td3. Black The test of Ufimtsev's order is may try to avoid this with S... cS 4 f4!. Then neither the Pirc (4 ... g6?! 6 dxcS 'ifas and 7... 'ifxcs since S tiJf3) nor Philidor (4 ... tiJbd7?! 7 exd6? allows a strong 7... tiJxe4. S tiJf3) are active enough. Ufimtsev But White can juggle the move preferred 4...'iYa5!, which enjoyed order, 4 f4 ..tg7 5 ..td3!? some vogue in the 1980s but remains slightly suspect. The starting point for most Pirc Defenses is 3 tiJc3 g6. Black looks for a chance to attack the center with ... cS. For example, 1 e4 d6 2 d4 tiJf6 3 tiJc3 tiJf6 4 tiJt3 ..tg7 5 ..te2 c5.
This avoids 5... c5 6 dxcS 'iYaS because e4 is already protected (7 cxd6!). Also bad is 6... dxcS 7 eS (7 ... tiJdS? 8 ..tbS+). If Black plays 5•.. 0-0 instead, White transposes into the main line with 6 tiJf3!. 104
Semi-Open Games
The downside may lie in the extra options Black enjoys after 5 .td3, such as S...lbc6 6 dS lbb4 or 6lbf3 .tg4. But there's been too little experience to tell.
CARO-KANN DEFENSE
The Caro-Kann contains transpo traps even after the super-quiet 1 e4 c6 2 d3. One is set by 2... dS 3 lbd2 lbd7 4 lbgf3.
Another refinement, this time by Black, occurs with 4 f4 .tg7 Slbf3 lba6!? This seems like an offer to transpose, after 6 .td3 0-0, into a variation that runs 4 f4 .tg7 5 lbf3 0-0 6 .td3 lba6 and is rated as favoring White after 7 0-0 c5 8 d5 and f4-f5. But Black has the extra option of 6 .td3 cst?~ We learned from a famous TalSmyslov game that 4... eS? S d4! favors White. We can also see that 4...lbgf6 S eS! is a better-than-usual French Defense (5 ... lbg4 6 d4 e6 7lbb3). If Black prepares ... lbgf6 with 4••:ilic7, White plays S exdS! cxdS 6 d4 and has a favorable form of the The idea is 7 dS .tg4 8 0-0 lbd7 or 7 .. :iic7 and ... c4 when his counterplay runs faster than usual. For example, 7 d5 'ikc7 8 'ike2 0-0 9 a3 e6! 10 dxe6 .txe6 was Bareev-Christiansen, Biel 1991, which went 11 .tc4 l:!.ae8 12 0-0 .txc4 13 'ikxc6 'ikc6! 14 e5 lbg4 with good play.
old Exchange Caro-Kann. For example, 6 ... e6 7 .td3 lbe7 8 0-0 g6 9 :e 1 iL.g7 10 lbfl lbc6 11 c3 0-0 12 iL.g5 and TiviakovDreev, Gothenburg 2005 continued 12 ... e5 13lbe3! lbb6 14 dxe5lbxe5 15 iL.f4 lbxf3+ 16 'iVxf3 'ikc6 17 lbc2 and lbd4 with clear superiority.
The flies in the 5... lba6 ointment The safest path to equality after are 6 e5 and 6 .tc4 0-0 7 e5. But 2 d3 is 2... dS 3 lbd2 eS 4 lbgf3 once again there's been virtually no .td6! (not 4 ...lbd7? 5 d4! which transposes to Tal-Smyslov). experience to judge them. 105
Semi-Open Games
Then 5 d4 resembles a Tarrasch French after 5•.. exd4! 6 exd5 cxd5.
But Black is a tempo (....td6) ahead of the Tarrasch version, 1 e4 e6 2 d4 d5 3 liJd2 c5 4 exd5 exd5 5 liJgf3 cxd4, and he should equalize.
Then 4 ttJgf3 .tg4 eases Black's game appreciably. White should pass with 4 c3. Then 4... dxe4 5 liJxe4 .tf5 6 liJg3 .tg6 reaches a Caro-Kann main line but with the addition of .. :iib6 and c2-c3. That makes it harder for White to develop his QB without hanging the b2-pawn. However, 7 h4 h6 8 liJh3 may make c2-c3 more useful than .. :iib6, e.g. 8... liJd7 10 liJf4 .th7 11 .tc4 e6 12 'ilie2, Lane-Pedersen, Trnava 1985.
You will often see GMs play 1 e4 c6 2 d4 d5 3 liJd2 and transpose into main lines after 3 ... dxe4 4 liJxe4. This order sidesteps the obscure 3 liJc3 b5, which can lead to a version of the ... exf6 Caro, after CARO: TWO KNIGHTS 4 a3 dxe4 5 liJxe4 liJf6 6 liJxf6+ exf6, that may be better for Black The 1 e4 c6 2 liJc3 d5 3 liJf3 than usual because 7 .tc4 is not order transposes into main (3 d4) possible. lines, after 3... dxe4 4 ttJxe4 ttJd7 5 d4 or 4... liJf6 5 liJxf6+ and 6 d4. But there's a more substantial However, there are two traps for the reason for 3 liJd2. White can meet unwary Black. 3••. g6 with 4 c3, as Anatoly Karpov One is notorious, 3••. dxe4 showed in a 1990 game with Deep 4 liJxe4 .tf5 5 liJg3 .tg6?, when Thought that went 4....tg7 5 e5! f6 White uses his extra knight move 6 f4 with a plus for White. with 6 h4 h6 7 ttJe5! .th7 8 'ilih5 g6 The drawback to 3 liJd2 is that the d-pawn is unprotected after 3••• 'iVb6.
9.tc4. The other trap, with 4...liJd7 5 .tc4 liJgf6 6 ttJeg5 e6 7 'iie2,
106
Semi-Open Games
wasn't known until 1954, largely because Aron Nimzovich's ... tiJd7 wasn't played much before then.
If d2-d4 has been inserted in place of tiJf3, Black obtains a solid game from 7...tiJb6. But here that would drop pawn to 8 tiJeS! and Black has to accept the slight inferiority of 7... tiJd5. The main Two Knights line is 1 e4 c6 2 tiJc3 dS 3 tiJf3 .i.g4 and then 4 h3 .i.xf3 S 'ilixf3. In the 1950s Black played 5... e6 until he found that 6 d4 dxe4 7 tiJxe4! was a sound gambit, e.g. 7... 'ilixd4 8.i.d3 tiJf6 9 c3 'ilid8 10 0-0 and .l:tdl. So he adopted S... tiJf6 because 6 d4 dxe4 7 tiJxe4 'iVxd4! is unsound.
Books used to dismiss this by citing a 1951 game that went 7 .i.d3 .i.e7 8 e5 tiJfd7 9 'ilig3 with an effortless attack. But Black improves with 7..• dxe4! 8 tiJxe4 tiJxe4 9 'ilixe4 tiJd7. For example, 10 c3 tiJf6 11 'iie2 'iVd5 equalizes after 12 0-0 .i.d6 13 ':'e 1 0-0 14 .i.c4 'iiia5 15 .i.d2 .l:tfe8 16 .l:tadl ':'ad8 17 .i.c1 'iih5!, Krai-Kamsky, U.S. Championship 2004. If White wants more after S... e6 6 d4 tiJf6 he may have to try the French-like 7 eS or 7 .i.gS but with his QB traded it should be a good French for Black.
But Karpov has exclusively played S... e6. Its benefits include keeping f6 free, e.g. 6 d3 tiJd7 7 i..d2 .i.c5 8 'ilig3 'ilif6!, Chemyshov-Lastin, Voronezh 2006.
Meanwhile, 4....i.hS (after 1 e4 c6 2 tiJc3 dS 3 tiJf3 .i.g4 4 h3) has been deeply analyzed, beginning with S g4 .i.g6 6 exdS cxdS 7 .i.bS+. When an opponent plays 4 ... .i.hS he is usually boasting that his memory is better than yours.
The main point of S... e6 is revealed by 6 d4 tiJf6!:
The mental hygiene alternative is S d4 e6 6 .i.d3.
107
Semi-Open Games
Then 5 tbf3 ~g4 6 tbc3 tbf6 transposes to a main Panov line. But the drawback is 5 cxd5! 'iixd5 6 tbfJ.
White's point is that 6••. dxe4 7 tbxe4 ~xf3 8 'iixf3 'iixd4 9 0-0 is a superior version of the 4 ... ~xf3 5 'iixf3 e6 6 d4 gambit. Black should probably decline with 6...tbf6. Then 7 e5 tbfd7 8 g4 ~g6 9 i.xg6 and 10 tbe2 offers mixed chances (Short-Khalifman, Merida 2001). CARO:PANOV
Then we've landed in an Alapin Sicilian (1 e4 c5 2 c3 d5 3 exd5 'iixd5 4 d4 tbc6 5 tbf3 cxd4?! 6 cxd4) in which the early exchange of c-pawns allows an annoying tbc3, e.g. 6... e5 7 tbc3 ~b4 8 ~d2 ~xc3 9 ~xc3 with an edge. More common is 4.••tbf6 5 tbc3. A key line begins 5••• tbc6 6 ~g5 e6. Now 7 c5 ~e7 8 ~b5 0-0 9 ~xc6! bxc6 10 tbf3 threatens 'iia4/tbf3-e5 and gives White an ideal version ofthe Panov bind. For example, 1O ... tbd7 11 ~xe7 'iixe7 12 0-0 e5 13 tbxe5 tbxe5 14 dxe5 'iixe5 15 l:tel 'iif6 16 'i'd2 l:tb8 17 b4 (18 ... l:txb4? 19 tbxd5), RathBrinck-Claussen, Vejle 1974.
Vasily Panov's 1 e4 c6 2 d4 d5 3 cxd5 exd5 4 c4 becomes a Since the Panov bind began Queen's Gambit Accepted after threatening the health of the Caro... dxc4. But it's a slightly inferior Kann more than 70 years ago, QGA and Black should delay Black has found ways to dodge it, ... dxc4 even at the risk of White such as 6... 'i'a5, 6... ~e6 and 6... e6 creating a favorable pawn wedge 7 c5 ~e7 8 ~b5 ~d7!. That's why with c4-c5. And on 4.•.tbc6?!: White may prefer 6 tbfJ to 6 ~g5.
108
Semi-Open Games
Then 6... e6? 7 cS! is good for him, e.g. 7... iJ..e7 8 iJ..b5 0-0 9 0-0 and 9 ... tLle4 10 iJ..xc6 tLlxc3 11 bxc3 bxc6 12 'iVa4. But 6 ... iJ..g4 renders easy equality following 7 iJ..e2? e6 8 c5 tLle4. Instead White's best answer to 6...iJ..g4 is 7 cxdS tLlxdS 8 ~3. This leads to a somewhat superior endgame that became famous when Bobby Fischer beat Max Euwe with it (8 ... iJ..xf3 9 gxf3 e6 10 'iVxb7 tLlxd4 11 iJ..b5+ tLlxb5 12 'iVc6+! <J;;e7 13 'iVxb5 tLlxc3 14 bxc3 'iVd5 or 14 ... 'iVd7). This is an important line because White can reach it in other orders and because Black's inferiority is so slight he often cooperates. Another Black strategy is quick castling after 4 c4 tLlf6 S tLlc3 and S... e6. Now on 6 iJ..gS:
White can tranpose, 6 ... tLlc6 7 cS, to that 4 c4 tLlf6 5 tLlc3 tLlc6 position considered earlier. Better is 6... iJ..e7! and then 7 tLln 0-0. On 8 iJ..d3 it's time for
8 ... dxc4! 9 iJ..xc4 a6 because this time it's a good QGA, e.g. 10 0-0 b5 11 iJ..b3 iJ..b7 12 l:.el tLlc6 13 a3 l1c8 or 11 iJ..d3 iJ..b7 12 'iVe2 tLlc6 13l1adl tLlb4 14 iJ..bl tLlbd5. In place of 6 iJ..g5 White can try 6 cS iJ..e7 7 iJ..bS+. But an alert Black will see through 7 ... tLlc6? which heads toward the bind after 8 tLlf3 - and play 7 ... iJ..d7! 8 iJ..xd7+ 'iVxd7 9 tLlf3 tLlc6. Instead of 6 iJ..g5 or 6 c5, White should consider an exchange of pawns on d5. But the immediate 6 cxdS tLlxdS and 7 iJ..d3 iJ..b4! has performed well for Black. Instead 6 tLln makes sense since 6 ... tLlc6 is once again inexact (7 c5!) and 6 ... iJ..b4 7 iJ..d3 dxc4 8 iJ..xc4 gives White a fairly good QGA, actually a Nimzo-Indian. But the main point is that against 6...iJ..e7:
White can play 7 cxdS! tLlxdS without having to defend c3 compared with 6 cxd5 tLlxd5 7 iJ..d3 iJ..b4. He obtains a slight edge with 8 iJ..d3 or 8 iJ..c4.
109
Semi-Open Games
If Black retakes with the pawn (7 cxd5 exd5) he has the worst of a nearly symmetrical position following 8 i.b5+! i.d7 (8 ... ~c6 9 ~e5) 9 'iib3 ~c6 10 0-0 or 9 •..i.xb5 10 'iWxb5+ 'iWd7 11 ~e5. When White does not play 4 c4 after 1 e4 c6 2 d4 d5 3 exd5 cxd5 Black's difficulties chiefly lie in his problem piece, the QB. White usually plays 4 i.d3, which stops ... i.f5 and makes ... i.g4 problematic. He stands well after 4 ... ~c6 5 c3.
CARO:PANOV ACCELERATED With 2 c4 White tries to bring about a Panov in which Black can't maintain a pawn on d5, nor can he reach the relative safety of the Fischer-Euwe endgame. Instead, Black often accepts a slightly worse isolated d-pawn middlegame, such as after 2•.. d5 3 exd5 cxd5 4 cxd5! ~f6 5 ~c3 ~xd5:
Black's eyes may light up when he sees 4 c3:
Now 6 i.c4 ~b6 7 i.b3 ~c6 8 ~f3 and 9 d4, or 6 ~f3 ~c6 7 i.b5 e6 8 0-0 i.e 7 9 d4 are promising for White. Thanks to his move order he avoided 6 d4 ~c6 7 ~f3 i.g4!, the This seems careless in view of endgame variation. 4•..i.f5. But the trap is sprung by White also gets the extra options 5 'iWb3!, attacking b7 and d5. of annoying checks, 5 i.b5+ and Following 5 ... b6? 6 i.b5+ or 5 ~a4+, which theory regards as 5 ... ~d7 6 ~f3 ~c6 7 i.b5! and difficult for Black. ~e5 Black is clearly worse. Black can sidestep these If Black is alert he'll realize that problems by meeting 2 c4 d5 after 4...~c6 White has nothing 3 exd5 with 3...~f6!' better than 5 i.d3, transposing into the main line.
Then 4 dxc6 ~xc6 is a sound gambit, better known in the Center
110
Semi-Open Games
Counter (1 e4 d5 2 exd5 liJf6 3 c4 c6 4 dxc6 liJxc6 and then 5 d3 e5 6 liJc3 i.f5 7 liJf3 i.b4 8 i.e2 e4!). If White stops 4 ... cxd5 with 4 'iVa4, Black can offer a different pawn, 4•.. e6! S dxe6 i.cS! 6 exti+ 'ii;xti or 6 liJf3 liJg4 7 d4 i.xd4 8 liJxd4 'iVxd4 9 'iVc2 liJa6 as in Huebner-Luther, Sauerbrucken 2002.
matter because White was assumed to be better in either. 'Practice shows that White's chances are somewhat better and now this defense is rare,' Alexey Suetin said of the Rubinstein in his 1983 book on the French. Svetozar Gligoric wrote that 3 liJc3 dxe4 4 liJxe4 liJd7 is best answered by S liJf3 liJgf6:
More exact is 3 cxdS!.
Then '6 i.g5!' reaches a good Burn, he said, since 6 ... c5? walks into 7 dxc5 liJxc5? 8 'iVxd8+ and 9 liJxf6.
Then 3...liJf6 4 dxc6 liJxc6 is less sound (5 liJc3 e5 6 liJf3 i.c5 7 i.b5). Black can avoid all of this by meeting 2 c4 with 2... eS. He is betting that a 1 e4 player will be temperamentally unsuited for an Old Indian (3 liJf3 d6 4 d4 liJbd7). FRENCH DEFENSE The French (1 e4 e6) variations that most often transpose into one another are the Rubinstein (2 d4 dS 3 liJc3 dxe4 or 3 liJd2 dxe4) and the Burn (3 liJc3 liJf6 4 i.gS dxe4). That used to be a 'who cares?'
But today many GMs distrust 6 i.g5 because 6... h6 and 6... i.e7 7 liJxf6+ liJxf6 8 i.d3 c5 seem to equalize. So White began to avoid i.g5 in a Rubinstein - and Black discovered that the Burn was his way to trick White into a i.g5 Rubinstein. That is 3 liJc3 liJf6 4 i.gS dxe4 S liJxe4 liJbd7 so that 6 liJf3 transposes to the last diagram. No better is 6 liJxf6+ liJxf6 7 liJf3 h6! (8 i.d2 c5! or 8 i.h4 i.e7 9 i.d3 c5 10 'iVe2 'iVa5+!).
111
Semi-Open Games
To avoid the Burn, White can play entirely different lines, such as 3 lbd2 or 3 lbc3 lbf6 4 e5. But Black found that 3.....te7 makes ... dxe4! stronger. For instance, 3lbd2 (or 3lbc3) ..te7 4 ..td3 dxe4 S lbxe4 lbf6:
Slbce2!? and then S... cS 6 c3lbc6 7 f4 'iYb6 8 lbo.
This is the same posItIon as 3 lbd2 lbf6 4 e5 lbfd7 5 c3 c5 6 f4 lbc6 7 lbdf3 ~6 once White plays lbe2.
Black can retake with a bishop on f6 and obtain an excellent Rubinstein, e.g. 6 lbxf6+ ..txf6 7lbo lbc6! 8 c3 eS. Or 6lbolbbd7 and 7 'iWe2lbxe4 8 ..txe4 cS! and then 9 0-0 cxd4 10 l:.dl lbc5 11 lbxd4 lbxe4 12 'iWxe4 0-0 (Becerra-Kaminski, internet 2005). Black managed to reach a position that could come about after 3 lbc3 dxe4 4 lbxe4 lbd7 5 lbf3 lbgf6 6 ..td3 ..te7 but only if White rejects 6lbxf6+!. FRENCH: STEINITZ
In this way White dodges other Tarrasch lines. When Vishy Anand wanted to reach the diagram in the 2000 FIDE championship finals he knew that if he chose 3 lbd2, his opponent, Alexei Shirov, wouldn't cooperate by playing 3... lbf6 because he preferred other moves. So Anand chose 3 lbc3!, reached the Tarrasch position he wanted and won. Quite a different situation arises after 3 lbc3 lbf6 4 eSlbfd7 slbo. White is ready to liquidate his center to obtain tactical chances, such as S••• cS 6 dxcS lbc6 7 ..tf4 ..txcS 8 ..td3 and 8... 0-07 9..txh7+.
A more open French results, e.g. The Steinitz Variation has become a favorite way of 8 ... f6 9 exf6 lbxf6 10 'iWe2 0-0 transposing into a sharp line of the 11 0-0-0 a6 12lbe5lbb4 13 ..tg5 b5 Tarrasch, 3 lbc3 lbf6 4 eS lbfd7 14 a3 lbxd3+ 15 lIxd3 'iWc7 16 f4 112
Semi-Open Games
as in Berg-Akesson, Gothenburg 2006.
7 d4, or try for more with 6 dxc3!?, e.g. 6... c5 7 h4 tbbc6 8 h5 h6 9 i.f4 'ikc7 10 'ii'd2 as in Rogers-Tondivar, Hoogoveen 2006.
If White likes this he can reach it while discouraging the Rubinstein and perhaps the Winawer as well. He begins with 1 e4 e6 2 tbf3!?
FRENCH: WINAWER As noted in the Introduction Black can sidestep many of the rare Winawer (3 tbc3 i.b4) lines by meeting 4 e5 with 4...tbe7 5 a3 i.xc3+ 6 bxc3 c5.
Francophiles regard that with contempt and will want more than 2•.. d5 3 tbc3 dxe4. And they'll see that Black is a bit worse after 3... d4 4 ttJe2 c5 5 c3. So they'll play 3 ...tbf6 and then 4 e5 ttJfd7 5 d4 gets White to where he wants to be. And on the natural 3... c5
The 'something to think about' quality of 4 ... tbe7 often prompts responses like 5 tbe2? c5 6 a3 i.xc3+ 7 tbxc3?, which favors Black after 7... cxd4 8 'ii'xd4 tbbc6 9 'ikg4 0-0 10 'fib5 d4!, ByrneUhlmann, Reykjavik 1968. And on 5 tbf3:
... then 4 exd5 exd5 5 d4! traps Black in one of Frank Marshall's lesser ideas (1 e4 e6 2 d4 d5 3 tbc3 c5 and 4 exd5 exd5 5 tbf3.) What should Black do after 2 ttJf3 d5 3 tbc3 if he doesn't like 3... ttJf6 ? His best bet may be 3.....tb4. It's unclear whether White should transpose to a Winawer, with 4 e5 ttJe7 5 a3 i.xc3 6 bxc3 and
Black can transpose (5 ... c5) into a normal Winawer. But also attractive is 5... b6, since White has missed his 'ikg4 opportunity, compared with 5 a3 i.xc3+ 6 bxc3 b6 7 'ikg4!. Chances should be equal after 5... b6 6 i.d2 i.xc3!? 7 i.xc3 i.a6
113
Semi-Open Games
8 ..txa6 ttJxa6 9 0-0 c5 (BokrosRustemov, Kircheim 2006). The counter-finesse to 4 ... ttJe7 is S ..td3. Then S... cS 6 a3 ..txc3+ 7 bxc3 seems to be a good version for Black of a main line Winawer in view of White's prematurely developed bishop (7 ... ttJbc6 8 ttJf3 c4 or just 7... c4 and ... ..td7-a4).
3 ttJc3 i.b4 4 e5 c5 5 a3 i.xc3+ 6 bxc3 ttJe7 7 dxc5) if White had an extra tempo.
But White improves with 6 dxcS! ttJbc6 7 ttJO. He's transposed into the 4 ... c5 5 dxc5 variation made famous by Fine-Botvinnik, AVRO But is dxc5 a good extra move? 1938. White's compensation is still That depends on positions like rated highly after 7•.. d4 8 a3 i.aS 9 b4 ttJxb4 10 axb4 ..txb4 in view 8 'i'g4 ttJd7 9 ttJO 0-0 10 i.d3. of 11 O-O!, rather than Fine's The thematic 9••.'i'c7 10 'i'xg7! 11 ..tb5+. :tg8 11 'i'xh7 ttJxeS favors White This is significant because today after 12 'i'h5 ttJxf3+ 13 'i'xf3 i.d7 White rarely gets there via Fine's 14 i.f4, Shamkovich-Gipslis, 4 e5 c5 5 dxc5 because theory urges Soviet Championship 1961. 5.. .'iic7!? or 5... ttJc6 6 ttJf3 d4!. In the main line Winawer, 4...cS S a3 i.xc3+ 6 bxc3 ttJe7, the A major virtue of 4 ... cS over 4 ... ttJe7 is that it gives Black an choices are 7 ttJf3, 7 a4 and 7 h4. extra option, the Armenian S a3 But Oleg Romanishin has made good use of7 ..td3!? ..taS. Then instead of the main line, 6 b4, the trickster may prefer This often transposes to the 6 dxcS!? This threatens 7 b4 and others but for the moment it retains transposes into a superior version of the option of 'i'g4. For example, Fine's line after 6... d4 7 b4 dxc3 7 ... 0-0 8 'i'g4 gets White to a 8 'iVg4!. popular 7 'i'g4 0-0 8 ..td3 position without having to deal with 7 'i'g4 Perhaps best is 6.....txc3+ 7 bxc3 'i'c7 and 7 ... cxd4. ttJe7. The result is a strange position that could come about from a For example, 7 i.d3 ttJbc6 normal Winawer (1 e4 e6 2 d4 d5 8 'i'g4 c4 9 ..te2 ttJf5 10 a4 'i'a5
114
Semi-Open Games
We've transposed into a version of Viktor Korchnoi's gambit (3 4Jd2 4Jf6 4 e5 4Jfd7 5 i.d3 c5 6 c3 4Jc6 7 4Jgf3 'iVb6 8 0-0) in which Black plays 8 ... i.e7 instead of grabbing the pawn.
11 i.d2 i.d7 12 'iWh5 h6 13 i.g4 as in Bagheri-Apicella, Chartres 2005.
The bishop move was once thought dubious but now it's a valid alternative, e.g. 9 dxeS 4JxeS 10 i.e2 4Jd7 11 l:te1 ~e7 12 ~e2 gS 13 ~a4 4JeS!, KabanovThe main drawback to 7 i.d3 is 7•.. b6. Why? Because often when Black tries to trade his bad bishop with ...b6/ ... i.a6, White's best way to cross him up is i.b5+ as in 7 4JO b6 8 i.b5+! i.d7 9 i.d3!. But after spending a tempo on 7 i.d3 there is little reason to play 7... b6 8 i.b5+ i.d7 9 i.d3. FRENCH: TARRASCH In the Tarrasch Variation (3 4Jd2) Black can use 3•••i.e7 as a waiting move. After 4 4Jgf3 4Jf6 S eS 4Jfd7 White has nothing better than 6 e3 eS 7 i.d3 4Je6 8 0-0 'iib6.
Kravstov, Tomsk 2003. In other words, 3 ...i.e7 can lure White into a gambit he never intended. If White likes the gambit he can try to reach it from other orders. For example, 3... c5 seems to rule it out. But 4 4Jgf3 and then 4 ... 4Jf6 5 e5 4Jfd7 6 c3 transposes. If Black wants to sidestep the gambit he has other good lines, including 4 ... 4Jc6 and 4 ... cxd4. There is also 3... 4Jc6 and 4 4Jgf3 4Jf65 e5 4Jd7 6 i.d3. Then 6... 4Jb4 is natural and after 7 i.e2 c5 8 c3 4Jc6 9 ~d3 'iVb6 10 0-0 we've reached the gambit again - although this time White may do better with 9 0-0 and 9 ... cxd4 10 cxd4 'iVb6 11 4Jb3. In the sharpest Tarrasch line, 3...4Jf6 4 eS 4Jfd7, White usually selects 5 f4 or 5 i.d3. If he prefers the latter, he may profit by delaying i.d3 with S e3.
115
Semi-Open Games
3 ... a6 4 lLIgf3 c5 S exdS exdS. Which is best for him? The drawback to the first order is that White can check at move five.
Both S c3 and S il.d3 reach a main line after S ... cS. However Black has an alternative strategy in S...b6 and 6 ... il.a6, to trade his bad bishop. That makes S c3 more accurate because if Black pursues the Btrading plan White will be a tempo ahead - S c3 b6 6 'iig4 (or 6 lLIh3) il.a6 7 il.xa6 - of S i..d3 b6 6 c3 il.a6 7 il.xa6. After 3•.•cS the major move order questions revolve around il.bS+ and the timing of exdS. Black has been doing well recently when he stops the check, e.g. 4 exdS exdS S lLIgf3 a6.
He can also reach this position via 4 lLIgf3 a6 S exdS exdS and
The drawback to 3 ... a6 is 4 eS. Some French players don't relish the prospect of shifting to an Advance Variation (4 ... cS S c3) in which they've played ... a6 in place of the usual Advance moves such as ... 'iib6 and ... lLIc6. In the third order, 3... cS 4 lLIgf3 a6, Black has to deal with S dxcS il.xcs 6 il.d3, which gives White a nice setup after 7 0-0 and 8 a3/9 h4. So it's up to Black to rate the worst-case scenarios in all three orders and pick the best of them. But how bad is that check? One school of thought says a trade of light-squared bishops is good for White and that 3 ... cS 4 exdS exdS S il.bS+ must be met by S... lLIc6. But Lev Psakhis insisted S... il.d7 is fine, arguing that a bishop trade helps Black. Max Euwe agreed and said S...lLIc6 6 lLIgf3 was so good for White that he should try to trick Black into that position with S lLIgf3 and ifS ... lLIc6, then 6 il.bS!. If Euwe is right about the pin, White may do better with another order, 3 •.• cS 4 lLIgf3 so that he can meet 4 •••lLIc6 with S exdS exdS 6 il.bS. The drawback is giving Black options such as 4 ... lLIf6 S exdS lLIxdS.
116
Semi-Open Games
But if White reaches his goal, the most common reply is 6... i.d6.
A trade of pawns is coming, either on c5 or d4. Two lines that often transpose are 7 dxe5 i.xe5 8 0-0 and 7 0-0 liJe7 8 dxe5. Mikhail Botvinnik called the second order inaccurate because Black can improve on 7 ... liJe7 with 7 ... cxd4!. Then 8 liJb3 9 liJbxd4 0-0 is equal.
liJe7
What has happened is we've transposed into 7 dxc5 i.xc5 8 0-0 liJe7 9 liJb3 i.d6 10 liJbd4 0-0. But in that order White has more options such as 10 .l:le 1, 10 i.g5 and 10 i.xc6+ bxc6 11 'ifd4. The 7 0-0 cxd4! order denies him the choice. FRENCH: ADVANCE VARIATION
The Advance Variation tabia used to be 3 e5 e5 4 e3 liJe6 5 liJf3 'ifb6 6 i.d3:
Following a Howard Staunton game in 1841, 6 ...i.d7 was considered best. But this was shown to be inexact in Nimzovich's celebrated victory over Georg Sa1we at Carlsbad 1911, when 7 dxe5! i.xe5 8 0-0 and b2-b4 Black's seriously hampered development. Black improves with 6 ... exd4! 7 exd4 i.d7. Then White has no convenient defense of d4 and the 8 0-0 liJxd4 gambit is regarded as not quite sound. These two basic positions, the Nirnzovich-Salwe one and the other with 6 ... cxd4, influence a lot of3 e5 theory. For example, you may see 4 .. .'iVb6, instead of 4 ... liJc6, and then 5 liJf3 i.d7. Black retains the option of ... i.b5, to get rid of his bad B. White has two strange but possibly good replies, 6 i.d3 i.b5 7 i.e2!? (7 ... liJc6 8 i.e3) and the devilish 7 dxe5 i.xe5 8 b4!?, based on trapping the bishop after 8 ... i.xf2+ 9 'ite2.
117
Semi-Open Games
7 .. :iixb2? with 8 ..te3! 'ifxal 9 'ifc2.
Black might avoid both replies by meeting 6 ..td3 with 6...tiJc6? But then 7 dxc5! puts us back in Nimzovich-Salwe.
Then 9... cxd4 10 cxd4? is bad because of 10... ..tb4+ and the queen can never be trapped.
He can try to repair matters with 6... cxd4 7 cxd4? tiJc6, reaching the Black-friendly position.
But 10 tiJxd4! gives White a big edge after 10 ... tiJxd4 11 ..txd4..ta3 12 ..tb5+ ~d8 13 0-0 or 1O ... il.a3 11 tiJb5!.
But thanks to his delay in ... tiJc6 White can do better with 7 tiJxd4!, e.g. 7... tiJe7 8 tiJd2 tiJbc6 9 tiJxc6 and ife2/tiJf3 with a slight edge.
Therefore the proper order is 6... cxd4 7 cxd4 tiJh6, reaching the last diagram. White should not play 8 ..txh6 because he is transposing into the poor 10 cxd4 line. Instead, he should go into one of the mediocre alternatives such as 8 tiJc3 or 8 b3 .
Because of 6 ... cxd4! in the last diagram, 6 ..te2 replaced 6 ..td3. Black's most common reply is 6 ... cxd4 7 cxd4 tiJge7. He has no major problems in lines that run 8 tiJc3 tiJf5 9 tiJa4 'ifa5+ 10 ..td2 ..tb4 11 ..tc3 b5! or 8 b3 tiJf5 9..tb2 ..tb4+. White's best may be 8 tiJa3 tiJf5 9 tiJc2 ..tb4+ 10 ~fl. But Black can avoid that if he reaches this:
David Bronstein called 5•..il.d7 a 'clever waiting move after 3 e5 c5 4 c3 tiJc6 5 tiJf3. White must reveal his intentions with 6 a3, 6 tiJa3 or a bishop move.
Then 8 tiJa3 allows 8.....txa3! 9 bxa3 tiJf5. But how does Black get here?
The wait pays off after 6 tiJa3? cxd4 7 cxd4 ..txa3! 8 bxa3 'ifa5+ with advantage.
The problem with 6 ..te2 tiJh6 is 7 .iLxh6!. White can punish
Another benefit is that by delaying ... 'iib6 Black ensures that
118
Semi-Open Games
dxc5 will not be forcing. That can be a problem after 5.. .'iVb6 because 6...lbge7? allows 7 dxc5! 'iVxc5 8 .ltd3, which favors White.
for a as 7 9 0-0 with bind.
knight in 6... c4 liJbd2 liJa5 8 h6 10 :e1 liJc8! prospects for a
lines such ..te2 liJe7 and ... liJb6 queenside
But in the diagram 6 ..te2 liJge7 Finally, if White feels that 7 dxc5 liJg6! is excellent for Black and so is 7 liJa3 cxd4 8 cxd4 liJf5 5... i.d7 means there's no pressure 9 0-0 i.xa3 10 bxa3 0-0 (Sax- on d4, he may play the natural Dreev, Tilburg 1992). 6..td3. If White opts for 6 a3, which is likely his best, Black benefits from his move order because b6 is free
But then 6... cxd4 7 cxd4 'iYb6! tricks him into the bad 5.. .'~'b6 6 i.d3 cxd4! position one more time.
119
Chapter Five: Double QP Openings But Black had better be ready for Most 1 d4 d5 games become Queen's Gambits after an 3 c4. immediate or delayed c2-c4. The reason White would delay is to rule out certain 2 c4 defenses and to weigh the option of a Queen's Pawn Game once he sees Black's next few moves. With 1 d4 d5 2 lbf3 he denies Black a few sharp lines such as the Albin Counter Gambit, offbeat Slavs and Queen's Gambit Accepteds like 1 d4 d5 2 c4 dxc4 3 lbf3 a6. Black can try to transpose into the latter with 2... a6, hoping for 3 c4 dxc4. But White can make ... a6 look like a wasted move. e.g. 3 i.f4 lbf6 4 e3 e6 5 i.d3 c5 6 c3 i.d6 7 0-0 lbc6 8 dxc5 i.xc5 9 lbbd2 i.d6 10 i.g3 0-0 11 e4 with a plus, Vladimirov-Iuldachev, Mumbai 2003.
This obvious move allowed a minor Swiss master, Hans Fahrni, to confound elite players a century ago. Fahrni-Spielmann, Barmen 1905 went 3...lbc6? 4 cxd5 'i'xd5 5 lbc3 'i'd8 6 d5 lbb8 7 e4 with obvious superiority. Even the natural 3...lbf6?! and 3... cxd4? favor White after 4 cxd5!.
Of course, there is nothing wrong with Black's game after 3... dxc4! or But 2lbf3 doesn't commit White to a QPG because he can shift gears 3 ... e6!. But to play those moves he with 3 c4. Suppose 2... c5, a good would need to know a lot about the move if White heads into something Queen's Gambit Accepted (3 ... dxc4 like 3 c3 lbf6 4 i.f4 lbc6 5 e3 4 e3 or 4 d5) or Tarrasch Defense 'ifb6!. QGD (3 ... e6 4 cxd5 exd5). 120
Double QP Openings
If there are so many pitfalls for Black, why isn't 2 ttJf3 played by every d-pawn player? The answer is White deprives himself of several options and gives Black some useful ones:
but landing in another opening, one he's never played or simply doesn't like.
How can that happen after such quiet moves? Well, suppose Black replies 3.•. cS. He had better be prepared for a Semi-Tarrasch (4 c4 White rules out the QGA lines of 1 d4 d5 2 c4 dxc4 3 e4 and 3 e3, as e6) or a Queen's Gambit Accepted well as QGD variations with ttJge2. (4 ... dxc4). He gives away his best chances of And if he prefers 3••. g6, he needs crushing the Tchigorin Defense to know a bit about the Gruenfeld in QGD (2 c4 ttJc6 3 cxd5 or 3 ttJc3) view of 4 c4!? Yefim Geller, who and the Baltic Defense QGD (2 c4 rarely played the Gruenfeld, once discovered as Black that 4.•• it.g7 .tf5 3 cxd5 or 3 ttJc3). S cxdS ttJxdS 6 it.e2 cS isn't an At the same time he grants Black easy position to handle. extra options such as the Ragozin and Semi-Tarrasch QGD variations, which are not possible after 2 c4 unless White cooperates. That's a lot of plusses and minuses to total up. Note that 1 ttJf3 dS 2 d4 reaches the same position as 1 d4 d5 2 ttJf3. This can be exasperating to an aggressive Black who likes to unbalance the position quickly with 1 ttJf3 d5 2 c4 d4 or 1 d4 ttJf6. That was the case when Lajos Portisch, an Indian specialist, faced a young Anatoly Karpov, who began 1 ttJf3. Portisch found himself in a relatively unfamiliar Slav Defense after 1...d5 2 d4! and lost. COLLE SYSTEM
Black's greatest danger after 1 d4 dS 2 ttJf3 ttJf6 3 e3 isn't the Colle
Soon after 7 e4! Geller was in muddy waters (7 ... ttJb6 8 d5 0-0 9 a4) and eventually lost. The other major peril for Black is a switch from the Colle, characterized by c2-c3, to the Colle's sibling, the Zukertort system, characterized by b2-b3 and J..b2. This is a problem because the best defenses against a Colle often involve ... ttJbd7 - which can be dubious in the Zukertort.
121
Double QP Openings
Here's how: After 1 d4 d5 2 ttJf3 ttJf6 3 e3 e6 4 iLd3 c5 5 c3 ttJc6 6 ttJbd2 White usually answers 6 ...iLe7 or 6 ... iLd6 with 7 0-0 0-0 8 dxc5! iLxc5 9 e4, with a good game. But if Black develops his QN at d7:
This time 8 dxc5? is met by 8... ttJxc5!. White's bishop is attacked and he can't carry out the vital e3-e4 break. He's already a bit worse. But in the Zukertort, 1 d4 d5 2 ttJf3 ttJf6 3 e3 e6 4 iLd3 c5 5 b3, Black stands well following 5...ttJc6! 6 iLb2 iLe7, e.g. 7 0-0 0-0 8 ttJbd2 'ilc7 9 ttJe5 ttJd7!. The danger to both players is tipping their hand. If Black had played ... ttJbd7 instead of ... ttJc6, he would be worse after 9 ttJe5!, e.g. 9... b6 10 f4 iLb7 11 c4 ttJe4 12 cxd5 exd5 13 ttJxe4 dxe4 14 iLc4 as in I. Rabinovich-Makogonov, Soviet Championship 1937. What this means is Black's QN should be watching White's QB and
vice versa. On 1 d4 d5 2 ttJf3 ttJf6 3 e3 e6 4 il.d3 it is inexact to play 4.•.ttJbd7:
White hasn't chosen between Colle and Zukertort yet, so 5 0-0 c5 6 b3! gives him somewhat the better chances. More precise is 4 ... c5. Then 5 b3 and 5 c3 commits White, and Black can respond 5... ttJc6! and 5... ttJbd7! respectively. The crafty response to 4.••c5 is 5 0-0, which was played by Jose Capablanca and Akiba Rubinstein. They waited for a Black commitment, 5... ttJbd7 6 b3! or 5... ttJc6 6 c3!. But they weren't tested by 5•••c4!, e.g. 6 iLe2 b5 7 b3 il.b7 8 a4 a6 9 axb5 axb5 10 ':xa8 iLxa8 11 bxc4 bxc4 12 c3 ttJbd7 with equality, Inkiov-Groszpeter, Copenhagen 1988. OTHER QUEEN PAWN GAMES The Trompowsky Attack (1 d4 ttJf6 2 iLg5) has a poor cousin in
122
Double QP Openings
1 d4 dS 2 i.gS. The lack of a i.xf6 threat makes Black's life easier, and books say 2 ... h6 3 i.h4 c6 and ... 'iib6 is good.
edge of a symmetrical position after 4 ..tf4. The easiest way to equalize against 3 c3 is likely 3.....tf5 4 "iVb3 'iic8 .
If Black doesn't like that he should beware of c2-c4 traps. One arises when he avoids doubled pawns with 2...ttJd7 and 3... ttJgf6. Then 3 c4!.
ALBIN COUNTER GAMBIT The main line of the Albin begins with 1 d4 dS 2 c4 eS 3 dxeS d4 4 ttJf3 tbc6. White characteristically fianchettoes his KB and plays tbbd2, 0-0, 'i'a4 and a2-a3. Black, meanwhile, develops his QB, usually at e6 or g4 and then decides between ...'iid7/ ... 0-0-0 and the alternative ... ttJge7/ ... 0-0. Albin aficionados have tried and failed to establish which QB move is correct. They should consider meeting S ttJbd2 with S•.. ..tfS!?
Now 3••• dxc4 4 ttJc3 is an inferior QGA and 3..• c6 is a dubious Slav, 4 cxdS! cxdS S ttJc3, because his QN should be at c6. For example, 5... ttJgf6 6 e3 g6 7 ttJf3 i.g7 8 i.d3 0-0 9 0-0 b6 10 .:tel i.b7 11 'iib3 with advantage in Skembris-Vasovsky, Skopje 2002. The Torre Attack is 1 d4 ttJf6 2 ttJf3 e6 3 i.gS. Normally Black can avoid it simply with 1...d5 since 2 ttJf3 ttJf6 3 i.g5 allows him quick equality with 3... ttJe4!. But there is a sneaky waiting move, 3 c3. Then on 3... e6 4 i.g5 an unwitting Black has been dragged into the Torre. If 3... c6 instead White has the usual slight
Unlike 5... i.e6 and 5... i.g4, this makes a threat, 6... ttJb4. That wins time for queenside castling after 6 a3 'iid7, under better circumstances than usual. The test of this is order is 7 b4 i.xb4!? (8 axb4 ttJxb4 9 "tWa4 ttJc2+ 10 ~dl ttJxal 11 'iixal c5). Declining the sack is harmless,
123
Double QP Openings
8 'ilVb3 ..ta5 9 e3 l::td8, SpaisKatsaris, Athens 1999.
1990s, White did well when he reached them via the Exchange
What happens after 7 g3 0-0-0 8 ..tg2 is virtually unexplored, e.g.
French or Petroff, so Black has a real decision at move three.
8... d3 9 e3 ..th3 10 ..txh3 "ifxh3 (11 b4 ..txb4!). White avoids this with S g3! and if S.••..te6 then 6lbbd2.
White makes it somewhat easier for him when he chooses 3 lbc3 because after 3... lbf6 he has nothing better than 4 e3, heading towards main QGA lines. But in those lines, lbc3 is better delayed, as we'll see shortly.
QUEEN'S GAMBIT ACCEPTED After Joseph Blackburne popularized 3 lbf3, following 1 d4 dS 2 c4 dxc4, White's alternatives almost vanished. But they're reappearing, mainly as transpo devices. Garry Kasparov played 3 e3, for example, when he wanted to avoid 3 lbf3 c6!? 4 e4 b5.
Another story entirely is the immediate 3 e4. It was once considered dubious because of 3 ... e5, since White cannot maintain a pawn on d4. But 4 lbO! revived this line and poses a move order problem.
Both 3 e3 and 3 lbc3 get Black thinking about whether to transpose into main lines with 3...lbf6 or to try for quicker equality with 3... e5. For example, 3 lbc3 e5 4 dxe5? 'iVxdl + is excellent. The key line is 3 lbc3 eS 4 e3 exd4 S exd4 lbf6 6..txc4.
A controversial position arises after 4 ... ..tb4+ 5 lbbd2 exd4 6 ..txc4 and by transposition from 4 ... exd4 5 ..txc4 ..tb4+ 6 lbbd2. Which is best for Black?
Theory used to call such positions boringly even. But in the
Around 1980 the answer was 4 .....tb4+ because 5 lbbd2 is punished by 5... c3! and S i..d2 i..xd2+ 6 'i'xd2 exd4 peters out to
124
Double QP Openings
equality after 7 'iVxd4 'iVxd4 or 7 l2Jxd4 'iVe7 8 f3 l2Jf6 9 iLxc4 0-0 and ...:d8. But if Black waits to check, 4..• exd4 5 iLxe4 iLb4+, then White had a choice between 6 iLd2 and the more aggressive 6 4:Jbd2!, the thinking went.
Black can sidestep that with 4..• e6 5 e4 iLb4. This an invitation to a Vienna QGD (6 iLgS! cS!) and one that should be accepted because 6 eS 4:JdS 7 iLd2 l2Jb6! makes it hard for White to regain his pawn.
Within 10 years, theory reversed itself and declared 4...iLb4+ to be second-best because White was getting good results from 5l2Je3!.
Another Black option after 4 l2Jc3 is 4•..l2Je6. Then S e4 Jt.g4 is one of the better lines of the Tchigorin QGD. It's as if White played I d4 dS 2 c4 l2Jc6 3 4:Jc3 dxc4 4 l2Jf3 l2Jf6 S e4 iLg4 rather than 4 dS!?
Today neither order is clearly best. It's a matter of which position Black most wants to avoid, 4 ... iLb4+ S l2Jc3! or 4 ... exd4 S iLxc4 iLb4+ 6l2Jbd2!.
But if Black played 2 ... dxc4 to stay in the QGA, he should consider the three alternatives to 3... l2Jf6. They not only avoid the gambit but speed his counterplay.
QGA: 3l2Jf3
The first is 3 ...e5. After 4 e3 Black can safely head towards a QGA tabia with 4 ... l2Jf6 S iLxc4 e6.
At any tournament you're likely to see Black rattle off his first half dozen moves in a QGA as if they didn't matter. But his third move is significant. The popular choice, 3.•.l2Jf6, allows White to offer a gambit, 4 l2Je3 and S e4, based on attacking the knight with e4-eS.
He also get an extra option, 4.•• exd4 with the idea of keeping the pawn, S exd4 iLe6!? The key line is 5 iLxe4 "ike7 (not S... dxe3?? 6 iLxf7+):
This was considered refuted by 6 'iVb3 because 6... iLe6? is met by 125
Double QP Openings
7 .t.xe6! 'ii'xc 1+ 8 \t>e2 with a fierce, probably winning, attack. However QGA and QGD lines are replete with 'little' moves that make a big difference, such as 6... e6!. Then 7 lLlxd4 a6 is equal and 7 exd4 lLlc6 8 0-0 a6 reaches a more normal QGA position but with the addition of the queen moves. That helps Black because he threatens 9 ... lLla5! and gets a fine position after 9 'tid! lLlf6, e.g. 10 lLlc3 .t.e7 11 d5 ttJa5! or 11 'tie2 0-0 12 l:tdllLlb4 as in WojtkiewiczYermolinsky, San Francisco 2002.
But that helps Black. He can transpose into Rubinstein's a2-a4 line of the main QGA, 4...lLlf6 5 e3 e6 6 .t.xc4 c5, which is currently considered too quiet for an edge.
Or he can try 5•...t.g4! 6 .t.xc4 e6 7lLlc3, a superior version of the pin That's two plusses for 3... c5!? avoiding the 3 .. .ti:Jf6 4 lLlc3/5 e4 variation. In the original pin order, gambit and offering Black a good 3 lLlf3 lLlf6 4 e3 .t.g4, White gets alternative to the QGA tabia. The the edge from 5 .t.xc4 e6 6 lLlc3 minus is 4 d5!? and then 4..• e6 particularly after 6... lLlc6 7 .t.b5!. But that's not possible in this 5 lLlc3 lLlf6 6 e4, which is unclear. order and 7... lLlc6 Black either The second alternative is 3••. a6, equalizes with ... e5 or does well which often transposes into the in complications such as 8 h3 .t.h5 third, 3... e6. It was considered an 9 0-0 .t.d6 10 .t.e2 0-0 11 g4 .t.g6 important finesse - to discourage 12 lLlh4 lLlb4, Hillarp Persson4 'iia4+ with 4 ... b5! - back in the C. Hansen, Malmo 2003. 1930s when White was getting a If there is a serious drawback to plus from 3 ... lLlf6 4 'tWa4+. 3... a6 (and 3... e6) it is 4 e4!? when Today the check seems harmless 4•..b5 5 a4 c6, as Kasparov has and 3... a6 is usually just a prelude played, is double-edged. to 4 e3 lLlf6 5 .t.xc4 e6, another tabia. QGA: 3•••lLlf6 There is also a risky extra option, 4 e3 b5, as well as Capablanca's The chief merit of3 ... lLlf6 is that 4 e3 e6 5 .txc4 b5!? and 6... i.b7. it stops 4 e4 and leads in the vast That's often enough to frighten majority of games to a tabia after White into 4 a4?!. 4 e3 e6 5 .t.xc4. The inevitable ... c5 126
Double QP Openings
allows Black to isolate White's dAnother version of the delayed pawn (... cxd4/exd4) or expand on ... c5 is 1 d4 d5 2 c4 dxc4 3 ttJO the queens ide with ... a6/ ... b5/... c4. ttJf6 4 e3 a6 5 ~xc4 b5, a But this also gives White a Capablanca idea. chance for dxc5, possibly followed by a trade of queens into a slightly favorable endgame. If Black is the higher-rated player he can avert an endgame by delaying ... c5, e.g. 5.•. a6 6 0-0 ttJbd7.
Then 7 'ii'e2 c5 8 dxc5 ~xc5 is harmless and Black can even get the edge after 7 a3?! b5 8 il.a2 il.b7
On the routine 6 ~b3~b7 7 0-0 e6 8 'ii'e2 Black can reach a roughly equal main line position with 8... ttJbd7 9 l:.dl c5. This order is a bit of a bluff because it's based on White heeding a QGA rule of thumb that says when the bishop is attacked by ... b5, it's almost always best to retreat to b3.
9 'i'e2 c5 10 dxc5 ttJxc5!.
This is an exception: 6 ~d3! and 7 a4! favorably probes Black's The problem with this order is queenside, e.g. 6 ... ~b7 7 a4! b4 7 a4!, creating a better-than-usual 8 ttJbd2 ttJbd7 9 0-0 e6 10 'ii'e2 c5 version of the Rubinstein line. 11 e4 cxd4 12 e5 ttJd5 13 ttJb3 ttJc5 Unlike the usual Rubinstein Black 14 ~g5! 'ii'd7 15 ttJfxd4, Topalovcannot develop his QN on c6, its Narcisco Dublan, Barcelona 2000. best square once the hole at b4 appears. He doesn't quite equalize White's strategy works because after 7... c5 8 ttJc3 cxd4 9 exd4 Ji.e7, his knight was at bI. If he violates e.g. 10 'i'e2 0-0 11 il.g5 ttJb6 another rule of thumb and brings it 12 il.b3 ttJbd5 13 ':'fel and ttJe5 as out early, we get a position like in Lechtynsky-Kantorik, Plzen 4... e6 5 Ji.xc4 c5 6 ttJc3 and 6... a6! 2003. 7 0-0 b5: 127
Double QP Openings
Experience with 8 i.d3 lLlbd7 shows that Black has an easy time, e.g. 9 'iVe2 i.b7 10 l:dl 'iib6 11 i.c2 l:d8 12 a4 b4 13lLlbl 'iic7 14 i.d3 cxd4 15 exd4 a5, BenitahFressinet, Paris 2004.
Now on 6 0-0 a6 7 dxc5 Black's best is 7.. :ifxdl and a draw is the typical result of 8 l:xdl i.xc5.
But 6 'iie2 makes dxc5 more of a threat, since it wouldn't allow a trade of queens. Then 6.•. cxd4 What happened: White was 7 exd4 lLlc6 avoids that and tries to tricked into a toothless line of punish White for endangering the Meran Variation of the Semi-Slav. d4-pawn. Compare it with 1 d4 d5 2 c4 c6 But another QGA rule of thumb 3 lLlf3 lLlf6 4 lLlc3 e6 5 e3 lLlbd7 6 i.d3 dxc4 7 i.xc4 b5 8 i.d3 a6 says Black shouldn't exchange on when he plays 9 O-O?! c5! 10 'iVe2 d4, as Wilhelm Steinitz habitually did, until he can safely threaten i.b7, not 9 e4!. the d-pawn or obtain other White improves in the diagram compensation. In this case, with 8 i.b3! when 8... i.b7 9 'iie2 accepting the gambit is risky, 80-0! lLlbd7 10 l:dl 'iib8 or 10 ...'iVc7 are lLlxd4 9lLlxd4 'iixd4 10 l:dl 'iib6 main QGA lines that have been 11 i.b5+ i.d7 12 lLlc3, Pelletiertested for decades without finding a Arencibia, Obeda 1998. significant edge. But declining it, 8... i.e7 9 lLlc3 Attention has turned instead to an 0-0 10 i.e3, transposes into the early 'iie2. Books used to say that somewhat discredited Steinitz the order of 0-0 and 'iie2 didn't Variation. As a result 6... cxd4 looks matter because - you guessed it suspect and there is no consensus they 'just transpose.' But it does on what Black should do instead matter because of the possibility after 6 'iie2. of dxc5 and e3-e4-e5, Semyon There are also good and Furman's plan. After 1 d4 d5 2 c4 dxc4 3 lLlf3 lLlf6 4 e3 e6 5 i.xc4 c5: bad transpositions when White 128
Double QP Openings
anticipates ... b5 with a bishop move, 6 0-0 a6 7 .i.b3.
Now 7...b5 8 a4! corrupts the Black queenside, e.g. 8... b4 9 tLlbd2 and 8....i.b7 9 axb5 axb5 10 lIxa8 .i.xa8 11 'iie2, 11 tLla3 or 11 tLlc3.
The reason they often transpose is that 8 tLle3 exd4 does not win a pawn. After 9lIdl! Black has been trapped in another Steinitz (9 ....i.e7 10 exd4 0-0 11 d5! exd5 12 tLlxd5 tLlxd5 13 .i.xd5 'iic7 14 'iie4) .
Max Euwe awarded 8 tLlc3 an Moreover on 7... .i.e7 White exclamation point for a different plays 8 'iie2 and prepares to get a reason. White can preserve his good version of Furman's line with bishop, after 8 tLle3 b5 9 .i.b3 .i.b7 9 dxe5!. 10 lIdl e4?!, with 11 .i.e2 tLlb4 If Black stops that with 8... cxd4 12 .i.bl!. This favors him when he he transposes into another dubious gets to push his e-pawn, e.g. Steinitz, 9 lIdl tLlc6 10 exd4 0-0 12 ... .i.e7 13 e4 tLld3 14 .i.xd3 cxd3 11 tLlc3. The same goes for 7... tLlc6 15 'i'xd3 b4 16 e5! as in 8 'iie2 cxd4 9 lId!. Moskalenko-Sulava, Aosta 1990. The correct order is 7•.• exd4! 8 exd4 tLle6. White doesn't get time for 'i'e2/lIdl because this time 9 'iie2 tLlxd4! is unsound. He has to settle for a tiny edge such as 9 tLlc3 .i.e7 10 lIel 0-0 11 a4 .i.d7, as in a Topalov-Kramnik match game in 2006. The final QGA issue concerns 6 0-0 a6 7 'i'e2 tLle6. Then 8 tLlc3 and 8 lIdl often transpose, e.g. 8 l:tdl b5 9 .i.b3 tLle6 10 tLle3.
In contrast, 8 ':dl b5 9 .i.b3 c4 10 .i.c2 tLlb4! equalizes for Black, e.g. 11 tLlc3 tLlxc2 12 'iixc2 .i.b7 13 e4? (13 d5!?) b4 14 e5 bxc3 15 exf6 gxf6 as in a celebrated Szabo-Euwe game won by Black. QUEEN'S GAMBIT DECLINED Players have been finding finesses in the QGD since Pierre St.
129
Double QP Openings
Amant showed that after 1 c4 Black A third way for Black to prompt could reach it with 1...e6! and ... d5. liJf3 is a bluff, 1 d4 d5 2 c4 c6. Even the recently popular order Actually it's two bluffs: Black is of 1 d4 liJf6 2 c4 e6 and then 3 liJf3 counting on White to heed the d5! is old. It became famous in the majority view that 3 liJf3 is best games of Boris Kostic and Heinrich against the Slav Defense. Then he can continue 3... liJf6 4 liJc3 e6, Wolf in the 1920s. hinting that he is armed to the That was the 'Nimzo-threat' order, based on White's fear of theoretical teeth with the latest 3 liJc3 .i.b4. As long as the Nimzo- Meran (5 e3 liJbd7 6 .i.d3 dxc4) Indian is performing well - as it is analysis. White often flees from now - this will be useful to Black. that, with 5 .i.g5, into main QGD He deprives White of liJge2, which lines - which a bluffing Black is often superior to liJf3, and of a wanted all along. tempo he might put to better use in He is taking two small risks. He orders like 1 d4 d5 2 c4 e6 3 liJc3 might end up in the rather liJf6. dull/drawish Exchange Slav Black has another way of (3 cxd5). And he can't reach the inducing liJf3: 1 d4 d5 2 c4 e6 best version of the Orthodox 3 liJc3 and now 3... i.e7. Defense (5 .i.g5 .i.e7 6 e3 0-0) because an unforced ... c6 doesn't work well with ... i.e7. A KramnikDeep Fritz match game in 2002 began 1 d4 liJf6 2 c4 e6 3 liJf3 d5 4liJc3 c6 5 i.g5 .i.e7?!.
Since White cannot play 4 i.g5 he usually picks the most natural developing move, 4 liJf3, and life goes on after 4... liJf6, as if 3... liJf6 4liJf3 i.e7 had been played. On the minus side, this order denies Black White was so surprised that he non-....i.e7 variations, such as the Cambridge Springs, Vienna and allowed equality after 6 e3 0-0 7 .i.d3 liJbd7 8 0-0 dxc4 9 i.xc4 Botvinnik. 130
Double QP Openings ~d5 10 ~xe7 tixe7 11 l:tc1?! ~xc3 12 l1xc3 e5. He forgot what
Black has closed some doors in the diagram but kept others has been known since the 1930s, open. The Tartakower-Makogonovthat 11 ~e4! favors White, e.g. Bondarevsky Variation and the 11...~5f6 12 '~g3 e5 13 ~f5 or Tarrasch Defense work badly with 12 ... c5 13 tie2 ~b6 14 ~b3 cxd4 ... ~bd7. But Black can still playa 15 ~xd4. Semi-Slav (5 e3 c6) or, after 5 ~g5, an Orthodox Defense (5 ... ~e7), a Witnesses described Krarnnik as Cambridge Springs (5 ... c6 6 e3 'visibly deflated' when he realized a tia5) or a Manhattan Variation computer, hardly a master of (5 ... ~b4). psychology, had tricked him into a This tabia also helps Black in a line whose theory he'd forgotten. Lasker Defense, 5 i.g5 h6 6 ~h4 ~e7 7 e3 ~e4, as Vlf Andersson ALTERNATIVE TABIAS showed. In the more familiar The QGD tabias we all know Lasker order, 1 d4 d5 2 c4 e6 3 ~c3 (and love, hate, etc.) typically ~f6 4 ~g5 i.e7 5 e3 0-0 6 ~f3 h6 appear after 1 d4 d5 2 c4 e6 3 ~c3 7 ~h4 ~e4, White can misplace the ~f6 4 ~g5 and then 4 ... ~bd7 or enemy bishop, 7 i.xf6!? i.xf6, and 4 ... ~e7. But there are others: hold a small pull. But in the Andersson order ~xf6? is bad because Black can
retake with the knight. A second virtue of this order appears after 5 ~g5 h6 6 i.h4 i.e7 7 e3 ~e4 8 ~xe7 tixe7 9 cxd5 ~xc3 10 bxc3 exd5 11 tib3:
This one can come from the Nirnzo-threat, 1 d4 ~f6 2 c4 e6 3 ~f3 and then 3... d5 4 ~c3 ~bd7. It also arises from 1 d4 d5 2 c4 e6 3 ~f3 ~d7!?, a means of discouraging the Catalan (4 g3 dxc4 and ... ~b6), and then 4 ~c3 ~gf6. 131
Double QP Openings
This is a standard anti-Lasker strategy. After Black defends the attacked d-pawn, with 11.. .c6 or 11..:iVd6, White tries to liquidate it favorably with 12 c4!. But here Black can defend it with the superior 11 •..ttJf6!. His pieces are better coordinated than normal after 12 c4 c6 13 i.d3 i.e6 (14 0-0 l:tc8 15 l:.abll:tc7 16 cxd5 i.xd5 as in P.Nikolic-Andersson, Leningrad 1987).
This tabia sets a trap: 4... cS?! falls into a Tarrasch line, S cxdS exdS 6 ttJc3, that's been known to
The chief drawback to the 1 d4 d5 2 c4 e6 3 ttJc3 ttJf6 4 ttJf3 ttJbd7 be bad since Mr. MacDonnell tabia is that White hasn't committed met Monsieur LaBourdonnais. A his QB yet. He can create threats century-plus of experience with with i.f4 and ttJb5. In most 1 d4 d5 6... i.e6 7 e4! and 6...ttJc6 7 i.xf6 lines the best response to i.f4 is gxf6 8 e3 confirms that. ... c5!. Butthat's not nearly as strong when a knight is at d7. Another virtue is White can favorably evade the Cambridge That means Fritz Samisch's S cxdS exdS 6 i.f4 should be good. Springs and Manhattan Variations The natural 6 ... i.e7? walks into by answering a check with ttJbd2. 7 ttJb5! i.b4+ 8 ttJd2 i.a5 9 'iiVa4. For example, 4... ttJbd7 5 e3 i.b4+ After 6... c6 White can build a good 6 ttJbd2! and then 6... 0-0 7 i.d3 h6 foundation for the middlegame with 8 i.h4 is a good Orthodox Defense, 7 'iiVc2 i.e7 8 h3 0-0 9 e3 l:te8 e.g. 8... dxc4 9 i..xc4 b6 100-0 i.b7 10 i.d3, a better-than-usual 11 'iiVe2 i.e7 12l:tfdl ttJe4 13 ttJxe4 Exchange Variation. i.xh4? 14 d5! with a big edge, Gabriel-Stangl, Altensteig 1993. Moving on to another tabia: Mihai Suba liked the fluid center Black gets in a Semi-Tarrasch, 1 d4 d5 2 c4 e6 3 ttJf3 ttJf6 4 ttJc3 c5 5 cxd5 ttJxd5!. But in 1977 Tigran Petrosian 'caught me out' with 4 i.gS!?, he recalled. Suba had to handle this unfamiliar position and lost.
But 4... i.b4+ is more exact. Then 5 ttJbd2 dxc4! is fine for Black and so is 5 i.d2 i.e7!. In 1934, Hans Mueller recommended 4... i.b4+ 5 ttJc3 dxc4 6 e4 c5 - and the Vienna Variation was born.
132
Double QP Openings
Black can play mind games with the immediate 4... dxc4!?
This pays off after s... lbbd7 6 e3 'iliaS? 7 a3! and 8 b4 with advantage. Or after 5•.. i.e7 6 e3 'Maybe I'm going to play the Vienna, by meeting S lbc3 with lbbd7, an Orthodox Defense in S... i.b4,' he says. 'But maybe I which he can meet 7 i.d3 0-0 8 0-0 intend S... c6, which means we'll get dxc4? with 9lbxc4! as Capablanca to see who memorized more of the used to do. Botvinnik Variation. And if I really For example, 9... lbdS 10 i.xe7 wanted to play the Vienna, why didn't I give the check at move lbxe7 11 l::tc1 lbg6 12 i.xg6! hxg6 13 e4 and Black cannot achieve the four?' needed ... eS break (CapablancaWhite often refuses to be drawn Tylor, Hastings 1930). into this guessing game and replies But whenever White plays lbbd2 S e3. Then S... cS is an offbeat QGA and S... bS leads to a very double- there is a chance either ... cS or edged and relatively unexplored ... i.d6/ ... eS will equalize quickly. line that runs 6 a4 c6 7lbc3 i.b4!. Here S... cS has less impact because it's a loss of tempo. But 5...lbbd7 Black's other option in the 1 d4 6 e3 h6 7 i.h4 i.d6! is fine, e.g. d5 2 c4 e6 3 lbf3 lbf6 4 i.g5 tabia 8 i.d3 0-0 9 0-0 eS! 10 i.g3 iie7 is 4••• c6, again hinting at a 11 cxdS cxdS 12 dxeSlbxeS, RivasBotvinnik (S lbc3 dxc4) but also a Yusupov, Minsk 1982. Cambridge Springs (s ... lbbd7 6 e3 'iliaS). White may be tempted by 5lbbd2.
The previous QGD tabias allow White to transpose into the tried and true Exchange Variation with cxdS. This one doesn't:
133
Double QP Openings
3 liJf3 liJf6 4 liJc3 liJbd7 S cxdS, said S ..igS was better. 'White can always exchange at a later stage when the 'economic climate' is more favorable for him,' he explained.
It can come about via 1 d4 liJf6 2 c4 e6 3 liJf3 d5 4 liJc3 dxc4. Black invites a QGA (after S e3), a Vienna Variation QGD (after S e4 or 5 ..igS) or even a Nimzo-Indian (after S ..igS ..ib4 6 e4 cS 6 a3!?). The deceptive choice is S 'it'a4+!?, trying to reach a superior QGA. In the normal QGA, after 1 d4 dS 2 c4 dxc4 3 liJf3 liJf6 Black can safely meet 4 'it'a4+ with 4 ... liJc6 or 4 ... c6. But 4 ... liJbd7 S liJc3 e6 is regarded as inferior. This means 5 'it'a4+ liJbd7?! traps Black in the bad 4 'it'a4+ line, e.g. 6 e4 a6 7 ..ixc4 c6? 8 'it'dl! ..ie7 9 0-0 0-0 10 a4 b6 11 ..id3 ..ib7 12 eS! and White soon had a winning attack in Kasparov-Short, London 1993. Better is 5.•. c6 6 'it'xc4 b5. QGD: EXCHANGE The first issue in the Exchange Variation is when to exchange. Today it's commonly done at move four. But Alexander Alekhine, commenting on 1 d4 dS 2 c4 e6
An example of that is I d4 dS 2 c4 e6 3 liJc3 liJf6 4 ..igS liJbd7 S e3 c6. Black is advertising his interest in the Cambridge Springs. If White can't remember the theory, 6 cxdS! makes sense. (A move later is too late, 6 liJf3 'tiaS 7 cxdS liJxdS!.) But exchanging on move three is too early. After 1 d4 d5 2 c4 e6 3 cxd5? exd5 4liJc3:
Black solves his main QGD problem, how to develop with QB, with 4... c6! and ..•..if5!, e.g. S 'tic2 ..id6 6liJf3 liJe7 7 ..igS ..ifS. It's also inaccurate to exchange at move four if White is already committed to liJf3. One of Petrosian's most impressive wins as Black began with 1 d4 liJf6 2 c4 e6 3 liJf3 dS 4 cxdS?! exdS S liJc3 c6
134
Double QP Openings
5 e3 0-0 6 cxd5? tDxd5! 7 .ixe7 'iixe7 8 tDf3 tbxc3 9 bxc3 b6! (10 .ie2 .ib7 11 0-0 c5 12 tDe5 Controlling the bl-h7 diagonal is tbc6 13 tDxc6? .txc6 14.tf3 ~ac8 vital in the Exchange. If Black gets and Black won in Alatortsevcontrol, he usually equalizes. But if Capablanca, Moscow 1935). he's played an early ... tDbd7, then White also has an easy time after cxd5! usually gives White enough tDf3 0-0 6 'ifc2 c6 7 e3 tbbd7 5 time to stop him, e.g. 1 d4 d5 2 c4 8 cxd5 tDxd5! 9 .ixe7 'iixe7 e6 3 tbc3 tbf6 4 .ig5 tbbd7 5 cxd5 exd5 6 e3 and .id3 or 'iic2 will (10 .te2 ~e8 11 0-0 ctJxc3 12 bxc3 make ... .if5 impossible. Salo Flohr, e5). 6 .ig5 .ie7 and ... .if5! (7 'iic2 g6 8 e3 .if5).
a QGD connoisseur, played cxd5 And bear in mind that a delayed immediately after ... tbbd7 even if cxd5 may be conflict with White's other early moves. Alekhine felt the that was at move eight. 'economic climate' was favorable The reason many GMs make the after 1 d4 d5 2 c4 e6 3 tbc3 tDf6 capture early (1 d4 d5 2 c4 e6 4 .ig5 .ie7 5 e3 tbbd7 6 l:.c 1 0-0 3 tbc3 tbf6 4 cxd5) is that once for 7 cxd5 because 7 ... tbxd5? .tg5 and ... .ie7 are played, Black 8 tbxd5 drops a pawn. can favorably swap two pairs of But after 7 ... exd5 8 .ltd3 c6 pieces after ... tbxd5. The result often resembles an Orthodox 9 'iVc2 ~e8 10 tDf3 he discovered Defense in which Black simplifies his QR would have been better placed at b I to support the minority with ... dxc4 and ...tbd5. attack ofb2-b4-b5. Take the case of 1 d4 d5 2 c4 e6 Another issue is when to bring 3 tbc3 tbf6 4 .ig5 .ie7: out White's KN. Consider 1 d4 d5 2 c4 e6 3 tbc3 tDf6 4 .ig5 c6 5 e3 tbbd7 6 cxd5 exd5 7 .td3 .ie7.
Black equalizes after 5 cxd5? tbxd5! 6 .ixe7 'ifxe7. Similar is 135
Double QP Openings
This was Smyslov-Keres, World Championship Tournament 1948. White played 8 tbf3. Black replied 8... 0-0. Today both moves are considered mistakes. Black should play 8...tbe4! because 9 tbxe4? dxe4 wins a piece and because 9 ..txe4 ..txg5! or 9 ..txe7 'iixe7 10 'iVc2 f5 are fine for him. If White wants to play tbf3 he should prepare it with 8 'iie2! so that 8 ... tbe4? loses a pawn to 8 ..txe7 and 9 tbxe4. By delaying tbf3 White retains the tbge2 option and can seize control of bl-h7. There is some further subtlety if Black pursues ... ..tf5 in another way, 1 d4 d5 2 e4 e6 3 tbe3 tbf6 4 exd5 exd5 5 ..tg5 ..te7 6 e3 e6 7 'ile2 tbbd7 and now 8 tbf3 tbf8 9 ..td3 tbe6 10 ..th4 g6.
He can meet the regrouping plan with fl-f3 and e3-e4 or g2-g4, e.g. 9 •••tbe6 10 ..th4 g6 11 0-0-0 tbg7 12 f3!, e.g. 12 ... ..te6 13 g4! or 12 ... tbf5 13 ..tfl 'iVa5 14 ~bl ..te6 15 h3 0-0-0 16 e4! with a fierce attack in Bronstein-Medina, Gothenberg 1955. But Black has another simplifying option after 8 ..td3 8•••tbh5 9 ..txe7 'fIxe7. This works because 10 ..txh7 is not a check but rather an unsound sack after 10 ... g6. Also 10 tbf3 and 10 g4 allow 10 ... tbf4. White should play 10 tbge2 and keep a small edge.
His idea is ...tbg7 and ... ..tf5. QGD EXCHANGE: ...tbbd7, Black is at least equal after 11 0-0 .•• e6 and ... h6 0-0 12 l:tabl a5 13 a3 tbg7! 14 b4 We've seen how an early ... tbbd7 axb4 15 axb4 ..tf5! 16 tbe5 ..txd3 impairs Black's chances for .....tf5!. 17 tbxd3 tbf5 18 ..tg5 tbd7 19 ..tf4 tbb6 and ... tbc4, as in a Najdorf But there are benefits to quickly connecting the knights. Black can game. meet ..txf6 with ... tbxf6, which is important when 'fIc2/..td3 is lined Black exploited a premature 8 tbf3 in that case. The careful up against the h7-pawn. transposer will prefer 8 ..td3! and then 8..•tbfS 9 tbge2.
It's also significant after 1 d4 d5 2 e4 e6 3 tbe3 tbf6 4 exd5 exd5
136
Double QP Openings
5 lbf3?! c6 6 i.g5 i.e7 7 e3 i.f5! and then S i.d3 i.xd3 9 'iVxd3 lbbd7!:
White was sloppy twice, at move five and again when he missed 9 i.xf6! i.xf6 10 'iVxd3, which misplaces enemy pieces. It would cost Black time to get his bishop to the superior b8-h2 diagonal and to find a good square for the knight.
c5!, with good play in MinoginaBelavenetz, Moscow 1990. White's counter-finesse S lbge2 liIeS 9 0-0:
IS
The absence of ... c6 means 9... lbrs can be met by 10 b4!. Usually White must invest a tempo, such as l:tb 1, to prepare this thematic push. Here it's tactically
But in the diagram Black's pieces are well-coordinated thanks to ... lbbd7! and he should have no difficulty equalizing, e.g. 10 0-0 0-0 11 l:tabl a5 12 a3 l:teS 13 b4 axb4 14 axb4 b5!.
based on 10... i.xb4 11 i.xf6 'iVxf6 12 lbxd5 or 11...gxf6 12 lbxd5! 'iVxd5 13 'iVa4.
For example, 1 d4 d5 2 c4 e6 3 lbc3 lbf6 4 cxd5 exd5 5 i.g5 i.e7 6 e3 0-0 7 i.d3 lbbd7 S'iVc2 l:teS and now 9 0-0-0 lbf8 10 Wb 1 i.e6 11 lbf3 a6 12 lbe5 l:tc8 13 h3
looking for an alternative to i.h4. For example, 1 d4 d5 2 c4 e6 3lbc3 lbf6 4 cxd5 exd5 5 i.g5 i.e7 6 e3 lbbd7 7 i.d3 c6 S 'YWc2 and now S... h6?
The final area of move order finesse in the Exchange involves ... h6. Black often inserts this Another move that Black often because of the danger of defers is ... c6. He can't play the 'iVc2/i.d3/i.xh7+. But he has to be freeing ... lbe4 without it, because careful that he gets the response the d-pawn would hang. But that is usually i.h4 - that he wants. offset by the possibility of ... c5 in one step. Consequently White should be
137
Double QP Openings
Chances would be roughly even after 9 i.h4 lbf8 10 lbge2 lbe6 and ...lbg5!. Also 9... 0-0 10 lbge2 l:te8 11 0-0 lbf8 and ...lbe4. However 9 i.f4! is much more promising after 9... 0-0 10 0-0-0 or 10 lbge2 :e8 11 0-0-0 and h2-h3/g2-g4. White should be looking to transpose from an Orthodox Defense in which ... h6 has been played to an Exchange. For instance, 1 d4 d5 2 c4 e6 3 lbc3 lbf6 4 i.g5 lbbd7 5 lbf3 h6. Now on 6 i.h4 i.e7 7 cxd5 Black has 7... lbxd5! with equality. But there is an immediate 6 cxd5!?
If Black ducks the challenge, 6... exd5 7 i.h4 i.e7 8 e3, he is slightly worse, e.g. 8 ... c6 9 'ilVc2 0-0 10 i.d3 l:te8 11 i.g3 lbf8 12 h3 i.e6 13 0-0 lb6d7 14lba4 a5 15 a3, LSokolov-Seirawan, Dutch Team Championship 2002. Accepting the sacrifice, 6... hxg5 7 dxe6 fxe6 8 lbxg5, grants White
ample compensation (8 ... i.d6 9 e4! as in Torre-P.Nikolic, Leningrad 1987). Another rendering is 1 d4 d5 2 c4 e6 3 lbc3 i.e7 4 lbf3 lbf6 5 i.g5 0-0 6 e3 lbbd7 7 'i'c2 and 7... h6. That provoked the routine 8 i.h4 from such QGD authorities as Viktor Korchnoi, Boris Spassky and Rubinstein. But their opponents missed a good 8 ... c5!, e.g. 9 :dl 'i'a5 10 cxd5 lbxd5 11 i.xe7 lbxe7 12 i.e2 lbd5! 13 0-0 lbxc3 14 bxc3 b6 and ... i.a6, Carlsen-Hansen, Skanderborg 2005. However in 1961 an amateur demonstrated 8 cxd5! is strong. His opponent was shocked into 8 ... lbxd5?, losing a pawn, because he feared 8... hxg5 9 dxe6 offered too much compensation for White. Kasparov revived the idea against Portisch who played 8... exd5 and was a little worse (9 i.f4 c5 10 i.e2).
QGD:ORTHODOX In the Orthodox Defense, 1 d4 d5 2 c4 e6 3 lbc3 lbf6 4 i.g5 i.e7 5 e3 0-0 6 lbf3 lbbd7, White's finesses begin with his QR. Its primary function, either at c1 or dl, is to discourage ... c5, Black's best freeing device.
138
Double QP Openings
13 .tbS cxd4 14 'ii'xd4liJf6 IS liJeS liJe8 16 .td7 (Bogolyubov-Mueller, Zurich 1934).
Black equalizes after 7 a3?! c5! and 7 'iib3 c5! (8 cxdS cxd4 9 liJxd4 liJcS! 10 'iic2 liJxdS, Poluljahov-Tregubov, Krasnodar 2000).
White would love to force his way to these positions. But if Black is careful with ... c6 he can only try to trick him. One way is an early 'iic2 in place of e2-e3. For example 1 d4 d5 2 c4 e6 3 liJc3 liJf6 4 .tg5 .te7 5liJf3 0-0 6 'ii'c2:
That explains 7 ltel. It usually prompts 7... c6, which safeguards the c- and d-pawns and prepares other freeing ideas such as ... liJe4. But if ... c6 has already been played, because Black uses an order such as 1 d4 d5 2 c4 c6 3 liJf3 e6 4 liJc3 liJf6 5 .tg5 liJbd7 6 e3 i.e7?!, there is no pressing need for ltel. Rubinstein showed that White can use his heavy pieces more effectively with 7 'iic2! 0-08 ltd1!. Then 8... cS transposes into a book position in which Black is a move behind and under strong pressure after 9 cxdS exdS 10 dxcS. And thanks to the Q+R tandem, the familiar liquidation, 8... dxc4 9 .txc4liJdS 10 .txe7 'iixe7 11 0-0 liJxc3 12 "ilxc3, doesn't allow ... eS and leaves Black with serious problems after 12 ... cS 13 "ila3 or
Now 6... c6?! 7 e3 liJbd7 8 ltd1! achieves White's goal. This order also opens a 0-0-0 option, e.g. 6 ... h6 7 .txf6! .txf6 8 0-0-0 and 9 e4 with good chances. And after 6...liJbd7 he can switch to an Exchange Variation since 7 cxd5 liJxd5? loses a pawn (8 liJxdS exdS 9 .txe7 "ilxe7 10 'iixc7 or 8 ... .txgS 9 liJxc7). But the problem with an early "ilc2 is that White's rook isn't in place yet so 6... c5! is good. The situation is similar when Black uses the 1 d4 d5 2 c4 e6 3 liJc3 .te7 order to delay .tgS and White voluntarily delays it further,
139
Double QP Openings
4 tbf3 tbf6 5 'iVc2. Then 5.•. 0-0 6 iLg5! h6 allows him to reach the promising 7 iLxf6 iLxf6 8 0-0-0.
Another natural move, 5•••tbbd7, allows an excellent Exchange Variation, 6 cxd5! exd5 7 iLf4! c5 8 e3 0-0 9 iLd3, since Black's QN is misplaced. But once again Black can strike with ... c5 before White has deterred it with l:.dl, e.g. 5... c5!? 6 cxd5 cxd4! 7 tbxd4 exd5 8 iLg5 tbc6 9 l:tdl h6 10 iLh4. 'iVa5 as in Rustemov-Ubilava, Olite 2006.
... dxe4 attacks it. White can avoid that by delaying either tbf3 or tbc3. For example, 1 d4 d5 2 c4 e6 3 tbf3 tbf6 4 iLg5 iLe7 5 e3:
Now 5...tbe4 has been under a cloud since W.Cohn-Leonhardt, Ostende 1907 went 6 iLxe7 'iVxe7 7 tbbd2! 0-0 8 iLd3 and 8... f5 9 0-0 tbd7 10 'iVb3 c6 11 lIac 1 gave White an excellent Stonewall Dutch.
However, 5... 0-0 puts the ball back in his court. He can continue to withhold tbc3 with 6 'iic2. Then he gets that ideal Orthodox Defense QGD: LASKER after 6... tbbd7 7 tbc3 c6 8 l:.dl!. A typical starting point of But once again 7••. c5! is good and Emanuel Lasker's defense is 1 d4 should equalize. d5 2 c4 e6 3 tbc3 tbf6 4 iLg5 iLe7 5 e3 h6 6 iLh4 0-0 7 tbf3 tbe4. QGD: CAMBRIDGE SPRINGS Unlike similar positions, the retreat White has various means of to g3 isn't promising here (8 iLg3?! avoiding the Cambridge Springs in iLb4 9 l:.cl c5). traditional orders. But most of them The trade-seeking ...tbe4 works best when White has one knight at c3, so ... tbxc3 is possible, and another at on f3 so that tbxe41
allow Black to fall back into a good Orthodox or Lasker Defense. If White plays 1 d4 d5 2 c4 e6 3 tbc3 tbf6 4 iLg5 tbbd7 5 e3 c6 6 a3:
140
Double QP Openings
Bogolyubov game won by Black and 13 'iVc2 exd4 14 exd4 tLl7b6 15 tLle5 it.e6 ended in a quick draw in Piket-Ivanchuk, Melody Amber 1993.
He can get a reasonable version of Rubinstein's Orthodox line, 6... it.e7 7 tLln 0-0 8 'iVc2 and ~d1. But instead, Black can transpose to a good Lasker with 7...tLle4 or 7... h6 8 it.h4 tLle4. Then a2-a3 plays little role. The strength of the main line Cambridge Springs, 6 tLln 'iVa5, rests today on whether White gets an edge from 7 cxd5 tLlxd5 8 'iVd2 it.b4 9 ~c1. Black has to decide when to play ... e5, ... h6 and ... 0-0 or he will be distinctly worse. The crucial tabia is:
This has performed well for Black, e.g. 13 e4 tLlf4 14 it.c4 tLlg6 15 a3 tLlxh4! was an Alekhine-
The hard part for Black is reaching the tabia. The 1983 Kasparov-Smyslov candidates match tested various orders and showed that 9•••e5?! is bad in view of 10 a3! it.xc3 11 bxc3 'iixa3 12 e4 with a dangerous attack. That kind of sacrifice is a major Black concern. Alekhine said Black's best order to reach the tabia was 9..•0-0 10 it.d3 e5 and then 11 0-0 h6 12 it.h4 ~e8. But this was based on his claim that 11 a3!? it.xc3 12 bxc3 'ii'xa3 is unsound. If a Cambridge Springs player has doubts about that, he should consider the slightly different 9••• 0-0 10 it.d3 h6.
Then 11 it.h4 .l:.e8 12 0-0 e5 arrives at the tabia and 12 a3 it.f8 is also safe. Black avoids 11 ... e5 12 a3!?, which may be a good version of the gambit.
141
Double QP Openings
But this order has a potential problem in 11 i.f4!? ttJxf4 12 exf4, which stops ... e5. For example 12 ... ttJf6 13 0-0 c5 14 dxc5 i.xc5 15 a3 i.e7 16 ttJe5 l:td8 17 'iie2 was promising for White in Hillarp Persson-Vera, Yerevan 1996. Some authorities say 9 ... h6 is best because 10 i.f4 is less promising here and 10 i.h4 0-0 11 i.d3 e5 reaches the tabia. But White might improve with 11 i.c4!? as in an Alekhine game that went 11...ttJxc3 12 bxc3 i.a3 13 l:tbl e5 14 i.g3.
QGD: BOTVINNIK
He means about 5 i.g5 dxc4 or 5 e3 ttJbd7 6 i.d3 dxc4. If White doesn't like his answers to those questions, he'll also be disappointed when he looks for alternatives and sees that 5 cxd5 exd5! is an inoffensive Exchange Variation (6 i.g5 i.e7 and ... i.f5!). And even ifhe knew Black's intent, White would be reluctant to play cxd5 at move 3 or 4 because he'd end up in another super-quiet line,
you
the Exchange Slav.
Therefore Black should pick the order that avoids what he wants to avoid. If he doesn't mind 9... 0-0 10 i.d3 h6 11 i.f4!?, that's his best bet. But ifhe's willing to accept one of the gambits, other orders are superior.
If
look
up
Mikhail
Botvinnik's variation in books you may find it under 1 d4 d5 2 c4 e6 3 ttJc3 ttJf6 4 ttJf3 c6 5 i.g5 dxc4. But the little secret of Botvinnik specialists is very few of them use that order. It grants White too many promising alternatives, including 4 i.g5 and 4 cxd5. It is other orders, such as 1 d4 d5 2 c4 c6 3 ttJf3 ttJf6 4 ttJc3 e6, that make the Botvinnik the weapon it IS.
'Do you really know as much about the Botvinnik as I do?,' Black is asking. 'Or about the Meran?'
So what a Botvinnik player is really saying in this order is, 'You don't have much choice. It's the Botvinnik for you or quick equality forme.' When White goes into 5 i.g5 dxc4 he justifies the lost pawn with 6 e4 b5 7 e5 h6 8 i.h4 g5 9 ttJxg5 hxg5 10 i.xg5, with mind-boggling complications. One of the less analyzed positions occurs when he retreats instead with 9 i.g3.
142
Double QP Openings
1 d4 d5 2 c4 e6 3 4:Jf3 4:Jf6 4 4:Jc3 dxc4 5 iLg5 iLb4 6 e4 c5, but the conservative 7 iLxc4 is the main line today. A resulting tabia is:
There's a lot happening and it's happening on both wings, e.g. 9... 4:JdS 10 4:Jd2 iLe7 11 iLe2 iLb7 12 4:Jde4 4:Jd7 13 h4 cS 14 hxgS 4:Jxc3 IS 4:Jxc3 cxd4 16 'i'xd4 .txgS 17 0-0 as in Hillarp PerssonHector, Gothenburg 2006, won by White.
Antique theory claimed White is much better (11...'i'xc4+ 12 'iitgl gxf6? 13 l:.c1 or 12 ... 4:Jd7 13 l:.c1). But 11 •.• gxf6! turned out to be fine If this seems attractive to you it's for Black and today his concern is worth adding 9 iLg3 to your the best route to the diagram. repertoire because you can get to One way is 7 iLxc4 cxd4 8 4:Jxd4 the diagram via other orders. For 'i'a5 after which 9 iLxf6 iLxc3+ instance after 1 d4 dS 2 c4 e6 3 4:Jc3 bxc3 'i'xc3+ 11 'iitn gets there. 10 c6 4 4:Jf3 4:Jf6 S iLgS a Black who doesn't like the Botvinnik or The other is 8...iLxc3+ 9 bxc3 'i'a5 Cambridge Springs will try S... h6. and then 10 iLxf6 'i'xc3+. Books used to say 'if 6 iLh4? then The advantage of the first order 6... dxc4! wins a pawn.' is that Black preserves his KB for dark square defense. That's However, 6 iLh4 dxc4 7 e4 is important in the second order when promising even after 7... gS 8 iLg3 White tries 10 iLbS+ instead of bS. Then White has various means 10 iLxf6. of seeking compensation and they Then 10 ... 4:Jbd7 11 iLxf6 'i'xc3+ include 9 eS, reaching the last 12 'iitfl gxf6 13 h4! and l:.h3 offers diagram. good chances. QGD:VIENNA But in the first order the check is There have been many attempts harmless, 8... 'i'aS 9 iLbS+ 4:Jbd7 to refute the Vienna Variation, 10 iLxf6 gxf6. 143
Double QP Openings
The problem with 8 ... 'iVa5, according to authorities going back to the 1930s, is '9 ..td2! and then 9... e5 10 l2Jc2.' But 9 ..td2 is so passive that 9... 0-0 makes White's pieces look misplaced (10 'iVe2 .:f.d8 11 l2Jf3 l2Jc6 12 a3 ..txc3 13 ..txc3 'iVh5, Summerscale-Wells, Millfield 2000). Instead, White may have to risk 10 l2Jc2 ..txc3 11 ..txc3 'iVg5! 12 'iVe2 'iVxg2 13 0-0-0.
The correct response is S...l2Jc6!. Then 6 ..tgS loses its punch to 6.....te7!. Black can also fall into the same trap via the Semi-Tarrasch, 1 d4 d5 2 c4 e6 3 l2Jc3 l2Jf6 4 l2Jf3 c5, if he plays 5 cxd5 exd5? 6 ..tg5! instead of 5 ... l2Jxd5!. SLAV DEFENSE
The early moves of the Slav conceal quite a bit of subtlety TARRASCHISEMI-TARRASCH because of the strong feelings Today's Tarrasch tabia usually players have about the main lines comes about via 1 d4 dS 2 c4 e6 the Open Slav and the Meran and 3 l2Jc3 cS 4 cxdS exdS S l2Jf3 l2Jc6 Tchigorin variations of the Semi6 g3 l2Jf6 7 ..tg2 ..te7 8 0-0 0-0 and Slav. As a result, a Slav is often a Black's d-pawn becomes a focus of silent battle between one player the middlegame after a pawn trade. trying to avoid, say, the Open Slav That's why Black may defend it and the other trying to avoid the quickly, with S•••l2Jf6?!. Meran, with success going to the better transposer.
His punishment is 6 ..tgS!. He has transposed into 1 d4 d5 2 c4 e6 3 l2Jf3 l2Jf6 4 ..tg5 when Black plays the sloppy 4 ... c5?! and allows 5 cxd5 exd5 6 l2Jc3.
Take the odd-looking 1 d4 l2Jf6 2 c4 c6 and 3... dS. It doesn't get much attention although it was used by Paul Keres, Yefim Bogolyubov and Alekhine. White often responds 3 l2Jc3 automatically and then realizes after 3 ... d5 that he's ruled out anti-Meran lines in which he can play l2Jbd2 or delay a QN move, as in 1 d4 d5 2 c4 c6 3 l2Jf3 l2Jf6 4 e3 e6 5 ..td3 l2Jbd7 6 0-0. Or suppose 1 d4 dS 2 c4 c6 and now 3 e3:
144
Double QP Openings
Black can transpose into a good version of the Schlechter Variation of the Gruenfeld (1 d4 d5 2 c4 c6 3 tZJf3 tZJf6 4 e3 g6). The Schlechter's stodgy reputation is based on lines with the more enterprising .if4 or .ig5. SLAV: EXCHANGE Why would White play such a quiet move? The reason is that an early e2-e3 is a good way of avoiding the high-maintenance Open Variation (1 d4 d5 2 c4 c6 3 tZJc3 tZJf6 4 tZJf3 dxc4). To get an advantage White has to play 5 a4 .if5 6 e3 or 6 tZJe5, or risk the 5 e4!? gambit. Either way he has to memorize a huge amount of book.
White can initiate the Exchange Variation at move three (as most players do today), at move four (as Botvinnik preferred) or later. The rationale for delay is that Black may prematurely commit his pieces before he knows that cxd5 is coming, e.g. 1 d4 d5 2 c4 c6 3 tZJf3 tZJf6 4 ttJc3 .if5? 5 cxd5 cxd5 6 ~3! favors White.
But if Black plays ... dxc4 after e2-e3 White can retake on c4 without spending a tempo on a2-a4. He transposes into a favorable QGA, as if Black played 1 d4 d5 2 c4 dxc4 3 e3 c6?! 4.ixc4. In case of 1 d4 tZJf6 2 c4 c6 3 e3, Black has a good alternative, 3... g6!. Then he transposes into one of his best versions of the King's Indian Defense, e.g. 4 tZJc3 .ig7 5 tZJf3 0-0 6 .id3 d6! 70-0 tZJbd7 8 'ilc2 e5 9 l:tdl 'ile7 10 .in e4 11 tZJd2 l:te8. There are two other drawbacks to e2-e3 in any Slav order. One is that ... .if5! is safer to play when White's QB is blocked in. Also,
The same goes for 4 ... ttJbd7 when 5 cxd5! cxd5 6 .if4 leaves his QN misplaced. White has the better pieces after 6... e6 7 e3 .ie7 8 h3 0-0 9 .id3, e.g. 9... tZJb6 10 0-0 .id7 11 ttJe5 l:tc8 12 'ilb3! .ie8 13 l:tfc1 tZJfd7 14 a4!, Reshevsky-Bernstein, U.S. Championship 1951.
145
Double QP Openings
In the normal Exchange, 3 cxd5 cxd5 4 lbc3 there is a small trap for the unwary Black who thinks this order allows 4•.. lbc6. He'll be surprised by 5 e4!. White enjoys a clear advantage after 5... dxe4? 6 d5 lbe5 7 'iVa4+ iLd7 8 'iVxe4. Black should play 5... lbf6, hoping White will allow the endgame variation of the Panov Caro-Kann (6 exd5 lbxd5 7 lbf3 iLg4 8 'iVb3 iLxf3 9 gxf3 etc.). But 6 e5 lbe4 7 iLd3 is much more promising, e.g. 7 ... lbxc3 8 bxc3 e6 9 lbf3 h6 10 0-0 lba5 11 .l::tbl iLd7 12 lbd2 and 'i'g4, Gaprindashvili-Murshed, Polanica Zdroj 1986. SLAV: 3lbc3 This was once preferred to 3lbf3 (after 1 d4 d5 2 c4 c6) because it does a better job of discouraging ... iLf5, as in 3 lbc3 lbf6 4 e3 iLf5?! 5 cxd5 cxd5 6 'iVb3. Today when White chooses 3 lbc3 it's usually for a different reason - to reach the Semi-Slav, after 3...lbf6 4 e3 e6 5lbf3, without allowing the Open Variation, 4 lbc3 dxc4. Books say the downside to 3 lbc3 is 3.•. dxc4 because ... b5-b4 will attack the knight.
But in practice Black often plays 3 lbc3 dxc4 with no intention of ... b5. He knows all about the 3lbc3 lbf6 4 e3 trick to avoid the Open Slav. 'But I'm the one doing the tricking, ' he thinks to himself, 'because I'll get there after 3... dxc4 4 lbf3 lbf6!.' White can avoid the Open with 4 e4!? But then 4... b5 has more justification (5 a4 b4 6 lba2 lbf6!). Instead he often prefers 4 e3 . There follows 4 ... b5 5 a4 b4 6lba2 e6 7 iLxc4lbf6 and since he has no e4-pawn under attack he can continue 8lbf3. What happened: This often comes about after 1 d4 d5 2 c4 c6 3 lbf3 lbf6 4lbc3 dxc4 when White does not try for advantage with 5 a4 or 5 e4!? but plays the innocuous 5 e3.
146
Double QP Openings
He wants to pull White into 6 ttJc3 .tb4!, a version of the Noteboom Variation in which White has pushed to e4. The pawn is more aggressive there but also more vulnerable. Both sides have chances after 7 .te2 ttJf6 8 'Yi'c2 .tb7 9 0-0 a6 10 eS ttJdS 11 ttJe4 ttJd7. Then S•.. bS 6 a4 b4 7 ttJa2 e6 8 .txc4 reaches the diagram, e.g. 8... ttJbd7 9 0-0 .tb7 10 .td2 as 11 ttJc1 .td6 12 ttJb3 cS 13 'i'e2 0-0 141Hdl 'i'b6. Black has the last laugh. He not only got into the Open Slav that White tried to avoid but even reached a quite equal version of it.
The thematic idea is 6 axbS cxbS 7 b3. But again Black should attack e4, e.g. 7 ... ttJf6! 8 eS ttJe4 9 bxc4 .tb4+ 10 ttJbd2 ttJc6 11 .td3 ttJcS!, with good chances for Black in Belyavsky-Kharlov, Ljubljana 2002.
More common after 1 d4 dS 2 c4 c6 3 ttJf3 is 3.•.ttJf6. Then 4 'Yi'c2!? allows White to sidestep the Meran SLAV: 3 ttJf3 by meeting 4... e6 with S g3. Black Even though 3 ttJf3 dxc4 has less finds his main choice is between a punch than the 3 ttJc3 dxc4 version, closed Catalan, which few Meran Black might like it because of players like, and a Stonewall Dutch tricks. When White tries to punish with S... ttJe4 and 6... fS, which even 3... dxc4 with 4 e4 bS S a4, he can fewer do. avoid transposition to the Slav The queen move can also be a Gambit, S... ttJf6, by means of good weapon against an Open Slav S.••e6!. player because of 4•.. dxc4 S e4 bS 6 b3!. White has good compensation after 6 ... cxb3 7 axb3 e6 8 .td2, e.g. 8... aS 9 .td3 i..b7 10 0-0 .te7 11 .tc3 (NakamuraHansen, Malmo 200S). There's no major drawback to 4 'i'c2. But 4.•. g6 is a better-thanusual version of the Schlechter Slav because ... iHS will gain a te~po. 147
Double QP Openings
If White passes up the various evasions of the Open Slav and continues 4 ttJc3, Black has an alternative to 4... dxc4 5 a4 it.f5 in the form of 5... e6!?
The natural temptation is to punish such a passive move with 6 e4. But 6 ...it.b4 7 e5 ttJd5 8 it.d2 b5 9 axb5 transposes to a line of the Slav Gambit that's not at all bad for Black (5 e4 b5 6 e5 ttJd5 7 a4 e6 8 axb5 it.b4 9 it.d2). This has a psychological benefit since White's preference for 5 a4 rather than 5 e4 indicates he's not as confident in gambit lines. The more common reply in the diagram is 6 e3. Then 6•••c5 7 it.xc4 gives us a QGA in which White has an extra tempo, a2-a4. But Kramnik, among others, likes Black's position because after 7 ... ttJc6 8 0-0 cxd4 9 exd4 it.e7 he gets some control of d5 by occupying the hole at b4 with his knight, e.g. 10 'ii'e2 0-0 II ':'dl ttJb4. Black seems to stand well , e.g. 10 it.g5 0-0 11 ':'e 1 it.d7, Navarra-Svidler, Turin 2006.
SLAV: THREE PAWNS When Black uses his first three moves to put pawns at d5, c6 and e6 he sacrifices development for transpositions. He can drag a reluctant White into the Botvinnik, Noteboom, and other sharp variations or even a better-thanusual Dutch, as well as into some Meran lines, while avoiding more dangerous Meran lines. Black may prefer 1 d4 d5 2 c4 c6 followed by 3 ... e6 to the other order, 2 ... e6 and 3 ... c6, because 2 ... c6 is more likely to prompt 3 ttJf3. In this way he won't have to face the Marshall gambit, 3 ttJc3 e6 4 e4. But another Black may prefer 2 ... e6 3 ttJc3 because it rules out the ttJbd2 option in the Semi-Slav.
The 3 ttJc3 order deprives White of 4 it.g5 and discourages 4 cxd5 because 4 ... exd5! and ... .tf5 equalizes. Black risks Marshall's 4 e4 but can try to bail out with 4 •.. it.b4 and then 5 exd5 cxd5 6 ttJf3 into a
148
Double QP Openings
relatively benign Panov Caro-Kann. However, 5 e5 or 5 'i'g4 are more of a challenge, e.g. (5 'i'g4 tt:Jf6 6 'i'xg7 l:tg8 7 'i'h6 dxe4 8 a3 i.xc3+ 9 bxc3 e5 as in GelashviliIzoria, Athens 2005). The chief alternatives at move four are a little-explored Catalan, 4 g3, and keeping the situation fluid with 4 tt:Jf3 or 4 e3. The flexibility of the Three Pawns Variation is illustrated by Black's options after 4 tt:JO: (a) 4... fS creates a Stonewall. Black may feel happy because he won't face the tt:Jh3-f4 and tt:Jbd2 maneuvers that are dangerous in normal Dutch orders. But White hasn't ruled out 5 i.f4!, which favors him.
This is White's way of saying, 'I'm guessing you don't want a Botvinnik because if you did you would have played 4 ... tt:Jf6 and then 5 i.g5 dxc4. But now you've got a problem if you want to avoid the Botvinnik. '
He means S..•f6 when 6 i.d2 cS 7 e3! is a dangerous gambit based on the weakened e6, e.g, 7... tt:Jc6 8 i.xc4! cxd4 9 exd4 tt:Jxd4 10 O-O!. For example, 1O ... tt:Jxf3+ 11 'i'xf3 'i'xd2 12 l:tadl 'i'xb2 13 tt:Jb5 or (b) 4•..tt:Jf6 invites a Semi-Slav (5 e3 tt:Jbd7) or a QGD (5 i.g5). 12 ... 'i'g5 13 .l:.fel i.e7 14 tt:Jb5 Black might expect the latter 'i'c5 15 .:tel, Glek-An. Bykhovsky, because if White wanted a Semi- Porto San Giorgio 1999. Slav he could have sought it A more conservative White will directly with 4 e3, without having prefer 4 e3. That rules out the to deal with 4 tt:Jf3 dxc4. Noteboom, Botvinnik, et al and invites a Semi-Slav, 4 ... tt:Jf6 5 tt:Jc3. (c) 4.•. dxc4 steers White towards Black can keep his cards concealed a Noteboom (5 a4 i..b4). The a bit longer with 4...tt:Jd7. Noteboom is so theory-heavy that He is hinting that on S tt:JO he'll only the brave and the booked up will play 4 tt:Jf3 confidently build an excellent Stonewall (S .•. fS!). Then 6 i.e2 tt:Jgf6 7 0-0 unless ... i.d6 is the kind of aggressive Unless White intends to meet position Black can only dream about in a Dutch Defense move 4 tt:JO dxc4 with S i..gS!?: 149
Double QP Openings
order, e.g. 8 a3 a5 9 l::tb1 fie7 10 i.d2 0-0 11 b4 ttJe4 12 bxa5 ':f6! 13 i.e1 l1h6 13 g3 g5 with a fine game, Polak-Stangl, Brunn 1993.
White has extra options, such as 8 .lif3 and ttJge2/a2-a4 in that line, or 7 .lid3 .lib7 8 ttJf3 a6 9 0-0 c5 10 .lie4 but they aren't very impressive.
White can counter with the finesse S i.d3, so that S... fS 6 cxdS! exdS? 7 i.xfS hangs a pawn and 6... cxd5 is an Exchange Slav in which Black's QN is misplaced. As a result, Black usually plays 5 ... ttJgf6 instead and 6 ttJf3 gets White to the Semi-Slav he wanted.
His best in the diagram may be simply to transpose to a Reynolds Variation of the Meran with 7 .lid3 .lib7 8 ttJf3 a69 e4! and then 9... c5 10 d5 c4 11 .lic2 ttJgf6!' But the super-sharp Reynolds isn't for everyone.
Alexey Dreev and Yevgeny Sveshnikov have met 5 .lid3 with a counter-finesse, S... dxc4 6 i.xc4
sidesteps. The other version occurs when White's KN comes out first, 1 d4 dS 2 c4 c6 3 ttJf3 e6.
bS.
Clearly the Three Pawns holds a huge number of feints and
After 4 e3 ttJd7 Black can again playa quasi-Meran without fear of e3-e4-e5, e.g. S .lid3 dxc4! 6 .lixc4 bS 7 .lid3 a6 and 8 0-0 ttJgf6 is a harmless line of the Meran.
The thematic idea is 8 ttJc3 i..b7 9 e4. But 9... cS 10 eS has less point when there's no knight on f6. And once again a Reynolds-like 10 d5 offers little following 10 ... exd5 This is an accelerated Meran 11 ttJxd5 ttJgf6, e.g. 12 ttJxf6+ with little experience to judge it. fixf6 13 0-0 .lid6 14 a4 c4 15 .lic2 For example, 7 i.e2 .lib7 8 ttJf3 a6 0-016 i.g5 fie617 ttJd4 fie5 18 f4 9 0-0 cS is a fine version of the fixd4+! 19 fixd4 .lic5, as in JelenMeran (10 d5 exd5 11 ttJxd5 ttJgf6 Sveshnikov, Nova Gorica 1998. 12 e4!? ttJxd5 13 exd5 i.d6, The major extra option White Scherbakov-Sveshnikov, Rostov enjoys in this order is 4 .ligS. 1993). 150
Double QP Openings
But the S liJbd2 order has two advantages based on being able to retake on c4 with the knight. First, he doesn't allow a Meran since S... liJbd7 6 ~d3 dxc4 7 liJxc4 bS?! 8 liJceS liJxeS 9 liJxeS ~b7 10 0-0 ~d6 11 f4 0-0 12 'iVf3, LitinskayaArakhamia, Womens Candidates 1988, is excellent. Then 4...liJf6 is a QGD. Black could prefer 4•.. ~e7 5 ~xe7 'ifxe7. Trading off the good bishop seems dubious but some might be happy to face 6 e3 liJf6 7 liJc3 0-0 8 :c 1 liJbd7 9 .te2 dxc4 10 .txc4 eS. SEMI-SLAV DEFENSE
Second, he doesn't allow Black's best version of Mikhail Tchigorin's system, 6... ~d6 7 e4 dxc4 (8 ..ixc4 eSt) because 8liJxc4! stops ... eS. Alekhine thought this trick was so strong that Black had to meet an early S liJbd2 with S... cS!? That's why White often inverts the order today - S ~d3 and if S... liJbd7 then 6 liJbd2. His reasoning is that Black's QN is misplaced at d7 in case of a later ... cS.
The Semi-Slav arises via the QGD, Slav or Three Pawns orders. White's first finesse is developing his QN on d2 instead of c3, as in Then Black's best may be to fall 1 d4 d5 2 c4 c6 3 liJf3 liJf6 4 e3 e6 back on the QGA in which White is and now 5 liJbd2. committed to a quiet liJbd2, as This allows him to reach Kramnik did in the 2006 world favorable book positions if Black championship match, S... dxc4 continues S... liJbd7 6 .td3 ~d6 6 ~xc4 cS. 7 e4! dxe4 8liJxe4. That's the same position he gets from S liJc3 liJbd7 Slightly different is another anti6 ~d3 .td6 7 e4!. Meran order, 5 ~d3liJbd7 6 0-0.
151
Double QP Openings
This employs psychology: Meran players usually don't want a QGA position, such as 6 ... dxc4 7 .Jixc4 c5, even if it's their best option. Instead, they will be attracted to the Meran-like 7... b5 8 .Jid3 a6 and take their chances after 9 a4. But this trickery backfires if Black adopts the Tchigorin plan of 6..•.Jid6!. Then 7 tDbd2 0-0 8 e4 e5 is equal. And on 7 tDc3 0-0 we get a that occurs more position commonly via 1 d4 d5 2 c4 e6 3 tDc3 tDf6 4 tDfJ c6 5 e3 tDbd7 6 .Jid3 .Jid6 when White plays the inexact 7 O-O?! and Black replies 7 ... 0-0. The difference is that now 8 e4 dxc4! 9 .Jixc4 e5 is good because Black is castled. However, if White had played 7 e4! In the 5 .Jid3 Nbd7 6 tDc3 .Jid6 order Black cannot respond 7 ... dxc4 8 .Jixc4 e5? 9 dxe5 tDxe5 10 tDxe5 .Jixe5 in view of 11 'i'xd8+ and .Jixfl with advantage.
13 gxh8('i'). His only good alternative is a highly complex variation (10 tDa4 c5 11 e5 tDd5) that may require him to sack at least a pawn to have any hope of advantage. For that reason many a White will avoid the issue by playing 9 0-0. Then 9 ... b4 10 tDe4 gets all the attention in books and the verdict is mixed. But 9.•• a6!? is the trickster's choice.
What has happened is Black reached a normal Meran (8 ... a6) in which the second-best 9 0-0 has been played and met by 9 ... .Jib7. This is a good weapon against a White who doesn't like the Reynolds Variation (10 e4 c5 11 d5). The point is that 10 e4 c5 11 e5 is like the old main Meran line but with 0-0 and ....Jib7 added.
In the main Meran line, 1 d4 d5 2 c4 c6 3 tDf3 tDf6 4 e3 e6 5 tDc3 tDbd7 6 .Jid3 dxc4 7 .Jixc4 b5 8 .Jid3, there is a deceptive order beginning with 8....Jib7 instead of This is a fairly new position and 8... a6. The natural 9 e4, which is the so far chances appear mixed after right move against 8 ... a6, is 11...cxd4 12 tDxb5 .JixfJ or answered by 9 ...b4!. 11...tDd5 12 a4 b4 13 tDe4 cxd4 Then White can enter a four-'ii 14 .Jig 5 'i'b8 (Dambachermiddlegame (!) after 10 e5 bxc3 Stellwagen, Dutch Team Champ11 exf6 cxb2 12 fxg7 bxal('ii) ionship 2006). 152
Chapter Six: Indians The most popular answer to 1 d4 today is 1...tbf6 and the most popular reply is 2 c4, opening the door to the Hypermodern and related defenses we call Indians. We tend to overlook 2 tbf3 because it usually 'just transposes.'
makes sense for a White who only likes the Tromp when he's allowed to double pawns, i.xf6, or to build a center with f2-f3, after 2... tbe4. For the Tromp player who has no interest in transposing to another opening, this is a critical position:
But it allows White to sidestep the Benko Gambit, Budapest Defense and a few other systems. In exchange, Black can enter a King's Indian without fear of the Samisch Variation or a Modern Benoni without concern about f2-f4 systems. Or he can offer a QGD (2 ... d5) in which White is committed to tbf3. Books also underestimate 2 ..tg5 but it becomes a significantly stronger weapon in the hands of a flexible thinker who can shift into a Torre Attack or a QGD when appropriate. For example, after 2 ... e6 White can back out of the normal Trompowsky (3 e3 or 3 e4) and enter the Torre with 3 tbf3. This
Theory claims Black can equalize with 4.. :iVb6. If that's true, how can White avoid this and how can Black obtain it? One route is 1 d4 tbf6 2 ..tg5 c5 and then 3 d5 tbe4 4 i.f4. But Black has to be concerned about 3 i.xf6 and 3 tbc3!? Both are dangerous and have their own body of must-memorize book.
153
Indians
More precise is 1 d4liJf6 2 i.g5 liJe4 and then 3 i.f4 c5 4 d5. Black's chances of reaching the last diagram are much greater because the B retreats to h4less often in this order. Moving on, let's suppose 4.. :ilVb6 is met by 5 i.el!. (Tromp players do things like that, they really do.) White is hoping to build a center with f2-f3 and e2-e4 that justifies his loss of time. But after 5 ••. e6 he either has to play a doubtful gambit, 6 c4 "ifb4+, or go into 6 f3 liJf6. The latter often leads to 7 c4 exd5 8 cxd5 c4:
5 dxc5!? Then 5.. :iia5+ 6 'i'd2!? 'i'xc5 7liJc3 favors him, e.g. 7... d5 8 0-0-0 i.e6 9 e4 or 7... d6 8 e4 a6 9 0-0-0, Berg-Ilincic, Budapest 2006. But Black has a course correction in 4•••'i'a5+! 5 c3liJf6. Then 6 dxc5 'i'xc5 7 e4 e5! is fine for him. White's best is 6 d5, after which 6 ... 'i'b6! 7 i.el (not 7 'i'd2liJxd5!) e6 8 c4 exd5 9 cxd5 c4 leaves us back at the diagram. One last point about the Trompowsky: After 1 d4 liJf6 2 i.g5 liJe4 3 i.f4 d5, rather than 3... c5, White can land in a favorable version of the Blackmar-Diemer Gambit, of all things. The Blackmar is dubious in its basic form, 1 d4 d5 2 e4 dxe4 3 liJc3 liJf6 4 f3 exf3 5 liJxf3. But with an extra tempo, 1 d4 liJf6 2 i.g5 liJe4 3 i.f4 d5 4 f3 liJf6 5 e4!? dxe4 6liJc3! exf3 7liJxf3:
This is significant for two reasons. First it's considered by some Trompmeisters to be Black's best defense. He gets good play after 9 e4 i.c5 10 liJh3 d6 11 liJd2 i.xh3 or 9 e3 i.c5 10 ~f2 0-0 11 i.xc4 l1e8. Second, it's another position that Black can bring about through misdirection. Suppose White tries to avoid it by means of 1 d4 liJf6 2 i.g5 liJe4 3 i.f4 c5 and now 4 f3 liJf6
It becomes downright respect-
able. Even Garry Kasparov played this (in a simul) - 7... i.g4
154
Indians
8 h3 i.xf3 9 'iixf3 c6 10 0-0-0 e6 11 i.c4 lLlbd7? 12 d5! cxd5 13 lLlxd5!, with a terrific attack in Kasparov-Cameiro, Sao Paolo 2004. TORRE ATTACK When Black plays ... d5 in the Tromp or Torre, White usually has the option of shifting to a QGD with c2-c4, and that's often good psychology against a 1...lLlf6 player.
3 f4 h6 4 i.xf6 'iVxf6 5 lLlc3 d6, White gets his best play from f2-f4, e.g. 6 'iVd2 c6 7 f4!. This means the Torre doesn't transpose well into a Tromp and that shift should be reserved for use against a Black who knows only the 1 d4 lLlf6 2 i.g5 lLle4 and 2 ... c5 lines of the Tromp. To minimize the danger of ... h6 White can try the delayed Torre, 1 d4 lLlf6 2 lLlf3 e6 3 c3, a favorite of Tony Miles.
For example, when Wolfgang Unzicker first faced the normal Torre, 1 d4 lLlf6 2 lLlf3 e6 3 i.g5 he responded 3 ... d5. His opponent, Tigran Petrosian, knew that Unzicker preferred the Slav to a normal QGD, so he shot back 4 c4!. Unzicker was in an unfamiliar position and it showed after 4 ... c6 5 1rc2 i.e7 6 e3 0-0 7 lLlc3 h6?! 8 i.f4lLlbd7 9 cxd5 cxd5?! 10 i.d3. White waits for Black to commit Petrosian won one of his greatest himself to a system of development games. (3 ... b6, 3 ... d5, 3 ... c5) before he In the main lines of the Torre, decides on whether to play i.g5. Black doesn't occupy the center. White has a reasonable Torre after Instead he goes after White's QB, 3... d5 4 i.g5 c5 5 e3. The GM such as via 1 d4 lLlf6 2 lLlf3 e6 preference, 3 ... b6, is also fairly 3 i.g5 h6. Then 4 i.xf6 'iixf6 5 e4 good for him after 4 i.g5. d6 transposes to a Trompowsky in which White has played lLlf3 For example, 4 ... i.b7 5lLlbd2 d5 instead of the more enterprising 6 e3 c5 7 i.d3 lLlbd7 8 0-0 ~c7 lLlc3. 9 i.xf6 lLlxf6 10 lLle5 and 11 f4 That's important because in the Tromp version, 1 d4lLlf6 2 i.g5 e6
gave White an aggressive setup in Timman-Tiviakov, Moscow 1993.
155
Indians
BENKO GAMBIT
can ignore 6 ... c5. Instead of 7 d5
The ... b5 sacrifice in a Benoni pawn formation existed long before the Benko Gambit, and there are some players who offer it only after starting with a King's Indian Defense or Benoni. In this way they benefit by knowing White's setup before risking a pawn. For example, 1 d4 l2Jf6 2 c4 g6 3 l2Jc3 JL.g7 4 e4 d6 S h3 0-0 6 l2Jf3 and 6... cS 7 dS.
b5!? White should castle since 7 0-0 cxd4?! lands Black in an inferior version of the Maroczy Bind. Another idea is to play ... a6 before ... c5, as in 1 d4 l2Jf6 2 c4 a6!? and 3 l2Jc3 cS. Then 4 dS bS! S cxbS g6 is a book Benko. There's an added benefit to this order compared with the normal Benko, 1 d4 l2Jf6 2 c4 c5. Then White can obtain a promising English Opening after 3 l2Jf3 cxd4 4 l2Jxd4 and if 4 ... a6 then 5 g3! d5 6 JL.g2. But after 1 d4l2Jf6 2 c4 a6 3l2Jc3 cS:
Now 7... bS!? 8 cxbS a6 is a nice version of the Benko for Black, e.g. 9 bxa6 'ifa5 10 l2Jd2 JL.xa6 11 JL.xa6 'ifxa6 12 'iVe2 e6, GheorghiuNemet, Suhr 1990. A similar case is I d4 l2Jf6 2 c4 c5 3 d5 g6 4 l2Jc3 d6 5 e4 JL.g7 6 l2Jf3 0-0 7 h3 a6 8 JL.d3 b5!, e.g. 9 cxb5 axb5 10 l2Jxb5 l2Ja6! and 11 ... l2Jb4.
The English that arises after 4 l2Jf3 cxd4 S l2Jxd4 is excellent for But in many other sequences Black in view of S•.. dS!. For White has a favorable way of example, 6 cxd5 l2Jxd5 7 l2Jdb5 declining or evading the gambit. In l2Jb4! or 7 'ifb3?! l2Jxc3 8 bxc3 e5! the Classical Variation King's 9 l2Jc2 l2Jd7 and ... l2Jc5 favored Indian, 1 d4 l2Jf6 2 c4 g6 3 l2Jc3 Black in Filgueira-D.Gurevich, JL.g7 4 e4 d6 5 l2Jf3 0-0 6 JL.e2 he Buenos Aires 2003. 156
Indians
However, there's always a dxc5 problem when ... a6 is played before ... c5. White gets the upper hand after 1 d4 tiJf6 2 c4 a6 3 tiJc3 c5 4 dxc5! and then 4 ... 'i!Va5 5 a3 'i!Vxc5 6 e4 d6 7 .lte3 or 4 ... e6 5 e4 .ltxc5 6 e5. A slightly different version is 1 d4 tiJf6 2 tiJf3 a6, used by Miles and Lev Alburt. Then 3 c4 c5 4 d5 b5 is a gambit-less Benko. This is mainly a psychological weapon, designed to upset a White who played 2 tiJf3 specifically to avoid the Benko. BENKO PROPER 'Benko doesn't know how to play the Benko,' said another expert on the opening, John Fedorowicz. He was referring to Pal Benko's reliance on the original move order (1 d4 tiJf6 2 c4 c5 3 d5 b5 4 cxb5 a6 5 bxa6 and now 5....ltxa6) long after the merits of 5 .•. g6! were recognized.
Black can retake the pawn later, when he knows whether it's better
to have a bishop or knight on a6. That's important when, for example, White fianchettos his QB, 5 ....ltxa6 6 g3 d6 7 .ltg2 g6 8 b3!. He has good chances of making the extra pawn count, e.g. 8 ... .ltg7 9 .ltb2 lLlbd7 10 lLlh3 0-0 11 0-0 tiJb6 12 a4. But after 5 ... g6! White can't play 6 g3 d6 7 .ltg2 .ltg7 8 b3?? lLlfd7. Fedorowicz pointed out that even if White puts his queenside house in order first, 6 b3 .ltg7 7 .ltb2 0-0 8 g3, Black can use the flexibility of his fifth move with 8 ••• lLlxa6! 9 .ltg2 .ltb7 and then 10 lLlh3 e6! to undermine d5. Then 11 tiJf4 tiJxd5 12 .ltxg7 cJi;xg7 13 tiJxd5 .ltxd5 14 .ltxd5 exd5 15 'i!Vxd5?? "iff6 costs material. The trickster's answer to the Benko begins with 4 tiJf3. He knows that a devoted Benko-ite doesn't want the queenside tension to be resolved by 4 ... bxc4 or 4 ... h4. White has a tiny positional edge, rather than a material one, in those lines, but it's often enough, e.g. 4 ... bxc4 5 lLlc3 d6 6 tiJd2! g6 7 tiJxc4 .ltg7 8 e4 0-0 9 .lte2 .lta6 10 0-0 lLlbd7 11 .ltd2 lLlb6 12 b3 tiJfd7 13 tiJa5! and lLlc6 as in Rashkovsky-Meshkov, Moscow 2002. The point of 4 tiJf3 is revealed when White accepts the gambit on the next move, 4 ••. g6 5 cxb5! a6 and 6 b6!?
157
Indians
12 ttJxe5 ltxb7 13 \i'a4+! ttJbd7 14 ttJc6 or 13 ... ttJfd7 14 ttJc6 'ifb6 15 i.f4!, Milov-De Vreugt, Ohrid 2001.
Then 6...'iWxb6 7 ttJc3 d6 8 e4 i.g7 is promising after 9 i.e2 i.g4 10 e5! or 9... 0-0 10 ttJd2 ttJbd7 11 ttJc4 'iic7 12 0-0 ttJb6 13 ltbl, Burmakin-Ostertag, Bad Worishofen 2006.
If White has doubts about 6 b6 he can still use 4 ttJf3 g6 5 cxb5 a6 with the idea of 6 e3!? This looks like a line of the Benko that was all the rage in the 1980s, 1 d4 ttJf6 2 c4 c5 3 d5 b5 4 cxb5 a6 5 e3, until Black found a good antidote in an attack on the d-pawn, 5... axb5 6 i.xb5 \i'a5+ 7 ttJc3 i.b7!. But in the 4 ttJf3 g6 5 cxb5 a6 6 e3 order:
White's point is that these positions more commonly come about via 4 cxb5 a6 5 b6 and then 5.. .'iixb6 6 ttJc3 d6 7 e4 g6 8 ttJf3 iLg7. But that order gives Black a very promising alternative in 5... e6!? with good center play. This option isn't possible after 4 ttJf3 g6 5 cxb5 a6 6 b6. Black gets little out of 6... axb5 Another benefit is that Black i.xb5 \i'a5+ 8 ttJc3 i.b7? 7 often replies 6... a5?!. He does this out of habit - because it's a good because White can simply castle move in the comparable 4 cxb5 a6 and keep his extra pawn (9 0-0 i.g7 5 b6 a5 position. Then Black can 10 a4 0-0 11 e4). playa quick ... iLa6 and decide later Instead, Black should play whether to retake on b6 with the 8...i.a6! 9 iLxa6 \i'xa6. Then 10 e4 queen or QN. d6 11 e5 needs more testing (11...dxe5 12 ttJxe5 i.g7 13 \i'd3, But in the delayed form White Balgojevic-S.Kasparov, Curto gets the upper hand from 7 ttJc3 2006). iLa6 and now 8 b7! i.xb7 9 e4 or 8... %:ta7 9 e4 iLxfl 10 ~xfl d6 And what happens if Black meets 11 e5!. For example, 11...dxe5 4 ttJf3 with 4... a6 instead of 4... g6 ? 158
Indians
This transposes to 1 d4 liJf6 2 liJf3 a6 3 c4 cS 4 dS bS. Then on 5 e3 two thematic ideas of the Benko, S... b4 and S... dxc4 make little sense because he's played ... a6.
And then 3 c4 exd5 4 cxd5 liJf6. The benefit of 1 d4 cS is this time 2 liJf3 allows Black to seize center ground with 2 ... cxd4 3 liJxd4 dS!, as Alexander Alekhine showed.
For example, S... bxc4 6 liJc3 d6 7 e4 g6 8 .ltxc4. Black has an extra . tempo compared with 4 liJf3 bxc4 S liJc3 d6 6 e4 g6 7 .ltxc4. But the extra move is ... a6, which is more harmful than useful.
But this order has a major minus since White hasn't committed himself to c2-c4. He can play a more open game with 3 liJc3. Then 3...liJf64 e4 exd5? 5 e5! favors him considerably and 4 ... d6 S .ltbS+ and 6 dxe6 less so.
A better irregular route to the Modern Benoni is 1 d4 e6 2 c4 c5 A Benoni is distinguished by a 3 d5. This avoids the Trompowsky White pawn on dS and a Black one and makes sure c2-c4 is played at cS. There are several kinds of before Black commits to ... cS. Benoni depending on whether Another benefit is evident after White supports dS with c2-c4 and 3 ... exd5 4 cxd5 d6 5 liJc3 g6 6 e4 whether Black tries to liquidate it .ltg7: with ... e6. THE BENONI FAMILY
Among the irregular versions is one that arose in the 1970s when White frequently met I d4 liJf6 2 c4 cS with 3 liJf3. Black began to use an anti-English order, 1 d4 c5, so that he can transpose to a Modern Benoni after 2 d5 e6.
Black can transpose into main lines after 7... liJf6.But in case of 7 f4 it's a bad decision, as 7...liJf6 8 .ltb5+! is known to favor White. Instead 7.. Ji'h4+! 8 g3 'ile7 has stood up well in tests (9 .ltg2 liJf6 10 liJge2 0-0 11 0-0 liJe8 12 .lte3 159
Indians
tiJd7 13 'iVd2 :b8 14 a4 a6 15 :abl f5, Skembris-Hamdouchi, Bastia 2002).
such as 1 d4 tiJf6 2 tiJo c5 3 d5 d6 and 1 d4 c5 2 d5 d6 3 e4:
Another virtue of this order is that on 7 tiJo Black has time for 7•.• a68 a4 .tg4. This is a common Benoni theme designed to gain control of e5. If Black had transposed with 7 ...tiJf6 White could avoid ....tg4xf3 by means of 8 tiJd2. But there is an obvious demerit to 1 d4 e6 2 c4 c5. Black has to have an answer to 2 e4.
Black often continues 3... g6 followed by ... i..g7 and ... tiJf6. But 3 ...tiJf6 4 tiJc3 g6 is more accurate.
SCHMID BENONI
This avoids 3... g6 4 tiJf3 .tg7 5 .tb5+ tiJd7 6 0-0 tiJf6 7 l1e 1 This became fashionable, in the when White can profit from order 1 d4 c5 2 d5 d6 3 tiJc3 g6, delaying tiJc3, e.g. 7... 0-0 8 a4! a6 after Lothar Schmid scored 9 .tf! b6 10 h3 tiJe8 11 tiJa3! and impressive wins with it in a 1954 tiJc4/e4-e5 with advantage, as correspondence tournament. For Mikhail Tal showed. example, 4 e4 .tg7 5 tiJf3 tiJf6 The bishop check is a recurring 6 .te2 tiJa6 7 0-0 tiJc7 8 tiJd2. By problem in Schmid lines. To avoid delaying ... 0-0 he had time for 8... a6 it Black can tweak the order, 1 d4 9 a4 .td7 10 tiJc4 b5 11 tiJb6 b4! c5 2 d5 tiJf6 3 tiJc3 g6 and then 12 tiJxa8 'iixa8 13 tiJbl tiJxe4 with 4 tiJo .tg7 5 e4 0-0. advantage in Hayes-Schmid. This also enables him to A benefit of the Schmid order is transpose into more normal Schmid that if White chooses 3 c4 instead of lines with 6 ... d6. Or he could go 3 tiJc3, Black has his choice of into sharp lines such as 6 i..e2 b5!? playing a KIDlBenoni hybrid or (7 e5 tiJg4 8 i..f4 b4! 9 tiJe4 d6). transposing to a Modem Benoni The dangers in this order come with 3 ... g6 4 tiJc3 .tg7 5 e4 tiJf6 from an advance of White pawns 6 i..e2 0-07 tiJf3 e6 and ... exd5, as such as the immediate 4 e4 i..g7 Schmid did. 5 e5!? or 4 tiJo i..g7 5 e4 0-0 6 e5 Another benefit is that Black can tiJg4 7 tiJg5!? and 6 d6 'iWb6 7 i..f4 experiment with different orders, 'iixb2. 160
Indians
KID/BENONI This hybrid is similar to the Schmid Benoni except that White is committed to c2-c4. This is usually due to the order 1 d4 ttJf6 2 c4 c5 3 d5 d6. The finesses generally concern what Black does with his e-pawn. For example, 4 ttJc3 g6 S e4 ..tg7 and 6 ttJf3 0-0 7 ..te2 e6 80-0. Now 8... exd5 9 cxd5 transposes into a very familiar Modem Benoni position that is rated somewhere between equal and plus-over-equal. But Black can delay further with 8...:e8!?
That leaves 9 ttJd2, which defends e4 and can reach a Modem Benoni tabia (9 ... exd5 10 cxd5). Black may feel happy that in this way he avoids the ~c2 line. That is, 1 d4 ttJf6 2 c4 c5 3 d5 e6 4 ttJc3 exd5 5 cxd5 d6 6 e4 g6 7 ttJf3 ..tg7 8 ..te2 0-0 9 0-0 lte8 and now 10 ~c2 rather than 10 ttJd2. But Black can delay further, 9 ttJd2 ttJa6!. He can spend the next few moves preparing ... b5 with ... ttJc7, ... a6, ... l1b8 while White doesn't have equally useful moves. Once again 10 dxe6? is dubious. After 10... ..txe6 11 f4 ..td7! White's vulnerable e-pawn becomes the main issue. e.g. 12 ..tf3 ..tc6 13 ttJd5 ttJb4 and 13 l1e 1 'iVb6 14 ttJb3 l:!ad8 15 ..td2 ttJb4 (Ardiansyah-Schmidt, Indonesia 1983) favors Black. The delayed Modem Benoni is still largely unexplored and the indications are that it's worth exploring.
This is more than a waiting move because he is preparing 9... exd5 10 cxd5? ttJxe4. On quiet moves such as 9 h3, Black gets a fine game with 9... exdS 9 exdS ..tfS. And if White tries to exploit the delay in ... exd5, by means of9 dxe6 ..txe6 10 ..tf4, Black has adequate compensation for the lost pawn after 10...ttJc6! 11 ..txd6 ttJd4. Or so experience indicates.
MODERN BENONI The Modem Benoni comes about from two main orders, 1 d4 ttJf6 2 c4 c5 3 d5 e6 followed by ... exd5/... d6 and ... g6, and 2 ... e6 and 3... c5 (4 d5 exd5 5 cxd5 d6). In both cases White can shift into the English Opening with ttJf3 instead of pushing d4-d5. A virtue of 1 d4 ttJf6 2 c4 cS is that it grants
161
Indians
Black the extra option of meeting 3 liJf3 with 3... cxd4 4 liJxd5 e5!?, one of the most enterprising antiEnglish lines.
i.a6 15 'iVb3! exf4 16 i.xf4 i.xc3 17 'i!Vxc3 'i!Vxe2 18 ltae1 with a powerful attack.
Objectively, a good solution to To avoid that White may prefer White's order is 6... d5!, a better3 e3. A BenkolBenoni specialist can than-usual Gruenfeld. But to play it become confused after 3... g6 4 liJc3 Black has to know complex theory i.g7 5 liJf3 0-0 6 i.e2. such as 7 dxc5 'ifa5 8 cxd5 ttJxd5!. That helps make 3 e3 a more credible weapon against a Benko/ Benoni-booked player. Today you're more likely to see the other Modem Benoni order, 1 d4 liJf6 2 c4 e6 and 3 ... c5 because it contains the Nimzothreat (3 liJc3 i.b4!). That often prompts 3 ttJf3 and enables Black, after 3 ... c5 4 d5, to avoid the Now 6... cxd4 7 exd4 d5 is a aggressive Modem Benoni lines in reversed Tarrasch QGD with an extra move for White. That sounds which White plays f2-f4 or f2-f3. more promising than it really If Black is also happy with is, e.g. 8 0-0 liJc6 9 i.e3 dxc4 Symmetrical English positions such 10 i.xc4 liJa5 with equal chances. as 3 liJf3 c5 4 liJc3 cxd4 5 ttJxd4, this is his most pleasant route to the Instead, a more Indian-like Modem Benoni. approach, 6... d6 7 0-0 b6?! 8 e4!, If White tries 1 d4 liJf6 2 c4 e6 drifts toward another poor Maroczy Bind (8 ... cxd4 9 liJxd4 and 10 f3) 3liJf3 c5 4 e3: or a Benoni or KID in which ... b6/ ... i.b7 only blocks Black's queenside counterplay. For example, 8... i.b7 9 d5! e5 10 liJe 1 liJe8 11 liJd3 suddenly becomes a dubious King's Indian. This was evident in BisguierKavalek, Tel Aviv 1964 after 11...f5 12 exf5 gxf5 13 f4 ~e7 14 i.e3 162
Indians
This is a tacit invitation to a Semi-Tarrasch QGD (4 ... d5). A true Benoni player will prefer Mihai Suba's 4... a6 5 tiJc3 "iic7 followed by ... d6/ ... tiJbd7 and eventually ... g6/ ... iLg7. All of those moves are useful to
7... iLd7 8 'Yib3 is nice for him and 7 ... tiJbd7 8 tiJfd2! imperils the g4-bishop, e.g. 8... a6 9 h3 b5 10 "iic2 iLh5 11 a4! b4 12 tiJc4 and iLf4. Or 11...iLg6 12 axb5 tiJxd5 13 bxa6 and White won in Dokhoian-Lukov, Plovdiv 1988.
Black in a Benoni. That means he'll be in great shape when White recognizes that his only chance for an edge lies in d4-d5, e.g. 6 iLe2 d6 70-0 tiJbd7 8 d5 exd5 9 cxd5 b5. Or 8 a4 g6 9 d5 exd5 10 cxd5 iLg7 11 e4 0-0 (Tukmakov-Suba,
Erevan 1980) - when Black has actually gained a move on a normal Benoni. White can try to exploit the Nirnzo-threat order, 1 d4 tiJf6 2 c4 e6 3 tiJf3 c5 4 d5 exd5 5 cxd5 d6, with 6 e4!?
Of course, Black has an easier time engineering ... b5 in this order. However, that push doesn't solve his problems in 6.•. g6 7 h3 iLg7 8 iLd3 0-0 9 0-0, e.g. 9 ... b5 10 :el! a6 11 a4 b4 12 tiJbd2 or 10 .. J~e8 11 iLxb5 :'xe4 12 tiJc3 :'xe1+ 13 "iixe 1 a6 14 iLa4 and iLf4, as in I.Sokolov - Papaioannou, Turin 2006. There's a simple counter to a delayed tiJc3. After 1 d4 tiJf6 2 c4 e6 3 tiJf3 c5 4 d5 Black should play 4... d6! rather than 4 ... exd5.
This is more than a one-move trap (6 ...tiJxe4?? 7 'iVa4+). White is holding his QN back until he knows whether it belongs on c3, a3 or d2.
White can't play 5 e4? tiJxe4 this time and 5 tiJc3 allows Black to transpose, with 5... exd5 6 cxd5, to normal lines.
He also benefits from tactics such as 6...iLg4 7 'iVa4+!. Then
This also gives White reason to think about 6 tiJxd5 tiJxd5 7 'iVxd5.
163
Indians
8 tiJd2) in which Black has been finessed into a ... tiJbd7 line, rather than one of the more popular ... tiJa6 or ... i.g4 variations. There's no downside to White's order - or at least we haven't discovered it yet.
But the backward d-pawn is not the weakling it seems after 7•..tiJc6, e.g. 8 e4 i.e6 9 "ii'dl i.e7 or 8 tiJg5 'ike7 8 'iVe4 h6 as in the forgotten game Bandal-Fischer, Meralco 1967.
In the Modem Benoni's main OlD-in to-KID line, 1 d4 tiJf6 2 c4 cS 3 dS e6 Some people actually like the 4 tiJc3 exdS S cxdS d6 6 tiJf3 g6, Old Indian. After 1 d4 tiJf6 2 c4 the move 7 tiJd2 was considered a virtual refutation when the opening they voluntarily play 2... d6 and was introduced in Nimzovich- 3... eS or 3••.tiJbd7 and 4... eS. But Marshall, New York 1927, e.g. others use the Old Indian order as a 7 ... i.g7 8 tiJc4 and 9 i.f4 or safe path into a King's Indian. The 7... tiJa6? 8 tiJc4 tiJc7 9 a4 i.g7 early ... e5 usually discourages White from sharper lines, such as 10 i.f4!. the Samisch Variation. That shrinks But the Benoni was rescued by the amount of theory Black has to the realization that 7 ... tiJbd7 8 tiJc4 know. tiJb6 is a good response. Some This finesse is a century old. annotators gave '7 tiJd2?!' and Janowsky-Tchigorin, Ostende 1907 '7 ... tiJbd7!' when those moves occurred in the 1972 Spassky- went 1 d4 tiJf6 2 c4 d6 3 tiJc3 Fischer match. Nevertheless, 7 tiJd2 tiJbd7 4 tiJf3 eS S e3. has transpositional value because after 7...tiJbd7:
White can play 8 e4!. Then he's reached the Classical Variation (usually occurring via 7 e4 i.g7
There followed S.•.i.e7 6 i.d3 and Black soon got a bad game (6 ... c6 7 0-0 'ikc7 8 i.d2 0-0 9 l:tc1 l:te8 10 tiJg5 tiJf8?! 11 'iVc2 h6 12 tiJge4 i.e6 13 d5! i.c8 14 f4!).
164
Indians
But when the diagram arose in another Ostende 1907 game, Black (Aron Nimzovich) deviated with 5... g6!. He tricked White into one of the most innocuous of KID variations and equalized after 6 .te2 .tg7 7 0-0 0-0 8 b3 :e8.
For instance, 7... exd4 8 tiJxd4 0-0 9 'iYd2 and O-O-O! or 7... h6 8 .te3 0-0 9 'i'id2 'it>h7 10 d5! a5 11 g4! tiJg8 12 h4 tiJc5? 13 h5 g5 14 tiJg3 with a big positional plus (Kohlmeyer-Lymar, Balaguer 2006).
White can insist on playing a Samisch setup, after say 4 e4 e5 5 d5 tiJc5 6 fJ. But the point of this order is revealed when Black changes his mind about the KID. After 6••• a5 7 .te3 .te7! was played in the 1954 USSR Championship, the tournament book said this was 'an openmg accuracy even M.l. Tchigorin knew. '
A sly move like 5 tiJge2 deserves a sly reply, 5... c6. Then 6 f3 .te7! transposes to a pretty good Old Indian, since 7 .te3 0-0 8 'iYd2 allows 8... d5! (9 cxd5 cxd5 10 exd5 exd4 11 .txd4 tiJb6). Or 8 tiJg3 d5! 9 cxd5 cxd5 10 exd5? .tb4 with an edge. The traditional counter to 1 d4 tiJf6 2 c4 d6 3 tiJc3 tiJbd7 is 4 tiJfJ but few players realize why. It stops 4... e5 5 .tg5 g6? because of 6 dxe5 dxe5 7 tiJxe5!. If instead Black chooses an Old Indian (5 ... .te7), it's an OlD that is quite good for White after 6 e3 and .te2!'iic2. Black's counter-finesse to 4 tiJf3 is simply 4.•. g6.
By avoiding ... g6 there is no object for the h2-h4-h5 attack. Black drew after 8 'ifd2 0-0 9 tiJge2 tiJe8 10 g4 h6! 11 0-0-0 .tg5! 12 .txg5 'ifxg5. White can be subtle, too. After 1 d4 tiJf6 2 c4 d6 3 tiJc3 tiJbd7 4 e4 e5 he can try 5 tiJge2 and wait for Black to commit his KB. Then 5... g6 6 f3 .tg7 7 .tg5! is a betterthan-usual Samisch.
Then if White continues in OlD mode (5 .tg5 .tg7 6 e3) he finds
165
Indians
himself in an offbeat KID, 6•.• h6 7 i.h4 g5 S i.g3 ttJh5, that is fairly equal. The downside of 4... g6 is that Black usually ends up, after S e4 or S g3, in a Classical or Fianchetto Variation of the KID in which he's relinquished some of his most aggressive weapons, involving For example, 7 i.e3 bxc5 ... ttJc6, and has made ... cS less S i.xc5 ttJc6 9 J:[c1 J:[bS 10 'iWd2 effective. But the upside of J:[e8! 11 ttJge2 d6 12 i.e3 e6 avoiding the Samisch, et aI, may be 13 ttJf4 d5!? 14 cxd5 exd515 exd5 more important. ttJb4 with a fierce initiative in On the other hand, if Black really Lomaya-Gufeld, Tbilisi 1961. wants to play the Old Indian, 1 d4 If White declines the gambit with ttJf6 2 c4 d6 3 ttJc3 ttJbd7 and 6 dS d6 7 i.e3 he transposes into 4 ttJf3 c6! is most precise. In that another Samisch gambit, 4... d6 S D way he can transpose to the main 0-06 i.e3 cS!? (7 dS). The counterOlD lines after S e4 eS and 6... i.e7 finesse to 4... 0-0 is S i.e3, which but he avoids 4 ... eS S i.gS! and allows White to transpose into a 6 e3. normal Samisch after S... d6 6 D. KING'S INDIAN DEFENSE After the traditional KID order, 1 d4 ttJf6 2 c4 g6 3 ttJc3 i.g7 4 e4, Black is not required to play 4... d6. With 4... 0-0 he tries to lure White into overextending himself, 5 e5 ttJeS. The pawn push hasn't recovered from Letelier-Fischer, Leipzig 1960 (6 f4 d6 7 i.e3 cS 8 dxcS ttJc6!). Aside from provocation, 4... 0-0 serves Black by creating a gambit, 5 f3 c5!? and then 6 dxc5 b6.
Black has to be careful about S i.gS because of the temptation in KIDs to kick the bishop as soon as it lands on gS. If he had played 4 ... d6 instead of castling, S... h6! (6 i.e3 ttJg4) is recommended before White can stop it with 6 'iWd2. However on 4... 0-0 5 i.g5 h6?! 6 i.e3 Black lacks 6... ttJg4. The result is an inferior Samisch, e.g. 6... d6 7 D! eS 8 dS c6 9 'iWd2 cxdS 10 cxdS ttJe8 11 0-0-0 ~h7 12 g4!, Ushenina-Siti, Kochin 2004.
166
Indians
11 0-0, e.g.ll ... a5 12 'ilVc2 "ikc7 13 l'itadl tiJc5 14 %:lfel tiJfd7 15 f4 tiJe6 16 "ikd2 tiJxd4 17 .i.xd4, Rogozenko - Dragomirescu, Timisoara 2006.
On the opposite side of the coin, Black can delay rather than accelerate castling. This gives him an extra tempo for development and denies White a kingside target, e.g. 1 d4 tiJf6 2 c4 g6 3 tiJc3 .i.g7 4 e4 d6 S f3 tiJc6!? and 6 .i.e3 a6 7 tiJge2 %:lbS.
What has happened is White reached a favorable position that usually comes about from 6 0-0 tiJbd7 7 tiJc3 e5 8 e4 c6 9 h3 %:le8 and ... exd4, a line Black gave up on decades ago. In the 6 tiJc3 order Black lost his chance to play better lines.
That tempo enables Black to meet S tiJc1 eS 9 dS tiJd4 10 tiJb3 with 10... cS! 11 dxc6 bxc6 with b-file play, e.g. 12 tiJxd4 exd4 13 .i.xd4 %:lxb2. After 9 tiJb3 Black can transpose into normal lines with 9... exd4 10 tiJxd4 0-0. White can also profit from delayed castling. In the Fianchetto Variation, 1 d4 tiJf6 2 c4 g6 3 g3 i.g7 4 .i.g2 0-0 sllJf3 d6 he can try Viktor Korchnoi's favorite 6 tiJc3 (instead of 6 0-0). Then 6...tiJbd7 7 e4 eS S h3 offers to transpose into the conventional ... c6 lines, such as 8... c6 9 .i.e3 exd4 10 tiJxd4 %:le8
The main drawback to tiJc3 in place of 0-0 is it may make ... c5 stronger, as we'll see in a few pages. But if White doesn't mind ... c5 lines, this order makes sense. KID: SAMISCH After 1 d4 tiJf6 2 c4 g6 3 tiJc3 .i.g7 4 e4 d6 S f3 0-0 Black usually chooses between breaking in the center with ... c5 or with ... e5. The first seemed harder to achieve until it was discovered in the 1980s that 6 .i.e3 c5 is a sound gambit. That told White he should consider other sixth moves, such as 6 tiJge2. It's a good waiting move since there are only a few Samisch lines in which White develops the knight elsewhere but there are several in which .i.g5 serves better than .i.e3, e.g. 6 tiJge2 eS?! 7.i.gS!.
167
Indians
On 7... c6 8 'iVd2 tiJbd7 9 dS! Black cannot easily play the thematic ... ~e8/... f5, e.g. 9 ... cxd5 10 ~xd5 or 9 ... ~b6 10 ~g3 'iVc7 11 l:.cl is awful (Dreev-Mork, Linares 1999). Well, then, how good is 6 ~ge2 cS 7 dS ? Black can seek a doubleedged Modern Benoni, 7 ... e6 8..tg5 exd5 9 cxd5. However, 8 ~g3 retains more options for White, since on 8 ... exd5 9 cxd5 his bishop may belong on f4 (e.g. 9 ... ~a6 10 ..te2 ~c7 11 0-0..td7 12 a4 a6 13 ..tf4! ~fe8 14 'ilVd2, with advantage in Khismatullin- N echepurenko, Serpukhov 2004).
Now 7 ..td3 can be answered by 7 ... cS!. Then 8 dxc5 dxc5 9 ..txc5 ~c6 is an improved version of the 6 ..te3 c5 gambit and 8 d5 e6! is a good Benoni. Black's main idea is that 7 'iVd2 can be met by 7 ... c6!. ECO considered 8 0-0-0 b5 and 8 ..th6 b5 to be equal - and among the few Byrne Variation lines that don't favor White. A similar idea was used by Tal in the 1961 world championship when he was trailing badly. He wanted to play an ... e5/... c6 line but was afraid of 5 ... 0-0 6 ..te3 e5 7 dxe5! dxe5 8 'iVxd8!, when Black has the worst of a likely draw.
Tal refined the order with 6... c6, feinting in the direction of ... a6/ ... b5. But after Botvinnik Against Robert Byrne's idea of replied 7 ..td3, he continued 7 ... e5! ... c6/ ... a6/ ... b5, a conventional view and got the opening he wanted . holds that ..td3 and kingside Before leaving the Samisch, we castling is a better policy than should mention S tiJge2. Books tell preparing a mating attack with 'iVd2 you that White's idea is 6 tiJg3 and and 0-0-0. Black can try to exploit they add that it only equalizes after S... eS 6 dS tiJa6. that via S f3 0-0 6 ..te3 a6. 168
Indians
However 6... ttJbd7 can be tested by 7 e5!? The key line was long regarded to be 7... dxe5 8 dxe5 ttJg4 9 e6 fxe6 10 0-0 ttJde5, and while experience indicates Black is equal (11 'iib3 ttJxf3+ 12 .txf3 ttJe5 13 .te2 ttJc6! and ... ttJd4), the positions that arise can be difficult to handle. That may be true after 7 ttJg3 h5 and 8... h4. But 7 f3! and 8 .tg5 reaches a better-than-usual Samisch.
The younger - and wiser sibling of 6... ttJbd7 is 6...ttJa6! and then 7 0-0 e5.
Black should ignore the books and play 5... 0-0 and then perhaps 6 .tg5 ttJbd7 7 'iid2 c5 since 8 d5 b5 9 cxb5 a6 is a good Benko (Damljanovic - Martinovic, Smederevska Palanka 1999). KID: CLASSICAL One database shows only eight examples of this played before 1988. In most cases Black will be able to transpose into 6 ... e5 7 0-0 ttJbd7 positions that are quite good for him, e.g. 8 l:te 1 exd4 9 ttJxd4 ttJc5 or 8... c6 9 d5 ttJc5! and 8 .te3 c6 9 d5 ttJg4 10 .tg5 f6 11 .td2 f5 That's a vote in favor of followed by ... ttJc5. 6...ttJbd7 and then 7 0-0 e5. This The benefits of having a knight at order also denies White the Petrosian Variation (6 ... e5 7 d5 and a6 show up when we don't .tg5 and ttJd2). White can't reach it transpose. White gets little from the after 6... ttJbd7 7 0-0 e5 because endgame, 8 dxe5 dxe5 9 'iixd8 8 d5 ttJc5 attacks the e-pawn l:.xd8, because 10 ttJd5 just loses the e-pawn and because the (9 .tg5 h6! 10 .th4? g5).
The most common Classical Variation order is 1 d4 ttJf6 2 c4 g6 3 ttJc3 .tg7 4 e4 d6 5 ttJf3 0-0 6 .te2. Then 6... e5 allows White to force a riskless, slightly favorable endgame, 7 dxe5 dxe5 8 'iixd8 l:.xd8 9 .tg5.
169
Indians
preparatory 10 ii..g5 is adequately Black can counter 6 ii..e3 with met by 10... c6 or 1O ... h6. In all of 6... c5. This is based on 7 dxc5 'i'a5 the endgame lines, tbxe5 is and 7 .te2? tbg4 8 ii..g5 cxd4 answered by ... tbxe4!. 9 tbxd4 'i'b6!. White's best may be 7 d5, reaching a double-edged Nevertheless, books recommend Benoni after 7... e6 8 h3 exd5 the orthodox 6 ... e5 7 0-0 tbc6 9 exd5. If that's the worst case, instead, with analysis that can run 6 ii..e3 may be worth White's while. 25 moves deep. If White doesn't like the mega-theory, he should KID: FIANCHETTO consider a Bent Larsen idea, 1 d4 The fianchetto order that has tbf6 2 c4 g6 3 tbc3 ii..g7 4 e4 d6 earned the greatest currency is 1 d4 5 tbf3 0-0 6 ii..e3. tbf6 2 c4 g6 3 g3 ii..g7 4 ii..g2 0-0 5 tbf3. There is more room for finesse when White delays pushing the c- or d-pawns, as in 1 d4 tbf6 2 tbf3 g6 3 g3 ii..g7 4 ii..g2 0-0 5 0-0:
Larsen's intent was to escape book (6 ... tbbd7 7 h3). But 6 ii..e3 can be used as a way to stay in book, but on another page. The reason is that when Black White can decide, after 5...d6, sees 6 .te3 he often figures 6 ... e5 whether to make it a KID, with 6 c4 7 dxe5 is a superior form of the or jump into the Pirc Defense 6 ii..e2 e5 7 dxe5 endgame. As a (6 tbc3 and 7 e4). Against a player result he prepares ... e5 with who rarely if ever adopts the Pirc, 6... tbbd7 or 6... tba6 instead. 6 tbc3 can be strong. But then 7 .te2! e5 8 0-0 allows White to transpose into normal ... tbbd7 or ... tba6 lines - having dodged the ... tbc6 mega-theory.
That was the case in LangewegVasiukov, Wijk aan Zee 1973: 6... tbbd7 7 e4 e5 8 a4 exd4 9 tbxd4 l:te8 10 a5! tbc5 11 l:tel tbg4?!
170
Indians
12 h3 4JeS 13 f44Jed7 14 .lte3 4Jf8 IS 'ti'd2 a6 16 b4! 4Jcd7 17 4Jb3 with a clear edge. But Black can hide his intentions with 5... c6!? Then on 6 4Jc3 d5! he should equalize (7 b3 .ltfS 8 iLb2 4Je4!). The ... dS idea is naturally stronger when Black hasn't spent a tempo on S... d6. If White has advanced his cpawn and not the d-pawn, as in 1 c4 4Jf62 4Jf3 g6 3 g3 .ltg7 4 .ltg2 0-0 5 0-0, then 5•.. c6 is what Tal called 'a slight psychological diversion.' Black gets a good version of the Neo-Gruenfe1d with 6 4Jc3 dS!, threatening ... d4, and then 7 cxdS cxdS 8 d4 4Je4!. White can try for a reversed version of that with 6 d4 so that his knight is the one in the center after 6 ... dS 7 cxdS cxdS 8 4JeS!. But then Black can go back to the KID with 6•••d6, transposing into a position pioneered by Lubos Kavalek.
comes about via 1 d4 4Jf6 2 c4 g6 3 g3 .ltg7 4 iLg2 0-0 5 4Jf3 d6 6 0-0 4Jbd7 7 4Jc3 eS 8 e4 c6 9 h3 'it'aS. The major difference is that Black has avoided a heavily analyzed alternative in the second order, 9 .lte3!? 4Jg4 10 .ltgS. This order also enables Black, after 7 4Jc3 'ii'aS, to attack the c-pawn (8 dS 'ii'b4). In recent years 8 h3 'ii'a6 has had some vogue. Joel Benjamin proposed a more accurate move order beginning with 7...'iib6. This transposes after 8 d5 'iVb4 or 8 h3 'it'a6. But 8 e4, which is a central line in the Kavalek order, is easier for Black to meet here with 8.•.iLg4!. Then 9 h3 .ltxf3 10 .ltxf3 4Jfd7 11 .lte3, which is played against the Kavalek, would hang the b-pawn. And 9 b3 cS! equalizes, e.g. 10 dxcS dxcS 11 .lte3 4Jc6 12 h3 :ad8, Tregubov-Izoria, Moscow 2006. Black can insert ... cS at a number of points in the first half dozen or so moves of a Fianchetto Variation. Some KID experts do it as soon as White declares his preference for g2-g3, e.g. 1 d4 4Jf6 2 c4 g6 3 g3 cS.
Then 7 4Jc3 'it'a5 and 8 h3 e5 9 e4 4Jbd7 is a main line that also
But others wait for 4Jc3, such as 1 d4 4Jf6 2 c4 g6 3 g3 iLg7 4 .ltg2 0-0 5 4Jf3 d6 6 4Jc3 and now 6.•. c5 7 d5 e6!.
171
Indians
White can try to punish Black's last move with 8 dxe6 because his pawns are loose after 8 ... fxe6 and because 8 ... .1i.xe6 is a sacrifice (9 l2Jg5 .1i.xc4 10 iLxb7).
Here 7 ... b5!? 8 cxb5 a6 transposes into the Fianchetto Variation of the Benko. That's a solid line for White but no threat to refute the Benko.
But after Korchnoi and Rafael Vaganian lost celebrated games to the sacrifice, theory said White should avoid this and merely transpose to a quiet Benoni (8 0-0 exd5 9 cxd5).
KID: .1i.g5
However, if White has played 6 0-0 instead of 6 l2Jc3, then 6 ... c5 7 d5 e6 8 dxe6! is strong because 8... iLxe6 9 l2Jg5 iLxc4 10 .1i.xb7 l2Jbd7 is rudely surprised by 11 l2Ja3!. This gave Korchnoi two easy victories when he introduced it (11...!:.b8 12 l2Jxc4 llxb7 13 l2Jxd6 with a clear extra pawn). This gives White a major reason for delaying l2Jc3. The downside is allowing Black a good version of the Benko Gambit, e.g. 1 d4 l2Jf6 2 c4 g6 3 g3 iLg7 4 iLg2 0-0 5 l2Jf3 d6 6 0-0 c5 7 d5:
In Yuri Averbakh's variation, 1 d4 l2Jf6 2 c4 g6 3 l2Jc3 .1i.g7 4 e4 d6 5 .1i.e2 0-0 6 .1i.g5, Black cannot play 6 ... e5? 7 dxe5 dxe5 8 'i'xd8 and 9 l2Jd5. His main choice is between 6 ... c5 7 d5 h6 8 .1i.e3 e6 and 6 ... h6 7 .1i.e3 c5 8 d5 e6, which transpose. Both orders have merit. Svetozar Gligoric preferred 6 ... h6 because after 7 .1i.h4 c5 8 d5 Black has a strong alternative in 8... a6 9 a4 'i'a5, since .1i.d2 is not possible. On 7 .1i.e3 e5 (or 7 ... l2Jbd7 8l2JO e5) he reaches a Classical Variation position in which the extra tempo, ... h6, is more asset than liability. Other authorities prefer 6 ... c5 7 d5 h6. But to play it Black needs to know a huge amount of theory about the pawn sacks that arise after
172
Indians
KID: FOUR PAWNS
8 i.f4 e6 9 dxe6. In contrast, 6... h6 7 i.f4 can be met by 7...t2Jc6 (8 12Jf3 12Jxd4! and 8 d5 e5!). That makes 6... h6 the mental hygiene order.
In the Four Pawns variation (1 d4 12Jf6 2 c4 g6 3 12Jc3 i.g7 4 e4 d6 S f4) Black usually challenges the center with ... c5 and White typically responds d4-d5. The timing of ... c5 doesn't matter unless White allows ... cxd4. After S••. cS 6 12Jf3!? cxd4 7 12Jxd4 12Jc6 8 i.e3:
Another i.g5 system is 5 h3 0-0 6 i.gS.
Then 6... h6 7 i.e3! and 'iVd2 favors White because Black cannot play 7... l2Jg4 and because this time ... h6 is more liability than asset, e.g. 7... c6 8 'iYd2 ~h7 9 i.d3 e5 10 d5 12Ja6 11 g4 with advantage, Paunovic-Illic, Belgrade 2004. Theory used to regard 6••• cS as good because 7 dS e6 8 i.d3 exdS 9 exdS is a somewhat lifeless Benoni. But 9 cxdS! transposes into a good one, e.g. 9... l:.e8 10 12Jf3 a6 11 a4 'iYa5 12 0-0 12Jbd7 13 :e1 'iYb4 14 'iYc2 c4 15 i.fl, BareevPredescu, Kitchener 2006. Note that since ... h6 and i.h4 haven't been played, all lines with ... 'iYa5 can favorably be met by i.d2, e.g. 6... c5 7 d5 a6 8 a4 'iYa5 9 i.d2! e6 10 i.d3 or 7... b5 8 cxb5 a6 9 a4 'iYa5 10 i.d2 with advantage.
Black can get dynamic play with 8 •.. l2Jg4!, e.g. 9 12Jxc6 bxc6 10 i.g1 e5! or 9 i.g1 12Jxd4 10 i.xd4 e5. This is better than 8...0-0 9 i.e2! which stops ... l2Jg4. White gets a somewhat favorable middlegame or a good endgame like 9 ... e5 10 12Jxc6 bxc6 11 fxe5 dxe5 12 i.c5 and 'iYxd8 and the related 9 ... l2Jxd4 10 i.xd4 e5 11 fxe5 dxe5 12 i.c5. The reason this is significant is that if Black plays 5... 0-0 instead of 5... c5, White can safely play 6 12Jf3 c5 7 i.e2!? cxd4 8 12Jxd4, forcing his way into the good line of the last paragraph.
173
Indians
Therefore 5 ... c5 is more precise. White can avoid it with a counterfinesse, S Jie2!. Then if Black assumes he's headed for a Classical Variation, 5 ... 0-0 6 ttJf3, he'll be surprised by 6 f4 c5 7 ttJf3, slipping into the 5 f4 0-0 6 ttJf3 c5 7 Jie2 line.
But 8 ttJrJ! transposes into a favorable line of the related ttJf31h2-h3 variation. For instance, 8 ... cxd5 9 cxd5 ttJa6 8 Jie2 ttJc5 9 ttJd2 and 9 ... ttJe8 10 ttJc4 a6?! 11 b4 ttJd7 12 Qd2 as in Aleksandrov-Zhelnin, Krasnodar 1997.
There is one further finesse. Black can meet 5 Jie2 with S... cS, so that 6 ttJf3 cxd4 7 ttJxd4 ttJc6 thwarts White.
But if White tries to reach this through the ttJf31h3 order, S ttJrJ 0-0 6 h3, he can be stopped after 6.•. eS 7 dS, when Black favorably opens the kingside, not the queenside, with 7... ttJa6! 8 Jie3 ttJhS and ... fS. Another curiosity of 5 ttJf3 0-0 6 h3 is that books say 6... cS is best, after which 7 dS e6 8 Jid3.
But 6 dxcS! favors White a bit, 6•.:iVaS 7 Jid2! 'iVxcs 8 ttJrJ 0-0 90-0 and now 9 ... b6?! 10 ::'cl Jib7 11 ttJd5! as in Uhlmann-Bindrich, Dresden 2001. KID: OTHER VARIATIONS An obscure variation that deserves more attention, if only because of its transpo possibilities, is 1 d4 ttJf6 2 c4 g6 3 ttJc3 Jig7 4 e4 d6 S h3 0-0 6 Jie3. Some KID experts say 6... eS is best and give 7 dS c6. Then 8 Jid3 b5!? leads to equal chances, they say.
And now 8... exdS 9 exd5 ::'e8+ 10 Jie3 Jih6 or 10 ... ttJh5 are considered unclear. But 9 cxdS! reaches a Benoni that many authorities say favors White. He is being rewarded for misdirection: if White tries to get to the 9 cxd5 position through a Benoni order there are various ways for Black to thwart him, such as I d4 ttJf6 2 c4 e6 3 ttJf3 c5 4 d5 exd5 5 cxd5 d6
174
Indians
6 ttJc3 g6 7 h3 and now 7... a6 8 a4 'fie7. MODERN DEFENSE The Modem Defense after 1 d4 g6 2 c4 i.g7 becomes a King's Indian after ... ttJf6. Black can use this route to dodge the Samisch among other KID lines. For example, 1 d4 g6 2 c4 i.g7 3 ttJc3 d6 4 e4 and now 4.•. e5.
More common is 5 ttJf3, which can become a Classical KID following 5 ... ttJd7, as Tigran Petrosian, among others, played. But if Black doesn't want to playa Classical he can exploit the 1 d4 g6 2 c4 i.g7 3 ttJc3 order with 3 ••• c5!?
That can become quite sharp after 4 d5 i.xc3+!? 5 bxc3 f5 or lead to a KIDlBenoni hybrid after 4 ... d6. Instead White usually plays Black has to be willing to risk an 4 ttJf3. But that allows 4 •.•cxd4 endgame but there are benefits. If 5 ttJxd4 ttJc6!, a fine version for White tries 5 i.e3, hoping for Black of the English because White something like 5 ... ttJd7 6 d5 or 6 f3, doesn't get the time he needs to set Black can misplace the bishop with up the Maroczy Bind. 5... exd4!? 6 i.xd4 ttJf6. If White is interested m the He will attack it with ... ttJc6, e.g. 7 ttJf3 ttJc6 8 i.e3 0-0 9 i.e2 l:te8 Classical, a safer route is 1 d4 g6 10 ttJd2 ttJd7! 11 0-0 ttJd4 12 i.d3 2 c4 i.g7 3 ttJf3 or 3 e4. Note that ttJc5, Lputian-Todorcevic, Cannes in the latter case, Black has an extra option in 3 ••. ttJc6. 1996. If, instead, 5 d5 in the diagram, Black gets a normal KID after 5... ttJf6. But another benefit of this order is 5..•ttJd7 6 i.e3 ttJe7! and .••f5, which theory smiles on.
The attack on the d-pawn turns out well if Black gets to secure his knight on d4, 4 d5 ttJd4 5 i.e3 with 5 ... c5 and then 6 dxc6 dxc6. This recapture isn't possible after ... d6.
175
Indians
But 3... tbc6 can tum out very badly after 4 tbf3! instead of 4 dS. Then a typical continuation such as 4••• d6 5 d5! tbb8 6 h3 heads for a Benoni (6 ... cS 7 i.d3) in which White enjoys two extra tempi, thanks to ... tbc6-b8. Life isn't easy for the Modem player.
But today we know ltJf3 fits in quite well with the center exchange because after, say, 7... cS 8 :blltJc6 White can favorably allow 9 dS! i.xc3+ 10 i.d2. And 4 ... dS no longer deserves an exclamation point.
A Black who can play both the KID and Gruenfeld may prefer the GRUENFELD DEFENSE more flexible 4 ... 0-0. If White plays When Bobby Fischer called S i.f4, which is designed to hold up 2 tbf3, after 1 d4 tbf6, 'a common ... eS after S... d6 6 e3, Black can error,' he was thinking of how it then go ahead with S... d5, as limits White's options in the Fischer did in his 'Game of the Gruenfeld as well as in the King's Century,' six years before he tried Indian. It makes the Exchange the normal Gruenfeld order. Gruenfeld harmless because of a In that order, 1 d4 ltJf6 2 c4 g6 pinning ... i.g4, he felt. That meant after 1 d4 tbf6 2 tbf3 g6 3 c4 i.g7 3 tbc3 d5, the most fertile area for traps lies in the Russian ('itb3) 4 tbc3: Variation. One of the tabias comes about after 4 tbf3 i.g7 S 'iVb3 dxc4 6 'Yi'xc4 0-0 7 e4 or alternatively, 4 'iVb3 dxc4 5 'Yi'xc4 i.g7 6 e4 0-0 7 tbf3. In an earlier era 4 'Yi'b3 was considered more accurate because 4 ... c6 S cxd5! favors White, whereas in the second order, 4 ltJf3 i.g7 5 'iVb3 c6 is a reasonable defense. That seems trivial today Black could play 4.•. d5! since 5 cxd5 tbxd5 6 e4 tbxc3 7 bxc3 but the case for 4 'iVb3 was allows him to corrupt the White strengthened in the 1963 world match when center with a timely ... i.g4. One of championship Ernst Gruenfeld's first games with Petrosian demonstrated that after the opening he created went 7 ... 0-0 4... dxc4 5 "iixc4 i.g7 6 e4 0-0 8 i.e2 cS 9 0-0 cxd4 10 cxd4 tbc6 White did not have to play 7 ltJf3 11 i.e3 i.g4! with advantage. but could try 7 i.e2!? 176
Indians
Annotators pass over this In practice Black usually replies because 7... a6 8 ltJf3 lands in the 6....td7, because it's forcing and Hungarian Variation and 7... ltJa6 because it leads to something 8 ltJf3 transposes to the Prins familiar, 7 'iib3 ltJb6 8 e4 0-0 Variation. Both have good 9 .te3 ..tg4. But this means he was reputations. But the point of 7 .te2 denied the Hungarian and Prins is White avoids Vasily Smyslov's lines. tried-and-true 7 ltJf3 .tg4. If that White also sets a trap, since the matters to you, it may be worth natural 6... .td7 7 'iib3 ltJb6 8 e4 risking 4 'iib3 dxc4 5 'iixc4 .te6!?, .tg4? is punished by 9 ltJg5!, Black's extra option. threatening 10 'iVxf7+. That wins time for 9 ... 0-0 10 h3! .tc8 (or But suppose you'd be happy in the Smyslov line and not in the 1O ....td7) 11 ltJf3. As usual in the alternatives. Then you should Gruenfeld, when Black has no consider 4 ltJf3 i.g7 5 cxd5 ltJxd5 pressure on the center he is worse. 6 'iia4+!?, an idea of Andor Lilienthal's. This is both a ' something-tothink-about' order and a good way to limit Black's options. White wants his opponent to think about 6... ltJc6 and worry about the consequences of7 e3 and 8 i.b5, as well as of 6... c6 7 e4!, e.g. 7... ltJb6 8 'iic2 .txd4 9 i.h6 and :dl or 8... 0-0 9 i.e3 .tg4 10 ltJe5!.
A potential drawback to this order is Black's extra option of 8... 0-09 i.e3 a5!? and ... a4. If you don't mind that - and you want to dodge the Prins or Hungarian 6 'iia4+!? suits you. A bit dicier is Wolfgang Uhlmann's similar 4 ltJf3 .tg7 5 'iVa4+. We again approach a Russian tabia, 5....td7 6 "iVb3 dxc4 7 'iVxc4! 0-0 8 e4.
177
Indians
tZJxc5 as in Goletiani-Neubauer, Port Erin 2005. Only tricksters should rely on 5 'ifa4+. GRUENFELD:EXCHANGE In the old Exchange Variation, a few issues stand out: Black's Sixth Move. After 1 d4 ttJf6 2 c4 g6 3 ttJc3 d5 4 cxd5 Black has an extra tempo, ... .itd7. ttJxd5 5 e4 ttJxc3 6 bxc3 Black But that limits his choices usually plays ... i.g7 and ... c5 in somewhat since 8... a6 9 e5! is a some order. They typically problem. His knight cannot retreat transpose. to d7, as in the usual Hungarian. But beginning in the 1990s a Ditto 8... ttJc6 9 e5!. difference was found in 7 .itb5+. If Notice that in either line, Black can clear d7 with 9....ite6. But 10 exf6! sacrifices the queen for three pieces. Black has been trapped in a line usually seen via 4 ttJf3 .itg7 5 ""3 dxc4 6 'ifxc4 0-0 7 e4 ttJc6 8 e5 .ite6? (8 ... ttJd7!) 9 exf6 .itxc4 10 fxg7 and 11 .itxc4. In the last diagram Black usually opts for 8 ... .itg4, reaching a Smyslov position once more. But that's probably a mistake because he can try to punish White's order with 8...ttJa6!? Then on 9 e5 ttJg4 10 i.f4 he gets to put his extra ... .itd7 to good use with 10...c5 11 d5 b5! as in Gerzhoy-Tiomin, Tel Aviv 2002. Another added option is 8... b5!, e.g. 9 ttJxb5 ttJxe4 10 'ifxc7 ttJc6 with good play. Or 9 ~3 c5! 10 dxc5 ttJa6 11 e5 ttJg4 12 .itf4
Black has played 6... i.g7:
White can meet 7...i.d7 with 8 .ite2 and transpose to an ttJf3 Exchange Variation in which Black's extra move, ... i.d7, blunts his pressure on d4, e.g. 8... c5 9 ttJf3 ttJc6 10 0-0 0-0 11 l:.b 1 cxd4 12 cxd4 b6 13 d5! with advantage (Almasi-Balusz, Gyula 1993). Instead, Black should meet 7 .itb5+ with 7... c6! after which 8 .ita4 begins a controversial
178
Indians
line championed by Alexander Belyavsky. However if he had played 6... c5 then on 7 i.b5+ he has nothing better than 7•.. i.d7, allowing White to transpose to the Exchange line with 8 i.e2!.
Then Lev Polugayevsky found a finesse, 10 :tel, instead of 10 0-0. His first point was that 10..:i!ic7 transposes into the Shamkovich plan after 11 O-O!. His second point was that the rook will defend the c4-bishop after
Instead, 7...tiJc6? 8 d5 is an old
10 ... cxd4 11 cxd4 so Black can't transpose into the ... ttJa5 line with trap that has caught a number of 11...i.g4 12 f3 ttJa5?? and he is GMs (8 .. .'ilfa5 9 'ifa4! 'ifxc3+ worse after 12 ... i..d7 13 0-0 I 0 ~e2 with a big edge in (13 .. ttJa5 14 i.d3 and 'ifd2/i.h6). Belyavsky-Mikhalchishin, Terme Either way White benefits from Zrece 2003, among others). This 10 l:kl. makes 6•..i.g7! more precise. The drawback is 10... cxd4 White Delays 0-0. A common 11 cxd4 'iVa5+. But White can make sequence is 4 cxd5 ttJxd5 5 e4 a virtue of having not castled by ttJxc3 6 bxc3 i.g7 7 i.c4 0-0 means of 12 ~f1 i..d7 13 h4! to 8 ttJe2 c5 9 i.e3 ttJc6. open the h-file. This, too, has scored some impressive victories and the merit of 10 .l:i.c I is still in dispute. Another wrinkle is to accelerate castling with 9 0-0, rather than 9 i.e3.
In the 1950s, 10 0-0 cxd4 11 cxd4 i.g4 (12 f3 ttJa5) was regarded as slightly favorable to White. Black began to try a plan of 10 .. :ifc7 and 11...l:td8 favored by Leonid Shamkovich and Isaac Kashdan. But after some impressive White is ready to transpose into Spassky victories with II l:tc1!, this main lines after 9...ttJc6 10 i.e3. In fell out of favor. this way he avoids Peter Svidler's 179
Indians
idea of meeting 9 j"e3 with 9 ... cxd4 10 cxd4 'iia5+ 11 j"d2 'it'd8!.
But with the c-pawns on the board, 12 j"d5 j"d7 13 ':bl 'it'c8
Experience with 12 j"c3 liJc6 is inconclusive. If White doesn't have anything better than 12 j"e3 'iia5+ he is headed towards a draw by repetition before the game is five minutes old. So far no downside to 9 0-0 has been found.
threatens the bishop with ... c4! and ... e6. Another problem with ... cxd4 lies in the line made famous in the 1987 world championship match, 13 j"xf7+ ':xf7 14 fxg4 ':xfl+ 15 ~xfl 'iib6 16 ~gl! 'it'e6.
Black delays ... cxd4. The old main line of the Exchange Gruenfeld was 4 cxd5 liJxd5 5 e4 liJxc3 6 bxc3 j"g7 7 i.c4 c5 8 liJe2 0-0 9 j"e3 liJc6 10 0-0 and then 1O ... cxd4 11 cxd4 j"g4 12 f3 liJa5. But in the late 1970s Jan Timman demonstrated the value of delaying an exchange of c-pawns, with 10 ... i.g4. Play often transposes following 11 f3 liJa5 and a later ... cxd4.
Then 17 "ii'd3 'iixg4 18 ':fl is regarded as a slight edge for White. But with the c-pawns on the board we reach a slightly different position after ~g1/ .. :ife6.
What's the point of 10... i.g4 11 f3 liJa5 ? Two stand out.
After 10 ... cxd4 11 cxd4 j"g4 12 f3liJa5 theory says White can try 13 j"d5 and get a nice game following 13 ... i.d7 14 l:tbl 'it'c8 15 'iVd3 and :t.fc 1.
Then 16 'it'd3 'it'c4! gives Black more vigorous queenside play, e.g. 17 'it'd2 'it'e6! 18 liJg3 liJc4 or 17 'it'xc4+ liJxc4 18 j"f2 cxd4 19 cxd4 e5. As a result, the more accurate 10 ... j"g4 is replacing 1O... cxd4. NEO-GRUENFELD
In the Neo-Gruenfeld, g2-g3 and ... d5 are played. If ... d5 comes first, Black can capture on c4. That may make recapturing hard, and this explains why 1 d4 liJf6 2 c4 g6 3 liJc3 d5 4 g3?! is rarely seen, e.g. 4 ... dxc4! 5 "ii'a4+ liJfd7! 6 "ii'xc4 liJb6 7 'it'd3 liJc6 8 liJf3 j"f5 9 e4 j"g4 with an excellent game for Black. More common is an order such as 1 d4 liJf6 2 c4 g6 3 g3 j"g7 4 j"g2.
180
Indians
7 ttJge2 eS 8 .te3? ttJg4 or 8 d5 ttJd4!. There's also independent territory to explore in 6 dS ttJaS 7 'iid3 cS, e.g 8 e4 d6 9 ttJge2 a6 and ... b5 as in Ilincic-Djuric, Vrnjacka Banja 1998.
Black often delays ... dS because he doesn't like 4 ... dS S cxdS ttJxd5 6 e4 ttJb6 7 ttJe2!. Instead, he waits, with 4... 0-0, for the easier-to-handle S ttJf3 dS! 6 cxdS ttJxdS. White's counter-finesse is S ttJc3, to rule out S... d5. Black can continue in KID mode (5 ... d6 or S... c5). But there is also S...ttJc6!? It says, 'You tried to avoid the Neo-Gruenfeld. But if you continue 6 ttJf3 now, I can transpose with 6... dS.'
The plot deepens if White responds 6 e4 because 6 ...d6! gets Black into one of those Fianchetto KID lines with an early e2-e4 that were discredited in the 19S0s, e.g.
Note that 8 ttJf3 d6 transposes to a Black-friendly KID that usually arises from 1 d4 ttJf6 2 c4 g6 3 g3 .tg7 4 .tg2 0-0 S ttJf3 d6 6 ttJc3 ttJc6 7 d5 ttJa5 when White plays 8 'iid3?! cS 9 0-0 instead of8 ttJd2!.
QUEEN'S INDIAN DEFENSE The Queen's Indian and NimzoIndian are opposite sides of the same coin, one designed if not minted by Aron Nimzovich. Before settling on his formula - 1 d4 ttJf6 2 c4 e6 followed by 3 ttJf3 b6 or 3 ttJc3 .tb4 - he experimented with an order that didn't depend on White's third move, 1 d4 ttJf6 2 c4 b6.
The benefit of this order is Black has sharper options after 3 ttJf3
181
Indians
If White can conquer his Nimzo.Jib7 4 g3 such as 4 ... e5!?, 4 ....Jixf3 and 4 ... c5 5 d5 b5!? And the fear, he can play the more ambitious downside? Books say White can 1 d4 tDf6 2 c4 e6 3 tDf3 b6 4 tDc3, win control of e4 with tDc3 and allowing 4 ... .Jib4. He takes that risk 'iVc2 or f2-f3. For example, 3 tDc3 in the hope that Black will prefer .Jib74 'iVc2 d5 5 cxd5 tDxd5 6 tDf3 4 ... .Jib7 and 5 'iVc2 .tb4 or 5 .tg5 e6 7 e4. h6 6 .th4 .tb4. But good players, including But then 5 a3! reaches the 4 a3 Fischer and Larsen have used .Jib7 5 tDc3 Petrosian system. The 2 ... b6!? and others as White, difference is that Black has lost his including Anatoly Karpov, usually chance for 4 a3 .ta6, a major transposed into normal lines with alternative. 3 tDf3, so there may be more plusses than minuses to it. That's why 4 tDc3 is more often In the conventional order, 1 d4 met by 4 •.• .Jib4. White usually tDf6 2 c4 e6. White is indicating his chooses between 5 .tg5 and 5 e3 feelings about the Nimzo-Indian and may transpose into one of the when he chooses 3 tDf3. Black can 4 .Jig5 and 4 e3 lines. For example, try to exploit that in lines such as 5 e3 .Jib7 6 .Jid3 tDe4 7 0-0 is that gambit again. 3 ..•b64 e3 .Jib7 5 .Jid3 .Jib4+!? In the old main line of the Queen's Indian, 1 d4 tDf6 2 c4 e6 3 tDf3 b6 4 g3 .Jib7, White begins a waiting game when he delays tDc3. Books say 5 tDc3 .tb4! equalizes. Play typically continues 5 .tg2 .Jie7 6 0-0 0-0. Then quiet moves such as 7 b3 or 7 l:e 1 allow Black to transpose into an equal Catalan with 7 ... d5!. White often replies 6 tDbd2 even though it offers no more than equality. His Nimzo-fear has hurt him since 6 tDc3! transposes into a line that runs 1 d4 tiJf6 2 c4 e6 3 tDc3 i..b4 4 e3 b6 5 .Jid3 .Jib7 6 tDf3 and offers excellent chances, even in the 6 ... tDe4 7 0-0 gambit.
The critical move is 7 tDc3, preparing to seize the center with 8 'ifc2 and 9 e4. Black usually occupies e4, 7... tDe4, to mechanically stop e2-e4. But if ... .tb4 equalized earlier, why not 7.••.Jib4, a tempo down?
182
Indians
a Queen's Gambit Declined with 4... d5 and then 5 a3 i.e7!:
This is rarely if ever mentioned in books but it's been played by Leonid Stein, David Bronstein and Boris Spas sky. White may be better after 8 i.d2 but not by a great deal (8 ... i.xc3 9 i.xc3 liJe4). Or 8... d6 9 :tel liJbd7 10 d5 liJc5 as in W.Schmidt-Gofshtein, Debrecen 1992 (11 liJh4 a5 12 'ifc2 :e8 13 :fe 1 i.xc3 14 i.xc3 exd5 15 i.xf6 'i'xf6 16 cxd5 g6 with equality).
It's as if the game began 1 d4 d5 2 c4 e6 3 liJf3 liJf6 and then 4 e3 i.e7 5 liJbd2?! and White got an extra a2-a3 as compensation. It isn't enough and 6 e3 0-0 7 i.d3 b6 8 0-0 i.b7 9 b3 liJbd7 10 i.b2 c5 is easy equality.
The main stem of the Bogo is 4 i.d2, not 4 liJbd2. Black usually protects the bishop (4 .. :Yi'e7, 4 ... a5, BOGO-INDIAN DEFENSE 4... c5), thinking that by keeping the According to Mark Taimanov's queenside situation in flux White version of an oft-told story, a will be reluctant to allow Nimzowomen's tournament game began like positions with liJc3. For 1 d4 liJf6 2 c4 e6 3 liJf3 i.b4+. example, 4 .. .'iVe7 5 liJc3?! b6 6 e3 White was surprised by the check i.b7 7 i.d3 and now 7... i.xc3! and asked, 'What opening are you 8 i.xc3 liJe4 is typical equalizing playing?' play. The reply: 'I'm playing the The 'something to think about' Nimzovich! What are you doing?' alternative is 4..•i.e7!? Black is The Bogo-Indian deserves more offering another routine QGD respect than that. After 4 liJbd2 (5 liJc3 d5 6 i.g5). Of course, he White hopes to win the two bishops could have transposed more with 5 a3. Experience indicates that directly with 3... d5. But he may Black's best chances lie in creating want White to consider 5 g3 as well. 183
Indians
His point is that White would see that S... b6 6 ii.g2 ii.b7 is a favorable version of the QID. But after S g3 Black has a good fallback - 5... d5! transposes into a good version of the Closed Catalan, one that usually arises from 1 d4 llJf6 2 c4 e6 3 g3 dS 4llJf3 .tb4+ S .td2 .te7 - and one a 3 llJf3 player may know little about. NIMZO-INDIAN DEFENSE The Nirnzo-Indian is related to an opening without a name, 1 d4 d5 2 c4 e6 3llJc3 .tb4!?
But Black can avoid transposing into Botvinnik-Capablanca (7 ... llJf6?!) with 7 ... llJc6 8 .td3 llJge7! and ... i.fS. For example, 9 'iVc2? c4 10 .te2 .tfS and 9 llJe2 c4 10 i.bl .tfS 11 a4 0-0 12 i.a3 l:.e8 with sufficient counterplay as in Khurtsidze-Tsereteli, Tbilisi 2006 . His KN can also be put to better use if White tries 4 "ika4+. This would transpose into a Ragozin QGD after 4 ... llJc6 S llJf3 llJf6 which is double edged after 6llJeS. But Black renders llJeS harmless with S... llJe7!? This move order is still very young and that's why many players sitting in White's chair will flee towards the familiar with 4 llJf3. But that limits him after 4 ... dxc4 to allowing the Noteboom QGD (S e3 bS) or trying to reach a Vienna after S e4 cS.
Black seems to be saying, 'If I'd played 1.. .llJf6 2 c4 e6 you might have avoided the Nirnzo with 3 llJf3 or 3 g3 or even 2 .tgS. But now I've gotten into a Nirnzo after all.' But this is not a normal Nirnzo because if Black doubles the c-pawns, White can undouble with cxdS, as we saw in BotvinnikCapablanca in the Introduction. That means the litmus test should be 4 a3 .txc3+ 5 bxc3. For instance, 5... c5 6 cxd5 exd5 7 e3.
NIMZO: LENINGRAD Alexander Alekhine said 'the most logical answer' to 1 d4 llJf6 2 c4 e6 3 llJc3 .tb4 was 4 .tgS because 4 ... dS is 'a rather unattractive variation' of the QGD. That is, 1 d4 dS 2 c4 e6 3 llJc3 llJf6 4 .tgS ii.b4?!, which favors White after S 'ii'a4+ or S e3. But the real test of 4 i.gS, the Leningrad Variation, is ... cS!. A main line is 4 .tg5 h6 5 .th4 c5 6 d5 .txc3+ 7 bxc3.
184
Indians
NIMZO: QUEEN MOVES After Gideon Stahlberg beat him with 4 'ifb3 three times in their 1934 match, Nimzovich declared the queen move 'almost a refutation of my variation.' It's not even a major line today. But it's so closely related to 4 'iic2 that Black has to be careful of being trapped in a On 7... d6 Black prepares a 4 'ifc2 variation that is either traditional dark-square blockade inferior or unfamiliar. strategy ( ... e5, ... 'ife7 and perhaps For example, 4 ... liJc6 is regarded ... g5). But that allows White to as sufficient to equalize against choose piece play over pawn 4 'ifb3 but unfavorable against structure with 8 dxe6!? since 4 'ifc2. Black can find himself in 8... fxe6 9 e3 and liJf3/i.e2 favors the wrong line after 4 'ifb3 liJc6 him. Better is 8... i.xe6 but 9 e3 sliJf3 dS. liJbd7 10 i.d3 liJe5 is unclear. Black can try to finesse with 7... eS and then 8... d6!, transposing into the blockade line. But White has a good extra option in 8 d6!. Experience shows that even if Black wins the advanced pawn White stands well. A superior order for Black is 4... h6 S i.h4 cS 6 dS d6!, delaying ... i.xc3+. This rules out d5-d6 and reaches the desirable blockade following 7 e3 i.xc3+! 8 bxc3 eS, with equal chances. For example, 9 'iVc2 'ife7 10 liJf3 g5 11 i.g3 liJh5 12 i.d3 liJd7 13 liJd2 liJdf6 14 0-0-0 ~d8!? and ... <3;c7, as in Guseinov-Ibragimov, Baku 2006.
Now 6 a3 i.xc3+? 7 'ifxc3 favors White. He's transposed into 4 'ifc2 liJc6?! 5 liJf3 d5 6 a3 i.xc3+. In that order, 6... dxc4?? would lose a piece. But with the queen on b3, 6... dxc4! 7 'iixc4 i.d6 is excellent for Black. A similar case arises after 4 'ifb3 c5 and 5 dxc5. Then 5... liJa6 6 a3 i.xc5 is an easy road to equality.
185
Indians
In contrast, 6....ltxc3+?! 7 'ifxc3 13 'ifxf5 g6 14 'tic2 tiJh5 with a tiJxc5 8 f3! is poor. This transposes quick draw in Shipov-Rozentalis, into 4 'iYc2 c5 5 dxc5 when Black Tromso 2006. plays the inferior 5... tiJa6?! (6 a3 That means an aggressive Black .txc3+ 7 'iYxc3 tiJxc5 8 f3!) rather can use this move order as a form of than 5... 0-0! 6 a3 .ltxc5. reconnaissance: He begins with The main difference between the 1 d4 tiJf6 2 c4 e6 3 tiJc3 ~b4 in two queen moves is that 4 'ifc2 hopes of reaching one of the doesn't attack the bishop but does sharp lines. If White adopts the prepare e2-e4. If Black stops that conservative 4 'ifc2 instead, Black push with 4... d5 what kind of QGD can retreat into a solid QGD do we have ? The answer is a good position with 4... d5 5 a3 i.e7 at one for Black after 5 .ltg5 dxc4 and little cost. even better after 5 tiJf3?! dxc4!. The newest finesses in 4 'tie2 Moreover, the cautious 5 e3 occur after 4... d5 5 exd5 e5!? This doesn't fit in well with 4 'ifc2. After is Veselin Topalov's gambit, based 5... c5, Black reaches comfortable on lines such as 6 dxe6 cxd4 4 e3 Nimzo territory following 7 exf7+ cJ;xf7 or 6....txe6 7 dxc5 6l2Jf3 tiJc6 7 a3 .ltxc3+ 8 bxc3 0-0. 0-0. Whether Black gets enough The real test of ... d5 in a Nimzo compensation is controversial but is usually a2-a3 so we have to our concern is whether White can avoid the messy issue entirely. evaluate 5 a3 .te7!? One try is 6 .ltg5, since 6... exd5 reaches a quiet line that more commonly arises after 4 'ifc2 d5 5 cxd5 exd5 6 .ltg5 c5.
White is better but only somewhat, e.g. 6 cxd5 exd5 7 .ltg5 c6 8 e3 .ltg4!? 9 tiJge2 tiJbd7 10 h3 .lth5 11 tiJg3 .ltg6 12 tiJf5 .ltxf5!
The resulting position may be good for White after 7 a3 ~xc3+ 8 bxc3, e.g. 8... h6 9 i.xf6 'tixf6 10 e3 0-0 11 tiJf3 tiJc6 12 .te2 ~e6 13 0-0 l:.fc8 14 'i'a4 with a modest edge in Rowson-Pert, British Championship 2006. There is, however, a complication in 6 i.g5 'i'xd5!?
186
Indians
White's goal is again to trap Black in an unfamiliar 4 'iVc2 line. 'What's this?' Black thinks to himself. 'Does he really believe 4 ... c5 5 d5 is good? Or is his idea to meet 4 ... d5 with 5 cxd5 exd5 6 'ifb5+?'
What happened is Black inveigled his opponent into a line that runs 4 'ifc2 d5 5 ~g5 c5 6 cxd5 'ifxd5, when he threatens 7... 'ifxg5, 7... cxd4 and 7... ttJe4. This usually leads to equal chances after 7 ~xf6 gxf68 dxc5 'ifxc5. Another way for White to deal 'I can't figure this all out so I'll with the new gambit is to answer 4 'ifc2 d5 5 cxd5 c5 with 6 dxc5. He just castle,' he may conclude. That's gets the upper hand after 6... 'fixd5 not a bad decision. There's nothing 7 a3! ~xc3+ 8 'ifxc3 0-0 9 f3!. To wrong with 4 ••• 0-0 5 a3 and avoid this Black can take his 5...~xc3+ 6 'fixc3 - provided that chances with 6... ttJxd5. Black knows the theory. But watch out when a trickster tries 6...exd5. Then on 7 ~g5 we transpose into a position that comes from 4 'ifc2 d5 5 cxd5 exd5 6 ~g5 when Black tries the ambitious 6... c5 and White fails to find 7 a3!. Instead he's played 7 dxc5, giving Black a fine game after 7••• h6! 8 ~h4 g5 9 ~g3 ttJe4 and then 10 .txb8 'iff6! 11 .tg3 ttJxc3 as in the sensational game I.Sokolov-Aronian, Turin 2006. Finally, few books even mention the third queen move, 4 'fid3.
What theory? The considerable amount of book that's been built up around 4 'i'c2 0-0 5 a3 i.xc3+ 6 'fixc3, into which he's unwittingly transposed. This is a problem if Black normally meets 4 'ifc2 with something other than 4 ... 0-0. Similarly, ifhe is unfamiliar with 4 'ifc2 d5 5 a3 ~xc3+, because he plays some other fourth move as Black, he will be a stranger in a strange land after 4 'iVd3 d5 5 a3! ~xc3+? 6 'fixc3.
187
Indians
NIMZO: SAMISCH
line, 4 a3 i.xc3+ 5 bxc3 0-0 6 e3 c5
After Fritz Samisch introduced 4 a3 attempts were made to refine it by delaying a2-a3 in favor of Nikolai Riumin's 4 D and Akiba Rubinstein's 4 e3, which are supposed to be more flexible. But a case can be made that 4 a3 i.xc3+ S bxc3 is the most flexible, because it allows White to choose between f2-D and e2-e3 once Black commits himself.
7 i.d3 tbc6 8 tbe2, with a quick e3e4 and f2-f4-f5 attack. But there's a better transposition than Samisch-into-Riumin or Samisch-into-Rubinstein. It is Rubinstein-into-Samisch, 4 e3 followed by a2-a3. It arises in several orders including 4 e3 cS S i.d3 0-0 and 6 a3 i.xc3+ 7 bxc3.
For example, after S...b6:
This is more promising than 4 a3 because Black's castled king is vulnerable, e.g. 7... b6 8 e4! i.b7 Black would be happy to see 6 e3 9 i.gS, as in the celebrated Keressince 6... i.b7 7 D c5! renders good Reshevsky, 1953 Candidates game counterplay. But White is better (9 ... h6 10 h4! d6 11 e5 dxe5 after 6 f3! and 7 e4 (or 6 ... d5 12 dxe5 i.e4 13 1:th3! with a ferocious attack). 7 i.g5!). Black often gets confused in the On the other hand, if Black chooses S... cS he would have good delayed Samisch, as in that game. play after 6 D d5 7 cxd5 tbxd5!. He thought he was transposing into This transposes into an unclear 4 D an earlier game from the same variation. tournament. That one went 4 e3 cS S a3, rather than 5 i.d3, and then White may do better with 6 e3!? S...i.xc3+ 6 bxc3 b6 7 i.d3 i.b7. For instance, 6 ... tbc6 7 i.d3 d6 8 tbe2 0-0 or 8 ... e5 9 e4 0-0 Black obtained excellent play transposes to another promising following 8 D tbc6 9 tbe2 0-0 188
Indians
10 0-0 lba5 11 e4 lbe8! with the idea of ...lbd6. Reshevsky would have been able to transpose into that game after 8lbe2 .tb7 or 8 f3 .tb7. But White crossed him up with 8 e4!. NIMZO: RUBINSTEIN The best known Nimzo-Indian tabia arises three moves after 4 e3.
commitments to pawn structure and can't be forced to trade off his bishop (as 4 ... c5 5 lbe2 and 6 a3 does). But 4 ... 0-0 does commit his king which is important if White shifts to the delayed Samisch, 5 a3 .txc3+ 6 bxc3. Castling also costs Black a tempo and prevents him from reaching Robert Huebner's dark-square blockade strategy. After 5 .td3 c5 6lbf3lbc6 he doesn't get a chance to see 7 0-0 .txc3! because White interrupts him with 7 d5!. A shrewd Black may be tempted to avoid all of this by means of 4...lbe4!? He is trying to trick White into one of the best versions of 4 ... b6.
'This position has been investigated by some masters, lovers of chess analysis, to the 3035 th (!) moves,' Vladimir Simagin wrote way back in 1954. Black's defenses have stood up well in this and similar positions and his main concern is move order. Some Nimzo experts reach the diagram via 4 ... 0-0, 5... c5, 6 ... d5 and 7... lbc6. Others prefer 5... d5 and 6 ... c5. Still others began with 4 ... c5 and then 5 ... lbc6 and 6 ... d5. And others have faith in 4 ... d5, 5 ... 0-0 and 6 ... c5. Today 4 ... 0-0 has the most supporters because Black makes no
Black succeeds after 5 "ifc2 f5 6lbe2 b6 or 6lbf3 b6 7 .td3 .tb7, both excellent 4 ... b6 lines for him. The main benefit in the 4 ... lbe4 order is that he doesn't have to deal with the enormous amount of 4 e3 b6 theory, as well as some specific dangers, such as the 5 iL.d3 iL.b7 6lbf3 lbe4 7 O-O!? gambit.
189
Indians
White can also meet the threat to c3 in the last diagram with 5 tbe2 instead of 5 'iYc2. If Black doesn't like the looks of 5... 'iYh4 6 g3 'iif6 he can try to reach a good 4... b6line once more with 5... f5 6 'iYc2 b6. TANGO
In a remarkably short time, 1 d4 tbf6 2 c4 tbc6 has spawned a lot
of theory, especially in sharp continuations beginning with 3 tbc3 e5 and 3 d5 tbe5. But in most games White flees towards the familiar with 3 tbf3. After 3... e6:
White can try 4 g3 after which 4... d5 makes it a Catalan. Such positions, with a blocked c-pawn, were once considered heretical but now have not-too-shabby reputations. After 5 i..g2 dxc4 the
position is considered unclear following 6 'iia4 tbd7! or 6 0-0 l:tb8!? Moreover the Black QN is well placed ifhe manages to transpose to a Bogo-Indian with 4 ... i.b4+ 5 i.d2 'iie7. But White should avoid that with 5 tbbd2! and 6 a3. This is a transposition from a normal Bogo, 1 d4 tbf6 2 c4 e6 3 tbf3 i.b4+ when 4 tbbd2 has been met by the weak 4...tbc6?! 5 g3 'iie7. If, instead, White plays 4 tbc3 in the diagram, Black can transpose into a rare Nimzo-Indian with 4... i.b4 5 'iic2. White can be reasonably happy with his prospects after, say, 5... 'iie7 6 a3 i.xc3+ 7 'iixc3 d6 8 b4 e5 8 dxe5 tbxe5 9 i.b2, Baburin-Scannell, Feugen 2006. But the main question is who has tricked whom? Did Black drag his opponent into a 4 'iic2 Nirnzo that he didn't want to play? Or did White get into a 4 'iic2 line in which Black has been denied the more popular 4... d5 or 4... c5 ? Once again, both players may have been tricked.
190
Chapter Seven: Dutch Defense and Flank Openings Everyone knows the first finesse of the Dutch Defense: Black can avoid the Staunton Gambit, 1 d4 f5 2 e4, by means of 1...e6 and 2 c4 f5. In the last few decades 1...e6 has also been used to reach a NimzoIndian or Queen's Indian (2 c4 liJf6) or a Benoni (2 c4 c5) without risk of facing a Trompowsky.
White hints he is amenable to a Tarrasch French (2 ... d5 3 e4). But his real intent may be a Colle (2 ... d5 3 liJgf3!) in which Black has blocked in his QB. Black can also find himself in an inferior Stonewall Attack after 2... c5 3 e3 liJc6 4 c3 liJf6 5 i..d3 d5 and 6 f4!.
Of course, there's another hazard. After 2 e4 Black has to be ready to defend a French (2 ... d5) or the more dangerous version of the English Defense (2 ... b6) or try 2 ... c5, as Bent Larsen did on occasion. And he has to know what to do after 2 liJd2!?
White can only dream of getting that position after 1 d4 d5 2 e3 and it's quite good, e.g. 6.•.i..d6 7 'iff3 i..d7 8 liJh3 'iib6 9 liJf2 0-0-0 10 0-0 ~b8 11 e4!, MarshallRubinstein, Vienna 1908. Today the threat of the Staunton Gambit doesn't scare many Dutch players. But some prefer 1.. .e6 anyway, to avoid other l...f5 lines, such as 2 liJc3 or 2 i..g5. They are an attempt to improve on the normal Staunton, 2 e4 fxe4 3 liJc3 liJf6 4 i..g5. After 1 d4 f5 2 liJc3:
191
Dutch Defense and Flank Openings
DUTCH: MAIN LINES Black often delays his biggest decision, ... d6 or ... d5, and looks for a good time for ... .tb4(+). For example, 1 d4 f5 2 c4 lLlf6 3 g3 e6 4 .tg2 .tb4+:
Black's choice is limited since 2... lLl f6 3 ~g5 threatens a ~xf6/e2-e3/ Trompowsky-like ~d3riVf3 setup that is not to the liking of many Dutch players. For example, 3... c6 4 ~xf6! exf6 5 e3 d5 6 ~d3 g6 7 'it'f3 ~e6 8 h3 ~g7 9 lLlge2 lLld7 10 g4! fxg4 11 hxg4 'it'e7? 12 l:txh7! as in KorchnoiKostjoerin, Havana 1966. The most common response to 2 lLlc3 is 2•.. d5, when 3 e4!? transposes to a good version of the Staunton after 3... fxe4? 4 f3 exf3 5 lLlxf3 but 4 'iWh5+! is better. That means 3... dxe4 is best. But this often transposes (4 f3 lLlf6 5 fxe4 fxe4) into 3... fxe4 4 f3 lLlf6 5 fxe4 dxe4. Then 6 ~c4 e5!? 7 ~g5! and 6... lLlc6 7lLlge2 e5 8 d5 are unclear. Bottom line: If Black can also play the French, avoiding all this with I d4 e6 may be his best route to the Dutch. Moving on to 1...f5:
Now 5 lLlc3?! is fine for Black, e.g. 5... 0-0 6 lLlf3 lLle4 7 'iic2 d6 8 0-0 .txc3! 9 bxc3 lLld7 10 l:tdl ~e8 11 lLld2 lLldf6!? as in HortShort, Prague 1990. Better is 5 ~d2! since an exchange of bishops tends to favor White. This is particularly true if Black adopts a Stonewall pawn structure, 5... ~xd2+ 6 lLlxd2 d5?, instead of 6... d6! and ... e5. Black has more success delaying ... .tb4+ in the hopes that White will voluntarily develop his knight on c3 and allow the pin. But that's only good when White is committed to lLlf3 or ~g2. The pin is premature after 1 d4 f5 2 c4 e6 3 lLlc3 .tb4?!, for example, because White hasn't ruled out Akiba Rubinstein's plan of e2-e3/lLlge2/a2-a3, which he used
192
Dutch Defense and Flank Openings
just as effectively against the Dutch as against the Nimzo-Indian.
control of e5 with ltJd3. But after ... d6 a knight maneuver to f4 runs into ... e5. In practice an order such as 1 d4 fS 2 g3 ltJf6 3 iLg2 e6 4 ltJf3 iLe7 S 0-0 makes S... dS attractive because White has already showed his hand with ltJf3. In contrast, after 4 ltJh3 iLe7 S 0-0 his best is S... d6!.
That means 4 'iic2 ltJf6 S e3! 0-0 6 iLd3 followed by ltJe2 and a2-a3, favors White, e.g. 6... d6 7 ltJe2 ltJc6 8 0-0 iLxc3 9 'iixc3 a5 10 b3 e5 11 iLa3! J:te8 12 d5 ltJe7 13 f4!, Yakovich-Gleizerov, Stockholm 2000.
Consequently, if Black commits his d-pawn too early, White knows what to do with his KN - 1 d4 fS 2 g3 e6 3 iLg2 ltJf6 4 c4 dS S ltJh3!. The same applies if White develops his knight at e2 rather than h3. Two of the 1951 world championship games reached this position.
The more accurate order, after 1 d4 f5 2 c4 e6 3 ltJc3, is 3••. ltJf6. This waits for White to make a decision, such as 4 g3 or 4 ltJf3, which would justify 4 ... iLh4!. If White chooses 4 e3?! instead he is starting the Rubinstein plan too early and Black should switch to the ... d6/ ... e5 pawn structure, e.g. Mikhail Botvinnik played 4 ...iLe7 5 iLd3 0-06 ltJge2 d6! with 6 ... d5?!, got the worst of7 ltJge2 c6 a good game. 8 b3 ltJe4 9 0-0 ltJd7 10 iLb2 ltJdf6 Once White decides to fianchetto 11 f3! and eventually lost. his KB, the focus turns to his KN. If it goes to e2 or h3, the conventional When David Bronstein was wisdom holds that Black is better Black he chose 6... d6!. White's off with ... d6 than with ... d5. A knight had little to do following knight stands nicely on f4 in a 7 ltJge2 c6 8 0-0 eS. In fact it never Stonewall because it watches the moved again and he was soon backward e-pawn and can reinforce worse. 193
Dutch Defense and Flank Openings
The situation becomes more difficult when Black seeks a Stonewall with ... ~d6. Then he can't use ... ~e7 as a waiting move. But its place as a 'pass' can be taken by ... c6.
Mark Taimanov, who praised 6 ... c6, explained that the Stonewall's defects are best exploited when White can play ttJf3-e5-d3 and ttJbl-d2-f3. He cited a game that went 6... d5,
After 1 d4 f5 2 g3 ttJf6 3 ~g2 e6 4 c4 d5 White stands well with 5 ttJh3! c6 6 0-0 ~d6 7 ~f4. But
instead of 6 ... c6, and turned in White's favor following 7 ttJbd2! c6 8 ~c2 ~e8 9 ttJe5 ttJbd7 10 ttJd3! ttJe4 11 ttJt3 g5 12 b4.
Black has 4 ... c6!?, which he believes will be a useful move after both 5 ttJh3 d6 and 5 ttJO d5.
However, once White is committed to ttJc3, Black has an easier game, Taimanov said.
For example, 5 ttJf3 d5! 6 0-0 ~d6 and 7 b3 'fie7 stops the
unfavorable trade of Black's good bishop (8 i..a3). If White insists on the trade, with 8 a4, Black gets adequate play from 8... a5! 9 ~a3 i..xa3 10 ttJxa3 0-0 11 ttJc2 b6 12 ttJce 1 i..b 7 13 ttJd3 ttJbd7 and ... c5 as in TenikashvilliRadjabov, Moscow 1998. A further bit of subtlety involves White's QN, as in 1 d4 f5 2 g3 ttJf6 3 ~g2 e6 4 ttJf3 ~e7 5 0-0 0-0 6 c4 and then 6... c6!. Black waits for 7 ttJc3, which makes 7 ... d5 more appropriate.
For example, 7 ttJc3 d5 8 b3 ttJe4 9 ~b2 ttJd7 10 e3 ~d6 11 ttJd2 ttJdf6. If White avoids 7 ttJc3 by means of 7 b3 Black has a good reply in 7 ... a5. He can answer 8 ~a3 with 8... ~xa3 9 ttJxa3 d6! or 9 ... ~e7 10 ttJc2 d6!. His bad bishop is no longer so bad. TRANSPO-DUTCH Black can set up a Dutch on his side of the board after 1 c4, 1 ttJO or 1 g3 by playing 1...f5. He can also make a good Dutch out of a
194
Dutch Defense and Flank Openings
doubtful English Defense, as in 1 c4 b6 2 d4 ~b7 3 liJc3 e6 4 a3 and now 4..• f5!'
pioneered by Carlos Guimard, 1 d4 d5 2 c4 e6 3 liJc3 ~e7 4 liJf3 and now 4••• f5.
White's a-pawn plays a minor role, e.g. 5 d5liJf6 6 g3liJa6 7 ~g2 ltJc5 8 liJh3 ~d6 9 ~e3 ~e5!, Karpov-Morozevich, Cannes 2002.
This sequence rules out the Staunton Gambit as well as the 1 d4 fS 2 liJc3 and 2 ~gS lines, and gets White to commit his knights, so that the liJbd2, liJh3 and b2-b3/~a3 options are lost.
But after I c4 fS or 1 liJf3 fS White can play d2-d3 rather than d2-d4. This takes e4 away from Black's KN and may prepare a powerful push of the e-pawn, e.g. 1 liJf3 fS 2 g3 liJf6 3 ~g2 e6 4 c4 dS S 0-0 c6 and now 6 d3 ~d6 7 liJc3 0-0 8 e4!.
But White can make a favorable course correction with 5 ~f4! liJf6 6 e3 c6 7 ~d3 0-0 8 'iic2. For example, 8... liJe4 9 g4! liJa6 10 a3 'iVaS 11 ~e2 liJxc3+ 12 bxc3 liJc7 13 cS liJe8 14 l:thg 1 ~h8 IS l:tg2 and White crashed through in LimaGiardelli, Sao Paolo 1993.
Black was victimized by his insistence on 3... e6 and ... dS. Ifhe's ENGLISH DEFENSE willing to handle a ... d6 Dutch, he should play 3... d6, preparing 4... eS. Most lines of the English If White deters him with 4 d4 Black can transpose to normal ... d6 lines Defense favor White at least with 4... e6. slightly and Black should be alert to transpositions to the Dutch, QID or Before leaving the Dutch we other openings. After 1 d4 e6 2 c4 should mention a little-known order b6 3liJc3 ~b7 4 e3: 195
Dutch Defense and Flank Openings
... i.b4 pin if he inserts a2-a3 first, such as after 1 d4 e6 2 c4 b6 3 a3. Black can reach the Petrosian QID with 3... ltJf6 4 ltJc3 i.b7 5 ltJf3. Another route to the same place is 3... i.b7 4ltJc3 ltJf6 5 ltJf3. The Petrosian QID is hardly a bad line. Why then would Black use one of these orders (and risk the Now 4... i.b4?! 5 ltJe2! ltJf6 is a complications of 5 d5)? The usual White-friendly Nimzo-Indian, e.g. reasons are to avoid the 6 a3 i.xc3+ 7ltJxc3 d5 8 b3. Trompowsky and the Fianchetto But 4... f5! is a good Dutch. The
lines of the QID.
right time for the pin would be 5 ltJf3 i.b4 since White can no longer break it favorably with ltJe2 and a2-a3. Better for White is 5ltJge2 ltJf6 6 d5 but after 6...ltJa6 this is one of the better lines of the English Defense for Black.
But a more adventurous Black will try 3...f5!? 4 ltJc3 ltJf6. White is better off here than he was in the position considered earlier, I d4 e6 2 c4 b6 3ltJc3 i.b7 4 e3 f5!. In this case he can benefit from a fianchetto, 5 g3 i.b7 6 d5!.
An irony is that many English Defenders rarely - if ever - begin with the characteristic ... b6. The reason is that I d4 b6 2 e4! i.b7 and its twin, I e4 b6 2 d4 i.b7, are considered better for White than when c2-c4 is played.
However, there is room for subterfuge after 5 g3 c6! 6 i.g2 d5:
The more precise order is 1 d4 e6, waiting for 2 c4 b6, they say. Against 1 e4 they avoid 1... b6 and go into something else entirely. Only in reply to I c4 or I ltJf3 do they begin 1... b6. Another aspect of the English Defense is that White can avoid the
Each side has an extra move compared with a normal Stonewall. But ... b6 is more useful than
196
Dutch Defense and Flank Openings
White's extra a2-a3. That's shown by 7 cxdS cxdS 8 .tf4 .td6 9 .txd6 'iixd6 10 ':'c 1 ttJc6 11 ttJf3 0-0 12 0-0 .tb7 13 e3 ':'ac8 with balanced play in RogozenkoMorozevich, Bundesliga 2000. Note how late Black waited to put his bishop on b7. He was holding back because after S dS it makes more sense to play S... .ta6!. Then 6 e3 exdS 7 cxdS .txfl and 7 ttJxdS .td6! are quite nice for him. If Black insists on the English Defense against 1 e4, a key question is how White defends the e-pawn after 1 e4 b6 2 d4 .tb7. On 3 ttJc3 e6 Black is ready for 4....tb4. Experience with 4 ttJf3 .tb4 S .td3 ttJf6 has varied widely but the results are better for Black than other 1 e4 b6 lines.
This retains the options of ttJc3, c2-c3 and even c2-c4, as in S... dS 6 eS ttJfd7 7 0-0 cS 8 c3 and 7... .te7 8 c4!, e.g. 8... dxc4 9 i.xc4 a6 10 ttJc3 bS 11 i.d3 ttJb6 12 i.e4 with an edge in Khalifinan-Bauer, internet 2004. CATALAN OPENING A Catalan is i.g2 attached to a Queen's Gambit. Curiously, the Catalan comes about more often these days via 1 d4 ttJf6 than 1 d4 dS, at least at the master level.
After 1 d4 dS 2 c4 e6 3 g3 Black More exact is 3 .td3! and 3... e6 equalizes faster, with 3 ... dxc4! 4 ttJf3. Then 4 ... cS S ttJc3 cxd4 4 ttJf3 cS, e.g. S 'i'a4+ i.d7 6 'i'xc4 6 ttJxd4 is an excellent Sicilian .tc6 7 .tg2 ttJd7 8 0-0 ttJgf6 9 ttJc3 Defense for him. bS 10 'i'd3 'iYb6, OpocenskyKotov, Prague-Moscow 1946. Black's best is 3... ttJf6!. The More commonly, a Catalan arises attacked e-pawn can't move (4 eS after 1 d4 ttJf6 2 c4 e6. But if White .txg2) and on 4 ttJc3 e6 Black has is concerned about the Benko, ... .tb4 on tap. Benoni, Budapest, et aI, he may try 2 g3 instead. White's most precise order may This transposes to a normal be 3 i.d3 ttJf6 4 'iie2! and then Catalan after 2 ... e6 3 c4. This order 4..• e6 5 ttJf3. 197
Dutch Defense and Flank Openings
allows Black to develop his QB outside his pawn center (2 ... d5 3 tbf3 c6 4 .lig2 .lif5). But that is no longer considered an easy road to equality after, say, 5 0-0 e6 6 b3 h6 7 .lib2 tbbd7 8 c4 .lid6 9 tbbd2 0-0 10 tbe5 and £2-f4, S.PolgarUshenina, Dresden 2006. There are two common I d4 tbf6 2 c4 e6 Catalan tabias and they often appear after 3 g3 d5 4 .lig2 and 4 tbf3 (or 3 tbf3 d5 4 g3). There are very minor differences between the tWo. But the third order, 1 d4 tbf6 2 c4 e6 3 tbc3 d5 4 g3, has plusses and minuses.
superior form of the Tarrasch QGD, 4... c5 5 cxd5 cxd4! 6 'iVxd4 tbc6 that worked well for Boris Spassky. Note that White can limit enemy resources by holding back the c- or d-pawn. For example, 1 d4 tbf6 2 tbf3 e6 3 g3 d5 4 .lig2 .lie7 5 0-0 and a later c2-c4 rules out ... .lib4+ and early ... dxc4 lines. He can also delay d2-d4, as with 1 c4 e6 2 g3 d5 3 tbf3 and then 3... tbf6 4 .lig2 .lie7 5 0-0 0-0 6 d4 is the closed Catalan. Of course, there is less chance of reaching a Catalan middlegame via these orders than after 1 d4, 2 c4 and 3 g3. Black's main decision in the Catalan is whether (and when) to capture on c4. If he wants to open the position the careful way is 1 d4 tbf6 2 c4 e6 3 g3 d5 4 .lig2 and now 4•••.lie7 5 tbfJ 0-0 6 0-0 dxc4.
For example, on 4 ... dxc4 5 'iWa4+ .lid7 6 'iWxc4 .lic6 7 tbf3 Black can unbalance matters with 7... .lixf3 and obtain good play against the isolated d-pawn. Better is 7 f3!, which has been considered favorable since Tinsley-Schlechter, Hastings 1895. But this order is imprecise because Black can transpose into a
In this way he avoids 'iVa4+ lines such 1 d4 tbf6 2 c4 e6 3 g3 d5 4 .lig2 dxc4 5 'iWa4+. This makes
198
Dutch Defense and Flank Openings
sense if he has no interest in 4 ... dxc4 S tbf3 lines such as S... a6, S... tbc6, S... i.d7 and S... i.b4+. The queen check can set a trap in delayed form, 4... i.e7 5 tbf3 dxc4 6 'i'a4+.
But this is a splendid time for 2•.. b6!, transposing to an English Defense in which White is committed to c2-c4 yet has denied himself the ambitious e2-e4/f2-f3 setups. A player who usually meets 1 d4 e6 2 c4 b6 with a setup such as 3 tbc3 .ltb7 4 a3 or 4 e3, can find Then 6... tbbd7?! doesn't seem himself in a suspect position after like an error. But what has 1 c4 e6 2 tbf3 b6 and then 3 e4 i.b7 happened is that Black has been 4 tbc3 i.b4 5 .ltd3 tbe7 6 a3?! tricked into S 'i'a4+ tbbd7 6 tbf3 i.xc3 7 dxc3 0-0 8 0-0 f5!, as in i.e7? (instead of6 ... cS! or 6 .. ..l:t.b8!) Ibragimov-Shabalov, U.S. Open which is a slight disadvantage for 2003. him. Quite a different story is 1 c4 g6 ENGLISH OPENING which retains a wide range of An amateur who plays 1 c4 options including the King's Indian, is often at a psychological Modern Defense and Dutch Indian. disadvantage when the game Black can wait for 2 tbc3 so that transposes into a I d4 opening 2... c5 gets him into a good line after because he typically chose 1 c4 to 3 tbf3 .ltg7 4 d4 cxd4 5 tbxd4 avoid main 1 d4 lines such as in the i.g7. If he plays 1 c4 cS instead, Slav, Gruenfeld and Nirnzo-Indian. White has the more promising It pays for him to be able to handle 2 tbf3!? and 3 d4. a QGD because 1... e6 2 tbc3 d5 Another family of finesses 3 b3? d4!, for example, is bad. White can try to stay in Reti-range begins with 1 c4 tbf6. White with 2 tbf3 and then 2 ... dS typically responds 2 tbc3 so that he can meet 2 ... e6 with the well3 b3. 199
Dutch Defense and Flank Openings
regarded 3 e4. This also enables him to dodge the Gruenfeld with 2 ... g6 3 e4 and transpose into the King's Indian after 3 ... d6 4 d4. Books devote a lot of attention to 2 12Jc3 d5. But perhaps the best aspect of the 1 c4 12Jf6 2 12Jc3 order for Black is he can transpose into a 1...c5 line with 2 ... c5! and then break in the center with 3 ... d5. As noted earlier, if he had played the more common order such as 1 c4 c5 White can break first with 2 12Jf3!? and 3 d4.
ENGLISH: 1... e5 After 1 c4 e5 Black typically chooses between open systems ( ... l2Jf6/ ... d5) and closed ones ( ... l2Jc6 and possibly ....tb4). The finesses begin with 2 g3. This will transpose into main lines after, say, 2 ... l2Jf6 3 .tg2 d5 4 cxd5 12Jxd5 5 12Jc3 or 2 ... l2Jc6 3 .tg2 g6 4 12Jc3. But in this way White avoids all the nasty ... .tb4 business.
After Anatoly Karpov made good use of 2 12Jc3 12Jf6 3 12Jf3 12Jc6 4 g3 Another virtue of 1 c4l2Jf6 is that .tb4 as Black in his 1987 world match, Vasser 2 12Jc3 e5 gets Black to a common championship position without having to concern Seirawan played 2 g3 against him. himself with 1 c4 e5 2 g3!? If Karpov was out of his repertoire so White insists on 2 g3 (after 1 c4 he played 2... g6?!, which he had 12Jf6) he has to deal with 2... c6! used only once before, when he was 3 .tg2 d5. 11. That game went 3 .tg2 .tg7, a reversed closed Sicilian in which the absence of an early ... l2Jc6 takes the sting out ofb2-b4-b5.
He either has to risk a gambit (4 12Jf3 dxc4/...b5) or defend the cpawn with 4 'ilVc2 or 4 b3, which allows Black to solve his opening problems with 4 ... e5!, e.g. 4 'ii'c2 e5 5 d3 h6! 6 12Jf3 .td6 7 0-0 0-0 8 12Jc3 d4, Socko-Belov, Hastings 2004.
But Seirawan's 3 d4! gave him either a distinct edge in space (3 ... exd4 4 'ii'xd4) or a superior endgame (3 ... d6 4 dxe5 dxe5
200
Dutch Defense and Flank Openings
5 'i'xd8+). Karpov chose the latter and lost.
against the OlD. For example, 8 e4 :e8 9 h3 i.f8 10 l:tel a6 11 d5?!
The drawback to 2 g3 is 2...tbf6 3 i.g2 c6 since 4 tbfJ e4 5 tbd4 d5 6 cxd5 can be met by Mikhail Tal's 6.••'i'xd5!. By attacking the knight
cxd5 12 cxd5 b5! 13 a3 tbb6 14 b3 i.d7 15 i.e3 'i'b8 with a fine game for Black (A.Sokolov-Sherwin, Swiss Team Championship 2001).
Black obtains excellent chances, e.g. 7 tbc2 'iih5 and ... i.h3.
In the main closed variation of the Reversed Sicilian, 1 c4 e5
The comparable 2 tbc3 tbf6 3 g3 c6 line never returned to fashion
2 tbc3 tbc6 3 g3 g6 4 i.g2 i.g7, White usually chooses among 5 d3, after it was met in the Tal-Botvinnik 5 e3 and 5 :b 1. But the finesses rematch by 4 tbfJ! e4 5 tbd4 d5 start earlier if Black plays 1 c4 g6 6 cxd5. Transposing to the other 2 tbc3 i.g7, as Kasparov did. This line with 6 ... 'i'xd5?? is clearly enabled him to shift to his favorite King's Indian after 3 d4 tbf6. impossible and 6 ... cxd5 7 d3! favors White. But what about 3 g3 ? That seems to narrow Black's options since the However, this order IS quite Gruenfeld is discouraged and the playable after 4... d6!. Neo-Gruenfeld, 3... c6 and 4 ... d5, can be stopped by 4 i.g2/5 e4. Instead, Kasparov responded 3...tbc6!?
White's only real chance for advantage lies in d2-d4. After 5 i.g2 Black can choose between a book KID position with 5... g6 6 d4 and a well-known OlD position after 5•••.te7 6 0-0 0-0 7 d4 tbbd7. The latter is a good choice because the fianchetto is one of the least ambitious weapons White has
White has missed his chance for d2-d4. Moreover, 4 i.g2 d6 5 tbfJ e5! puts him in a closed variation with tbf3 and d2-d3, which theory calls fairly even.
201
Dutch Defense and Flank Openings
ENGLISH: ..•ii.b4 The reversed version of 1 e4 c5 2 tLlf3 tLlc6 3 ii.b5 has more nuance in the English because Black can play ... ii.b4 at move two, three, four or five. An instructive example is 1 c4 e5 2 tLlc3 tLlc6. White can try to punish Black's last move with 3 e3 and d2-d4-d5. Then 3 ...tLlf6 is an excellent waiting move.
Now 4 tLlf3 ii.b4! is a wellguarded line for Black. He has transposed into 1 c4 e5 2 tLlc3 tLlf6 3 tLlf3 tLlc6 and then 4 e3 ii.b4. Theory indicates that after 3 tLlf3 tLlc6 White's only serious chance for an edge lies instead in 4 g3!. That was a case of White tipping his center plans with his third move. He does better with 1 c4 e5 2 tLlc3 tLlf6 3 g3 or 3 tLlf3. But Black can anticipate that with 2 ... .tb4. He wants to know how White will retake after 3 ... .txc3. That's important information because the conservative dxc3 is recommended in some 2 tLlc3 tLlf6
3 tLlf3 tLlc6 4 g3 ii.b4 lines. But it's harmless after 1 c4 e5 2 tLlc3 ii.b4 3 g3 ii.xc3 4 dxc3?! because Black can make better use of his KN than transposing with ... tLlf6, e.g. 4••• d6 5 ii.g2 tLle7! 6 tLlf3 tLlbc6 with good chances. There are other English lines in which bxc3 is preferable. But in this case, on 4 bxc3 Black takes advantage of the absence of ... tLlf6 by playing ... f5 or ... f6.
For example, 4•.• d6 5 ii.g2 tLlc6 6 d3 f5 or 6 d4 tLlge7 7 tLlf3 0-0 8 0-0 f6! (9 ii.a3 ~h8 and now 10 d5? tLla5 11 'ti'a4 b6 12 c5 tLlxd5 13 cxd6 cxd6 14 tLlxe5 ii.b7 favored Black in Akesson-Sisatto , Stockholm 2006). The drawback to 1 c4 e5 2 tLlc3 ii.b4 is, of course, 3 tLld5 which both avoids ... ii.xc3 and gains time. But if you believe recent experience indicating that 3 ....td6 or 3 ....tc5 equalize, then 2 ... ii.b4! is an improved order for a Black who likes the positions that arise after 1 c4 e5 2 tLlc3 tLlf6 3 g3 ii.b4 or 3 tLlf3 tLlc6 4 g3 ii.b4.
202
Dutch Defense and Flank Openings
It may also pay to accelerate ... i.xc3 in that 3 tiJf3 tiJc6 4 g3 i.b4 line. Play usually continues 5 i.g2 0-0 6 0-0. Since 6... .l:te8 7 tiJd5! favors White Black's task is to time ... i.xc3 and ... e4 accurately.
Books consider 6... e4 at length and hardly mention 6..•i.xc3!?
After 7 bxc3 .l:te8 Black obtains a good version of 6....l:te8, in which he has dodged 7 tiJd5. And on 8 d3, which threatens to gain a positional edge with 9 e4, Black has 8 •.. e4!. If White retakes with the d-pawn Black may do best to avoid ... e4, which could transpose to a difficult 6... e4 7 tiJel i.xc3 8 dxc3 line. Instead he should maintain his darksquare pawn structure, e.g. 7 dxc3 d6 and then 8 tiJel i.e6 9 b3 'iVd7 and ..•i.h3. REVERSED DRAGON The mirror image of the Accelerated Dragon arises after 1 c4 e5 2 tiJc3 tiJf6 3 tiJf3 tiJc6 4 g3 d5 5 cxd5 tiJxd5 as well as after
3 g3 d5 4 cxd5 tiJxd5. Once more White's choice between the two depends on what he most wants to avoid. The first order allows Black alternatives such as 4 ... i.c5 and 4 ... tiJd4 that are quite reasonable. In contrast, after 1 c4 e5 2 tiJc3 tiJf6 3 g3, there is no ... tiJd4 at all, and 3•.. i.c5?! is premature because White has not committed himself to tiJf3.
This means he can get the upper hand with 4 i.g2 tiJc6 5 e3! followed by tiJge2 and d2-d4. For example, 5... a6 6 tiJge2 d6 7 d4 exd4 8 exd4 i.a7 9 h3 0-0 10 g4, as in Moskalenko-Diaz, Badalona 2000, among others. The main objection to the second order is 3 ... i.b4, which some English players regard as more promising for Black than 4 ... i.h4 is in the four-knights order. When the Dragon occurs in the four-knights order, White decides whether to delay d2-d3, as in
203
Dutch Defense and Flank Openings
1 c4 e5 2 liJc3 liJf6 3 liJf3 liJc6 4 g3 d5 5 cxd5liJxd5:
ENGLISH: l ... c5 Many 1 c4 players consider l...c5 the most unwelcome reply - not because of its fighting qualities but rather because of the drawishness of symmetry. If Black simply imitates his opponent, 2 liJc3 liJc6 3 g3 g6 4 .tg2 i..g7, White has a hard time developing a significant advantage.
In fact 5 e3 has long been regarded by GMs as 'a tacit draw offer,' as Andras Adorjan put it, There used to be a debate over because of the liquidating 5... e6 whether 6 d3 is necessary to avoid 6 liJge2 liJge7 7 d4 cxd4 8 liJxd4 6 .tg2 liJxc3 7 bxc3 e4. The issue d5! and then 9 cxd5 liJxd5 seemed to hinge on the soundness 10 liJxd5 exd5. Another potential of 8 liJd4. That question has not problem with 5 e3 is the been resolved but it now seems unbalancing and highly doub1e8 liJgl! f5 9 f3 favors White, edged 5... i..xc3 6 bxc3 f5. as does 8....tf5 9 'iVa4 'iVd5 10 liJh3 White can avoid both lines with (Yrjo1a-Westerinen, Finnish Cham- 2 g3!?, delaying liJc3. If Black pionship 1995). copies him, 2... g6 3 .tg2 .tg7, he That suggests 6 .tg2 is right provided White wants to quickly push his b-pawn, such as 6 ... liJb6 7 0-0 .te7 8 a3 or 8 .l:lbl. If he doesn't, 6 d3 is easier to handle.
can try 4 e3 and if 4... e6 then 5 d4 d5 6 dxc5! or 5... cxd4 6 exd4liJe7 7 d5! and 6... d5 7 c5. More 5liJe2!:
Note also that 1 c4 e5 2liJc3liJc6 would transpose into the fourknights lines following 3 liJf3 liJf6. But this doesn't happen often because theory says Black is doing well after 3 liJf3 f5 and White is doing better after 3 g3. If Black wants the four-knights, 2 ... liJf6 is the way to go. 204
common
IS
4... liJc6
Dutch Defense and Flank Openings
apparent in Seirawan-Dlugy, World Open 1989, which went 6 'ilVb3 tbc6 7 .txc6+ bxc6 8 tbf3 f6 9 'i'a4. Black was lost soon after 9 ... 'ilVd7 10 d3 e5 11 i.e3 tbb5? 12l::i.cl tbd4 13 0-0 h5? 14 h4 .
This enables White to transpose safely into normal lines (5 ... e5 6 tbbc3 or 5...tbf6 6 tbbc3) as ifhe had played 2 tbc3 and Black had passed up both the symmetrical ... e6/ ... d5 and the unbalancing ... .txc3.
If Black wants the reversed And if Black tries to restore Maroczy setup he has a sneaky way symmetry with S•.• e6, White can to reach it beginning with 1 c4 tbf6 break out with 6 d4 and then and then 2 tbc3 dS 3 cxdS tbxdS: 6... cxd4 7 exd4 dS 8 cSt? (Grischuk-Malakhov, Moscow 2006). That modest 2 g3 can pose problems for a Black who likes to play the 'improved Tarrasch Defense,' 1 c4 c5 2 tbf3 tbf6 3 tbc3 e6 4 g3 tbc6 5 .tg2 d5 6 cxd5 tbxd5!. In the 1 c4 cS 2 g3 order, he discovers that 2...tbf6 3 .tg2 dS 4 cxdS tbxdS S tbc3 puts immediate pressure on d5.
This variation is almost always treated as some offbeat Gruenfeld, with analysis of 4 g3 met by 4 ... tbxc3. However, Black also has 4... cS! and then 5 i.g2 tbc7.
If Black responds with the natural S... e6 he reaches his goal after 6 tbf3 b6 or 6...tbc6, as has This setup is more often created happened in dozens of master by 1 c4 c5 2 tbc3 tbf6 3 tbf3 d5 games. But 5... e6?? 6 tbf3?? is a 4 cxd5 tbxd5 and 5... c5. But in that double blunder since 6 tbxdS exdS order White can avoid the Maroczy 7 'ilVb3! wins a pawn, as in the with 2 tbf3!? and 3 d4. Vladimir Kramnik debacle cited in In the last diagram White has the Introduction. another chance to avoid the bind, If an 'improved Tarrasch' player with 4 tbf3. But he has to know realizes the problem in time, he can what he's doing after 4...cS S d4 or recover with S...tbc7! and seek a S e4. reversed Maroczy Bind. But that's Closely related to those positions not an easy system to play for the are the ones arising from 1 c4 cS first time as Black. That was 2 tbf3 tbf6 3 g3. This innocent205
Dutch Defense and Flank Openings
looking order was prepared by When White focuses on d5 with Kramnik for his world champion- 2 tbc3, rather than 2 g3 or 2 tbf3, ship match with Kasparov. On the character of the game changes 3...tbc6 he could respond 4 tbc3 slightly. Black can fall into the and reach a Symmetrical Four Kramnik trap through a back door, Knights. That would be a moral 2 ... tbf6 3 g3 d5 4 cxd5 tbxd5 victory for Kramnik because he 5 i.g2 and now 5... e6?? 6 tbxd5 liked the Four Knights, while cxd5 7 'iib3. Kasparov, like many other players, One of the drawbacks to 2 tbc3 is avoided it (by means of 1 c4 c5 that 2 ... g6!? allows Black to 2 tbc3 tbf6 3 tbf3 d5 or 2 ...tbc6 concentrate firepower on d4. His 3 tbf3 tbd4). point is that 3 tbf3 i.g7 4 d4 cxd4 A key position arises after 3 ... d5 5 tbxd4 tbc6 stops the Maroczy 4 cxd5 tbxd5 5 i.g2. Bind:
The natural 5... g6? is faulty because 6 d4! cxd4 7 "iVxd4 attacks Black's KR. Black is also worse after 6 ... i.g7 7 e4 tbc7 8 d5. What has happened is we've reached a Neo-Gruenfeld (1 d4 tbf6 2 c4 g6 3 g3 d5 4 i.g2 i.g7 5 tbf3) in which Black has played 5... c5?! (and 6 cxd5 tbxd5 7 e4!) instead of the superior 5... 0-0 or 5... dxc4.
The d4-knight is under attack before White can smooth out his development with 6 e4 and 7 i.e3. If the knight retreats to b3 or c2 then 6 ... i.xc3+! puts White's queenside under severe pressure.
Trading that bishop was once thought to be foolhardy but that view was overthrown by examples like 6 tbc2 i.xc3+ 7 bxc3 tbf6 8 f3 But in the diagram Black can "iVa5 9 i.d2 "iVa4 10 e4 d6 and avoid all of this with the simple ...tbd7-c5, with excellent chances 5•..tbc6!. in Agzamov-Taimanov, Riga 1975. 206
Dutch Defense and Flank Openings
There is a possible drawback to 1 c4 c5 2 liJc3 g6 in the rarely played 3 d4. Then 3•.. cxd4 4 'iixd4 attacks the KR. This leads to 4 ... liJf6 5 i.g5 liJc6 6 'iid2, which is hard to evaluate, when compared with vaguely similar positions, such as 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4 4 liJxd4 g6 5 c4 liJf6 6 liJc3 liJxd4 7 'iixd4 d6 8 i.g5 i.g7 9 'iVd2.
prefer 6 ... liJa5, e.g. 7 e4 liJf6 8 h3 0-0 9 .td3 a6 10 0-0 b5 11 cxb5 c4 12 i.c2 axb5 13 liJxb5 liJb3! (lzoria-Savchenko, internet 2006). Despite that, and support for 6 ... i.xc3+ 7 bxc3 liJa5 as well, theory doesn't trust Black's system.
But if there is no knight on c6, the pawn push is weak, as BisguierKarpov, Skopje 1972 showed. After There are two other ways of 1 c4 c5 2 liJc3 g6 3 liJf3 i.g7 4 e3 trying to reach the last diagram. liJf6 5 d4 Black continued 5•.• 0-0 Bobby Fischer met 1 c4 on occasion 6 .te2 cxd4! 7 exd4 d5! with an with 1...g6 2 liJc3 i.g7 3 liJf3 c5. excellent game, as we saw with But Black has to be willing to play 1 d4liJf6 2 c4 c5 3 e3. a KID or Modem (3 d4). RETI OPENING Fischer also met 1 c4 c5 2 liJc3 with 2.•.liJc6 and then 3 liJf3 g6. This rules out a 'iixd4 recapture after d2-d4/ ... cxd4. Black gets where he wants to go following 4 d4 cxd4 5 liJxd4 i.g7. But White can play 4 e3! and threaten 5 d4 and 6 d5, which is stronger when it attacks a knight on c6.
Kramnik made 1 liJf3 famous as a purely transpositional device. After 1...d5 he reaches normal Queen's Gambits with 2 d4 and 3 c4. But after 1...liJf6 he almost always replies 2 c4. In this way he avoids the NimzoIndian by meeting 2 ... e6 3 liJc3 ?-. \lUI Ab4 with 4 wc2!?, rather than 4 d4. He also dodges normal Gruenfeld positions, meeting 2 ... g6 with 3 liJc3 d5 4 cxd5 liJxd5 5 'iVa4+ or 3 ... i.g74 e4!. That saves gallons of midnight oil. Note that 1 c4liJf6 2 liJf3 would get White to the same position as 1 liJf3 liJf6 2 c4. But 1 c4 runs the risk of 1...e5!?
Fischer was willing to play 4 ... i.g7 5 d4 d6 6 d5 liJe5. Others
Of all the major first moves, 1 liJf3 grants Black his best
207
Dutch Defense and Flank Openings
- 1...f5 tries to exploit White's opportunity for 1... b6 since White cannot build a center with £2-f3 and early commitment of his KN. cannot easily transpose to an But as we saw in 1 c4 f5 White can e2-e4/d2-d4 position. After 1 1LJf3 try for e2-e4, even starting with 2 d3. b62 e4 ~b7 31LJc31LJf6!:
The attack on the e-pawn interrupts White's center-building, 4 e5 lLJd5 5 lLJxd5 .txd5 6 d4 e6, e.g. 7 c4 ~b4+ or 7 .te2 c5 8 c4 ~xf3! 9 ~xf3 lLJc6 10 d5 lLJxe5 11 ~e4 f5 with a fine game in Shchekachev-Bauer, ClermontFerrand 2003. Black's other options are more elastic: - l...c5 is a favorite of Sicilianistas (2 e4). Since 1 1LJf3 players generally prefer closed lines, the typical result is a Symmetrical English (2 c4) or a King's Indian Reversed (2 g3). - 1. .. d6 seems to say, 'I'm offering a KID this way (2 d4 lLJf6 and 3 ... g6) because I don't want to end up in a Pirc after 1...g6 2 e4.' If White tries to trick him into a Pirc, 1... d6 2 e4, he can find himself in a Sicilian,2 ... c5.
White's goal is to transpose into a favorable version of the Lisitsyn (2 e4) Gambit after 2.••lLJf6 3 e4 fxe4 4 fxe4 lLJxe4 5 ~d3 lLJf6 6 lLJg5 and lLJxh7! or ~xh7!. The original Lisitsyn runs 2 e4 fxe4 3 lLJg5 lLJf6 4 d3 exd3 5 ~xd3 but Black gets the edge with 3... d5! in that version. If Black is willing to playa ... d6 Dutch, he should reply 2 ... d6, which avoids the gambit as well as 2 ... lLJf6 3 ~g5. But if Black wants a Stonewall, then 2 d3 is a major problem for him. - 1. .. g6 keeps more Black options open. Against 2 e4 he can shift to a hyper accelerated Dragon Sicilian with 2 ... c5 or head toward a safe Pirc, with 2... d6 or 2 ... ~g7, because White has ruled out £2-f4 and £2-f3 lines.
208
Dutch Defense and Flank Openings
After 2 d4 or 2 c4 Black retains the option of various kinds of Modem Defenses, Englishes and the Dutch Indian (2 ... f5). If he meets 2 d4 with 2... c6 the results include a Neo-Gruenfeld (3 g3 d5), a Modem (3 e4 d5), or some nondescript Indian (3 c4 d6 4 tbc3 i.g7 or 4 ... i.g4).
- 1... e6 is, next to 1. .. tbf6, the most elastic response to 1 tbS. It's often used by a Dutch loyalist (2 d4 f5) who is trying to trick White into a French (2 e4 d5!) or a Sicilian (2 e4 c5). RETI: 1...d5 After 1...d5 the Hyperrnodemist will play 2 c4 or 2 g3. But the transpo-minded will look at 2 e3.
Taimanov used this to get into a Slav (2 ... c6 3 c4 tbf6 4 tbc3 e6 5 d4) or QGA (2 ... tbf6 3 c4 dxc4 4 i.xc4 e6 5 d4, dodging ... i.g4 lines). White enjoys many of the same opportunities as he would after 2 c4 c6 3 d4 tbf6 but without having to find an answer to 2 c4 d4!?
More common after 1 tbf3 d5 is 2 c4. The simplest defense is 2... dxc4. Then 3 e3 becomes a book QGA in the vast majority of games (3 ... tbf6 4 i.xc4). Can White do better? Richard Reti liked 3 tba3. But 3 ... c5 4 tbxc4 tbc6 and ... e5 transposes into a kind of English (1 c4 c5 2 tbs tbf6 3 g3 d5 4 cxd5 tbxd5 5 i.g2 tbc6 6 0-0 e5 7 d3 i.e7) in which White has chosen a suspect knight development, 8 tbbd2 and tbc4. After 3 tba3 fell out of fashion, White tried 3 'i'a4+. The simplest response to that is 3...tbd7 and ... a6/ ... b5, seeking a Catalan.
For example, 4 'i'xc4 e6 5 g3 a6 6 i.g2 b5. And on 4 g3 a6 5 tbc3 e6 6 i.g2 ltb8!. The threat of 7 ... b5 leads to 7 'i'xc4 b5 8 'i'b3 i.b7. Then White has nothing better than to transpose into an equal Catalan with d2-d4. That is, 1 d4 tbf6 2 c4 e6 3 g3 d5 4 i.g2 dxc4 5 'i'a4+ tbbd7 6 tbS a6 7 tbc3 ':b8 8 'i'xc4 b5 etc.
209
Dutch Defense and Flank Openings
If Black doesn't like these positions he should consider meeting 2 c4 with 2•••c6 since the Slav Defense (3 d4) is holding up well these days. White can remain true to the spirit of Reti with 3 b3. Today's theory says Black can equalize whether he adopts Lasker's defense ( ... i.f5) or Capablanca's ( ... i.g4). A major reason is that 3 b3 rules out 'ifb3, the attack on b7 that is so troublesome to Black in ... d5/ ... i.f5 openings.
4lbf3 e4 5lbd4 i.c5! 6lbxc6 dxc6 7 e3 i.f5 8 'ifc2 'ife7 9 i.e2 0-0-0 and White was virtually lost after 10 f4? Theory concluded that White should play lbe2 and f2-f4 instead. But another point of view says Larsen's real error was playing c2c4 before lbf3. Instead of 3 e3 and 3 c4 lbf6 4 e3, which are played most often today, White may prefer 3lbf3.
The modem treatment of the Reti is to delay both b2-b3 and c2-c4. White hints he is willing to play a King's Indian Reversed instead, with 1 lbf3 d5 2 g3 lbf6 3 i.g2. Then on 3•.•c5 4 0-0 e6 he can act in the center with 5 d3 and lbbd2/e2e4. The KIR enjoys a pretty good reputation against ... e6. If, however, Black brings out his QB, with 3•••i.f5. White can switch from the KIR back to the Reti with 4 c4. In this way he avoids 2 c4 d4 and 2 ... dxc4. NIMZO-LARSEN OPENING Bent Larsen helped make 1 b3 e5 2 i.b2 lbc6 famous by attacking the e-pawn with lbf3. Then after ... e4 he occupied the central dark squares with pieces and assaulted ... d5 with c2-c4. But this strategy was almost retired in 1970 by a Larsen-Spassky miniature, which went 3 c4 lbf6
Black defends the e-pawn with ... f6 in some 1 b3 lines but here 3••. f6?! 4 e4! and 5 i.c4 can't be good for him. Moreover, 3... d6, which is fine following 3 e3, allows White to punch at e5 with 4 d4, e.g. 4•••e4 5 d5! exf3 6 dxc6 bxc6 7 gxf3 lbf6 8 'ifd4! with advantage. For example, 8... i.f5? 9 e4 i.g6 10 lbd2 i.e7 11 'ifa4! 'ifd7 12 i.h3! resigns, Hansen-Bergfalk, correspondence 1992, or 8... i.e7 9 l:tgl c5 10 'ife3 0-0 11 lbc3 g6 12 0-0-0 l:te8 13 'iff4 and 14 e4. Note also that in similar positions ... i.d6 defends the e-pawn well -
210
Dutch Defense and Flank Openings
1 b3 e5 2 ~b2 ~c6 3 e3 ~f6 4 ~b5 ~d6!' But it fails here, 3 ~f3 ~d6 4 ~a3! and ~c4 or ~b5.
without reaching the Spassky position (5 ... .i.xc5?? 6 ~xc6 and
Black most natural response to 3 ~f3 is 3 .••e4 and then 4 ~d4 ~f6. The absence of c2-c4 plays a role in S e3 ~cS 6 ~xc6 dxc6:
If White likes such positions he should be happy (after 1 b3 e5 2 i.b2 ~c6) with 3 e3 d6 because of 4 ~f3, rather than the bookendorsed 4 .i.b5.
~xg7).
This is better for White than Larsen-Spassky after 7 d4!? ~e7 8 c4 and perhaps after 7... exd3 8 ~xd3 as well. If instead S...~xd4 6 ~xd4 dS then 7 c4 is timed well and transposes into positions Larsen had shown to be good. For instance, 7 ... ~f5 8 ~c3 dxc4 9 ~xc4 ~d6 10 'ifc2 0-0 11 f4! c6 12 0-0 l:e8 13 ~dl 'ife7 14 ~f2, Welling-Minasian, Capelle la Grande 1996. Nevertheless books say that after 1 b3 eS 2 ~b2 ~c6 the best move is 3 e3. They often give 3... dS followed by 4 ~b5 ~d6. Yet here again ~f3 is an attractive idea 4 ~f3 e4 S ~d4 ~xd4 6 ~xd4 transposes into the above line, again
Now 4... e4 will bring about the desired position, a tempo ahead, after S ~d4 ~xd4 6 .i.xd4 when Black inevitably plays ... dS. Instead 4 ... ~f6 S d4! favors White after 5... exd4 6 ~xd4 i.e7 7 ~b5, e.g. 7... i.d7 8 .i.xc6 bxc6 9 'iif3! d5 10 ~f5 as in HodgsonPercerias, Breda 1989. White has transposed into a good 3 e3 d6 4 ~e2 .i.e7 5 d4 exd4?! 6 ~xd4 line. And if White is happier facing 1 b3 d5 than 1...e5, he should consider 1 ~f3 so that 1.. .d5 2 b3! gets him closer to the middlegame he wants.
211
212
Index of Opening Variations (numbers refer to pages)
Albin Counter Gambit
123
Alekhine's Defense
94
Alekhine's Four Pawn Attack
94
Alekhine's 4 lDf3
95
Benko Gambit
156
Benoni Defense
159
Bishop's Gambit
16
Bishop's Opening
19
Black Knights Tango
190
Blackmar-Diemer Gambit
154
Bogo-Indian Defense
183 14
Breyer Gambit Caro-Kann Defense
105
Caro-Kann, Panov Variation
108
Caro-Kann, Panov Accelerated
110
Caro-Kann, Two Knights Variation
106
Catalan Opening
197
Center Counter Defense
97
Center Counter, 2 ... lDf6
99
213
Index of Opening Variations
Center Game
18
Colle System
121
Danish Gambit
19
Dutch Defense
191
Dutch by transposition
194
English Defense
199
English Opening
199
English, Reversed Dragon
203
English, 1.. .c5
204
English, 1.. .e5
200
English, ... i.b4
202
Evans Gambit
31
Four Knights Game
28
French Defense
111
French Defense, Advance Variation
117
French Defense, Burn Variation
111
French Defense, Rubinstein Variation
111
French Defense, Steinitz Variation
112
French Defense, Tarrasch Variation
115
French Defense, Winawer Variation
113
Giuoco Piano
29
Goring Gambit
32
Gruenfeld Defense
176
Gruenfeld, Exchange Variation
178
Gruenfeld, Russian Variation
176
214
Index of Opening Variations
Hungarian Defense
17
King's Gambit
14
King's Gambit Declined
15
King's Indian, Classical Variation
169
King's Indian, Fianchetto Variation
170
King's Indian, Four Pawns Attack
173
King's Indian, Samisch Variation
167
King's Indian, h3
174
King's Indian, iog5
172
King's IndianlBenoni
161
Maroczy Bind
64
Max Lange Attack
29
Modem Benoni Defense
161
Modem Defense, 1 d4
175
Modem Defense, 1 e4
99
Modem, Gurgendize Variation
101
Neo-Gruenfeld Defense
180
Nimzo-Indian Defense
184
Nimzo-Indian, Leningrad Variation
184
Nimzo-Indian, Queen-move variations
185
Nimzo-Indian, Rubinstein Variation
189
Nimzo-Indian, Samisch Variation
188
Nimzo-Larsen Opening
210
Old Indian Defense
164
Petroff Defense, 3 d4
22 215
Index of Opening Variations
Petroff Defense, 3 lbxe5
22
Philidor Defense
20
Pirc Defense
102
Ponziani Opening
24
Pseudo-Lopez
20
Queen's Gambit Accepted
124
Queen's Gambit Accepted, 3 ... lbf6
126
Queen's Gambit Declined
129
QGD, Botvinnik Variation
142
QGD, Cambridge Springs Defense
140
QGD, Exchange Variation
134
QGD, Lasker Defense
140
QGD, Orthodox Variation
138
QGD, TarraschlSemi-Tarrasch
144
QGD, Vienna Variation
143
Queen's Indian Defense
181
Queen's Pawn Game
120
Reti Opening
207
Ruy Lopez, Anti-Marshall Variation
46
Ruy Lopez, Arkhangel Variation
44
Ruy Lopez, Berlin Defense
34
Ruy Lopez, Bird Defense
37
Ruy Lopez, Classical Defense
34
Ruy Lopez, Cozio Defense
35
Ruy Lopez, Exchange Variations
42
216
Index of Opening Variations
Ruy Lopez, Fianchetto Defense
35
Ruy Lopez, Marshall Gambit
46
Ruy Lopez, Main Line
49
Ruy Lopez, Modern Steinitz
39
Ruy Lopez, Moeller Defense
44
Ruy Lopez, Old Main Line
51
Ruy Lopez, Open Defense
42
Ruy Lopez, Steinitz Defense
37
Schmid Benoni
160
Scotch Gambit
32
Scotch Game
25
Semi-Slav Defense
151
Semi-Slav, Meran Variation
152
Sicilian Defense, Alapin Variation
60
Sicilian, Chameleon Variation
55
Sicilian, Classical Variation
76
Sicilian, Closed Variation
55
Sicilian, Dragon Variation, Accelerated
72
Sicilian, Dragon Variation, Classical
69
Sicilian, Dragon Variation, Fianchetto
65
Sicilian, Dragon Variation, Levenfish
69
Sicilian, Dragon Variation, Yugoslav Attack
65
Sicilian, English Attack
74
Sicilian, Four Knights Variation
92
Sicilian, Grand Prix Attack
58
217
Index of Opening Variations
Sicilian, Kan Variation
87
Sicilian, Keres Attack
74
Sicilian, Moscow Variation
63
Sicilian, Najdorf Variation
84
Sicilian, Rauzer Variation
78
Sicilian, Rossolimo Variation
63
Sicilian, Scheveningen Variation
73
Sicilian, Sozin Variation
81
Sicilian, Sveshnikov Variation
92
Sicilian, Taimanov Variation
87
Sicilian, .. :"b6
90
Sicilian, 4 'ii'xd4
56
Sicilian, 5... J.d7
84
Sicilian, 5 ... lLlbd7
83
Slav Defense
144
Slav, Exchange Variation
145
Slav, Three Pawns Variation
148
Staunton Gambit
192
Stonewall Attack
191
Torre Attack
155
Trompowsky Attack
153
Two Knights Defense
29
Vienna Game
16
218