You are on page 1of 220

Transpo Tricks in Chess

Finesse your Chess Moves and Win

Andrew Soltis

First published in the United Kingdom in 2007 by
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

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 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


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


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 In 99 of 100 games Black
reach the middlegame they want continues 5... ttJd6 or 5... .te7 and
to play. It's rarely the same 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 calcul- Since it's a footnote we take little
ating ability or your iridescent notice. After all, 5 ... a6 'just
personal charm. What you can use transposes.' That tells us it doesn't
is trickery - the trickery of really matter because it only leads
transposition. to some other opening, something
irrelevant on that page.
Consider this opening: 1 e4 e5 But 5... a6 has been a valuable
2 ttJO ttJc6 3 .tb5 ttJf6 4 0-0 ttJxe4 weapon for players from Paul
5 d4, the main line of the Ruy Morphy to Vasily Ivanchuk. Its
Lopez's Berlin Defense. power lies in how it gets Black


where he wants to go - to the Open We don't judge transpositions by

Defense - and avoids what he wants the same standard as we do other
to avoid - the Exchange Variation moves. An original opening idea, a
(3 ... a6 4 .txc6). 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

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
This is a 'tabia,' that is, a standard
transpose into a favorable variation
starting point in a major opening. In
which would not be reached in any
this case it's a tabia that has served
normal manner,' Reuben Fine
as the launching pad for thousands
of Dragon Sicilians, which often
continue 10 ... ttJe5 and .. J:tc8/
In truth, there is a normal
... ttJc4.
manner, the discredited 4 a3 .txc3+
5 bxc3 d5?!, which allows White to But in a 1997 game, Anand-
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
version of 4 a3. 11 .tb3 Black could transpose into


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 equanim-
Objectively, 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
objective strength. 15 moves. Players who lose their
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 But that meant choosing
policy, 12... a6? In the Dragon such 6 i.xc6?!, which turned out to be a


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
How it gets your opponent to not reduce his options. Rather they
think increase them, giving him more to
The real battle of the opening
begins when you can force your There really isn't much value, for
opponent out of his book know- example, to 1 d4 c6 and then
ledge. Only then does he risk 2 c4 d5 compared with the normal
making errors and spending costly route, 1 d4 d5 2 c4 c6. But this
minutes. order, employed by Anatoly Karpov
among others, gets White debating
That was case when Vishy Anand
with himself over whether he
was Black in a speed playoff game
knows more about the Caro-Kann
at a big-bucks event in 1994. In the
(2 e4) than the Slav. Or what he
Petroff Defense, 1 e4 e5 2 ~f3 ~f6
would do about 2 c4 b5!? A strange
3 d4 ~xe4, his opponent failed to
move order can do that.
play 4 .ltd3, the move considered
virtually automatic. He played A 'something to think about'
4 ~xe5!? instead. move can cost your opponent more


than minutes. It can prompt a bad And there's a fourth way of

decision. The most drastic recent evaluating a crafty transposition,
example of that befell Vladimir by:
Kramnik in what was then the most How it preserves your mental
important game of his life, the final health
game of a 1994 Candidates match.
We all have to deal with an ever-
His 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
The familiar path is 4 ... c5 and
good for Black.
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
several years. Black.


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'
This is another book POSItIon
move order.
from the Keres Attack - but White
Every transpositional trick has to has lost a move because he spent
be judged by a balance sheet: What two tempi to get his h-pawn to h4.
are the benefits? What are the A lost tempo should make a huge
drawbacks? Which order counts difference.
more? Some, like 4 ... liJe7, may
But what mattered most was
have more plusses than minuses. Black's unfamiliarity with Keres-
Other's like Kramnik's 5... e6?? are like positions. After 9...liJde5?!
disastrous. 10 i.e3 b5?! she was worse and
But the vast majority are after 11 liJxc6 liJxc6 12 'ifd2 i.b7
somewhere in between. The bottom 13 f4 ~c7 14 'ifflliJb8 15 a3liJd7
line is not whether they give you a 16 f5 liJe5? 17 fxe6 fxe6 18 i.h3
superior position but whether it's a she was lost. White's objectively
position you want to play. Ideally, bad 8 g5? worked brilliantly as a
it's also one your opponent wants to transposition. It was really 8 g5!.
Consider 1 e4 c5 2 liJf3 liJc6 3 d4 'IT JUST TRANSPOSES'
cxd4 4 liJxd4 liJf6 5 liJc3 d6 and
Some opening positions can
now 6 h3 e6 7 g4. This transposes
come about from two, three or more
into a conservative version of the logical orders. Books - and even
Keres Attack, in which White relies experienced players - tend to
on a slow buildup with i.g2, rather dismiss an alternative route because
than g4-g5. 'it just transposes'. But each route
However, Andreikin-Kosteniuk, is likely to have benefits and
Tomsk 2006 went 7... a6 8 g5liJd7 liabilities that a shrewd transposer
9 h4. knows to evaluate. Take this


position from the Taimanov realizes Black isn't going to allow

Sicilian. It's White's move and 10 e5. So he'll be aware that Black
when it was analyzed, by Mark is planning to transpose, with
Taimanov, in the 1972 edition of 9....tc5! to the diagram, and may
ECO he claimed Black wins. 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
In fact, White has a good reply, three, 8... .te7 may be Black's best
10 .!Df5!. The critical line is route to the diagram.
10....!De7!, which forces 11 .!Dxg7+
~f8 12 .txc5 'i'xc5+ 13 ~h1 THE LURE OF THE
~xg714 e5. 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
The second route is forcing. with an unfamiliar move, they are
Black attacks the h-pawn with strongly, even irrationally, tempted
S....td6. White's instinct is to meet to look for a way to reach a
the threat with a threat, 9 f4. But he recognizable position.


Laszlo Szabo fell victim to that 2 ttJf3. He knows that books

temptation in the 1953 candidates recommend 2 d4!. But about half of
tournament when his game with all 1 e4 ttJc6 games continue 2 ttJf3
Paul Keres began 1 d4 d5 2 ttJf3 because White wants to transpose to
ttJf6 3 c4 dxc4 and then 4 ttJc3 and the familiar (2 ... e5).
4 a6.
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

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,
That was unfamiliar to him. But
1 e4 c5 2 ttJf3 ttJc6 3 d4 cxd4
he remembered getting a very good
4 ttJxd4 ttJf6 5 ttJc3 g6.
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
This order gives Black all the
player takes the easy way out of an
tactical benefits of an Accelerated
unfamiliar position and transposes
Modem Dragon without allowing
to a recognizable one. He may do
the Maroczy Bind as the Acceler-
this even if he suspects - or knows
ated does. But there is a big minus -
for certain - that he is playing an
6 ttJxc6 sentences Black to either a
inferior move.
dubious middlegame (6 ... bxc6 7 e5)
You've seen this happen when a or a no-fun endgame (6 ... dxc6
1 e4 player meets 1...ttJc6 with 7 'ifxd8+).


Nevertheless the trickster will Examples like that should tell

weigh the risk of being squeezed to you to beware those words 'it just
death in a Maroczy Bind against the transposes.' There is usually a plus,
risk of White playing 6 tiJxc6!. The a minus, an extra option to consider.
odds are on his side. A database If you don't, you are allowing your
check finds White played 6 tiJxc6! opponent to control the direction of
less than a third (327 of 1,000 the opening. And the price may be
games) of the time. Even super- ending up in the middlegame he
GMs like Bobby Fischer and wants.
Mikhail Tal played the inferior
6 i.e3 to return to a position they

Chapter One:
Double KP Openings

Trickery begins with the oldest Breyer Gambit, 1 e4 e5 2 f4 exf4

openings, like the King's Gambit 3 'fif3 d5 4 exd5. But the Breyer is
Accepted. Today's theory says the harmless and Black can even try the
KGA (1 e4 e5 2 f4 exf4 3ltJf3) is no sacrificial 4 ... ltJf6 5 J.b5+ c6!?,
longer dangerous because of 3... d5 recommended by Yacov Estrin.
and then 4 exd5 ltJf6 with equal
chances. This order can upset a Bishop's
Gambiteer. There are few 1 e4 e5
If that makes Black happy, he 2 f4 exf4 3 .tc4 players who are
should consider the 'something to comfortable in KGA positions,
think about' order of 1 e4 e5 2 f4 d5 which they would reach by playing
3 exd5 exf4!?, as played by Anatoly 4 ltJf3 in the diagram.
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
What is there to think about? normal Bishop's Gambit, that is
Well, for starters, White wonders 1 e4 e5 2 f4 exf4 3 J.c4 ltJf6 4 ltJc3
whether he should head into the c6 5 d4 d5 6 exd5. But it takes a lot
book line with 4 ltJt3 ltJf6 or try of clock time to figure that out over
4 'ift3, which transposes into the the board.

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
(8 g3 i.h3! 9 :f2 liJf6! or 9 gxf4 In contrast with the normal
'i'd7! 10 :f2liJf6). Allgaier, 1 e4 e5 2 f4 exf4 3 liJf3 g5
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.
Bobby Fischer's try was 3... d6
It may be his best but after 5 d4 g5
and then 4 i.c4 h6! 5 d4 g5 6 0-0
6 g3! we've reached a position that
i.g7 7 c3 liJc6 reaches the tabia.
often occurs in KGA lines and the
The problem with 3... d6 is that evaluation is at best fuzzy.
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-for-
confusing one, 4liJc3!? Black tabia. This can confuse White

Double KP Openings

unless he knows that 4 d4 gS and BISHOP'S GAMBIT

S ttJc3! d6 6 g3! reaches the fuzzy
If your opponent is a Bishop's
Gambiteer, you have good reason to
That leaves us with 3. g5. The fear he's better prepared than you, a
Romantic-era continuation was lot better. The careful transposer
4 .tc4 g4 but 4.ttJc6! clears a path will answer 1 e4 e5 2 f4 exf4 3 .tc4
to the tabia. with the forgotten favorite of
Mikhail Tchigorin, 3... ttJc6.

Black invites S 0-0 .tg7 6 d4 h6

and so on. Once White plays 4 .tc4 Black can then meet the natural
he can't insert h2-h4 effectively 4 ttJf3 with 4.. g5!, reaching that
because S h4? g4! 6 ttJgS good form of the KGA that we
boomerangs badly after 6... ttJeS! considered in 3 ttJO gS 4 .tc4
7 .tb3 h6 8 d4 hxgS 9 dxeS .tg7. ttJc6!. This is a psychological plus
The best answer to 3 ... gS is because, as noted before, Bishop's
supposed to be 4 h4, which rules Gambiteers are often uneasy in a
out the tabia after 4 ... g4 S ttJeS. KGA.
However, the analysis of that The bonus is that White has some
position is immense and White may natural but bad responses, such as
be tempted by the confuse-move 4 ttJc3? which allows a strong
4 d4. Then 4 ... g4 S .txf4!? gxf3 4 ... 'iVh4+ S ~f1 .tcS. Whether
6 'iVxf3 is an unusual Muzio-like White has an edge after 4 d4 ttJf6
line that served Alexander S eS dS! is murky (6 .tb3 ttJe4).
Morozevich well.
The safe response is 4 d4 h6 VIENNA GAME
since S h4 .tg7 6 g3 g4! is There are two types of Vienna
promising for Black. But S ttJc3! players. The first don't want to
and S... d6 6 g3 puts us in the fuzzy memorize volumes of Ruy Lopez
position again. theory. The second is booked up

Double KP Openings

and prepared to inflict volumes of defend - any- cramped -posi tion

Vienna theory on you, such as with reputation but it's based on White
1 e4 eS 2 lbc3 lbf6 3 ii.c4 lbxe4 expanding with c2-c3 and d2/d4 or
4 'iih5. '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
This looks prohibitively passive. 7 0-0 lbd7 8 d5 lbcb8 (Movsesian-
But White has to be careful since Efimov, Pula 1997).
3 f4 exf4 and then 4 lbf3 h6! S d4
There's much more experience
g5 is that fuzzy KGA again. with the other 'beginner's move,'
White's best is probably 4 d4 or 2 .ii.c5, which has been put to good
4 ii.c4. But few Vienna players are use by Karpov, Bent Larsen and
familiar or comfortable with KGA Vasily Smyslov. Books used to
lines in which Black gets to play claim Black is already worse in
... 'iih4+. view of 3 lbf3 lbc6 4 lbxe5! lbxe5
5 d4.
Well, what about 3 ii.c4 ? Then
3 ... lbc6 leaves White to decide But Black has a good alternative
in 3 .. d6! since 4 d4 exd4 S lbxd4
whether to allow 4 d3 lbaS!? or lbe7 or 5 ... lbc6 heads toward a
plunge into another unfamiliar reasonable version of the Scotch
KGA (4 f4 exf4 5 lbf3 h6 or 5 ... g5 Game after 6 ~e3, 6 lbb3 or
6 h4 g4 7lbgS lbeS). 6lbxc6 'tWf6.
Instead, he may assume that For example, 5 ... lbc6 6 ~e3
simple development, 4 lbf3, will ii.b6 7 lbdS lbf6 8 lbxb6 axb6 9 f3
punish 2 ... d6. That transposes into a 0-0 10 ii.c4 d5 !. Or 10 c4 l:te8
Hungarian Defense after 4...ii.e7!? lllbc2lbd7 12 'i'd2lbde5 13 ~e2
The Hungarian has a stodgy, I-can- ~e6 14 b3 f5 with good play for

Double KP Openings

Black in Zarnicki-Bianchi, Buenos This often has the effect of

Aires 1989. scaring Black into another defense,
such as 5 ... iLe7 6 ttJc3 0-070-0-0-
White can stay in Vienna mode
with 3 f4. Then 3... d6 is a King's that is, transposing into the less
Gambit Declined in which White ambitious 5 ttJc3 iLe7 6 iLd2 0-0.
has forsaken c2-c3, his most But this is primarily a bluff
ambitious plan in the KGD, and because if Black meets 5 iLd2 with
books disagree about his chances 5... iLb4! White has nothing better
for advantage. than transposing into the book line
There is also 3 g3, which with 6 ttJc3!.
transposes into a main line of Louis The other finesses in the Center
Paulsen's variation. Black's only Game arise if Black is afraid of a
minus is relinquishing the book- line that hardly anyone knows
recommended defense (I e4 e5 about. This is 5 e5!?, rather than
2 ttJc3 ttJf6 3 g3 d5!). So 2... iLc5 5 ttJc3 or 5 iLd2. That seems to just
has no major drawbacks or benefits lose a pawn but 5... ttJg4 6 'ife4
and is more of a 'something to think ttJgxe5? allows 7 f4.
about' move.
To play 4 ... ttJf6 with confidence
Black should know a little about
crazy lines like 6... d5 7 exd6+ iLe6
Most players who meet 1 e4 with and then 8 iLa6 or 8 iLe2! ttJf6
1... e5 know that 2 d4 exd4 3 'ir'xd4 9 dxc7 'ir'xc7.
ttJc6 4 "iVe3 is neutralized by
But he can save himself a lot of
4..ttJf6 5 ttJc3 ii.b4 6 iLd2 0-0.
worry and/or study by picking
That's about all that anyone
another fourth move. One is
remembers. But White has a
4 ..iLb4+, with the idea of reaching
confuse-move in 5 iLd2.
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.

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
But there's another wild card for dxc3 5 tbxc3. This forgotten
Black to worry about, 5 'iVg3 and variation is quite playable after
then 5...tbf6 6 e5, which Paul Keres 5...tbc6 6 i.c4 d6 7 tbg5 .Jl.xc3+
said favors White, or 5... iu6 6 tbc3 8 bxc3 tbe5 and now 9 ..tb3 h6
tbge7 7 .td2 as in Ziemacki-
10 f4! hxg5 11 fxe5 'iie7 12 ifd5
Fatyga, correspondence 1992
dxe5 13 ..txg5!.
(7 ... d6 8 0-0-0 .te6 9 f4).
The main benefits to 1 e4 e5 2 d4
So Black has to decide whether
the drawbacks of 4 ... .tb4+ or exd4 3 lbf3 are forcing Black to
4 ... .te7 outweigh the merit of start thinking at move three and
avoiding the rare 4 ... tbf6 5 e5!? preventing him from playing a
Petroff, e.g. 3...lbc6 4 lbxd4 or
DANISH GAMBIT 3... .tc5 4 tbxd4 tbc6.
The Danish is a blood relative of Black's chief alternative is
the Goring and Scotch Gambits. 3 ...lbf6. Then 4 e5 is another
The Danish has the poorer Petroff line, I e4 e5 2 tbf3 lbf6 3 d4
reputation of the three but White exd4 4 e5, that Black has been
can try to improve with 1 e4 e5 2 d4 avoiding for years (by playing
exd4 and now 3 tbf3!?, rather than 3... tbxe4!). It's not a bad line but
the Danish 3 c3. few Black players will know
the theory and that makes 3 tbf3
a worthy weapon against a

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

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. positions (2 ... 4Jf6 3 d3 4Jc6 44Jf3)
is that White avoids the Petroff.
He benefits from 2 .i.c4 if he
likes 2... 4Jc6 3 4Jc3 4Jf6 4 d3 or
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
4Jf6 3 d3 and the natural 3....i.c5: key issue is how to reach this tabia.

Then 4Jc3 gets White into a The traditional order, 1 e4 e5

Vienna without having to face 2 4Jf3 d6 3 d4 4Jd7, named after
2 4Jc3 4Jf6 3 .i.c4 4Jc6 4 d3 .i.b4 or the U.S. Civil War major James
4... 4Ja5. Black retains those options Hanham, allows White to win the
with the more precise 34Jc6!. two bishops, 4 .i.c4 c6 5 0-0 .i.e7
There's a reason they say 'Knights 6 dxe5! dxe5 7 4Jg5 .i.xg5 8 'ii'h5.
before bishops.'
Aron Nimzovich helped popular-
The 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.
It has many of the benefits of a If Black is concerned about
regular Lopez, minus the 30 tons of 4 dxe5 he should consider offering

Double KP Openings

an endgame, because even fewer gives up the center, such as in 1 e4

Whites will trade queens after 1 e4 e5 2 liJf3 d6 3 d4 exd4 and then
d62 d4 e5!? Emanuel Lasker used 4liJxd4 g6 or 4 ... liJf6 5liJc3 iJ.e7.
this version in a world champion-
White can try to exploit this with
ship match against David ('I detest
Henry Bird's 4 iJ.c4!?
endgames') Janowsky. Today's
players, like Michael Adams, are
more confident on the Black side of
1 e4 d6 2 d4liJf6 3liJc3 e5!?

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
Now 4 liJf3 liJbd7 transposes to met 1 e4 e5 2 liJf3 liJc6 3 d4 exd4
Nimzovich's line and 4 dxe5 dxe5 4 c3 dxc3 5 liJxc3 with 5 ... d6
5 'ii'xd8+ ~xd8 is a slightly worse 6 iLc4.
endgame, e.g. Vaisser-Rontaine,
The test of 4 iJ.c4 is 4...liJf6. In
French Championship 2006 went
the first ECO Lev Polugayevsky
6 iLc4 i.b4! 7 iJ.d2 ~e7 8 f3 c6.
gave 5 liJg5 d5 as favoring White
Black was soon equal (9 liJge2 (6 exd5 h6 7 liJf3 iJ.b4+ 8 c3 dxc3
liJbd7 10 liJc 1 b5! 11 iJ.e2 iJ.d6 9 'ii'a4+ and wins).
12 a4 b4 13 liJdl a5), then better
On the same page he considered
(14 liJe3 liJb6 15 liJd3 g6 16 b3
1 e4 e5 2 liJf3 d6 3 d4 liJf6 4 liJg5!?
liJfd7 17 c3 bxc3 18 iLxc3 f6
exd4 5 iJ.c4 d5 and called it equal.
19 liJb2 liJc5) and eventually won.
It's the same position.
Note that the 1...d6 orders have
Better, after 4 iLc4liJf6 5liJg5, is
the added benefit of cutting down 5...iJ.e6. This dates back to Paul
White's 1 e4 e5 options, such as
Rudolf von Bilguer, who thought
2 f4, 2 iLc4, 2 liJc3 and so on.
6 liJxe6 fxe6 7 iJ.xe6 'ike7 (8 iLc8!
There's a separate chapter in 'ikxe4+ 9 'ike2) was even. To
Philidor theory in which Black modem eyes this looks like a slight

Double KP Openings

White edge. Therefore 4 iL.c4!? is brothers popularized 1 e4 eS 2 ttJf3

more than just a confuse-move. 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 Their 19th Century colleagues
3 iL.c4 with the idea of 3... ttJd7 tried 1 e4 eS 2 ttJf3 ttJf6 3 d4 exd4
4 0-0 iL.e7 S d4. In this way he and 4 iLc4 in an effort to get into a
avoids 3 d4 exd4 and traps Black in Two Knights (4 ... ttJc6), which is
the inferior Hanham line. terra incognita to a Petroff player.
But the critical line is 4 ... ttJxe4!, the
Better is 3...iLe7! 4 0-0 ttJf6
dubious Prince Yurosov Gambit.
which leaves White with a tiny edge
after S 1::te I 0-0 6 c3 ttJbd7 7 d4. These days 3 d4 ttJxe4 is more
More ambitious is 3... ttJf6!? 4 ttJgS common, and theory says the
dS S exdS, which reaches a Two chances are roughly balanced after
Knights Defense in which Black is 4 iL.d3 dS S ttJxeS. Yaacov Murey's
a tempo behind. discovery, 4 ... ttJc6, enables Black
to avoid this. It's based on S .txe4
dS 6 .td3 e4, which appears to be
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
PETROFF DEFENSE into the main line as if 4 .td3 dS
had been played.
The Petroff is so annoyingly
solid that a frustrated White will be This order gives Black an extra
sorely tempted to transpose into option, 4... d6!? S ttJf3 dS. He's
another opening. The Paulsen transposed into the 'other' Petroff,

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
uncomfortable. The best challenge to S i.d3 is
5... d5 and then 6 'iVe2 'ile7. A
The 'other' Petroff is the most typical case was Romero Holmes-
popular 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
Exchange French in which h2-h3 is A later Petroff tabia begins with
an extra move. 3 ltJxe5 d6 4 ltJo ltJxe4 5 d4 d5
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

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!.

Double KP Openings

Garry Kasparov said 6.. tbc6 chooses between undermining the

7 0-0 i.e7 8 :tel i.g4 was fine for e4-knight with 8 c4 or attacking it
Black. 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
with 8... tbf6!? Then 9 tbc3 is Then 9 c4 tbb4 10 cxd5 tbxd3
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.
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
Alexey Shirov preferred 6... i.e7! There doesn't seem to be
7... tbc6 - and Yusupov plays both. anything in 5 i.c4 tbxe4!, so
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.

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 Now 7... d5 8 exd5 cxd5 9 0-0
'i'e8 (threat of ... ttJxd4!) 9 d5 ttJd8 0-0 10 i.g5 is played with almost
11 c4 c6 11 i.a4 b5! as in Manik- robotic uniformity and is very
Mokry, Olomouc 1998. slightly in White's favor or even,
depending on whom you read. But
But if, on the other hand, Black
there are several diversions starting
knows his way in a sharp line of the
with 8...'iWe7+!?
Scotch Gambit (1 e4 e5 2 ttJf3 ttJc6
3 d4 exd4 4 c3 ttJf6!? 5 e5 ttJe4) he Theory has gone back and forth
can transpose into it from the about the value of 9 'iVe2. In any
Ponziani by means of 3... ttJf64 d4 case, against an opponent who likes
exd4!. to keep queens on the board,
8... 'iWe7+ has obvious merits.
SCOTCH GAME But if White is happy in an
White can get into a major ending, he can try to force it by
Scotch variation via the Four meeting 8... cxd5 with 9 ~e2+!? so
Knights (1 e4 e5 2 ttJf3 ttJc6 3 ttJc3 that 9... 'i'e7 transposes. There are
ttJf6 4 d4 exd4 5 ttJxd4). If that's a extra benefits when an endgame-
position he wants to play, this is the hating Black plays 9 ... i.e6

Double KP Openings

(10 i.b5+! tbd7 11 i.c6 d4 12 These days Black is more

'ife4) or 9 ... i.e7 (10 i.g5 0-0 concerned about the standard
11 O-O-O!? with better-than-usual Scotch order, 1 e4 eS 2 tbo tbc6
chances). 3 d4 exd4 4 tbxd4 and then 4 ...tbf6
5 tbxc6 or 4 ... i.c5 5 tbxc6. For that
Black can anticipate this and rule
reason, he can try to reach the Four
out endgames through another route
Knights version with 4..i.b4+!?
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
Black's point is 9 eS 'ife8! and with an extra c2-c3 for White.
White is the one surprised, e.g. Who benefits? White gets added
10 f4? tbg4 11 'ife2 f6 (12 exf6 protection of d4. But he loses one of
gxf6 13 h3 Xg5 14 bxg4 i.xg4! the major assets of the Scotch, the
15 iLxh7+ <Jiig7 and he resigned knight hop from c3 to d5. Recent
in Candela-Korneev, Ponferrada practice indicates Black is OK after
7 tbfS i.xe3 8 tbxe3 tbf6 because
If White spots that and heads White cannot defend the e-pawn
instead towards the tabia, with with tbc3 (and 9 tbd2 allows the
9 exdS, Black can use an extra equalizing 9... d5). Or 7 i.c4 d6
option, 9...'ifxdS!. This hits at g5 8 0-0 tbf6 9 tbd2 tbe5! 10 i.e2 0-0
and g2 and after 10 i.xf6 "iVe6+ and 11 h3 :e8 as in Goloshchapov-
11..:iVxf6 he stands well. Brodsky, Hoogoveen 2006.

Double KP Openings

And what about the standard the queen move is even playable. Its
order, 4 llJxd4 .tcS S .te3, fate depends on how we evaluate
threatening llJxc6 ? Much midnight this tabia.
oil has been spent on 5.. :iVf6. The
mental hygiene alternative is

The final word on 7 ... ..txc3+

8 llJxc3 or 8 bxc3 is still to be
written. But equally important is
Now 6 llJc3 d6 and 7... llJf6 or finding White's best way to the
6... llJf6 transpose to a solid Four diagram. There are four routes.
Knights/Scotch line. For example,
(a) S llJc3 ..tb4 6 llJdbS 'iYxe4+
Sveshnikov-Mi. Tseitlin, Budapest
7 ..te2 gets there. But Black does
1989 went 6 llJc3 d6 7 g3 llJf6
much better with 6 .....ta5! and ... a6,
8 .tg2 0-0 9 llJxc6 bxc6 10 .txb6
e.g. 7 .te2 a6 8 llJd4 ..txc3+ 9 bxc3
axb6 with equality.
llJf6 10 llJxc6 dxc6 11 e5 llJg4
Another trendy line is 4 ... llJf6 (Renteria-Miktov, Minneapolis
5llJxc6 dxc6 5 e5. Black can avoid 2005). For that reason White should
that with 4:i'f6. Then he is trying prefer:
to reach, not avoid, the 5 .te3 .tc5
(b) SllJc3 ..tb4 6 ..te2! and then
position. This order has the added 6.:i'xe4 7 llJdbS. If Black varies
benefit of dodging another difficult with 6 ... llJf6 7 0-0 .txc3 then
line, 4 ....tc5 5 llJb3.
8 llJf5! is potent, e.g. 8... 'iYxe4
The best try at punishing 4 ... 'i'f6 9 .td3 'iVg4 10 f3! ~a4 11 bxc3
is 5 llJb5 .tc5 6 'i'e2. But the 0-0 12llJxg7! with a winning attack
verdict is out on 6....tb6 or 6... 'i'd8 in Karjakin-Malinin, Sudak 2002.
(and ... a6).
But White may opt for 5 llJb5
At one time the fate of the Scotch because it is more forcing and seeks
seemed to hinge on the soundness something better than the diagram.
of 4..'i'h4. Today there's doubt that For example:

Double KP Openings

(c) S tbbS 'fixe4+ 6 .te2 i.b4+ . tbd8-e6/ ..cS sets up a dark-

7 tblc3 gets to the diagram. But square blockade. This is Black's
White can improve with 7 .td2!, best way to play for a win
with elaborate lines following in the Four Knights, since Akiba
7... .txd2+ 8 'iVxd2 'fie5 9 f4!, for Rubinstein's 4 ... tbd4 allows a
instance. draw-minded White 5 tbxd4 exd4
6 e5.
(d) Black can try to improve with
S....tb4+ since 6 tblc3 'iVxe4+ There are ways for White to
7 .te2 transposes. But again the avoid the Metger - and still play to
bishop interpolation, 6 .td2!, is win - and other ways for Black to
better and reaches ( c) after seek it. White's sidestep is 7 tbe2,
6... 'iVxe4+ 7 i.e2. which has been revived by Daniel
Campora, who showed that White's
The main problem with 5 tbb5 is
edge in a !1ymmetrical position
S...tcS and then 6 'iVe2 tbd4, when
(7 ... tbe7) is something he can work
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
FOUR KNIGHTS GAME .tb6 13 .l:tel h6 14 .tc2 'fic7
15 i.e3 i.e6 16 'fic 1! ':ad8 17 a4
The main Four Knights line for a5 18 tbf5 d5 19 tbxg7!, Campora-
most of the last century was
Baron Rodriguez, Lanzarote 2003.
Johannes Metger's 'unpin', after
1 e4 eS 2 tbf3 tbc6 3 tbc3 tbf6 Black can avoid this by capturing
4 .tbS .tb4 S 0-0 0-0 6 d3: 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

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

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
Frank Marshall and Smyslov. avoid. If he still feels that way,
5....txd4! and 6 lDxd4lDxd4 7 ..tg5
GIUOCO PIANO! d6 8 f4 'iie7 or 8... .te6 is the
TWO KNIGHTS antidote. But Yakov Estrin argued
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
With 3... .tc5 Black is avoiding Two Knights. He can head towards
certain lines, such as 3 ... lDf6 the Pseudo-Lopez with 4 d3, 4 c3 or
4 lDg5, which have good 4 0-0 while retaining other options.
reputations, or even the question-
able 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 The rationale for 4....tb6 is that
often seen as an alternate route to .. .'ike7 is not as useful as ... ..tb6 in
the Pseudo-Lopez (4 0-0 lDf6 5 d3). the Pseudo-Lopez. White can
There is also 5 d4!? change his mind after 4...'iie7 and

Double KP Openings

switch to S d3 with idea of hitting Another move order issue in the

the vulnerable queen later on with Two Knights concerns the periodic-
ttJe3-dS or ttJfS. The minus to this ally revived 4 d4 exd4 S eS dS
order is that after 4 ... ..tb6 S d3 6 ..tbS ttJe4 7 ttJxd4.
Black will have lost a tempo if he
feels ... a6 and ... ..ta7 are necessary
later on.
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 It was once assumed Black must
Vienna order. However, he can get reply 7.....td7, Paul Morphy's
to it in the Two Knights Defense, move, after which 8 ..txe6 bxe6
1 e4 eS 2 ttJf3 ttJe6 3 .te4 ttJf6 ifhe 9 0-0 ..teS and 10 ..te3 is fairly
avoids 4 ttJc3? ttJxe4! and plays balanced. But recent experience
4 d3! and then 4.....teS S ttJe3. suggests 10 f3! gives White the
upper hand.
In a standard Two Knights order,
1 e4 eS 2 ttJf3 ttJe6 3 .te4 ttJf6 To avoid that, Black can employ
4 ttJgS, there are two offbeat but Adolf Anderssen's 7.....teS!? Then
related lines, 4... dS S exdS ttJd4 and 8 .te3 .td7 9 ..txc6 bxc6 10 f3
S... bS. They don't look at all alike. would be met by 10 .. .'ifh4+ 11 g3
Yet they transpose after S... ttJd4 ttJxg3 !. Instead, White usually
6 c3 bS! 7 .tfl! and S... bS 6 .tfl! replies 10 0-0.
ttJd4 7 c3. The point is he transposed into
the equal 7... ..td7 line in which
Black's choice should depend on
10 ..te3 is played and lost the
which line is likely to provoke a bad
promising 10 f3 option.
response by White, such as S... bS
6 dxc6?! bxc4 7 ~e2 h6 8 'ilxeS+ The risk of this finesse is
JLe7 which favors Black. Hans allowing 7 ... .tcS 8 ttJxc6!? and
Berliner revived S... bS in his world then 8 ... ..txf2+ 9 ~fl or
postal championship run, explain- 9 ~e2!? ..tg4+ 10 ~fl. White also
ing that S... ttJd4 'is less clear' after gets an extra option after 8 ..te3
6 ttJc3. JLd7 9 ..txe6 bxe6 when he tries

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
For example, Sveshnikov-Zaja,
6 d4!. Black can't reach the safety
Bled 2001 went 10 ... liJxd2
of the Lasker Defense, 6... i.b6
11 'i'xd2 0-0 12 liJb3! i.b6
(7 0-07 d6), in view of 7 dxe5!.
13 0-0-0 'i'e7 14 l::the1 a5 15 a3 a4
16 ..tg5! 'i'e6 17 liJd4 'i'g6 18 f4 Therefore he has to enter one of
and White eventually won. the main Evans lines, such as
6... exd4 7 O-O! d6 8 cxd4 i.b6,
On balance, 7 ... i.d7 appears where White's claim to compensat-
better than 7... i.c5. And Morphy ion has withstood the test of time -
beats Anderssen once more. a century and a half and counting.
Once Black decides on S...i.aS a
EVANS GAMBIT key question is whether White
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
bishop. There are two ways for it to Compare that with 6 'ib3:
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?
six. Consider S.. i.cS: walks into 8 d5! liJ-moves 9 Vi'a4+.

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
The real challenge to 6 'ifb3 is d5!.
6...'fif6 7 d4 tDxd4!. Then 8 tDxd4
exd4 9 0-0 has been tested a bit in SCOTCWGORING GAMBIT
correspondence (9 ... dxc3 10 e5
'fixe5! 11 .txf7+ ~f8) without a These two gambits are closely
conclusion. If White likes the looks related and may reach Giuoco and
of this, 6 'tWb3 would be much better Two Knights positions. The Scotch
for him than 6 0-0 and perhaps order (1 e4 e5 2 tDo tDc6 3 d4 exd4
superior to 6 d4, too. 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 question-
able 4 ....txb4 5 c3 .tc5.

White's bid for cleverness 1S The simplest reply is 4...tDf6

worse. It is 5 0-0 after 4 ... .txb4. which transposes into the Two

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 His idea is not 6 c3 dxc3?! -
that 5 c3 dxc3 6 tDxc3 transposes because that transposes into an
into what was long considered his unattractive Goring.
best Goring defense (1 e4 e5 2 tDf3
tDc6 3 d4 exd4 4 c3 dxc3 5 tDxc3 Better is 6... ~g4 with unclear
~b4! 6 ~c4). play. For example, 7 'ii'b3 'ii'd7
8 cxd4 ~xf3 9 dxc5 tDd4 10 'ike3
That conclusion is in recent
'ii'g4 11 'iVg5 with a complex
dispute. However White doesn't
endgame in Wist-Rausis, corres-
have to get involved in the debate
pondence 1994.
since in the Scotch order 6 bxc3!
favors him (6 ... ~e7? 7 'ii'd5). One final point is that 4... ~c5
Well, then, what about the third can also transpose, after 5 c3 tDf6 to
option in the last diagram, 4...~c5 ? a Giuoco Piano. That's a safer
That is a not-so-veiled invitation to choice than 5... dxc3?! 6 tDxc3 d6,
the Lange (5 0-0 tDf6 6 e5 d5). another unclear Goring line (7 ~g5
Compared with 4... tDf6, Black loses tDge7 8 tDd5 f6 9 ~xf6!?).
the anti-Lange possibility. But he So the bottom line of the Scotch
benefits by not having to face Gambit, 4 ~c4, is that 4... ~b4+?!
the 5 e5!? line, which Yevgeny is to be avoided, 4 ... tDf6 is the safe
Sveshnikov has proven to be choice and 4 ... ~c5 is the most
dangerous. double-edged. And that's one of the
Another virtue is that after more definitive conclusions we can
4..~c5 5 0-0 Black has an extra make about those trap-happy 1 e4
5... d6 option. e5 openings.

Chapter Two:

There are far more tricks in the Black should know the counter-
move order of the Ruy Lopez than finesse, 4 ... a6!. Then 5 .i.xc6 dxc6
in any other I e4 e5 opening and 6 ttJxe5 'iWd4 regains the pawn.
they often start as early as move White's best is 5 .i.a4!, but he
four, as in the Berlin Defense. The has to know the WorralllWormald
bona fide Berliner is content to head Variation, (usually reached by 3 ... a6
into the endgame of 1 e4 eS 2 ttJf3 4 .i.a4 ttJf6 5 'iWe2), into which he's
ttJc6 3 .i.bS ttJf6 4 0-0 ttJxe4 5 d4 transposed.
ttJd6 6 ..txc6 dxc6 7 dxe5 ttJf5
8 'iixd8+. He might be upset by CLASSICAL DEFENSE
4 'iWe2!? 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 There's a consensus that White is
DNA. at least slightly better after 6 d4 and


his main concern is what he wants 11 exd6 cxd6 12 l:.d1 as in

to avoid: 4 0-0 will sidestep 4 c3 Matulovic-Ilincic, Sombor 2004.
f5!?, for example. But 4 c3 will
There's psychology behind 4 0-0
avoid 4 0-0 ttJd4.
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 What's happened is we've landed
preference for 4 0-0 when he in a Four Knights Game (1 e4 e5
expects Black to continue with a 2 ttJf3 ttJc6 3 ttJc3 ttJf6 4 ..tb5) in
normal Berlin. In this way Black which Black has played a rare
avoids an early c2-c3/d2-d4 in view sideline, 4 ... ..tc5!? and White
of3 ... ttJf6 4 c3? ttJxe4!. replied 5 0-0.
The other two orders start with That's fine for Black - if he
3...tc5 - 4 c3 ttJf6 5 d4 ..tb6 6 0-0 knows the theory (5 .. 0-0! 6 ttJxe5
0-0 and 4 0-0 ttJf6 5 c3 0-0 6 d4 ttJxe5 7 d4 ..td6!).
..tb6. Books give much more
attention to the first order because it Alexander Be1yavsky was once
grants White an extra option, surprised by Ljubomir Ljubojevic
6 'iWe2. But if he intends 6 0-0 in this move order and was
anyway, and he doesn't remember outplayed after 6.. J~e8?! 7 ttJf3
the complex 4 c3 f5 theory, the ttJxe4 8 d4 ttJxc3 9 bxc3 ..tf8
second order is best. 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
4 c3? ttJxe4! mentioned above. But DEFENSES
this one turns out to be bad after
6 'iWe2!, e.g. 6... f5 7 d3 ttJf6 8 d4 When Black develops his KB on
..te7 9 dxe5 ttJe4 10 ..tc4! d5 g7 he is playing a reversed Vienna,


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
can push his d-pawn faster. There's a back-door route to a
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
More dangerous is 4 c3: 4 0-0 g6 5 c3 Jtg7 6 d4 exd4 7 cxd4
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, This becomes an old Steinitz
to a Modem Steinitz Defense (3 ... a6 Defense, in fact one of the better
4 i.a4 d6 and ... g6). 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 There must be a drawback to the

with 5 .tc4. That's a pretty solid Cozio order or everybody would be


playing it. It is 3...tbge7 4 tbc3! After S 0-0 the most popular

which gets very complex after move in recent years is S...i.cS.
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.


Henry Bird's defense (1 e4 eS

2 tbO tbc6 3 i.bS tbd4 4 tbxd4
exd4) looks absurd: Black moves Now 6 d3 c6 7 i.c4 dS! heads
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 STEINITZIANA
100-0 i.cS! 11 l:.e1+ tbe7 is quite
nice for Black. Wilhelm Steinitz's solution to the
White's error was committing Lopez problem was maintaining a
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
sees Black's fifth move. 3 tbf6 4 0-0 i.e7 S tbc3 d6 6 d4
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 ne1and
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.


White is better but Black has And now 5... exd4! 6 ltJxd4 ltJf6
play along the b-file and from ... c5!/ 7 0-0 iJ.e7. Then 8 iJ.xc6 or 8 l:Ie 1
... iJ.c6, e.g. 10 iJ.f4 c5 11 ltJD iJ.c6 0-0 9 iJ.xc6 heads towards the
12 'it'd3 .!tJh5! 13 iJ.e3 l:te8 14 l:Iadl double-edged ... bxc6 tabia that
'ifb8 15 iJ.c1 'iib7, Narciso Dublan- should make Black happy.
Mizoev, Salou 2006.
Unfortunately for him there's a
counter-counter-finesse. White can
Experience shows that White
move up the capture on c6, as
should try to force Black to retake
Jackson Showalter showed more
on c6 with the bishop, thereby
than a century ago, 3... d6 4 d4 .td7
eliminating ... c5! counterplay. He
and then 5 iJ.xc6 iJ.xc6 6 .!tJc3. This
does that with 3... d6 4 d4 iJ.d7
transposes to the favorable line after
5 .!tJc3 .!tJf6 6 iJ.xc6! since 6... bxc6? 6 ...ltJf6 7 'it'd3 exd4 8.!tJxd4.
7 dxe5 drops a pawn.
Of course, Black can accelerate
Instead 6... iJ.xc6 7 'it'd3 renews his plan, too, with 4 d4 exd4, as
the dxe5 threat and favors White Alexander Onischuk does. But
following 7 ... .!tJd7 8 iJ.e3 exd4 there's a difference because White
9 iJ.xd4! or 7... exd4 8 .!tJxd4 iJ.d7 can retake with his queen, 5 'it'xd4
9 iJ.g5 iJ.e7 10 0-0-0 and 8... g6 iJ.d7 6 iJ.xc6, which has been
9 iJ.g5 iJ.g7 10 .!tJxc6 bxc6 known to favor White since
11 e5 dxe5 12 'it'D 'it'd6 13 l:tdl Morphy's day.
(Lupulescu-Pessi, Bucharest 2004). To resuscitate the Steinitz Black
needs a new idea, perhaps the little-
Black's counter-finesse lies in explored 5....!tJge7!? in the last
moving up ... exd4 before iJ.xc6 so order, or something that confuses
that he can retake with the b-pawn. White about the timing of iJ.xc6. He
He accomplishes this after 3... d6 might try the Berlin move order -
4 d4 iJ.d7 5 .!tJc3: 3... .!tJf6 4 0-0 and then 4 ... d6 or


4 ... i.e7 5 :tel d6. This has the most obvious improvement is that
benefit of denying White his most by inserting 3... a6 4 i.a4 before
aggressive plans such as 'iVd3/0-0-0 4 ... d6, Black makes d2-d4 a trifle
and avoiding 4 d4. risky. After 1 e4 e5 2 lbf3 lbe6
But White should be aware after 3 i.b5 a6 4 i.a4 d6 5 d4 Black can
3lbf6 4 0-0 d6 5 d4 i.d7 that he reply 5...b5! 6 i.b3 lbxd4 7 lbxd4
has 6 i.xe6! i.xe6 7 lbe3.

The liquidation that follows Now 8 'ifxd4? is the Noah's Ark

7. lbxe4 8 lbxe4 i.xe4 9 :tel Trap (8 ... c5 9 'i'd5 i.e6 10 'iVc6+
favors him (9 ... f5 10 lbg5 or i.d7 11 'i'd5 c4 and wins). But
9 ... i.f5 10 dxe5 d5 11 lbd4). 8 e3! and then 8... dxc3 9lbxc3lbf6
Instead, Black usually transposes 10 0-0 is a worthwhile gambit.
with 7... exd4 8 lbxd4 into a ... i.xc6
line, rather than ... bxc6!. 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


and it's not clear if that or 6... tbf6 7 .txc6 and get a favorable old
7 tbc3 tbd4 equalize. Steinitz position. '
The most common continuation He means 7....txc6 8 %:tel exd4
after 5 0-0 .td7 is 6 d4. 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 Smagin-
Babu, London 1990.
This looks like an innocuous
Theory prefers 6...tbf6!. Then
route to quiet lines, such as 6 c3
both players can claim they won the
tbf6 7 !:tel 0-08 h3 .td7. But there
battle of the opening. Black says,
are differences between 5... .te7 and
'J tricked you into a Steinitz
Defense Doubly Deferred without
allowing you a chance to use your One is that Black is better
best weapon.' situated to play 6... .tg4 compared
Translation: This position can be with the somewhat dubious 5 0-0
reached via 3... a6 4 .ta4 tbf6 5 0-0 .tg4.
d6 6 d4 .td7 but only if White For example, 6 c3 .tg4 7 d4 b5
passes up 6 .txc6+!. 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


(11 g4 .tg6 12 d4 h5) in Browne- followed by ...ti:Je7. The forgotten

Portisch, Milan 1975. 5....te7 deserves to be remembered.
Another difference is Black can Black can also mix the Modern
try to upgrade the Siesta Variation Steinitz with the Cozio if he plays
by meeting 6 c3 with 6... f5!? 7 exf5 1 e4 e5 2lbf3lbc6 3 .tb5 a6 4 .ta4
.txf5 8 d4 e4. and now 4...lbge7.
More of a test of 5... .te7 is 6 d4. Alexey Suetin recommended
Then 6...b5 7 .tb3 lbxd4 8 lbxd4 '5 c3!' and cited one of his games
exd4 9 c3 is another center gambit. that led to a 5...b5 6 .tb3 d5 7 'ife2!
After 9 ... lbf6! 10 cxd4 i.b7 11 f3 0- edge.
o 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).
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

Moreover after 4 ... lbge7 5 0-0

... allows White to set up a bind Black can safely transpose into the
after 8... c5 9 lbc6! 'ifd7 10 lba5! fianchetto line (5 ... g6 6 c3 d6).
(10 ... lbf6?! 11 c4). 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.



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

Eduard Gufeld beat Mikhail Tal

in a game that went 1 e4 e5 2 ttJf3 This can transpose to a Steinitz
Defense Doubly Deferred line
ttJc6 3 Jlb5 a6 4 Jla4 ttJf6 5 Jlxc6
(7 :el d6) that Akiba Rubinstein
dxc6 6 d3 ttJd7?! 7 ttJbd2 Jle7
liked to playas Black and is similar
8 ttJc4 Jlf6? and then 9 'iVe2 c5
to the better forms of the old
10 Jld2 0-0 11 g4 with a winning
Steinitz after 8 d4 exd4 9 ttJxd4
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! White might improve with
and ttJg4). 7 ttJxe5 ttJxe4 or 7 d4 ttJxe4 8 dxeS
0-0 (9 :e 1 dS 10 ttJd4 Jld7 11 f3
A major benefit of a delayed ttJgS as in I.Sokolov-S.Nikolic,
Jlxc6 is that Black finds it harder to Pancevo 1987) but there's too little
defend his e-pawn. After 3 ... a6 experience to tell.
4 Jla4 ttJf6 S Jlxc6 dxc6 he cannot
play... f6 as he would after 4 Jlxc6
dxc6. However, 6... Jld6! is a good Before getting to the main Open
developing defense and Black line, 1 e4 e5 2 ttJf3 ttJc6 3 Jlb5 a6
stands well after 7 ttJbd2 Jle6. 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!?:


This tries to lure White into i..e3, lbbd2) and at Black's ( ... i.e7,
8 :el and then 8... d5 9 dxe5 .te6, ...lbc5, ... lba5) can be mixed in a
a dead even position that could variety of orders, making trans-
come about after 7 ... d5 8 dxe5 .te6 positions inevitable.
if White plays the inept 9 l:.el?! In the first edition of ECO,
.te7. Viktor Korchnoi endorsed the
The best try at punishing 7 ... .te7 conventional wisdom that 9 c3 .te7
is 8 dxe5. Experience with 8...lbc5 10 .te3 and 9 c3 .te7 10 lbbd2
9 i.d5 .tb7 (or 9 ...lbe6) has been were so favorable that Black should
far from convincing, e.g. 9 ... .tb7 avoid them with 9 ... .tc5!. But
10 lbc3 0-0 11 .tf4 lbe6 12 .tg3 Andrei Sokolov showed how White
f5!. can trick his opponent into a
favorable line with 9 .te3!? and
If Black likes these positions he then 9 ....te7 10 c3!. This is very
can get to them even if White effective against booked-up 9 c3
adopts a standard way of avoiding .tc5 opponents.
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
Then 6 dxe5! lbxe4! 7.tb3lbc5 19 'ilxf3 c5 20 'ilg3 lbc4 21 b3
8 .td5 i.e7 9 0-0 transposes to the lba3 22 e6! 'ifc6 23 .tf5! and wins,
previous line. Kamsky-Piket, Dos Hermanas
The tabia of the Open Defense 1995.
arises after 1 e4 e5 2 lbf3 lbc6 There is no obvious antidote to
3 .tb5 a6 4 .ta4 lbf6 5 0-0 lbxe4 9 .te3 but attention has shifted to
6 d4 b5 7 .tb3 d5 8 dxe5 i.e6. The 9 lbbd2. It allows White to answer
moves at White's disposal (c2-c3, 9 ... .tc5 with 10 lbxe4 dxe4


11 it.xe6 with at least a small White 8 ... l:tb8 9 axbS axbS, he can reach a
edge. That often prompts 9 ... it.e7, much improved Open Defense,
after which 10 c3! is another good 10 dxeS it.e6 11 c3.
But 8 a4? was knocked into a
The downside to 9 tiJbd2 is that footnote by the Lasker-Schlechter
White has diminished control of d4, match of 1910 when Black
compared with 9 it.e3 and 9 c3. equalized with 8... tiJxd4!'
That's significant after 9..tiJcS!.
The appeal of the tweaked move
Then White would be happy to order, 8 dxeS it.e6 9 a4, is that
see 10 c3 it.e7 11 it.c2 it.g4 because 9 ... l:tb8 allows White to reach that
that transposes to yet another good ideal version of 8 a4 following
9 c3 line (9 c3 tiJcs 10 it.c2 it.g4 10 axbS axbS 11 c3. For instance,
11 tiJbd2 it.e7). For example, 12 tiJd4! tiJxeS (12 .. :iVd7
12 .:tel 'ifd7 13 tiJf1 lId8 14 tiJe3 13 it.c2!) 13 f3 tiJcs 14 it.c2 it.d7
it.hS IS tiJfS. IS b4 with advantage, Alekhine-
But Black has been doing better, Rohacek, Munich 1941.
after 10 c3 with 10. d4. That has So what's wrong with 9 a4 ?
become a main line, with book Vintage theory says 9..tiJaS is the
analysis stretching past move 20. answer. But 10 axbS axbS 11 tiJd4!
A trickster will prefer yet another
it.cs 12 c3 0-0 13 it.c2 is hardly
convincing. More testing is needed
ninth move, 9 a4!?
(say of 9 . b4 10 as, threatening
it.a4) before we can properly
evaluate 9 a4!?

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
White is trying to slip into a possibly ... bS/ ... i.b7. Up until
forgotten line that begins 1 e4 eS fairly recently they seemed like
2 tiJf3 tiJc6 3 it.bS a6 4 it.a4 tiJf6 distant cousins unworthy of our
S 0-0 tiJxe4 6 d4 bS 7 it.b3 dS and attention:
now 8 a4, instead of 8 dxeS. If
The Moeller Attack (1 e4 eS
White gets control of an open a-file,


2 ~f3 ~c6 3 i..bS a6 4 i..a4 ~f6 commonly come from 5 'iVe2 b5

S 0-0 i..cS) was considered too 6 i..b3 i..c5 7 0-0 and lead to
flaky because of 6 c3 i..a7 7 d4 or roughly equal chances.
6 ~xeS ~xeS 7 d4. If White abandons c2-c3/d2-d4
The Arkhangel, S.. bS 6 i..b3 and plays 7 d3 Black can safely
i..b7, enjoyed 15 minutes of fame revert to the old Arkhangel, 7... 0-0
in the 1960s. That expired when 8 a4 i..b7. That's an 'old' position
White obtained an edge on both that occurs after 5 0-0 b5 6 i..b3
wings with 7 .l:tel i..c5 8 c3 d6 9 d4 i..b7 when White passes up the
i..b6 10 i..g5 and a2-a4. critical lines such as 7 .:tel. Again
But then came the 'New Black gets a free ride in the
~ ~ opening.
Arkhangel,' S...bS 6 .... b3 .... cS.
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 The difference between new and
example, on 7 .l:tel he avoids old is revealed by 9 .:tel. Then
7... i..b7 - which transposes to the 9 ... i..b7? transposes to an inferior
unfavorable old Arkhangel - and 'old' position. But after 9...i..g4! we
plays 7... ~g4! in the Moeller spirit. are 'new.' White can't support his
That's good for him, e.g. 8 d4? center on the current squares and he
~xd4 9 ~xd4 'ii'h4! or 8 .l:te2 ~d4 has no edge after 10 d5 ~e7 11 h3
9 ~xd4 i..xd4, threatening i..d7! and ... c6.
10 ... 'ii'h4.
What about a2-a4, the usual
If White concludes that he should antidote to an early ... b5 ? That
be playing 7 'i'e2 instead, he is shouldn't trouble Black in the last
merely transposing into one of the diagram, e.g. 9 a4 i..g4! 10 axbS
Worrall/Wormald tabias that axbS 11 .l:txa8 'iVxa8 12 h3 i..hS


has perfonned well, e.g. 13 "ifd3 face the more critical 8...l:tb8 and
exd4 14 ii.g5 ii.xf3 15 ii.xf6 8... ii.g4 lines.
ii.xg2!, Pavlovic-Carlsen, Reyk-
Despite that, the .. .ii.c5 family is
javik 2006.
respectable today - because so far
Is there nothing but good news Black has found more good
for Black in the New Arkhangel? transpositions than White.
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
Then 7...ii.b7 most dangerous Ruy Lopez
defense. That adds huge signific-
ance 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 anti-

Marshall 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 With 7 ... d6 Black commits
10 ii.gS!, a favorable 'old' position. himself to a main-line variation
What happened: We've (8 c3 0-0 9 h3 etc.). But 7... 0-0
transposed into 7 c3 d6 8 a4!? ii.b7 allows him to threaten to play the
9 d4 ii.b6 10 ii.g5. This helped Marshall Gambit, 8 c3 d5!?
White because he didn't have to It's just a threat because he can


back out with 8... d6, tranposing into Alexander Grischuk among others.
the main line. Yet the threat is so
worrisome that today's strongest And what about other anti-
GMs skirt the issue with modest, if Marshall lines? The simplest begins
not cowardly anti-Marshall moves with 8 d3. Then 8... d5 is considered
like 8 d3 or 8 h3. unsound since c3 is available for a
knight (9 exd5 tiJxd5 10 tiJxe5
Why then should anyone play tiJxe5 11 :xe5 c6 12 l:te 1 .i.d6
7... d6 if it enables White to play 13 tiJc3!).
more aggressively without risk?
The only downside to 7... 0-0 is Black typically meets 8 d3 with
supposed to be that 8 a4 is stronger. 8...d6. Following 9 c3 White will
It is - but not significantly, after eventually play d3-d4. This ensures
8 a4.i.b7. he can execute the tiJf1 maneuver
That's why top GMs have been that is so important to the Lopez.
hunting for an aggressive way to He plans 10 tiJbd2, 11 tiJf1, 12 tiJg3
sidestep the Marshall. One idea is and eventually d3-d4.
8 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


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,
reasonable system for White. be worthwhile because allowing the
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 little room for transposition
with the comment that 8 ....tb7 following 8 c3 d5 9 exd5 tbxd5
keeps the gambit alive. That is, 9 c3 10 tbxe5 tbxe5 11 l1xe5 c6. An
d5! 10 exd5 tbxd5 11 tbxe5 tbxe5 exception occurs in Fischer's idea
12 l1xe5 tbf4! is a prohibitively of 12 g3.
strong attack.
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 Marshall-
and most Lopez players do - then
he's scored a minor victory.


This is designed to bypass the Arthur Bisguier once explained

mega-theory of 12 d4 .td6 in Chess Review why he used 5.. b5
13 .:tel 'iVh4 14 g3 'iVh3. 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 Variation. Second, White 'couldn't
.:te6 18 'iWfl 'ifh5 19 .txd5. He can be sure I would transpose into the
main line and he felt compelled to
even reach the same positions with
analyze the complexities of 7 ttJg5. '
12 i.xd5 exd5 13 d4 i.d6 14 .:tel
Opinions vary about the soundness
'iVh4 15 g3 'ifh3.
of7 ... d5 8 exd5 ttJd4.
Which is best? Perhaps the last
order because - aside from the extra White eventually selected 7 e3
14 .:te3 option - it bypasses another and Black transposed after 7... i.e7
Geller line, 12 d4 i.d6 13 .:te 1 'iVh4 8 h3 ttJa5 9 i.e2 e5 10 d4. But by
14 g3 'iVh3 15 i.e3 i.g4 16 'iVd3 then he had gained more than an
ttJxe3!? 17 .:txe3 c5. hour (!) on the clock.



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.


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
7ltJg5? O-O! is pointless. hanging but no one who has faced
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
but there is little experience to ~xe4! and played llltJbd2.
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.

Ruy Lopez

This is not just a trap, 13 dxe5 Instead he usually plays 14 i.b1

dxe5 14 liJxe5? 'iie7!. It deserves and then 14... 0-0 15 liJf1.
the serious testing it hasn't gotten.
That transposes to a position that
OLD MAIN LINE normally comes from 5 0-0 ~e7
The way the main Ruy tabia was 6 :el b5 7 ~b3 0-0 8 c3 d6 9 h3
reached for much of the 20th liJa5 10 .tc2 c5 11 d4 'iic7
century was 5 0-0 ~e7 6 :e1 b5 12 liJbd2 cxd4 13 cxd4.td7 14 liJf1
7 ~b3 d6 S e3 and now SliJa5 :ac8 when White passes up
9 .te2 e5. This was considered a 15 liJe3! and 15 .l:!e2! and settles for
finesse because it avoided 8... 0-0 15 ~b1.
9 d4, at a time when that line was
regarded as potent. The 8... liJa5 In this order White didn't get that
order began to disappear in the choice and the result is a roughly
1950s but was revived by Bent equal position (15 ... l:.fe8 16 liJg3
Larsen and Oleg Romanishin. liJc6).
Why was it revived? One reason
is that by delaying castling Black Another plus of the Old Main
can meet 10 d4 'iie7 11 h3 with Line is that Black isn't castled after
a quick ... l:tc8 that threatens the 8... liJa5 9 .tc2 c5 10 d4 'iic7 11 h3
c2-bishop earlier than usual. liJe6. That means 12 d5? - which is
For example, 11 ...~d7 12 liJbd2 a good move when liJbd2 and ... 0-0
exd4 13 exd4 :eS. Since White are added - turns out well for Black
hasn't had time for liJf1, he can't after 12 ...liJbS 13 liJbd2.
defend the bishop with :e2 or liJe3,
as he does in the orthodox main Instead of 13 ... 0-0? 14 a4 ...


use the 11.. ..id7 12 ltJbd2 cxd4

13 cxd4 :c8! 14.ib 1 0-0 trick.

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,
... he can attack with 13 ... g5!
... ltJd7 or ...ltJc6, with or without a
(14 ltJxg5 :g8), with advantage
trade of c-pawns.
according to Aleksandr Matanovic.
Often he adds ... :e8 so that after
The drawback to the Old Main
... .if8 he smoothes out his
Line is that White saves a tempo by
development and puts pressure on
not playing h2-h3. But that may be
e4, e.g. 11...'ifc7 12 ltJbd2 cxd4
temporary. For instance, 11 ltJbd2
13 cxd4 .ib7 14 ltJf1 :ac8 and
0-0 12 ltJf1?, rather than 12 h3!,
15 ...:fe8.
allows 12 ... cxd4 13 cxd4 .ig4! with
excellent play. More economical is Yaacov
Murey's gambit idea, 11 ...:e8!?
Tal was plainly frustrated with
It's based on 12 dxe5 dxe5
the Old Main Line when Larsen
13 ltJxe5 .ib7 with compensation
adopted it against him in 1991. Tal
for the sacrificed pawn.
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
This was the final game between 14 'iff3 .if8 15 .if4 :e6 16 ltJg4
these two great warriors and it ltJc4 17 ltJxf6+ :xf6 18 tiJd2 tiJe5,
ended in a draw after Black got to Chandler-Hebden, Millfield 2000.


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 Zaitsev's original move order is
Ruy occurs when Black meets the inexact if Black wants to avoid a
tabia's 9 h3 with 9...:e8 10 d4 draw as well as the complications of
i.b7. He prevents the liJd2-fl 9...:e8 10 a4 liJa5 11 .ta2. More
maneuver that is so valuable in the precise is 9....tb7! first and then
Ruy, as noted earlier. 10 d4 :e8, transposing.

Chapter Three:
Sicilian Defense

Transposition tricks punish White has a strong sacrifice, 8 fS!

routine moves more often in the gxfS 9 tbh4 fxe4 10 dxe4 followed
Sicilian than in any other opening. by tbf5 or 'i'h5, e.g. 10 ... ~e6
That might be expected in sharp 11 tbf5 .ltxf5 12 exf5 f6 13 tbc3
Sicilian lines but it also applies to 0-0 14 tbd5, Korolyev-Varlamov,
the closed variations. correspondence 1981.
For example, White can try to If Black chose 6... tbf6 instead,
save a tempo compared with the Korolyev was content to transpose
traditional closed order, 1 e4 cS to a normal closed with tbc3, since
2 tbc3 tbc6 3 g3, with 2 g3. The many players think ...tbf6 is one of
Russian postal champion Sergey the less effective closed defenses.
Korolyev showed the benefits of
The drawback to 2 g3 is 2... dS!.
this following 2...tbc6 3 .ltg2 g6
White doesn't have time for 3 d3
4 d3 .ltg7 S f4 d6 6 tbf3.
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

Sicilian Defense

CHAMELEON SICILIAN liJc6 3 liJge2 e6 4 g3.

This elastic system is based on Black indicates his willingness to
White's ability to delay a decision play a Taimanov or Scheveningen
of whether to open the center with variation and White is hinting at a
d2-d4. It works because the closed line. But if Black naively
sophisticated Black moves of an responds 4... g6?, expecting 5 .tg2
open line, such as ... a6 and ...'iic7, .tg7
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.

... 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
Black's setup would serve him
in normal Closed Sicilian orders
well after 5 d4 cxd4 but not in the
such as 1 e4 c5 2 liJc3 liJc6 3 g3 g6
game's continuation, 5 .tg2 .tb7
4 .tg2 .tg7 because he has too
6 d3! liJc6 7 0-0 d6 8liJg5. Despite
much firepower focused on d4.
his opening expertise in the open
Sicilian, Lajos Portisch appeared And suppose Black abandons
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.

Sicilian Defense

There are three basic versions of players are not familiar with. And
the Chameleon: White can bring his 5... a6 is trickier than you might
knights to f3 and c3. Or he can play think - 6 'ii'd3 tbc6 7 ..tf4 'ii'c7 8 0-
2 tbe2 and 3 tbbc3. Or he can begin 0-0 tbe5?? 9 tbxe5 dxe5 10 ..txe5!
with 2 tbc3 and then 3 tbge2. 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).
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
Then 5...tbc6 6 ..tb5 is a position new and complicated territory after
(usually reached via 2 tbf3 d6 3 d4 4 d4 cxd4 5 'ii'xd4! (5 ...tbf6 6 e5
cxd4 4 'iixd4) that most 2 ... e6 tbc6 7..tb5).

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
7 :b 1 0-0 8 b4 a6 9 d3 b5! as in Perhaps Black's best policy after
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.
far 3... e5! has performed well. He delays ... e5 until White has
given up on tDd5/i.c4, such as 4 g3
Jose Capablanca and Paul Keres e5!.
tried to solve the ... e5 problem with
the second Chameleon version, There are slight differences in the
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
pieces well. has a choice between transposing
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
and invites the complications of That avoids the 6 i.g5 Najdorf
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 Pons-
Topalov, Leon 2006.
The problem with the Najdorf-
wannabe order is Black has to
justify ... a6 after 4 g3!.

Sicilian Defense

Now 4 ... e5 is not nearly as Then 4 g3 .ltg7 is an innocuous

effective when White's f-pawn is Closed Sicilian in which White's
unblocked. He is virtually a move KN is once again misplaced. Better
ahead of normal closed Sicilians is 4 d4 cxd4 5 tDxd4 i.g7. If Black
after 5 .ltg2 tDc6 6 d3 g6 7 f4 .ltg7 has any Dragon instincts he'd be
8 0-0 tDge7 9 .lte3 0-0 10 'i'd2 happy here because he enjoys the
tDd4 11 litf2. benefits of the accelerated order
without allowing White his
Note that Black also gets an extra strongest weapon, the Maroczy
option in both the 2 tDc3/3 tDge2 Bind.
and 2 tDc3/3 tDf3 orders. After But White can also feel happy
2 ... tDc6 and 3 ... d6 4 d4 exd4 with this move order, because he
5 tDxd4 he can try 5... e5. discouraged Black from a slew of
other openings he might have
This has the benefit of reaching a prepared, such as the Taimanov or
Sveshnikov Variation (6 tDdb5 a6 Sveshnikov.
7 tDa3 b5 8 tDd5) but with the
possibility of 8 ...tDce7!?, which has GRAND PRIX ATTACK
tested well. However, White also The Grand Prix was once
has an extra option, 6 tDf5!?, and considered a patzer's opening. But
this entire 5 ... e5 line must be called no less a Sicilian authority than
unclear. Yevgeny Sveshnikov said that
allowing it by answering 2 f4 with
The biggest problem with all the 2 ... tDc6 was a serious mistake.
Chameleon orders is the accelerated 'Already Black is worse,' he wrote.
form of the Dragon.
White's basic plan in the attack is
After 1 e4 c5 2 tDf3 tDc6 3 tDc3 tDc3 and f2-f4 followed by tDf3 and
g6!: posting his KB on c4 or b5. If Black

Sicilian Defense

could count on White not trans- What about a more flexible - and
posing, his best order would be 1 e4 very common - order such as 1 e4
c5 2 lLlc3lLlc6 3 f4 e6 4lLlnlLlge7: 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,
But the greatest strength of the
e.g. 5...lLlge7 6 i.g2 and if 6... g6
Grand Prix today is its trans-
then 7 d4! exposes the d6 hole
potential. White can switch to an
open Sicilian, 5 d4 cxd4 6 lLlxd4.
The result is perfectly good version Better is 4... a6. But 5 d4 cxd4
of the Taimanov Variation. But 6 lLlxd4 lLlf6 would land Black in a
there's quite a bit of theory for strange Najdorf. That is, 1 e4 c5
Black to know - and if he's not a 2 l'bf3 d6 3 d4 cxd4 4 lLlxd4 l'bf6
Taimanov player, he's been tricked 5 l'bc3 a6 and 6 f4 l'bc6?!, e.g.
out of his repertoire. 7 i.e3 i.d7 8 'iYn g6 9 l'bd5! ':c8
10 l'bxf6+ exf6 11 c3 is excellent
This kind of switch happens
for White (F 0 garasi -Ftacnik,
because Black has to make pawn
Austria 2006).
and piece commitments. That
provides White with the inform- Most likely the best answer to
ation he needs to decide whether to 1 e4 c5 2 l'bc3 l'bc6 3 f4 d6 4 l'bf3 is
transpose into another Sicilian line. neither 4... e6 nor 4... a6 but 4... g6.
In the last diagram for example, Black should be happy with 5 i.b5
White could play 5 g3, creating a i.d7, a familiar Grand Prix
closed Sicilian in which Black has position. But he had better know a
given up on the recommended lot of Dragon theory as well
... lLlf6 and ... e5 defenses. because of 5 d4!?

Sicilian Defense

The Grand Prix is a particularly The trickster can try to exploit

good weapon against Najdorf, Kasparov's order with the
Scheveningen and Modem Dragon paradoxical 5 .tb5+ .td7 6 .tc4!?
players because they'll meet 1 e4 c5 His point is that Black's bishop gets
2 liJc3 with 2.. d6 in the hopes of in the way of supporting ... d5.
transposing (3 liJf3 liJf6 4 d4) or
There are only a few examples of
punishing White's order (3 liJf3
e5!). 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
... liJd4, a strong idea when the pawn And there is one more headache
is at d7. for Black after 3... g6. White can
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
(7 ... dxe5 8 'i'xd8+ ~xd8 9 fxe5!
liJxe5 10 .tf4 liJed7 11 0-0-0).


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
He also invites a Dragon, 5 d4 second move before he commits
cxd4, and can meet 5 .tc4 himself to c2-c3. That's significant
with 5 ...liJc6 6 0-0 e6 followed because after 2 liJf3 liJc6 3 c3!?
by kicking the bishop with Black has lost some of his best 2 c3
... liJge7/ ... d5 options.

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 And now 6 exd5 allows
way (3 ...ttJf6 4 e5 ttJd5 5 d4 cxd4 6... ttJxd5!. What's happened is
6 cxd4 b6 7 i.c4!). Black slipped into a harmless Semi-
Tarrasch 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,
e.g. 5 d4 i.g4 6 i.e2 e6?! 7 h3 White does better with 6 e5 even
i.h5 8 c4 'i'd6 9 g4 i.g6 10 d5! though 6...ttJbc6 and 7...ttJf5 is
with advantage. quite a respectable version of the
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
after 2 ttJf3 a6, when White tries to That's why you'll see 1 e4 c5
prove that ... a6 is irrelevant. 2 ttJO e6 3 c3 d5 4 e5 being played,
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!?

Sicilian Defense

MAIN ALAPIN LINES unclear, especially after trendy

After 2 e3 tLlf6 play usually continuations such as 7 .te2 'fIe7
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
7 .Jtc4 dxe5 and 7... tLlb6 8 .i.b3 d5. 6 tLlf3 e6 7 tLlc3 'fId8 8 i.c4
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

But White can cross him up with setups with .i.d3 because of 6 .id3
6 .i.xd5! exd5 7 d3 and claim a .i.g4! (7 .i.e3 cxd4 8 cxd4 i.xf3
slight edge. The more accurate 9 gxf3 e6).
route to the fianchetto line is 4 tLlf3 Another benefit of 5... tLlc6 is that
b6! 5 ..tc4 ..ib7. 6 .i.e2, which prepares c3-c4, can
The drawback to 4 tLlf3 tLle6 be handled by 6.. exd4! 7 exd4 e6
5 .Jte4 is Gedeon Barcza's idea, with rough equality. White cannot
5... tLlb6! 6 ..ib3 e4. That's not take with his knight on d4 because
necessarily bad for White. It's just g2 is hanging.

Sicilian Defense

In contrast, 6... e6 7 0-0 cxd4 8 i.g5! and 0-0-0.

8lbxd4! lbxd4 9 cxd4 followed by
However, 7 ... h6!? leaves him
lbc3 or i.f3 favors White.
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.
open Sicilians with both 1 e4 c5 As a result White has been
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
Another virtue is that 4 d4 cxd4 has found a drawback to 7 i.g5!.
5 'ifxd4 i.d7 6 i.xc6 i.xc6 is a
main line that can also occur after The easiest way to deal with 1 e4
2... d6 3 d4 cxd4 4 'ifxd4 and then c5 2lbn d6 3 d4 cxd4 4 'ilVxd4 with
4... lbc6 5 i.b5 i.d7 6 i.xc6 i.xc6. the prospect of c2-c4 is simply to
avoid it by means of 3 ...lbf6!
If Black spends his home time on (4lbc3 cxd4!).
that position, he can be prepared for
all the major i.b5(+) and 'ilVxd4 Top GMs such as Vishy Anand
variations. have tried to find an edge for White
in 4 dxc5 but failed after 4...lbxe4
Note that in the 6 i.xc6 i.xc6
5 cxd6 lbc6! 6 lbbd2 lbxd6.
position White's only bid for
advantage or even for the initiative There's a more direct, but little
is supposed to be 7 lbc3 lbf6 known, way for White to seek the

Sicilian Defense

Maroczy Bind, 1 e4 cS 2 tiJf3 d6

3 d4 cxd4 4 tiJxd4 tiJf6 S .td3!?

For example, 8 ....tg7 9 .tb2 0-0

10 0-0 "ikc7 11 tiJc3 (Fogarasi-
Starostits, Scanno 2005). So far a
simple antidote to 5 .td3!? hasn't
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 MODERN DRAGON
and avoid c2-c4. Today we know VARIATION
that the pawn structures after The Modern Dragon was born as
5 iLd3 e6 6 c4 or 5... g6 6 c4 aren't a counter-finesse. When the Dragon
necessarily bad for Black. But they began scoring victories in the 19th
can be very difficult to handle if century it commonly came about
you're unfamiliar with them - as a via 1 e4 c5 2 tiJf3 tiJc6 3 d4 cxd4
Modern Dragon, Scheveningen or 4 tiJxd4 tiJf6 5 tiJc3 d6 6 .te2 g6.
Najdorfplayer usually is. But White found an effective
deterrent in 6 .tg5 and then 6... g6?!
Black has two forcing replies to 7.txf6!.
. 5 .td3. One is 5... e5 when 6 tiJf3
and 7 c4 reaches a good version of Kurt Richter, who helped launch
6 .tg5, also popularized the refined
the Kalashnikov Variation for
order of 1 e4 cS 2 tiJf3 d6 3 d4 cxd4
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. The price Black pays IS
Unlike the 6 .td3 Najdorf we'll committing himself to a Dragon
consider later, Black hasn't spent a before .te2. This gives White more
tempo on ... a6. But experience aggressive options, including the
indicates White stands well after Yugoslav Attack - .te3/fl-f3J'iid2/
7 c4 g6 8 b3: 0-0-0 - and some quiet ones like the

Sicilian Defense

fianchetto system, 6 g3. Even the More common is 6 .te3. It sets a

latter contains pitfalls after 6 g3 small trap (6 ... lDg4?? 7 .tb5+) and
.tg7 7.tg2: 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
This trap is important to Then on 9 .tc4 he can transpose,
recognize because it arises from with 9... 0-0, into main lines. That's
other move orders, including 1 e4 useful during times, like these,
c5 2 lDf3 lDc6 3 d4 cxd4 4 lDxd4 when Black is faring better against
lDf6 5 lDc3 d6 6 g3 g6 when 7 .tg2 .tc4 Yugoslavs than against non-
.tg7 8 lDxc6! transposes into what .tc4 ones. Thanks to this order he
White wants and 7... lDxd4! 8 'iixd4 can temporarily keep his king in the
.tg7 is what Black wants. 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
punished by 6... 'i'b6 7 .te3 'iixb2. Another situation arises when
But that allows 8 lDdb5! with White plays .tc4 but delays 'iid2,
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
make 6... 'i'b6 the move deserving This order tries to tempt Black
ofa '?!'. into 8... 'i'b6. That looks good in

Sicilian Defense

view of 9 ii.b3? liJxe4 or 9... liJg4. Then 10 exd5 liJxd5 11 liJxc6

However 9 liJf5! 'ifxb2 10 liJxg7+ bxc6 12 ii.d4 is heavily analyzed
is such a promising sack that and remains controversial and
masters abandoned 8... 'ifb6?! in the 12 liJxd5 cxd5 13 'ifxd5 is risky.
1960s. White should avoid the alternative
Black's best policy after 8 ii.c4 is 10 liJxe6 bxe6 11 exdS.
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.

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
For instance 13 hxg6 hxg6 13 a3 l::tfd8 14 'ife2 l::txd1+
14 liJdS liJxdS 1S ii.xdS l::te8 15 'ifxd1 l::td8 16 'i'e2 liJd5
16 'ifd2 ':xe2! 17 'ifxe2 ii.xd4 17 liJxd5 'ifxd5 18 c7? ii.xb2+!
18 0-0-0 ii.g7 19 'itb1 'ib6 20 g4 19 'itxb2 'i'a2+ and wins.
e6 as in Perez Candel-Guseinov,
If Black doesn't play 9... d5, he
Kusadasi 2006.
has two natural orders to reach a
tabia. One is 9...liJxd4 10 ii.xd4
DRAGON: 90-0-0 ii.e6 11 'itb1! 'i'e7 (not 11...'i'a5
12 liJd5! 'i'xd2 13 liJxe7+).
In the Yugoslav tabia (1 e4 eS
Experience says White is better but
2 liJt3 d6 3 d4 exd4 4 liJxd4 liJf6
not enough to discourage Black
S liJe3 g6 6 ii.e3 ii.g7 7 t3 liJe6
from seeking it.
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
after the forcing 9...dS. be skittish about it because ...

Sicilian Defense

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
.. .instead of 10 'it'b 1 ttJxd4 winning .
11 i.xd4 "ikc7, transposing, White
This underlines a danger to both
can try 10 ttJxe6 fxe6 11 g3! and
players in a Modem Dragon - the
12 i.h3, which puts e6 under fire.
possibility of a 9 0-0-0 line
But there's an upside to 9... i.e6. suddenly turning into a risky 9 g4
Black steers clear of 9 ... ttJxd4 position and vice versa. Or a
10 i.xd4 i..e6 11 ttJd5!?, which 9 0-0-0 line transforming into a
many Dragonistas don't like to 9 i.c4 position.
face. So it's up to him to choose
which line he wants to avoid more, That's the case with 9 0-0-0
the 11 ttJd5!? one or the 9... i.e6 ttJxd4 10 i.xd4 "iia5 and 11...i.e6.
10 ttJxe6 one. This looks like just another routine
Whatever the route to that tabia Dragon. But Black's order is faulty
White usually continues 12 g4 or because 11 i.c4! will transpose,
12 h4. A clever transposer may after 11 i.e6 12 i.b3, into that bad
prefer 12 i.b5!? The bishop is 1950s line.
headed for b3, its best square on the
But doesn't that just lose time? 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 9 g4 was dealt a serious blow
m the 1995 world championship

Sicilian Defense

match. What Team Kasparov came DRAGON: 9 il..c4

up with was 9.. il..e6! 10 0-0-0
Some Dragon experts, like
tbxd4 11 il..xd4 'ilVa5!.
Sergey Tiviakov, have circum-
vented this hyper-theoretical line by
meeting 6 il..e3 il..g7 7 f3 tbc6
S'ilVd2 with!?:

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 Black is willing to transpose into
non-il..c4 lines after 9 0-0-0 0-0 or
9 0-0-0 tbxd4 10 il..xd4 il..e6
9 g4 0-0. The obvious drawback is
11 ~bl! when l1..:iWc7 is played
9 tbxe6 fxe6. But so far the pawn
because 11 ... Via5? stumbles into
weakness hasn't proven significant
12 tbd5!.
in games that went 10 0-0-0 .a.c8
An extra tempo is worth 11 il..c4 'ilVd7 or 10 il..c4 ct;f7 11 a3
platinum in the Dragon and this was (else ... tba5) .a.c8 12 il..a2 tbe5 and
apparent in the first Anand- ...tbc4.
Kasparov game. After 12 <itbl.a.fcS Black can try to improve on this
Black was already threatening with S..tbxd4 9 il..xd4 il..e6, when
... l:.xc3/ ... tvxa2+. His attack would White is denied both tbxe6 and
be faster after 13 a3 ':'abS 14 h4 b5 il..c4. But this fails to 10 il..b5+!.
15 h5?! b4!. Then 1O... il..d7? 11 il..c4 gets Black
into the position he was trying
There doesn't seem to be a to avoid - and a move down to
downside to 9 g4 il..e6! as there is in boot. There's been virtually no
the comparable 9 0-0-0 il..e6 experience with 1O... ct;f8 (11 0-0-0
position because 10 tbxe6 fxe6 'ilVa5) so we'll have to let the
11 g3 is, of course, illegal. tricksters test it for us.

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

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
The reason is that this order halts play 6 f4 to see if Black falls into it,
the quick attack of 10 h4? with and then shift into the Classical line
10 'i'b4! 11 i.b3 ~xd4 and after 6 ... ~c6 7 .te2 .tg7 8 .te3
12 .txd4? ~xe4. White has to play 0-0.
the 12 'i'xd4 ending and, while it's But he's outsmarted himself
not dreadful, it's certainly not what because 9 O-O?! walks into 9'i'b6!
he had in mind. again.
The way to avoid that kind of
surprise as well as other liberating
When White plays .te2 he ideas such as ... d5, ... ~xd4 and
usually continues fl-f4 and .te3. ... ~g4 is to retreat the knight
An early tabia is 1 e4 c5 2 ~f3 d6 from d4 to b3. Siegbert Tarrasch
3 d4 cxd4 4 ~xd4 ~f6 5 ~c3 g6 routinely gave ~b3 a question mark
6 .te2 .tg7. in the Sicilian. But a case can be
There are few .te2 lines in which made in the Modem Dragon for
6 .te2 .tg7 7 ~b3 rather than
White castles long, so 7 0-0 would
7 0-0.
seem to best retain his options.
However, the g I-b6 diagonal is a The reason is the natural 7... 0-0
problem after 7... 0-0 8 .te3 ~c6: allows White a sharp extra option,

Sicilian Defense

8 g4!?

. White may be gun-shy about 8 f4

Black's king posItion makes because of 8.. :ii'b6. But this time
8...tbc69 g5 tbd7 10 h4 or 10 ..te3 White has a strong 9 e5!, which
more dangerous than if he had Sammy Reshevsky once allowed
played 7 ...tbc6. (9 ... dxe5?? 10 tbxc6 wins).
Some players don't like tbb3
After 8... 0-0 instead, White
because it allows Black to develop
reaches the Foltys Attack with
... ..te6 without fear of tbxe6. But
the bishop faces more jeopardy 9 tbb3 and then 9... ..te6 10 g4. As
from pawns - 7 0-0 tbc6 8 tbb3 for Black, his way of dodging it
..te6? favors White after 9 f4! and is 9... tba5. Then 10 0-0 ..te6!
10 f5!, e.g. 9 .. J:tc8 10 f5 ..td7 transposes into a position normally
11 g4. reached via 9... ..te6 10 0-0 tba5 that
is considered safe enough.
In truth, tbb3 serves too many
good purposes and the main
question is when to play it. It is This order allows Black to meet
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
This order avoids 8 tbb3 ..te6 der Wiel, Dutch Team Champion-
9 f4 'iYc8!, which stops g2-g4. Then ship 1996.
White has nothing better than to
transpose into the older lines with White, too, gets an extra option
10 0-0 0-0 that work well for Black from this order and it is 10 e5.

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

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
doesn't fit into any traditional If that's true, 6 ... ltJxd4!? IS a
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
with the vanilla move 6 i.e2. after all these years.

Sicilian Defense

ACCELERATED DRAGONS lines with ~hllf2-f4 in which .i.e3

The Accelerated Dragon, 1 e4 cS is delayed.
2 liJf3 liJc6 3 d4 cxd4 4 liJxd4 g6, Some books claim 6 liJb3 is
becomes a Modem Dragon after refuted by 6 ... .i.xc3+!? (7 bxc3liJf6
... d6. The benefits of the 8 .i.d3 d5). But many GMs don't
Accelerated order lie in limiting believe it: Anatoly Karpov has
White options and enabling Black played 6 liJb3 anyway and Gata
to shoot for ... d5!. Kamsky and Bent Larsen passed up
6 ... .i.xc3+ when they had the
White's knight is immediately
chance as Black.
attacked after S liJc3 .i.g7. If he
wants to fianchetto, he has to try Nevertheless if White wants to
something like 6 liJde2liJf6 7 g3. 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
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
tl> not 8 ... d6 but 8 ... d5! with excellent
Now 7 ... 0-0 8 ... g2 d6 9 0-0
chances. The same goes for 8 g4
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 But there is a rarely mentioned
retains the .i.g5 option as well as downside to 6 ... liJf6. It grants

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 No, not when White develops
sleep at night. ii.g2, for example. One of the
Because of the ways that the best features of the modem
Accelerated move order degrades Scheveningen order, 1 e4 c5 2 ltJf3
White's options it would be a d6 3 d4 cxd4 4 ltJxd4 ltJf6 5 ltJc3
perfect defense to 1 e4 - if it e6, is that after 6 g3:
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 Black can dispense with ... 'iic7
play 1 e4 cS 2 ltJf3 g6 3 d4 ii.g7 and obtain equal play from 6...ltJc6
Black has to feel comfortable in 7 ii.g2 ii.d7!, e.g. 8 0-0 a6 9 a4 ii.e7
other openings, including a Schmid 10 ii.e3 0-0 11 ~hl ltJeS 12 f4
Benoni (4 dS) and a Panov Caro- ltJc4, Malakhov-Vogt, Feugen
Kann after (4 c3 cxd4 S cxd4 dS! 2006. No better is 8 ltJdbS ~8
6 exdS!) as well as 4 dxcS!? 9 ii.f4 ltJeS and ... a6.

Sicilian Defense

The Scheveningen order is also 5 ... a6 6 ~c4 e6 7 ~e3?! ~e7. For

effective against White players who example, 8 'iWe2 b5 9 ~b3 b4 or
try to escape to a more exciting 9 ~d3 .tb7 10 a3 tiJbd7 11 0-0
Sicilian with 6 ~c4 or 6 ~g5. They tiJc5 12 f3 0-0 and Black stood
expect 6 ... tiJc6 or 6 ... a6, reaching splendidly in Ulko-Ulybin,
familiar territory. Moscow 2005.
But Black has a better move,
6 ... ~e7!. Then both sides risk being KERES AND ENGLISH
trapped in an inferior Najdorf or ATTACKS
other Sicilian.
Two major alternatives to quiet
For example, 6 ~c4 ~e7! 7 ~b3 Scheveningens involve g2-g4.
and 7 ... a6? 8 f4 favors White, e.g. There is the Keres Attack, with an
8 ... b5 9 e5! with advantage or immediate 6 g4, and the English
8... 0-0 9 'iWf3. (In the Najdorf move Attack, with ~e3, f2-f3 and 'iWd2
order this happens with 5 ... a6 added.
6 ~c4 e6 7 .tb3 when Black plays
7 ... ~e7?! instead of7 ... b5!.) For decades it was thought the
Keres Attack could only come
Black can get better counterplay about from the traditional order.
from ... tiJa6-c5 rather than ... a6, e.g. There's no point to g2-g4-g5 if
7... 0-0 8 ~e3 tiJa6! 9 f4 tiJc5. Black hasn't put a knight on f6, the
If White is alert he may try thinking went. As a result a
7 ~e3 instead of 7 ~b3. He 1967 Shakhmatny Bulletin article
wouldn't mind reaching a book recommended 1 e4 c5 2 tiJf3 d6
Sozin Variation (7 ... tiJc6). But 3 d4 cxd4 4 tiJxd4 tiJc6 5 tiJc3 e6.
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.

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
g5 is an excellent Keres. Avoiding Another example, is 6.....te7 7 g5
... tiJf6 is better but not necessarily tiJfd7 8 h4 tiJc6 9 ..te3 when
equal (6 ... a6 7 ..te3 tiJge7 8 tiJb3). 10 ..tc4! gives White a favorable
Velimirovic Attack.
The sad truth is there is no
simple anti-Keres route to the From the Sozin order (1 e4 c5
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
move alternatives. - 7... a6 8 'i'e2 ..te7 9 g4? 'i'c7?
(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
5 tiJc3 e6 6 g4, a consensus has Another dangerous attacking
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
tiJfd7 8 ..te3 b5 9 a3: The pawn move seemed superior
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:

Sicilian Defense

10 g4! dxe4 11 'iVf2 'iVc7 12 g5 with


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
If White adopts the f2-f4 plan it in the 1880s. But there already is
Black obtains quick counterplay a Paulsen Variation so 'Classical'
(7 f4 b5! 8 'iVf3 .tb7 9 .td3 ttJbd7 may have to do.
10 a3 ':c8).
Black has a choice of two orders,
But 7 g4 is a good response. 2 ... d6 and 5 ... ttJc6 versus 2... ttJc6
Then the natural 7... h6 transposes and 5 ... d6. Today his decision
into a kind of Keres in which White usually hinges on how he feels
has added the useful move .te3 about 2 ... ttJc6 3 .tb5 as opposed to
while Black has added the slow 2 ... d6 3 .tb5+, as well as how he
... a6. As a result White has good would deal with 3 c3 or 3 ttJc3 .
chances whether he continues a la
The Paulsens used the Classical
Keres (8 h4), in English Attack
to reach the Dragon, Scheveningen
style (8 f3), or in a TaVTopalov
and other lines that came to
hybrid with 8 f4.
prominence long after they were
There is a drawback to 7 g4 gone, including Isaac Boleslavsky's
inasmuch as 7. e5 8 ttJf5 g6 forces 6 .te2 e5. Today it's also used as a
White to sack material (9 ttJg3? waiting policy. After 6 .te2, for
.txg4). But 9 g5! gxf5 10 exf5, the example, Black can play a
Bela Perenyi line, has been Scheveningen or a Dragon more
prohibitively strong. safely (or 6 ... ttJxd4!? 7 'iVxd4 g6 as
noted earlier).
But there is no absolutely correct
English move order after 6 .te3 a6, In the majority of games White
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.

Sicilian Defense

This move can be a problem for a l:.e8 14 i.b3 a6 15 llJd4 llJa5

Black who uses the Classical to 16 llJf5 llJxb3 17 axb3 llJxe4
dodge the Keres Attack and reach 18 llJh6+ gxh6 19 l:txe4 i.xe4
the Scheveningen (6 i.e2 e6). 20 llJxe4 i.f8? 21 i.g5 with mixed
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
Other tricky Classical options are
ambitious plan in 7 i.e3 and 'if01 6 i.e3 and 6 0, which are both
0-0-0. This transposes into 1 e4 c5
aggressive and waiting moves.
2 llJf3 d6 3 d4 cxd4 4 llJxd4 llJf6 They can transpose into a Dragon
5 llJc3 e6 6 f4 llJc6 7 i.e3 when (6 ... g6) or an English Attack (6 ... e6)
Black has lost the chance for the - but if Black really wanted to play
sharper 6 ... a6 and ... b5/ ... i.b7. a Dragon he would have chosen
A Black who prefers to meet 5 ... g6 and if he wanted an English
6 i.e2 with the Boles1avsky he would have played 5... e6.
Variation may also have a problem He loses some options this way.
with 6 f4. Books recommended Theory prefers ... a6 to ... llJc6 in the
6 ... e5 because 7 llJxc6 bxc6 8 fxe5 English but after 6 i.e3 e6 Black is
llJg4! gives him excellent dark- already committed to ... llJc6. Also,
square play whether White accepts 6 ... e6 7 i.c4! will be annoying to a
the pawn (9 exd6 i.xd6) or permits Black who usually meets 6 i.c4
9 ... llJxe5. with Benko's 6 ... 'iib6.
However, 6 ... e5 is tested by There are two clouds on 6 i.e3 's
7 llJO! followed by i.c4. An horizon. One is 6 ... e5, which
example of a Boleslavsky player transposes into a line of the
finding himself in unfamiliar Boleslavsky that is considered even
surroundings is Salmensuu-Atalik, (7 llJb3 i.e7 8 i.e2) or into a
Groningen 1999: 7 ... i.e7 8 i.c4 Sveshnikov Variation that's rated
0-0 9 0-0 exf4 10 i.xf4 i.g4 good for Black (7 llJdb5 a6 8 llJa3
11 'iid2 i.h5 12 l:.ae1 i.g6 13 ~h1 b5).

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.
White can still reach Yugoslav Both moves enjoy good reputations.
Dragon and English Attack
That's why Black may prefer
positions (6 ... g6 7 .te3 or 6 ... e6
6...'it'b6 and then 7 lbb3 e6 8 'it'd2
7 .te3) but he avoids 6 .te3 lbg4.
.te7 9 0-0-0 0-0, transposing. A
Experience with 6 ... e5 7 lbb3 .te7
third route to the tabia is 6... e6
8 .te3 .te6 9 lbd5 and 7 ... .te6
7 'it'd2 'it'b6 8 lbb3 .te7 9 0-0-0
8 lbd5 is mixed.

RAUZER VARIATION Which order is best? The latter

two threaten ... 'it'xd4 as well as
In the Classical order, 6 .tg5 ... 'it'xb2. But they share the demerit
discourages a switch to the Dragon of allowing White to double pawns
(6 ... g6 7 .txf6) or the Boleslavsky with .txf6.
(6 ... e5 7 .txf6 'it'xf6? 8 lbd5). In addition, White can answer
Black usually replies 6 ... e6 and 6.. e6 7 'it'd2 'it'b6 with 8 0-0-0 (and
experience has shown that White's transpose after 8 ....te7 9 0, for
only way to seek an edge is 7 'it'd2 example). The only way to punish
and 8 0-0-0. this is 8 ... lbxd4 9 'it'xd4 'it'xd4
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 He preferred 6..'it'b6. But that
might reach this position, one of the may have its own problem,
best Rauzer tabias for Black. 7.te3!?

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
Declining the pawn is bad wants to go with 8liJxd4 9 .xd4
(7 ...a5? 8 liJb3 .c7 9 liJd5! or 0-010 f4.
8...d8? 9 i.e2 e6 10 0-0, a
The problem with the second
Scheveningen a tempo or two
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
best may depend on whether Black White also has a promising
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
Another tabia is: d5!).

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

Sicilian Defense

His idea is to answer the natural This order is significant because

10 Jic4 with 1O ... liJxd4 11 'iixd4 on 9 f4 Black transposes not into
Jic6 and reach a promising position the bad line (9 ... 0-0?) but into a
(12 .l:thel 'iia5). very good one, 9... Jid7!. This
comes about via 7 'iid2 a6 8 0-0-0
That's a line that commonly Jid7 9 f4 Jie7 when White has
comes about via 9 ... liJxd4 bypassed 9 D.
10 'iixd4 'iia5 11 Jic4 Jid7 when
White passes up 12 e5! for the less If he insists on 9 f3 Black
ambitious 12 ::thel. 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 circum-
Sutovsky-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
bxc6 12 fxe5. the Rauzer.

Sicilian Defense

SOZIN VARIATION Pal Benko's great contribution to

6 Jtc4 theory was realizing that
Joseph Blackburne used 6 Jtc4
against the Classical Variation so
6...'iib6 7 lbb3 e6 transposes to a
fairly good Scheveningen for
that 6 ... g6 could be punished by
Black, even though he'll probably
7 lbxc6 bxc6 8 e5!. Later Os sip
retreat his queen to c7 at the cost of
Bernstein demonstrated the finesse
a tempo. One of the secrets of the
of6... Jtd7.
Sozin is that getting into the right
line is often worth a tempo.

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'
The subtle reply is 7 Jtb3, which Sozin. (It would occur after
waits for Black to choose between 6 Jtc4 e6 7 Jte3 a6 if Black had the
7... e6 and 7 ... g6. But after 7 ... g6 extra tempo .. :~i'a5.)
8 .ie3 Black has 8 ... lbg4! and
against 8 1'3 there is 8 ... lbxd4! But even a move down, White
9 'i'xd4 Jtg7 with excellent play, has good chances following 10 0-0
e.g. 10 Jte3 0-0 11 'i'd2 b5 12 Jth6 Jte7 11 f4, theory says. In fact,
.ixh6 13 'i'xh6 b4 14 lbd5 e6! GMs often gives the tempo back,
15 lbe3 a5 16 h4 a4 17 Jtc4 d5!, say with 11...'i'c7 12 .tb3 0-0 to
Shivaji-Motylev, Minneapolis reach a book line, 13 'i'f3 lbxd4
2005. 14 Jtxd4 b5, with equal chances.

Sicilian Defense

In the orthodox 6 .tc4 e6 White .td7 10 .tb3 ""8 11 g4! :c8 12 gS

chooses between the old 0-0 and ttJe8 13 h4 ttJaS 14 g6! as in
f2-f4-fS plan and Dragolyub Velimirovic-Milic, Belgrade 1965,
Velimirovic's attack with .te3, one of the earliest successes of
'i'e2, 0-0-0 and g2-g4-gS. Black, White's strategy.
meanwhile, has to pick either quick
Instead, Black does better against
development ( ... .te7, ... 0-0) or
the Velimirovic when he begins
more ambitious but slow-
queenside operations, e.g. 7 i.e3 a6
developing counterplay (... a6, ... bS,
.. .'iVc7). 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 What this means is each side
is crucial. Queenside play didn't should be flexible enough to exploit
fare well against the old Sozin when the plan his opponent chooses.
6 .tc4 e6 7 0-0 was young. After White's best waiting move is
7... a6 8 .te3 'i'c7 9 .tb3 ttJa5 10 f4 7.tb3.
Then on 7....te7 8 .te3 0-0

Black often got crushed by 11 fS

... he can head toward the
ttJxb3 12 cxb3! (l2 ....te7 13 :c1 Velimirovic with 9 "iie2!, having
'iVd7 14 'i'f3 and IS eS or IS fxe6
the benefit of knowing where
fxe6 16 'iVh3).
Black's king lives.
An antidote was found in quick
The best waiting reply is 7... a6.
development, 7....te7 8 .te3 0-0
If White makes a commitment with
9 f4? d5!.
8 f4, then 8... .te7 9 .te3 0-0 heads
But that policy doesn't fare well into an old Sozin tabia that offers
against the Velimirovic, e.g. 7 .te3 chances for both sides (10 0-0
.te7 8 'iVe2 0-0 and now 9 0-0-0 ttJxd4! 11 .txd4 bS!).

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
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.
But there are no other 5... ttJbd7
RARE FIFTH MOVES tricks and plenty of drawbacks: On
6 i.g5 Black's best is probably
We know a lot about 1 e4 c5 6 ... a6, reaching a Najdorf (5 ... a6
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,
of its relative novelty.' Balogh-Bilek, Budapest 2002.

Sicilian Defense

The more successful alternative the unclear 7 ltJdb5 i.c6 8 i.xf6.

is 5...i.d7, a specialty of Viktor
Bottom line: 5... i.d7 has a few
Kupreichik and Aloyzas K veinys.
advantages over the Classical order
but Black has to risk the worst-case
scenario, a somewhat questionable
English Attack or Rauzer.

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
This works as a waiting move in on which of the nearly a dozen
cases such as 6 i.e2 g6, when Black reasonable alternatives White picks
has dodged the Yugoslav Attack, or at move six. The move orders of
6 ... e6, when he got a Scheveningen some of them, such as 6 i.e2, 6
without encountering the Keres or i.g5 and 6 f4, are too complex to
English attacks. consider here. Among the others:
The principle drawback to (a) 6 i.d3
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).
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
This virtually unbooked move
6 f4.
has great transpo value. After 6... e6
For the sake of originality, GMs 7 f4 or 7 0-0 White gets quite
prefer meeting 6 i.g5 with 6 ... e6. reasonable Scheveningen positions
That offers White a choice of a very in which i.d3 replaces i.e2, e.g.
rare Najdorf line (7 f4 a6), another 7 f4 ltJc6 8 ltJf3 i.e7 9 0-0 and
Rauzer (7 "iVd2 ltJc6 8 f4 h6) and "iVel-g3. If Black never plays

Sicilian Defense

Scheveningen positions, because he 2004. That can't be the best line for
meets 6 ~e2 with 6 ... e5, then Black after 6 ~d3. But what is?
6 ~d3 could be a problem. (b) 6 a4
And if Black answers 6 ~d3 with This is White's best waiting
6... e5 White has a good retreat move in the Najdorf. Against 6... e5,
square at e2, e.g. 7 ttJde2 ~e7 he can continue 7 ttJf3 so that
8 0-0 0-0 9 f4 ttJbd7 10 ttJg3 7....te7 8 ~c4 0-0 9 ~g5 or 7... h6
(10 ... exf4 11 ttJf5! ttJe5 12 ttJxe7+ 7 ~c4 will put him a tempo ahead
'ilixe7 13 ~xf4 h6 14 ~g3! and of 6 ~e3 e5 7 ttJf3 and 6 ~e2 e5
.th4, Gallagher-Karjakin, Panomo 7 ttJf3 lines.
Instead, Black may transpose to a
The drawback to 6 ~d3 would Scheveningen, Boleslavsky or
seem to be 6.. ttJc6, since the Dragon after 6 ... ttJc6. Evidence
attacked knight cannot be indicates the a-pawn advances help
maintained on d4 (7 ~e3 ttJg4!). White in the Dragon. For example,
However after 7 ttJxc6 bxc6 8 0-0: 6..ttJc6 7 .te2 g6 8 0-0 ~g7 9 ~e3

White can continue ttJa4/c2-c4!

as he does in the Sozin line that runs The hole at b6 gives White
1 e4 c5 2 ttJf3 ttJc6 3 d4 cxd4 chances for a queenside bind here
4 ttJxd4 ttJf6 5 ttJc3 d6 6 .tc4 'ilib6 after 10 'ilid2 ttJg4 11 ~xg4 ~xg4
7 ttJxc6 bxc6 8 0-0 and ttJa4 with a 12 ttJd5! lic8 13 a5.
bright outlook.
Also good is 10 f4 since
In the diagram he can answer 1O ... 'ilib6, which instantly equalizes
8... e5 with 9 b3 .te7 10 .tb2 0-0 when the a-pawns are on their
11 ttJa4. Then 11...~e6 12 'ilie2 original squares, can be met by the
'ilic7 13 c4 ttJd7 14 c5! favored him strong 11 a5! ttJxe5 12 e5. If Black
in Dgebaudze-Wemmers, Belgium has to settle for 10 ... ~d7 then

Sicilian Defense

11 ttJb3 and ttJd5 guarantees a So Black turned to 7...b5. Then

positional edge. White found 8 0-0 was good since
8 ... iLb7 9 :tel! ttJbd7 10 .tg5
Another idea after 6...ttJc6 is
7 iLe2 e5. Books recommend ttJc5? 11 iLd5! is a sound sack, as
8 ttJxc6 and show how it only Fischer showed.
equalizes. Better is 8 ttJb3! iLe7 To preempt that Black began to
90-00-0 and now 10 iLg5!: 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
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
This favors White after 10...iLe6 position that White wants and it
11 iLxf6 and ttJd5. This is different avoids the 7 i.b3 ttJbd7 one that
from the normal Boleslavsky Black wants.
(5 ... ttJc6 6 i.e2 e5 7 ttJb3 i.e7)
If Black plays 7...ttJbd7 anyway
when Black equalizes immediately then 8 i.g5!:
after 80-00-09 i.g5 ttJxe4!.
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,
7 i.b3, was considered best This transposes to a 6 i.g5
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
9.. :ilc7 10 f5 ttJc6 11 i.e3. Leonid Yudasin have tried

Sicilian Defense

8... 'ii'c7?!. But 9 .txe6! fxe6 11 liJd5, as in Topalov-van Wely,

10 liJxe6 has been crushing Black Wijk aan Zee 2007.
since a Keres brilliancy (l0 .. .'iic4 The crafty response to 6 .te3 is
11 liJd5 rJif7 12 .i.xf6 rJixe6
~ 6...liJc6.
13 .&.c3! and wins). To play this line
Black may have to risk 8... h6 9 .th4
g5 10 .i.g3 liJe5.
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
(d) 6.te3 an English Attacker to prepare
g2-g4. On 7 'iVd2 Black has an
White is ready for a quiet active reply in 7... liJxd4 8 .txd4 e5
Scheveningen or an English Attack. and ...te6. And if White offers a
Some GMs, like Michael Adams, Scheveningen, with 7 .te2, Black
would rather play against 6 .te3 e5 can force his way to a Classical
than 6 .te2 e5 so they begin with Dragon, 7... g6, in which White has
6 .i.e3. If 6... e6, then 7 .i.e2 allows been deprived of the Yugoslav
them to transpose to 6 .te2 e6 Attack and .tg5 lines. The price he
7.i.e3. pays is being committed to ... a6!?
Some English Attackers prefer
6 f3 and 6... e6 7 .i.e3. In this way TAIMANOVIKAN VARIATION
they avoid 6 .i.e3 liJg4. But they The standard starting position of
also surrender options such as the Taimanov Variation can be
6 .i.e3 e6 7 g4 and 6 .i.e3 e5 7 liJb3 reached in two different ways, and
.i.e7 8 'ii'd2!? Mark Taimanov used both of them,
In the last line White delays 2-f3 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4
so that he rules out ... d5 and keeps 4 liJxd4 e6 and 2... e6 3 d4 cxd4
the 2-f4 option, as in 8 'ii'd2 .i.e6 4 liJxd4 liJc6.
9 0-0-0 liJbd7 10 f4. He can also The first order discourages the
trick Black into a bad line after King's Indian Reversed (3 d3),
8... 0-0 9 0-0-0 b5? 10 f3! .te6 which is promising against 2 ... e6.

Sicilian Defense

The second order eliminates the The timing of the three key
Rossolimo, in view of 2 ... e6 Taimanov moves, ... a6, ... lLlc6 and
3 .tb5? a6!. Black also may choose ... 'iVc7, has turned out to be more
an order based on how he intends to complicated than it seemed a
meet 3 lLlc3 or 3 c3. For example, decade or so ago.
1 e4 c5 2 lLlf3 e6 3 lLlc3 is a good
Only recently was 1 e4 cS 2lLlo
Chameleon for White but 2 ... lLlc6
lLlc6 3 d4 cxd4 4 lLlxd4 e6 S lLlc3
3 lLlc3 e5! isn't.
a6 tested by the English Attack, for
Both orders, as well as a third example. This has proven to be
used by Salo Flohr, 1 e4 cS 2 lLlo dangerous after 6 .te3 lLlf6 7 g4 or
lLlc6 3 d4 cxd4 4 lLlxd4 'iVc7!?, 6... 'iVc7 7 'iVd2 lLlf6 8 0-0-0.
have the drawback of allowing a
This provides an argument in
Maroczy Bind. For instance, 4... e6
favor of S. 'iVc7.
S lLlbS threatens lLld6+ and usually
buys time for 5... d6 6 c4!.
Black can bailout of the bind
with S..lLlf6!?

Now on 6 ..te3 Black can

respond more vigorously with
6lLlf6 7 'iVd2 ..tb4! and then 8 f3
lLlxd4 9 .txd4 e5 or 9 'iVxd4 ..txc3+
10 bxc3 e5.
If White defends the attacked e-
pawn with 6 lLllc3 then 6 ... d6 The drawback to this order may
7 .tf4 e5 8 .tg5 is the Sveshnikov be 7 lLldbS 'iVb8 8 f4 although
Variation. So is 6 .tf4 e5 7 .tg5 d6 Black gets a tempo back after 8... d6
8lLllc3. and ... a6.
Of course, Black has a simpler Fischer, who liked lLlb5/c2-c4
route to the Sveshnikov if that's his lines as White, reached the
wish, and 5... lLlf6 makes most sense Taimanov as Black through another
when Black seeks a Taimanov and order, Ilya Kan's 1 e4 c5 2 lLlf3 e6
is upset to see 5 lLlb5. 3 d4 cxd4 4 lLlxd4 a6. After SlLlc3

Sicilian Defense

he transposed (S ... ~c6) into the and now S c4 ~f6 6 ~c3 .i.b4!) but
S... a6 Taimanov. 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 a2-a3 to rule out ... i..b4 and to
appears when White fianchettos, anticipate ... b5-b4. A popular order
5 ~c3 fie7 6 g3 and then 6....i.b4!: 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
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?!.

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.
Then 6... a6 7 ..tg2 d6 8 0-0 lLlf6
9 liel! allows White to carry out his Now 10 lLlb3, 10 lLlde2 and
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 THE ...~6 VARIATIONS
punished by 6 g3! and that S... a6 is There is a family of lines built
a superior waiting move (6 g3 d6!).
around ... 'iib6 and ... e6, often with
But Black has another finesse ... ..tcS. They all use pressure on the
even if he's committed to ...'fic7 b6-f2 diagonal, at least at the start,
and ... a6. but often are used to reach the
Scheveningen. The family members
(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
This can come about after 4 lLlxd4 ~6.
5 :ii'c7 6 g3 a6 7 .i.g2 and then The last two, popularized by
7.. d6 8 0-0 ..te7!? (Erling Michael Basman and Kveinys

Sicilian Defense

respectively, can transpose into one 7 i.d3 i.e7 8 0-0 0-0 9 i.e3 "fic7
another, e.g. 1 e4 cS 2 tLln e6 3 d4 10 f4 d6 11 iif3 a6 12 a4 or 12 g4.
cxd4 4 tLlxd4 iVb6 and now S tLlc3 Theory says White is favored a bit
i.cs 6 i.e3 tLlc6. 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!
His point is that 7 tLld4 tLlf6
12 'i'c4 d6.
8 tLle3 e6 is a Taimanov Variation
Black's most accurate route to with i.e3. That has a much better
the diagram is Kveinys' . One reputation today thanks to 'iVd2 and
reason is that 4... i.c5 has not done 0-0-0 but it may not be something
well when met by the surprising White feels comfortable with,
5 .te3 'i'b6 6 c3!. For example, especially ifhe normally plays i.g2
6... 'i'xb2 7 'i'b3 'i'xal? 8 tLlc2 or against the Taimanov.
7 ... 'i'xb3 8 axb3. Or 6 ... tLlc6
This order may also confuse
7 tLld2! tLlxd4 8 tLlc4.
White since 7 tLlSc3 and then 7... e6
The three other orders often 8 tLld2 bS! is fine for Black. If
reach Scheveningens after White Black has any doubts about that, he
retreats his knight to b3. The benefit should prefer order (b) because
to Black is avoiding the English and 6 tLldbS a6 7 i.e3? is a mistake
Keres Attacks. For example, in the (7 .. :iVa5 wins material) and 7 tLla3
Kveinys order, play may continue e6 8 tLle4 "fie7 9 i.e3 bS has never
5 tLlb3 tLlc6 6 tLlc3 liJf6 and then done well for White.

Sicilian Defense


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.

This changed in the 1970s when After 6 ... d6 Black says, 'I
Black adopted the Four Knights as a outsmarted you. I reached a
way of transposing, after 6 ... d6 Scheveningen in which your a2-a3
7 .tf4 e5 8 .tg5 a6, to what was is a wasted tempo.'
dubbed the Sveshnikov Variation.
White replies, 'Not at all. My
This denies White options of the goal was to avoid both the
normal Sveshnikov, 1 e4 c5 2 liJf3 Sveshnikov and 6 ... ..tb4. The added
liJc6 3 d4 cxd4 4 liJxd4 liJf6 5 liJc3
benefit is that I'll play 7 ..tc4! and
e5 6 liJdb5 d6 7 .tg5 a6, such as get a Sozin in which you've denied
6 liJf5, 6 liJdb5 d6 7 liJd5 and 7 a4. yourself the Benko line.'

Of course, there's a lot of theory Or Black can play 6 ... e5 and say,
to know in the Four Knights as 'I'm the one who got what I
well, if White decides not to join wanted, a Sveshnikov. In this
him on the road to the Sveshnikov. version you can't play 7 liJdb5 d6
But Black can reduce that 8 .tg5 because your knight has no
significantly by shifting to a retreat aft er 8... a6"..
Scheveningen with 6 ... d6.
To which White replies,
That is usually second-best. But 'Nonsense, I'm the one who came
it avoids heavy analysis such as out ahead because after 7 liJb3 it's a
6 g3 d5 7 .tg2 ..tb4 8 exd5 liJxd5 or Sveshnikov with an extra tempo for
6 .te2 .tb4 7 0-0 .txc3 8 bxc3 me. That tempo sops
t ... A~b4'.. '
liJxe4 9 .td3 d5 10 ..ta3. Instead,
There is no consensus on the best
Black is roughly equal after 6 g3
reply to 6 a3 but 6 ... d6 can't be bad,
d6!? or 6 .te2 d6!?
especially since 7 ..tc4?! really isn't
There's a who-tricked-whom a good Sozin after 7 .. :ikc7! 8 ..ta2
quality to 6 a3. a6 9 0-0 liJxd4 10 'iVxd4 liJg4.

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
6 ~db5 d6 7 i.f4 e5 8 ~d5!?: 11 'ii'f3.

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 Zaitsev-
Sveshnikov game). But so do the
main lines of the Sveshnikov
Variation. The Sicilian is never easy
for Black - or for White.

Chapter Four:
Semi-Open Games

Black reveals something about can lead to a Steinitz French (4 f4

himself when he answers 1 e4 with e6 5 d4). But the alternatives are
a move other than 1...e5 or 1...c5. somewhat questionable, such as
An Alekhine's specialist, for 3... liJe4 or 3... d4 4liJce2liJg4.
example, almost certainly likes
positional imbalances and quick If White is going for a bigger
counterplay. He may have a edge in the Alekhine's, the
problem with 1 e4 liJf6 2 liJc3!? traditional choice is the Four Pawns
because the Vienna Game (2 ... e5) is Attack, 1 e4 liJf6 2 e5 liJd5 3 c4
not likely to give him what he liJb6 4 d4 d6 5 f4.
The standard response is 5... dxe5
Moreover, few Alekhine's
6 fxe5liJc6. It sets some traps based
players are prepared for a French
on a liJf3/ ... .tg4 pin. For example,
(2 ... e6 3 d4 d5) or a Pirc (2 ... d6
7liJf3 hopes for 7... .if5? 8 d5! liJb4
3 d4), and a Sicilian (2 ... c5?! 3 e5)
is dubious. 9liJd4 with advantage but 7... .tg4!
is excellent for Black.
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
Black may be reluctant to play Viktor Korchnoi did with 5....tf5
the best move, 3... liJfd7! because it 6liJc3 e6.

Semi-Open Games

12 i.e3ltJc6 gets Black to where he

wants to go. When this order was
introduced in a 1960 Geller-
Korchnoi game White became
confused and was worse after
11 ~f4? ltJc6 12 exf6 ..txf6.


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
another trap. for exd6, since ... cxd6 allows him to
grab space with d4-d5!.
But after 9 i.e3! ~e7 White has
reasonable chances in either the But on the immediate 5 c4 ltJb6
well-trod 10 i.e2 0-0 11 0-0 f6 or 6 exd6 Black has 6... exd6!. That's
the enterprising 10 d5!? (1O ... ltJb4 a satisfactory version of the
11 ltJd4 i.g6 12 a3). Alekhine's Exchange Variation,
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
Four Pawns tabia smoothly after Another question concerns
9 i.e2 0-0: h2-h3. In most cases Black will
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

Semi-Open Games

Black wants to set up a target on If the bishop were still on g4

e5 (8 0-0 dxe5 9 dxe5 lbd7 10 'iVe2 Black would stand well with ....tf6.
'tic7). He benefits from White's But here it's en prise and that gives
sloppiness because 8 c4 lbb6 White time for 14... i.g6 15lbd4!.
9 exd6 allows the favorable
9 lbxc4!.
His pieces dominated after
15 ... ..tf6 16 lbxc6 bxc6 17 'i'xc6,
If White intends to put the as in A.Sokolov-Arzhenkov, Bor
question to the bishop, a better 2000 (17 ... i.xb2 18 ::tadl 'i'a5 19
order is 6 O-O! i.e7 7 h3. Then
'tixd6 'iVxa2 20 ..tf3 ::tac8 21 c5! a5
7.....txf3 8 i.xf3 c6 9 c4 lbb6 loses
22 c6).
material to 10 exd6 and 11 c5. The
reason Black needs 8 ... c6 is that A separate branch of the 4 lbf3
8... dxe5 fails to 9 c4! lb-moves tree is 4... g6. Today 5 .tc4 and
10 i.xb7. 5... lbb6 are considered best. But the
Moreover 7...i.f5 permits White old 5... c6!? has a new use thanks to
to carry out the space-grabbing plan transposition.
under ideal conditions. For example After 6 0-0 White seems to be
8 c4 lbb6 9 lbc3 0-0 10 i.e3 lbc6 getting a good Exchange Variation
and now 11 exd6 cxd6 12 d5! exd5 in view of 6.....tg7 7 exd6! exd6
13 cxd5! lbe5 14 lbxe5! dxe5 8 ::te1+.
15 "iVb3 with advantage.
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.

Semi-Open Games

Black has been equalizing, e.g. 11...0-0 12 b3 .tf6 13 .tb2 as

8 c3 lbd7 9 lbxd7 ..txd7 10 lbd2 14 0-0 with a space edge, Emms-
0-0 11 lbf3 ..te6 12 'iie2 lbc7! Baburin, Isle of Man 1997.
13 iH4 ..txc4 14 'iixc4 lbe6 This possibility wouldn't arise in
IS ..teS 'iidS, Sarbok-Carlsen, a normal Exchange (4 c4 lbb6
Gausdal 2003. S exd6 exd6 6 lbf3) if Black is
The point of the 4... g6 S ..tc4 c6 careful, e.g. 6.....te7 7 lbc3 0-0
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 Caro-
Kann 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
For example, 8 ..te2 ..te7 9 d5! Now 6 0-0 ..tg4! has the benefits
and 9... ..txf3 10 ..txf3 lbeS 11 ..te2 of a superior Caro-Kann (7 d4 e6
lbexc4? 12 ..txc4lbxc4 13 'iia4+ or 8 .:tel lbbd7).

Semi-Open Games

But Black is the one most likely The same idea succeeds,
to be outfoxed in the Center however, after 9 0-0-0 tiJd5 in view
Counter even when he thinks he is of 10 ~xd5 cxd5 11 'iib5+ tiJd7!
the fox. After 4 tiJf3 he might be (12 'iVxb7 l:.b8 13 'iVc6 ~a3).
tempted to seize the center with White also derives some extra
4 ... e5. But then 5 d4! traps him in a options from 9 tiJe5 !, such as
discredited old line that used to 9 ..tiJbd7 10 tiJxd7 tiJxd7 11 g4
arise after 4 d4 e5?! 5 tiJf3!. ~g6 12 h4, which has been
The major finessing of the Center dangerous the few times it's
Counter begins after 4 d4 tiJf6 occurred.
5 tiJf3 (or 4 tiJf3 tiJf6 5 d4). The The most dangerous situation for
waiting move 5 ... c6 anticipates Black is when both ~c4 and 'iie2
White most dangerous strategic are played early on so that the
idea, d4-d5, and avoids queen traps strategic d4-d5! break is in the air.
(6 tiJe5 tiJbd7 7 tiJc4 'iWc7!). For example, 1 e4 d5 2 exd5 'iVxd5
The more common reply to 5... c6 3 tiJc3 'iVa5 4 d4 tiJf6 5 ~c4 c6
is 6 ~c4. This reaches a tabia after 6 ~d2 ~f5 and now 7 'iVe2! allows
6... ~f5 7 iLd2 e6 8 'ife2 ~b4 when White's bishops to take command
White chooses between 9 tiJe5 and after 7 ... e6 8 d5! cxd5 9 tiJxd5 'iVd8
castling. The two often transpose - 10 tiJxf6+.
9 tiJe5 tiJbd7 10 0-0-0 and 9 0-0-0 The best try to exploit White's
tiJbd7 10 tiJe5. order is 5 .. ~g4!?, rather than
5 ... c6.

But the more accurate knight

move cuts down Black's options. Then 6 tiJf3? tiJc6! and 7.. 0-0-0
The simplifying 9 tiJe5! tiJd5? is gives Black the most aggressive
dreadful (10 iLxd5! cxd5 11 'iWb5+! setup he can get in the Center
'iVxb5 12 tiJxb5 iLxd2+ 16 <;i;xd2 Counter, e.g. 7 ~b5 0-0-0 8 ~xc6
tiJa6 17 tiJd6+ and wins). bxc6 9 0-0 e5 10 'ifd3 exd4 11 tiJe2

Semi-Open Games

.txf3 12 'iixf3 'iid5, David- A more subtle order is 3 d4

Kosteniuk, Dresden 1999. 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.
If you're going to seek a big edge
against 1 e4 d5 2 exd5 tbf6 you This rules out 4 ... .tg4, and
probably have to master the transposes into a main White-
complex 3 .tb5+ lines. But if you friendly line after 4 ... g6 5 tbf3.
just want a reasonable middle game Black might prefer 4....tf5 on the
you can save midnight oil with 3 c4. grounds that he tricked White into
Black's best is 3... c6 4 d4 cxd5, one of the quietest 3 .tb5+ lines
transposing into the Panov (3 ... .td7 4 .te2!? tbxd5 5 d4 .tf5 ).
Variation of the Caro-Kann.
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.
The virtue of 3 tbf3 and then MODERN DEFENSE
3...tbxd5 4 d4 lies in avoiding the
Black often uses 1.. .g6 and
sharpest versions of the Portuguese
2....tg7 as a form of reconnaiss-
Gambit (3 d4 .tg4!?). If Black
ance. He waits for White to show
insists on meeting 3 tbf3 with
his cards before he decides whether
3....tg4, White seems to get the
to transpose into a Pirc Defense
upper hand with 4 .tb5+! in view
with ... tbf6.
of 4... c6 5 dxc6 tbxc6 6 h3 .th5
7 .txc6+ bxc6 8 d3 or 4... tbbd7 This makes sense when, for
5 h3 .th5 6 tbc3. example, White has given up his

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
the Modem with 4 ... c6, 4 ... ttJc6 or
4 ... a6.
But the Pirc option soon expires.
After 4 f4 c6 5 ttJf3 ttJf6?:

For example, 5 d5 ttJd4 6 ttJbl!?,

to trap the knight, leads to double-
edged play like 6... c6 7 c3 ttJb5
8 .tg2 cxd5 9 exd5 ttJc7 10 ttJe2
This is a bad Pirc because ... c6
ttJf6 11 0-00-0 12 c4 b5!, Raetsky-
doesn't fit in with 6 .td3 0-0 7 0-0
Sakaev, St. Petersburg 1999.
b5 and then 8 e5! dxe5 9 fxe5 ttJd5
10 ttJxd5 'iVxd5 11 a4. A more precise White will prefer
Black was clearly worse after 4 ttJge2!. Then 4ttJf6 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
Pirc. g2-g3 variation.

White has to make a key decision The .tc4 lines tell a similar story.
at move four in the Modem. In the Pirc (1 e4 d6 2 d4 ttJf6 3 ttJc3
Suppose he wants to devleop his g6) White has had some success
KB at g2. The fianchetto is carried with 4 .tc4 .tg7 5 'iVe2. But in the
out smoothly in the Pirc (1 e4 d6 Modem he faces 4 .tc4 ttJc6!?

Semi-Open Games

Then 5 d5 is inconsistent and on Aside from Modem-into-Pirc,

other moves Black can shift into a Black can also try Modern-into-
pleasant Pirc, such as 5 ttJf3 ttJf6 Dragon by inserting ... c5. For
6 0-0 i.g4 or 5 i.e3 ttJf6 6 f3 0-0 instance, I e4 g6 2 d4 i.g7 3 ttJc3
7 ttJge2 e5. and 3 ... c5 seeks an Accelerated
Dragon (4 ttJf3 cxd4 5 ttJxd4 ttJc6)
The moral is that White needs without having to deal with the
ttJc3 in the Pirc order to defend e4 Maroczy Bind.
but he should delay it in the Modern
if he wants the option of meeting In practice he succeeds about one
... ttJc6 and/or ... i.g4 with c2-c3!. third of the time because I e4
The favored order is 3 ttJf3 and then players generally avoid the Schmid
3... d6 4 i.c4 ttJf6 5 'iie2. 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 MODERN: 3... c6
seeks the Classical setup, since When Black plays ... c6 before
4.ttJf6 5 ttJc3 transposes to a ... d6 he creates problems for a
heavily trafficked Pirc. 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
This avoids some offbeat lines
Ambrus-Eliseyev, St. Petersburg
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 But ... d5 and ... 'ifb6 aren't the
White's bishops. By delaying ttJc3 only ideas behind 3... c6. White has
White also discourages ... a6/ ...b5 to be wary of being trapped in an
because there is no knight for ... M unfamiliar Pirc if Black follows
to attack. with ... d6 and ... ttJf6, e.g. 4 ttJf3:

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 And then 3 c6 4 ttJf3 dS S eS.
considered fairly even. Thanks to Unlike the normal Gurgenidze,
the 3 ... c6 order Black has dodged when his knight is misplaced at c3,
the sharper Pirc lines, based on here he can support his center with
f2-f4 and/or .tg5. c2-c3. And on 3 ... d6 4 ttJf3 White
has a promising Pirc. 'I felt like a
The sternest test of 3 ... c6 is 4 f4. broken man after 3 f4,' Julian
Then 4 ... d6 5 ttJf3 ttJf6? is a bad Hodgson once said when it was
Pirc, as noted above, and 5 ... .tg4 is played against him.
a somewhat dubious Modern. Black
often prefers 4 ... d5, seeking a Nevertheless Black has a good
Gurgenidze Variation. response in 3 ... cS!, which is good
here because 4 c3? does a poor job
There are several routes to a of supporting the center (4 ... cxd4
Gurgendize tabia, including 1 e4 g6 5 cxd4 'iib6!).
2 d4 .tg7 3 ttJc3 c6 4 f4 ~6!?
S ttJf3 dS and then 6 eSt, since PIRC DEFENSE
6 exd5 .tg4! would grant Black
plenty of compensation for a pawn. The standard Pirc order is 1 e4
d6 2 d4 ttJf6. White's e-pawn is
Which route to the tabia is best?
attacked and he is more or less
The only thing that seems certain is
limited to 3 .td3, 3 ttJd2, 3 f3 and
Black should avoid 1 e4 c6 2 d4 d5
the main line, 3 ttJc3.
3 ttJc3 g6 if he really wants a
Gurgenidze. White has too many The first two moves are attempts
excellent alternatives to 4 e5 and to reach Yefim Geller's quiet
5 f4, such as 3 cxd5 and 3 e5. system in the Modern Defense (1 e4

Semi-Open Games

g6 2 d4 i.g7 3 c3 d6 4 lbf3 and (12 dxe5 lbxe5 13 lbxe5 'ixe5

i.d3/lbbd2). If he succeeds, it's a 14 lbf3 ~5 15 e5 i.f5) in
small victory since Pirc players are Rublevsky- Azmai parashvili,
often uncomfortable in Modem internet 2004.
lines. With 3 f3 instead, White is
On 3 i.d3 Black can avoid the signalling he's willing to play a
Geller system with 3... eS. Then 4 c3 Samisch King's Indian. An example
of what can happen to a strong GM
dS!? is the active way of disturbing
who likes the Pirc as Black but not
White's plan. The more solid way is
the KID was A.Rodriguez-Kuzmin,
4...lbc6 Slbf3:
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 'id2 'iia5? 15 .l::i.d8 16 i.e2
i.e6 17 0-0 'iib4 18 b3 and White
Black can try to avoid that fate
with 3 ... eS. Then he's offering
White a tiny-edge endgame, 4 dxe5
Who is more uncomfortable dxe5 5 'iixd8+. But 4 lbe2!
here? A Black who likes the sharp followed by 5 c4 and 6 lbbc3 heads
tactics of a Pirc may not like to the Samisch.
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

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
the Pirc with 4 ... g6. Dragon. On 6 d5 0-0 it's a Schmid
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 Wells- The 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!.

Semi-Open Games

The downside may lie in the CARO-KANN DEFENSE

extra options Black enjoys after
The Caro-Kann contains transpo
5 .td3, such as S...lbc6 6 dS lbb4
traps even after the super-quiet 1 e4
or 6lbf3 .tg4. But there's been too
c6 2 d3. One is set by 2... dS 3 lbd2
little experience to tell.
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 Tal-
Smyslov 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
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 old Exchange Caro-Kann.
or 7 .. :iic7 and ... c4 when his For example, 6 ... e6 7 .td3 lbe7
counterplay runs faster than usual. 8 0-0 g6 9 :e 1 iL.g7 10 lbfl lbc6
For example, 7 d5 'ikc7 8 'ike2 11 c3 0-0 12 iL.g5 and Tiviakov-
0-0 9 a3 e6! 10 dxe6 .txe6 was Dreev, Gothenburg 2005 continued
Bareev-Christiansen, Biel 1991, 12 ... e5 13lbe3! lbb6 14 dxe5lbxe5
which went 11 .tc4 l:!.ae8 12 0-0 15 iL.f4 lbxf3+ 16 'iVxf3 'ikc6
.txc4 13 'ikxc6 'ikc6! 14 e5 lbg4 17 lbc2 and lbd4 with clear
with good play. 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
experience to judge them. transposes to Tal-Smyslov).

Semi-Open Games

Then 5 d4 resembles a Tarrasch

French after 5.. exd4! 6 exd5 cxd5.

Then 4 ttJgf3 .tg4 eases Black's

game appreciably. White should
pass with 4 c3. Then 4... dxe4
But Black is a tempo (....td6) 5 liJxe4 .tf5 6 liJg3 .tg6 reaches a
ahead of the Tarrasch version, 1 e4 Caro-Kann main line but with the
e6 2 d4 d5 3 liJd2 c5 4 exd5 exd5 addition of .. :iib6 and c2-c3.
5 liJgf3 cxd4, and he should
equalize. That makes it harder for White to
develop his QB without hanging the
You will often see GMs play 1 e4 b2-pawn. However, 7 h4 h6 8 liJh3
c6 2 d4 d5 3 liJd2 and transpose may make c2-c3 more useful than
into main lines after 3 ... dxe4 .. :iib6, e.g. 8... liJd7 10 liJf4 .th7
4 liJxe4. This order sidesteps the 11 .tc4 e6 12 'ilie2, Lane-Pedersen,
obscure 3 liJc3 b5, which can lead Trnava 1985.
to a version of the ... exf6 Caro, after
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 The other trap, with 4...liJd7
3 'iVb6. 5 .tc4 liJgf6 6 ttJeg5 e6 7 'iie2,

Semi-Open Games

wasn't known until 1954, largely

because Aron Nimzovich's ... tiJd7
wasn't played much before then.

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
If d2-d4 has been inserted in
improves with 7.. dxe4! 8 tiJxe4
place of tiJf3, Black obtains a solid
tiJxe4 9 'ilixe4 tiJd7.
game from 7...tiJb6. But here that
would drop pawn to 8 tiJeS! and For example, 10 c3 tiJf6 11 'iie2
Black has to accept the slight 'iVd5 equalizes after 12 0-0 .i.d6
inferiority of 7... tiJd5. 13 ':'e 1 0-0 14 .i.c4 'iiia5 15 .i.d2
The main Two Knights line is .l:tfe8 16 .l:tadl ':'ad8 17 .i.c1 'iih5!,
1 e4 c6 2 tiJc3 dS 3 tiJf3 .i.g4 and Krai-Kamsky, U.S. Championship
then 4 h3 .i.xf3 S 'ilixf3. In the 2004.
1950s Black played 5... e6 until he
If White wants more after S... e6
found that 6 d4 dxe4 7 tiJxe4! was a
6 d4 tiJf6 he may have to try the
sound gambit, e.g. 7... 'ilixd4 8.i.d3
French-like 7 eS or 7 .i.gS but with
tiJf6 9 c3 'ilid8 10 0-0 and .l:tdl.
his QB traded it should be a good
So he adopted S... tiJf6 because French for Black.
6 d4 dxe4 7 tiJxe4 'iVxd4! is
unsound. Meanwhile, 4....i.hS (after 1 e4
c6 2 tiJc3 dS 3 tiJf3 .i.g4 4 h3) has
But Karpov has exclusively
been deeply analyzed, beginning
played S... e6. Its benefits include
with S g4 .i.g6 6 exdS cxdS
keeping f6 free, e.g. 6 d3 tiJd7
7 i..d2 .i.c5 8 'ilig3 'ilif6!, 7 .i.bS+. When an opponent plays
Chemyshov-Lastin, Voronezh 4 ... .i.hS he is usually boasting that
2006. his memory is better than yours.

The main point of S... e6 is The mental hygiene alternative is

revealed by 6 d4 tiJf6!: S d4 e6 6 .i.d3.

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.
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
White's point is that 6. dxe4
~xc3 9 ~xc3 with an edge.
7 tbxe4 ~xf3 8 'iixf3 'iixd4 9 0-0 is
a superior version of the 4 ... ~xf3 More common is 4.tbf6 5 tbc3.
5 'iixf3 e6 6 d4 gambit. A key line begins 5 tbc6 6 ~g5 e6.

Black should probably decline Now 7 c5 ~e7 8 ~b5 0-0

with 6...tbf6. Then 7 e5 tbfd7 8 g4 9 ~xc6! bxc6 10 tbf3 threatens
~g6 9 i.xg6 and 10 tbe2 offers 'iia4/tbf3-e5 and gives White an
mixed chances (Short-Khalifman, ideal version ofthe Panov bind. For
Merida 2001). example, 1O ... tbd7 11 ~xe7 'iixe7
12 0-0 e5 13 tbxe5 tbxe5 14 dxe5
CARO:PANOV 'iixe5 15 l:tel 'iif6 16 'i'd2 l:tb8
17 b4 (18 ... l:txb4? 19 tbxd5), Rath-
Vasily Panov's 1 e4 c6 2 d4 d5 Brinck-Claussen, Vejle 1974.
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.

Semi-Open Games

Then 6... e6? 7 cS! is good for 8 ... dxc4! 9 iJ..xc4 a6 because this
him, e.g. 7... iJ..e7 8 iJ..b5 0-0 9 0-0 time it's a good QGA, e.g. 10 0-0 b5
and 9 ... tLle4 10 iJ..xc6 tLlxc3 11 bxc3 11 iJ..b3 iJ..b7 12 l:.el tLlc6 13 a3
bxc6 12 'iVa4. But 6 ... iJ..g4 renders l1c8 or 11 iJ..d3 iJ..b7 12 'iVe2 tLlc6
easy equality following 7 iJ..e2? e6 13l1adl tLlb4 14 tLlbd5.
8 c5 tLle4. In place of 6 iJ..g5 White can try
Instead White's best answer to 6 cS iJ..e7 7 iJ..bS+. But an alert
6...iJ..g4 is 7 cxdS tLlxdS 8 ~3. Black will see through 7 ... tLlc6? -
This leads to a somewhat superior which heads toward the bind after
endgame that became famous when 8 tLlf3 - and play 7 ... iJ..d7! 8 iJ..xd7+
Bobby Fischer beat Max Euwe with 'iVxd7 9 tLlf3 tLlc6.
it (8 ... iJ..xf3 9 gxf3 e6 10 'iVxb7 Instead of 6 iJ..g5 or 6 c5, White
tLlxd4 11 iJ..b5+ tLlxb5 12 'iVc6+! should consider an exchange of
<J;;e7 13 'iVxb5 tLlxc3 14 bxc3 'iVd5 pawns on d5. But the immediate
or 14 ... 'iVd7). 6 cxdS tLlxdS and 7 iJ..d3 iJ..b4! has
performed well for Black.
This is an important line because
White can reach it in other orders Instead 6 tLln makes sense since
and because Black's inferiority is so 6 ... tLlc6 is once again inexact
slight he often cooperates. (7 c5!) and 6 ... iJ..b4 7 iJ..d3 dxc4
8 iJ..xc4 gives White a fairly good
Another Black strategy is quick QGA, actually a Nimzo-Indian.
castling after 4 c4 tLlf6 S tLlc3 and
S... e6. Now on 6 iJ..gS: But the main point is that against

White can tranpose, 6 ... tLlc6

White can play 7 cxdS! tLlxdS
7 cS, to that 4 c4 tLlf6 5 tLlc3 tLlc6
without having to defend c3
position considered earlier.
compared with 6 cxd5 tLlxd5 7 iJ..d3
Better is 6... iJ..e7! and then 7 tLln iJ..b4. He obtains a slight edge with
0-0. On 8 iJ..d3 it's time for 8 iJ..d3 or 8 iJ..c4.

Semi-Open Games

If Black retakes with the pawn CARO:PANOV

(7 cxd5 exd5) he has the worst of a ACCELERATED
nearly symmetrical position With 2 c4 White tries to bring
following 8 i.b5+! i.d7 (8 ... ~c6 about a Panov in which Black can't
9 ~e5) 9 'iib3 ~c6 10 0-0 or maintain a pawn on d5, nor can he
9 ..i.xb5 10 'iWxb5+ 'iWd7 11 ~e5. reach the relative safety of the
When White does not play 4 c4 Fischer-Euwe endgame.
after 1 e4 c6 2 d4 d5 3 exd5 cxd5 Instead, Black often accepts a
Black's difficulties chiefly lie in his slightly worse isolated d-pawn
problem piece, the QB. White middlegame, such as after 2.. d5
usually plays 4 i.d3, which 3 exd5 cxd5 4 cxd5! ~f6 5 ~c3
stops ... i.f5 and makes ... i.g4 ~xd5:
problematic. He stands well after
4 ... ~c6 5 c3.

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 Then 4 dxc6 ~xc6 is a sound
the main line. gambit, better known in the Center

Semi-Open Games

Counter (1 e4 d5 2 exd5 liJf6 3 c4 matter because White was assumed

c6 4 dxc6 liJxc6 and then 5 d3 e5 to be better in either.
6 liJc3 i.f5 7 liJf3 i.b4 8 i.e2 e4!).
'Practice shows that White's
If White stops 4 ... cxd5 with chances are somewhat better and
4 'iVa4, Black can offer a different now this defense is rare,' Alexey
pawn, 4.. e6! S dxe6 i.cS! 6 exti+ Suetin said of the Rubinstein in his
'ii;xti or 6 liJf3 liJg4 7 d4 i.xd4 1983 book on the French. Svetozar
8 liJxd4 'iVxd4 9 'iVc2 liJa6 as in Gligoric wrote that 3 liJc3 dxe4
Huebner-Luther, Sauerbrucken 4 liJxe4 liJd7 is best answered by
2002. S liJf3 liJgf6:
More exact is 3 cxdS!.

Then '6 i.g5!' reaches a good

Then 3...liJf6 4 dxc6 liJxc6 is less Burn, he said, since 6 ... c5? walks
sound (5 liJc3 e5 6 liJf3 i.c5 into 7 dxc5 liJxc5? 8 'iVxd8+ and
7 i.b5). 9 liJxf6.

Black can avoid all of this by But today many GMs distrust
6 i.g5 because 6... h6 and 6... i.e7
meeting 2 c4 with 2... eS. He is
7 liJxf6+ liJxf6 8 i.d3 c5 seem to
betting that a 1 e4 player will be
equalize. So White began to avoid
temperamentally unsuited for an
i.g5 in a Rubinstein - and Black
Old Indian (3 liJf3 d6 4 d4 liJbd7).
discovered that the Burn was his
way to trick White into a i.g5
The French (1 e4 e6) variations That is 3 liJc3 liJf6 4 i.gS dxe4
that most often transpose into one S liJxe4 liJbd7 so that 6 liJf3
another are the Rubinstein (2 d4 dS transposes to the last diagram. No
3 liJc3 dxe4 or 3 liJd2 dxe4) and better is 6 liJxf6+ liJxf6 7 liJf3 h6!
the Burn (3 liJc3 liJf6 4 i.gS dxe4). (8 i.d2 c5! or 8 i.h4 i.e7 9 i.d3 c5
That used to be a 'who cares?' 10 'iVe2 'iVa5+!).

Semi-Open Games

To avoid the Burn, White can Slbce2!? and then S... cS 6 c3lbc6
play entirely different lines, such as 7 f4 'iYb6 8 lbo.
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:

This is the same posItIon as

3 lbd2 lbf6 4 e5 lbfd7 5 c3 c5 6 f4
lbc6 7 lbdf3 ~6 once White plays
In this way White dodges other
Black can retake with a bishop
Tarrasch lines. When Vishy Anand
on f6 and obtain an excellent
wanted to reach the diagram in the
Rubinstein, e.g. 6 lbxf6+ ..txf6
2000 FIDE championship finals he
7lbo lbc6! 8 c3 eS. knew that if he chose 3 lbd2, his
Or 6lbolbbd7 and 7 'iWe2lbxe4 opponent, Alexei Shirov, wouldn't
8 ..txe4 cS! and then 9 0-0 cxd4 cooperate by playing 3... lbf6
10 l:.dl lbc5 11 lbxd4 lbxe4 because he preferred other moves.
12 'iWxe4 0-0 (Becerra-Kaminski, So Anand chose 3 lbc3!, reached
internet 2005). the Tarrasch position he wanted and
Black managed to reach a
position that could come about after Quite a different situation arises
3 lbc3 dxe4 4 lbxe4 lbd7 5 lbf3 after 3 lbc3 lbf6 4 eSlbfd7 slbo.
lbgf6 6 ..td3 ..te7 but only if White White is ready to liquidate his
rejects 6lbxf6+!. center to obtain tactical chances,
such as S cS 6 dxcS lbc6 7 ..tf4
FRENCH: STEINITZ ..txcS 8 ..td3 and 8... 0-07 9..txh7+.
The Steinitz Variation has A more open French results, e.g.
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

Semi-Open Games

as in Berg-Akesson, Gothenburg 7 d4, or try for more with 6 dxc3!?,

2006. e.g. 6... c5 7 h4 tbbc6 8 h5 h6 9 i.f4
If White likes this he can reach it 'ikc7 10 'ii'd2 as in Rogers-Tondivar,
while discouraging the Rubinstein Hoogoveen 2006.
and perhaps the Winawer as well.
He begins with 1 e4 e6 2 tbf3!?
As noted in the Introduction
Francophiles regard that with
Black can sidestep many of the rare
contempt and will want more than
Winawer (3 tbc3 i.b4) lines by
2.. d5 3 tbc3 dxe4. And they'll see
meeting 4 e5 with 4...tbe7 5 a3
that Black is a bit worse after 3... d4
i.xc3+ 6 bxc3 c5.
4 ttJe2 c5 5 c3.
The 'something to think about'
So they'll play 3 ...tbf6 and then quality of 4 ... tbe7 often prompts
4 e5 ttJfd7 5 d4 gets White to where responses like 5 tbe2? c5 6 a3
he wants to be. And on the natural i.xc3+ 7 tbxc3?, which favors
3... c5 Black after 7... cxd4 8 'ii'xd4 tbbc6
9 'ikg4 0-0 10 'fib5 d4!, Byrne-
Uhlmann, 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.)
Black can transpose (5 ... c5) into
What should Black do after a normal Winawer. But also
2 ttJf3 d5 3 tbc3 if he doesn't like attractive is 5... b6, since White has
3... ttJf6 ? His best bet may be missed his 'ikg4 opportunity,
3.....tb4. compared with 5 a3 i.xc3+ 6 bxc3
b6 7 'ikg4!.
It's unclear whether White
should transpose to a Winawer, with Chances should be equal after
4 e5 ttJe7 5 a3 i.xc3 6 bxc3 and 5... b6 6 i.d2 i.xc3!? 7 i.xc3 i.a6

Semi-Open Games

8 ..txa6 ttJxa6 9 0-0 c5 (Bokros- 3 ttJc3 i.b4 4 e5 c5 5 a3 i.xc3+

Rustemov, Kircheim 2006). 6 bxc3 ttJe7 7 dxc5) if White had an
The counter-finesse to 4 ... ttJe7 is extra tempo.
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).
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
1938. White's compensation is still But is dxc5 a good extra move?
rated highly after 7.. d4 8 a3 i.aS That depends on positions like
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
A major virtue of 4 ... cS over S a3 i.xc3+ 6 bxc3 ttJe7, the
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

Semi-Open Games

11 i.d2 i.d7 12 'iWh5 h6 13 i.g4 as We've transposed into a version

in Bagheri-Apicella, Chartres 2005. 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.

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!, Kabanov-
The main drawback to 7 i.d3 is Kravstov, Tomsk 2003. In other
words, 3 ...i.e7 can lure White into
7.. b6. Why? Because often when
a gambit he never intended.
Black tries to trade his bad bishop
with ...b6/ ... i.a6, White's best way If White likes the gambit he can
to cross him up is i.b5+ as in 7 4JO try to reach it from other orders. For
b6 8 i.b5+! i.d7 9 i.d3!. But after example, 3... c5 seems to rule it out.
spending a tempo on 7 i.d3 there is But 4 4Jgf3 and then 4 ... 4Jf6 5 e5
little reason to play 7... b6 8 i.b5+ 4Jfd7 6 c3 transposes. If Black
i.d7 9 i.d3. wants to sidestep the gambit he has
other good lines, including 4 ... 4Jc6
and 4 ... cxd4.
In the Tarrasch Variation (3 4Jd2)
Black can use 3i.e7 as a waiting There is also 3... 4Jc6 and 4 4Jgf3
move. After 4 4Jgf3 4Jf6 S eS 4Jfd7 4Jf65 e5 4Jd7 6 i.d3. Then 6... 4Jb4
White has nothing better than 6 e3 is natural and after 7 i.e2 c5 8 c3
eS 7 i.d3 4Je6 8 0-0 'iib6. 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.

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.
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
Both S c3 and S il.d3 reach a of the usual Advance moves such as
main line after S ... cS. However ... 'iib6 and ... lLIc6.
Black has an alternative strategy in
In the third order, 3... cS 4 lLIgf3
S...b6 and 6 ... il.a6, to trade his bad
a6, Black has to deal with S dxcS
il.xcs 6 il.d3, which gives White a
That makes S c3 more accurate nice setup after 7 0-0 and 8 a3/9 h4.
because if Black pursues the B-
So it's up to Black to rate the
trading plan White will be a tempo
worst-case scenarios in all three
ahead - S c3 b6 6 'iig4 (or 6 lLIh3)
orders and pick the best of them.
il.a6 7 il.xa6 - of S i..d3 b6 6 c3
il.a6 7 il.xa6. But how bad is that check? One
After 3.cS the major move order school of thought says a trade of
questions revolve around il.bS+ and light-squared bishops is good for
the timing of exdS. Black has been White and that 3 ... cS 4 exdS exdS
doing well recently when he stops S il.bS+ must be met by S... lLIc6.
the check, e.g. 4 exdS exdS S lLIgf3 But Lev Psakhis insisted S... il.d7
a6. 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
He can also reach this position Black options such as 4 ... lLIf6
via 4 lLIgf3 a6 S exdS exdS and S exdS lLIxdS.

Semi-Open Games

But if White reaches his goal, the

most common reply is 6... i.d6.

Following a Howard Staunton

game in 1841, 6 ...i.d7 was
considered best. But this was shown
A trade of pawns is coming, to be inexact in Nimzovich's
either on c5 or d4. Two lines that celebrated victory over Georg
often transpose are 7 dxe5 i.xe5 Sa1we at Carlsbad 1911, when
8 0-0 and 7 0-0 liJe7 8 dxe5. 7 dxe5! i.xe5 8 0-0 and b2-b4
seriously hampered Black's
Mikhail Botvinnik called the development.
second order inaccurate because
Black can improve on 7 ... liJe7 with Black improves with 6 ... exd4!
7 exd4 i.d7. Then White has no
7 ... cxd4!. Then 8 liJb3 liJe7
9 liJbxd4 0-0 is equal. convenient defense of d4 and the
8 0-0 liJxd4 gambit is regarded as
What has happened is we've not quite sound.
transposed into 7 dxc5 i.xc5 8 0-0
These two basic positions, the
liJe7 9 liJb3 i.d6 10 liJbd4 0-0.
Nirnzovich-Salwe one and the other
But in that order White has more with 6 ... cxd4, influence a lot of3 e5
options such as 10 .l:le 1, 10 i.g5 theory. For example, you may see
and 10 i.xc6+ bxc6 11 'ifd4. The 4 .. .'iVb6, instead of 4 ... liJc6, and
7 0-0 cxd4! order denies him the then 5 liJf3 i.d7.
choice. Black retains the option of
... i.b5, to get rid of his bad B.
FRENCH: ADVANCE White has two strange but possibly
VARIATION good replies, 6 i.d3 i.b5 7 i.e2!?
(7 ... liJc6 8 i.e3) and the devilish
The Advance Variation tabia 7 dxe5 i.xe5 8 b4!?, based on
used to be 3 e5 e5 4 e3 liJe6 5 liJf3 trapping the bishop after 8 ... i.xf2+
'ifb6 6 i.d3: 9 'ite2.

Semi-Open Games

Black might avoid both replies 7 .. :iixb2? with 8 ..te3! 'ifxal

by meeting 6 ..td3 with 6...tiJc6? 9 'ifc2.
But then 7 dxc5! puts us back in Then 9... cxd4 10 cxd4? is bad
Nimzovich-Salwe. because of 10... ..tb4+ and the queen
He can try to repair matters with can never be trapped.
6... cxd4 7 cxd4? tiJc6, reaching the But 10 tiJxd4! gives White a big
Black-friendly position. edge after 10 ... tiJxd4 11 ..txd4..ta3
But thanks to his delay in ... tiJc6 12 ..tb5+ ~d8 13 0-0 or 1O ... il.a3
White can do better with 7 tiJxd4!, 11 tiJb5!.
e.g. 7... tiJe7 8 tiJd2 tiJbc6 9 tiJxc6
Therefore the proper order is
and ife2/tiJf3 with a slight edge.
6... cxd4 7 cxd4 tiJh6, reaching the
Because of 6 ... cxd4! in the last last diagram. White should not play
diagram, 6 ..te2 replaced 6 ..td3. 8 ..txh6 because he is transposing
Black's most common reply is into the poor 10 cxd4 line. Instead,
6 ... cxd4 7 cxd4 tiJge7. He has no he should go into one of the
major problems in lines that run mediocre alternatives such as
8 tiJc3 tiJf5 9 tiJa4 'ifa5+ 10 ..td2 8 tiJc3 or 8 b3 .
..tb4 11 ..tc3 b5! or 8 b3 tiJf5 9..tb2
..tb4+. David Bronstein called
a 'clever waiting move after 3 e5
White's best may be 8 tiJa3 tiJf5 c5 4 c3 tiJc6 5 tiJf3. White must
9 tiJc2 ..tb4+ 10 ~fl. But Black can reveal his intentions with 6 a3,
avoid that if he reaches this: 6 tiJa3 or a bishop move.

Then 8 tiJa3 allows 8.....txa3! The wait pays off after 6 tiJa3?
9 bxa3 tiJf5. But how does Black cxd4 7 cxd4 ..txa3! 8 bxa3 'ifa5+
get here? with advantage.
The problem with 6 ..te2 tiJh6 is Another benefit is that by
7 .iLxh6!. White can punish delaying ... 'iib6 Black ensures that

Semi-Open Games

dxc5 will not be forcing. That can for a knight in 6... c4 lines such
be a problem after 5.. .'iVb6 because as 7 liJbd2 liJa5 8 ..te2 liJe7
6...lbge7? allows 7 dxc5! 'iVxc5 9 0-0 h6 10 :e1 liJc8! and ... liJb6
8 .ltd3, which favors White. with prospects for a queenside
But in the diagram 6 ..te2 liJge7 bind.
7 dxc5 liJg6! is excellent for Black Finally, if White feels that
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 But then 6... cxd4 7 cxd4 'iYb6!
likely his best, Black benefits from tricks him into the bad 5.. .'~'b6 6
his move order because b6 is free i.d3 cxd4! position one more time.

Chapter Five:
Double QP Openings

Most 1 d4 d5 games become But Black had better be ready for

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
This obvious move allowed a
3 lbf3 a6.
minor Swiss master, Hans Fahrni,
Black can try to transpose into to confound elite players a century
the latter with 2... a6, hoping for ago. Fahrni-Spielmann, Barmen
3 c4 dxc4. But White can make ... a6 1905 went 3...lbc6? 4 cxd5 'i'xd5
look like a wasted move. e.g. 3 i.f4 5 lbc3 'i'd8 6 d5 lbb8 7 e4 with
lbf6 4 e3 e6 5 i.d3 c5 6 c3 i.d6 obvious superiority. Even the
7 0-0 lbc6 8 dxc5 i.xc5 9 lbbd2 natural 3...lbf6?! and 3... cxd4?
i.d6 10 i.g3 0-0 11 e4 with a plus, favor White after 4 cxd5!.
Vladimirov-Iuldachev, Mumbai
2003. Of course, there is nothing wrong
But 2lbf3 doesn't commit White with Black's game after 3... dxc4! or
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).

Double QP Openings

If there are so many pitfalls for but landing in another opening, one
Black, why isn't 2 ttJf3 played by he's never played or simply doesn't
every d-pawn player? The answer is like.
White deprives himself of several How can that happen after such
options and gives Black some quiet moves? Well, suppose Black
useful ones: replies 3.. cS. He had better be
White rules out the QGA lines of prepared for a Semi-Tarrasch (4 c4
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
.tf5 3 cxd5 or 3 ttJc3). discovered as Black that 4. it.g7
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
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
Soon after 7 e4! Geller was in
1 ttJf3 d5 2 c4 d4 or 1 d4 ttJf6. That
muddy waters (7 ... ttJb6 8 d5 0-0
was the case when Lajos Portisch,
9 a4) and eventually lost.
an Indian specialist, faced a young
Anatoly Karpov, who began 1 ttJf3. The other major peril for Black
Portisch found himself in a is a switch from the Colle,
relatively unfamiliar Slav Defense characterized by c2-c3, to the
after 1...d5 2 d4! and lost. Colle's sibling, the Zukertort
system, characterized by b2-b3 and
COLLE SYSTEM J..b2. This is a problem because the
best defenses against a Colle often
Black's greatest danger after 1 d4 involve ... ttJbd7 - which can be
dS 2 ttJf3 ttJf6 3 e3 isn't the Colle dubious in the Zukertort.

Double QP Openings

Here's how: After 1 d4 d5 2 ttJf3 vice versa. On 1 d4 d5 2 ttJf3 ttJf6

ttJf6 3 e3 e6 4 iLd3 c5 5 c3 ttJc6 3 e3 e6 4 il.d3 it is inexact to play
6 ttJbd2 White usually answers 4..ttJbd7:
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:

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,
This time 8 dxc5? is met by
and Black can respond 5... ttJc6!
8... ttJxc5!. White's bishop is
and 5... ttJbd7! respectively.
attacked and he can't carry out the
vital e3-e4 break. He's already a bit The crafty response to 4.c5 is
worse. 5 0-0, which was played by Jose
Capablanca and Akiba Rubinstein.
But in the Zukertort, 1 d4 d5 2
They waited for a Black
ttJf3 ttJf6 3 e3 e6 4 iLd3 c5 5 b3,
commitment, 5... ttJbd7 6 b3! or
Black stands well following
5... ttJc6 6 c3!.
5...ttJc6! 6 iLb2 iLe7, e.g. 7 0-0 0-0
8 ttJbd2 'ilc7 9 ttJe5 ttJd7!. But they weren't tested by
5c4!, e.g. 6 iLe2 b5 7 b3 il.b7
The danger to both players is
8 a4 a6 9 axb5 axb5 10 ':xa8
tipping their hand. If Black had
iLxa8 11 bxc4 bxc4 12 c3 ttJbd7
played ... ttJbd7 instead of ... ttJc6, he
with equality, Inkiov-Groszpeter,
would be worse after 9 ttJe5!, e.g.
Copenhagen 1988.
9... b6 10 f4 iLb7 11 c4 ttJe4 12 cxd5
exd5 13 ttJxe4 dxe4 14 iLc4 as in
I. Rabinovich-Makogonov, Soviet OTHER QUEEN PAWN
Championship 1937. GAMES

What this means is Black's QN The Trompowsky Attack (1 d4

should be watching White's QB and ttJf6 2 iLg5) has a poor cousin in

Double QP Openings

1 d4 dS 2 i.gS. The lack of a i.xf6 edge of a symmetrical position after

threat makes Black's life easier, and 4 ..tf4. The easiest way to equalize
books say 2 ... h6 3 i.h4 c6 and against 3 c3 is likely 3.....tf5 4 "iVb3
... 'iib6 is good. 'iic8 .
If Black doesn't like that he ALBIN COUNTER GAMBIT
should beware of c2-c4 traps. One
arises when he avoids doubled The main line of the Albin begins
pawns with 2...ttJd7 and 3... ttJgf6. with 1 d4 dS 2 c4 eS 3 dxeS d4
Then 3 c4!. 4 ttJf3 tbc6. White character-
istically 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 Unlike 5... i.e6 and 5... i.g4, this
can avoid it simply with 1...d5 since makes a threat, 6... ttJb4. That wins
2 ttJf3 ttJf6 3 i.g5 allows him quick time for queenside castling after
equality with 3... ttJe4!. 6 a3 'iid7, under better circum-
stances than usual.
But there is a sneaky waiting
move, 3 c3. Then on 3... e6 4 i.g5 The test of this is order is 7 b4
an unwitting Black has been i.xb4!? (8 axb4 ttJxb4 9 "tWa4
dragged into the Torre. If 3... c6 ttJc2+ 10 ~dl ttJxal 11 'iixal c5).
instead White has the usual slight Declining the sack is harmless,

Double QP Openings

8 'ilVb3 ..ta5 9 e3 l::td8, Spais- 1990s, White did well when he

Katsaris, Athens 1999. reached them via the Exchange
What happens after 7 g3 0-0-0 French or Petroff, so Black has a
8 ..tg2 is virtually unexplored, e.g. real decision at move three.
8... d3 9 e3 ..th3 10 ..txh3 "ifxh3 White makes it somewhat easier
(11 b4 ..txb4!). White avoids this for him when he chooses 3 lbc3
with S g3! and if S...te6 then because after 3... lbf6 he has
6lbbd2. nothing better than 4 e3, heading
towards main QGA lines. But in
ACCEPTED those lines, lbc3 is better delayed,
as we'll see shortly.
After Joseph Blackburne popul-
arized 3 lbf3, following 1 d4 dS 2 Another story entirely is the
c4 dxc4, White's alternatives immediate 3 e4. It was once
almost vanished. But they're considered dubious because of
reappearing, mainly as transpo 3 ... e5, since White cannot maintain
devices. Garry Kasparov played a pawn on d4. But 4 lbO! revived
3 e3, for example, when he wanted this line and poses a move order
to avoid 3 lbf3 c6!? 4 e4 b5. 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

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?

Around 1980 the answer was

4 .....tb4+ because 5 lbbd2 is
Theory used to call such punished by 5... c3! and S i..d2
positions boringly even. But in the i..xd2+ 6 'i'xd2 exd4 peters out to

Double QP Openings

equality after 7 'iVxd4 'iVxd4 or Black can sidestep that with

7 l2Jxd4 'iVe7 8 f3 l2Jf6 9 iLxc4 0-0 4.. e6 5 e4 iLb4. This an invitation
and ...:d8. to a Vienna QGD (6 iLgS! cS!) and
one that should be accepted because
But if Black waits to check,
6 eS 4:JdS 7 iLd2 l2Jb6! makes it
4.. exd4 5 iLxe4 iLb4+, then White
hard for White to regain his pawn.
had a choice between 6 iLd2 and
the more aggressive 6 4:Jbd2!, the Another Black option after
thinking went. 4 l2Jc3 is 4..l2Je6. Then S e4 Jt.g4 is
one of the better lines of the
Within 10 years, theory reversed
Tchigorin QGD. It's as if White
itself and declared 4...iLb4+ to be
played I d4 dS 2 c4 l2Jc6 3 4:Jc3
second-best because White was
dxc4 4 l2Jf3 l2Jf6 S e4 iLg4 rather
getting good results from 5l2Je3!. than 4 dS!?
Today neither order is clearly But if Black played 2 ... dxc4 to
best. It's a matter of which position stay in the QGA, he should consider
Black most wants to avoid, the three alternatives to 3... l2Jf6.
4 ... iLb4+ S l2Jc3! or 4 ... exd4 They not only avoid the gambit but
S iLxc4 iLb4+ 6l2Jbd2!. speed his counterplay.
The first is 3 ...e5. After 4 e3
QGA: 3l2Jf3 Black can safely head towards a
QGA tabia with 4 ... l2Jf6 S iLxc4 e6.
At any tournament you're likely
He also get an extra option,
to see Black rattle off his first half
4. exd4 with the idea of keeping
dozen moves in a QGA as if they
the pawn, S exd4 iLe6!? The key
didn't matter. But his third move is
line is 5 iLxe4 "ike7 (not S... dxe3??
significant. The popular choice,
6 iLxf7+):
3..l2Jf6, allows White to offer a
gambit, 4 l2Je3 and S e4, based on
attacking the knight with e4-eS.

This was considered refuted by

6 'iVb3 because 6... iLe6? is met by

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 But that helps Black. He can
threatens 9 ... lLla5! and gets a fine transpose into Rubinstein's a2-a4
position after 9 'tid! lLlf6, e.g. line of the main QGA, 4...lLlf6 5 e3
10 lLlc3 .t.e7 11 d5 ttJa5! or 11 'tie2 e6 6 .t.xc4 c5, which is currently
0-0 12 l:tdllLlb4 as in Wojtkiewicz- considered too quiet for an edge.
Yermolinsky, San Francisco 2002. Or he can try 5...t.g4! 6 .t.xc4 e6
That's two plusses for 3... c5!? - 7lLlc3, a superior version of the pin
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!.
5 lLlc3 lLlf6 6 e4, which is unclear. But that's not possible in this
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 Persson-
4 'iia4+ with 4 ... b5! - back in the
C. Hansen, Malmo 2003.
1930s when White was getting a
plus from 3 ... lLlf6 4 'tWa4+. If there is a serious drawback to
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
QGA: 3lLlf6
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

Double QP Openings

allows Black to isolate White's d- Another 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.

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
Then 7 'ii'e2 c5 8 dxc5 ~xc5 is thumb that says when the bishop is
harmless and Black can even get the attacked by ... b5, it's almost always
edge after 7 a3?! b5 8 il.a2 il.b7 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, Topalov-
cannot 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:

Double QP Openings

Experience with 8 i.d3 lLlbd7 Now on 6 0-0 a6 7 dxc5 Black's

shows that Black has an easy time, best is 7.. :ifxdl and a draw is the
e.g. 9 'iVe2 i.b7 10 l:dl 'iib6 typical result of 8 l:xdl i.xc5.
11 i.c2 l:d8 12 a4 b4 13lLlbl 'iic7
But 6 'iie2 makes dxc5 more of a
14 i.d3 cxd4 15 exd4 a5, Benitah-
threat, since it wouldn't allow a
Fressinet, Paris 2004.
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
i.b7, not 9 e4!. did, until he can safely threaten
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, Pelletier-
tested 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
Furman's plan. After 1 d4 d5 2 c4 There are also good and
dxc4 3 lLlf3 lLlf6 4 e3 e6 5 i.xc4 c5: bad transpositions when White

Double QP Openings

anticipates ... b5 with a bishop

move, 6 0-0 a6 7 .i.b3.

The reason they often transpose

is that 8 tLle3 exd4 does not win a
pawn. After 9lIdl! Black has been
Now 7...b5 8 a4! corrupts the
trapped in another Steinitz (9 ....i.e7
Black queenside, e.g. 8... b4 9 tLlbd2
10 exd4 0-0 11 d5! exd5 12 tLlxd5
and 8....i.b7 9 axb5 axb5 10 lIxa8
tLlxd5 13 .i.xd5 'iic7 14 'iie4) .
.i.xa8 11 'iie2, 11 tLla3 or 11 tLlc3.
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!. 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! In contrast, 8 ':dl b5 9 .i.b3 c4
8 exd4 tLle6. White doesn't get time 10 .i.c2 tLlb4! equalizes for Black,
for 'i'e2/lIdl because this time e.g. 11 tLlc3 tLlxc2 12 'iixc2 .i.b7
9 'iie2 tLlxd4! is unsound. He has to 13 e4? (13 d5!?) b4 14 e5 bxc3
settle for a tiny edge such as 9 tLlc3 15 exf6 gxf6 as in a celebrated
.i.e7 10 lIel 0-0 11 a4 .i.d7, as in a Szabo-Euwe game won by Black.
Topalov-Kramnik match game in
The final QGA issue concerns DECLINED
6 0-0 a6 7 'i'e2 tLle6. Then 8 tLlc3
and 8 lIdl often transpose, e.g. Players have been finding
8 l:tdl b5 9 .i.b3 tLle6 10 tLle3. finesses in the QGD since Pierre St.

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
Wolf in the 1920s. can continue 3... liJf6 4 liJc3 e6,
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
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 Kramnik-
Deep 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
Botvinnik. 7 .i.d3 liJbd7 8 0-0 dxc4 9 i.xc4

Double QP Openings

~d5 10 ~xe7 tixe7 11 l:tc1?! Black has closed some doors in

~xc3 12 l1xc3 e5. He forgot what the diagram but kept others
has been known since the 1930s, open. The Tartakower-Makogonov-
that 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
line whose theory he'd forgotten. This tabia also helps Black in a
Lasker Defense, 5 i.g5 h6 6 ~h4
~e7 7 e3 ~e4, as Vlf Andersson
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.

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 This tabia sets a trap: 4... cS?! falls
1987). 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
That means Fritz Samisch's favorably evade the Cambridge
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,
Moving on to another tabia: Gabriel-Stangl, Altensteig 1993.
Mihai Suba liked the fluid center
Black gets in a Semi-Tarrasch, 1 d4 But 4... i.b4+ is more exact.
d5 2 c4 e6 3 ttJf3 ttJf6 4 ttJc3 c5 Then 5 ttJbd2 dxc4! is fine for
5 cxd5 ttJxd5!. But in 1977 Tigran Black and so is 5 i.d2 i.e7!. In
Petrosian 'caught me out' with 1934, Hans Mueller recommended
4 i.gS!?, he recalled. Suba had to 4... i.b4+ 5 ttJc3 dxc4 6 e4 c5 - and
handle this unfamiliar position and the Vienna Variation was born.

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
wanted to play the Vienna, why For example, 9... lbdS 10 i.xe7
didn't I give the check at move lbxe7 11 l::tc1 lbg6 12 i.xg6! hxg6
four?' 13 e4 and Black cannot achieve the
needed ... eS break (Capablanca-
White often refuses to be drawn Tylor, Hastings 1930).
into this guessing game and replies
S e3. Then S... cS is an offbeat QGA But whenever White plays lbbd2
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, Rivas-
Botvinnik (S lbc3 dxc4) but also a Yusupov, Minsk 1982.
Cambridge Springs (s ... lbbd7 6 e3
The previous QGD tabias allow
White to transpose into the tried and
White may be tempted by true Exchange Variation with cxdS.
5lbbd2. This one doesn't:

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
An example of that is I d4 dS
2 c4 e6 3 liJc3 liJf6 4 ..igS liJbd7
S e3 c6. Black is advertising his
It can come about via 1 d4 liJf6 interest in the Cambridge Springs.
2 c4 e6 3 liJf3 d5 4 liJc3 dxc4. If White can't remember the theory,
Black invites a QGA (after S e3), a 6 cxdS! makes sense. (A move later
Vienna Variation QGD (after S e4 is too late, 6 liJf3 'tiaS 7 cxdS
or 5 ..igS) or even a Nimzo-Indian liJxdS!.)
(after S ..igS ..ib4 6 e4 cS 6 a3!?). But exchanging on move three is
The deceptive choice is too early. After 1 d4 d5 2 c4 e6
S 'it'a4+!?, trying to reach a superior 3 cxd5? exd5 4liJc3:
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, Black solves his main QGD
London 1993. Better is 5.. c6 problem, how to develop with QB,
6 'it'xc4 b5. with 4... c6! and ....if5!, e.g. S 'tic2
..id6 6liJf3 liJe7 7 ..igS ..ifS.
QGD: EXCHANGE It's also inaccurate to exchange
The first issue in the Exchange at move four if White is already
Variation is when to exchange. committed to liJf3. One of
Today it's commonly done at move Petrosian's most impressive wins as
four. But Alexander Alekhine, Black began with 1 d4 liJf6 2 c4 e6
commenting on 1 d4 dS 2 c4 e6 3 liJf3 dS 4 cxdS?! exdS S liJc3 c6

Double QP Openings

6 .ig5 .ie7 and ... .if5! (7 'iic2 g6 5 e3 0-0 6 cxd5? tDxd5! 7 .ixe7
8 e3 .if5). '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 Alatortsev-
control, he usually equalizes. But if
Capablanca, Moscow 1935).
he's played an early ... tDbd7, then
cxd5! usually gives White enough White also has an easy time after
time to stop him, e.g. 1 d4 d5 2 c4 5 tDf3 0-0 6 'ifc2 c6 7 e3 tbbd7
e6 3 tbc3 tbf6 4 .ig5 tbbd7 5 cxd5 8 cxd5 tDxd5! 9 .ixe7 'iixe7
exd5 6 e3 and .id3 or 'iic2 will (10 .te2 ~e8 11 0-0 ctJxc3 12 bxc3
make ... .if5 impossible. Salo Flohr, e5).
a QGD connoisseur, played cxd5 And bear in mind that a delayed
immediately after ... tbbd7 even if cxd5 may be conflict with White's
that was at move eight. other early moves. Alekhine felt the
'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
pieces after ... tbxd5. The result But after 7 ... exd5 8 .ltd3 c6
often resembles an Orthodox 9 'iVc2 ~e8 10 tDf3 he discovered
Defense in which Black simplifies his QR would have been better
with ... dxc4 and ...tbd5. placed at b I to support the minority
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

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
He can meet the regrouping plan
for him.
with fl-f3 and e3-e4 or g2-g4, e.g.
9 tbe6 10 ..th4 g6 11 0-0-0 tbg7
If White wants to play tbf3 he
12 f3!, e.g. 12 ... ..te6 13 g4! or
should prepare it with 8 'iie2! so
12 ... tbf5 13 ..tfl 'iVa5 14 ~bl ..te6
that 8 ... tbe4? loses a pawn to
15 h3 0-0-0 16 e4! with a fierce
8 ..txe7 and 9 tbxe4.
attack in Bronstein-Medina,
Gothenberg 1955.
By delaying tbf3 White retains
the tbge2 option and can seize But Black has another
control of bl-h7. There is some simplifying option after 8 ..td3 -
further subtlety if Black pursues 8tbh5 9 ..txe7 'fIxe7. This works
... ..tf5 in another way, 1 d4 d5 2 e4 because 10 ..txh7 is not a check but
e6 3 tbe3 tbf6 4 exd5 exd5 5 ..tg5 rather an unsound sack after 10 ... g6.
..te7 6 e3 e6 7 'ile2 tbbd7 and now Also 10 tbf3 and 10 g4 allow
10 ... tbf4. White should play
8 tbf3 tbf8 9 ..td3 tbe6 10 ..th4 g6.
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
axb4 15 axb4 ..tf5! 16 tbe5 ..txd3 We've seen how an early ... tbbd7
17 tbxd3 tbf5 18 ..tg5 tbd7 19 ..tf4 impairs Black's chances for .....tf5!.
tbb6 and ... tbc4, as in a Najdorf But there are benefits to quickly
connecting the knights. Black can
meet ..txf6 with ... tbxf6, which is
Black exploited a premature important when 'fIc2/..td3 is lined
8 tbf3 in that case. The careful up against the h7-pawn.
transposer will prefer 8 ..td3! and It's also significant after 1 d4 d5
then 8..tbfS 9 tbge2. 2 e4 e6 3 tbe3 tbf6 4 exd5 exd5

Double QP Openings

5 lbf3?! c6 6 i.g5 i.e7 7 e3 i.f5! c5!, with good play in Minogina-

and then S i.d3 i.xd3 9 'iVxd3 Belavenetz, Moscow 1990.
White's counter-finesse IS
S lbge2 liIeS 9 0-0:

White was sloppy twice, at move

five and again when he missed
9 i.xf6! i.xf6 10 'iVxd3, which The absence of ... c6 means
misplaces enemy pieces. It would 9... lbrs can be met by 10 b4!.
cost Black time to get his bishop to Usually White must invest a tempo,
the superior b8-h2 diagonal and to such as l:tb 1, to prepare this
find a good square for the knight. thematic push. Here it's tactically
But in the diagram Black's pieces based on 10... i.xb4 11 i.xf6 'iVxf6
are well-coordinated thanks to 12 lbxd5 or 11...gxf6 12 lbxd5!
... lbbd7! and he should have no 'iVxd5 13 'iVa4.
difficulty equalizing, e.g. 10 0-0
0-0 11 l:tabl a5 12 a3 l:teS 13 b4 The final area of move order
axb4 14 axb4 b5!. 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
For example, 1 d4 d5 2 c4 e6 looking for an alternative to i.h4.
3 lbc3 lbf6 4 cxd5 exd5 5 i.g5 For example, 1 d4 d5 2 c4 e6 3lbc3
i.e7 6 e3 0-0 7 i.d3 lbbd7 S'iVc2 lbf6 4 cxd5 exd5 5 i.g5 i.e7 6 e3
l:teS and now 9 0-0-0 lbf8 10 Wb 1 lbbd7 7 i.d3 c6 S 'YWc2 and now
i.e6 11 lbf3 a6 12 lbe5 l:tc8 13 h3 S... h6?

Double QP Openings

Chances would be roughly even ample compensation (8 ... i.d6 9 e4!

after 9 i.h4 lbf8 10 lbge2 lbe6 and as in Torre-P.Nikolic, Leningrad
...lbg5!. Also 9... 0-0 10 lbge2 l:te8 1987).
11 0-0 lbf8 and ...lbe4.
Another rendering is 1 d4 d5
However 9 i.f4! is much more
2 c4 e6 3 lbc3 i.e7 4 lbf3 lbf6
promising after 9... 0-0 10 0-0-0
5 i.g5 0-0 6 e3 lbbd7 7 'i'c2 and
or 10 lbge2 :e8 11 0-0-0 and
7... h6. That provoked the routine
8 i.h4 from such QGD authorities
White should be looking to as Viktor Korchnoi, Boris Spassky
transpose from an Orthodox and Rubinstein.
Defense in which ... h6 has been
played to an Exchange. For But their opponents missed a
instance, 1 d4 d5 2 c4 e6 3 lbc3 good 8 ... c5!, e.g. 9 :dl 'i'a5
lbf6 4 i.g5 lbbd7 5 lbf3 h6. Now 10 cxd5 lbxd5 11 i.xe7 lbxe7
on 6 i.h4 i.e7 7 cxd5 Black has 12 i.e2 lbd5! 13 0-0 lbxc3 14 bxc3
7... lbxd5! with equality. b6 and ... i.a6, Carlsen-Hansen,
But there is an immediate Skanderborg 2005.
6 cxd5!?
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).
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 In the Orthodox Defense, 1 d4 d5
10 i.d3 l:te8 11 i.g3 lbf8 12 h3 2 c4 e6 3 lbc3 lbf6 4 i.g5 i.e7 5 e3
i.e6 13 0-0 lb6d7 14lba4 a5 15 a3, 0-0 6 lbf3 lbbd7, White's finesses
LSokolov-Seirawan, Dutch Team begin with his QR. Its primary
Championship 2002. function, either at c1 or dl, is to
Accepting the sacrifice, 6... hxg5 discourage ... c5, Black's best
7 dxe6 fxe6 8 lbxg5, grants White freeing device.

Double QP Openings

13 .tbS cxd4 14 'ii'xd4liJf6 IS liJeS

liJe8 16 .td7 (Bogolyubov-Mueller,
Zurich 1934).
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
Black equalizes after 7 a3?! c5! .te7 5liJf3 0-0 6 'ii'c2:
and 7 'iib3 c5! (8 cxdS cxd4
9 liJxd4 liJcS! 10 'iic2 liJxdS,
Poluljahov-Tregubov, Krasnodar
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
Now 6... c6?! 7 e3 liJbd7 8 ltd1!
such as 1 d4 d5 2 c4 c6 3 liJf3 e6
achieves White's goal. This order
4 liJc3 liJf6 5 .tg5 liJbd7 6 e3
also opens a 0-0-0 option, e.g. 6 ... h6
i.e7?!, there is no pressing need for
7 .txf6! .txf6 8 0-0-0 and 9 e4 with
good chances.
Rubinstein showed that White
And after 6...liJbd7 he can switch
can use his heavy pieces more
to an Exchange Variation since
effectively with 7 'iic2! 0-08 ltd1!.
7 cxd5 liJxd5? loses a pawn
Then 8... cS transposes into a book
(8 liJxdS exdS 9 .txe7 "ilxe7
position in which Black is a move
10 'iixc7 or 8 ... .txgS 9 liJxc7).
behind and under strong pressure
after 9 cxdS exdS 10 dxcS. But the problem with an early
"ilc2 is that White's rook isn't in
And thanks to the Q+R tandem,
the familiar liquidation, 8... dxc4 place yet so 6... c5! is good.
9 .txc4liJdS 10 .txe7 'iixe7 11 0-0 The situation is similar when
liJxc3 12 "ilxc3, doesn't allow ... eS Black uses the 1 d4 d5 2 c4 e6
and leaves Black with serious 3 liJc3 .te7 order to delay .tgS and
problems after 12 ... cS 13 "ila3 or White voluntarily delays it further,

Double QP Openings

4 tbf3 tbf6 5 'iVc2. Then 5.. 0-0 ... dxe4 attacks it. White can avoid
6 iLg5! h6 allows him to reach the that by delaying either tbf3 or tbc3.
promising 7 iLxf6 iLxf6 8 0-0-0. For example, 1 d4 d5 2 c4 e6 3 tbf3
tbf6 4 iLg5 iLe7 5 e3:

Another natural move, 5tbbd7,

allows an excellent Exchange Now 5...tbe4 has been under a
Variation, 6 cxd5! exd5 7 iLf4! c5 cloud since W.Cohn-Leonhardt,
8 e3 0-0 9 iLd3, since Black's QN is Ostende 1907 went 6 iLxe7 'iVxe7
misplaced. 7 tbbd2! 0-0 8 iLd3 and 8... f5
But once again Black can strike 9 0-0 tbd7 10 'iVb3 c6 11 lIac 1 gave
with ... c5 before White has deterred White an excellent Stonewall
it with l:.dl, e.g. 5... c5!? 6 cxd5 Dutch.
cxd4! 7 tbxd4 exd5 8 iLg5 tbc6 However, 5... 0-0 puts the ball
9 l:tdl h6 10 iLh4. 'iVa5 as in back in his court. He can continue
Rustemov-Ubilava, Olite 2006. 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 c5). traditional orders. But most of them
The trade-seeking ...tbe4 works allow Black to fall back into a good
best when White has one knight at Orthodox or Lasker Defense. If
c3, so ... tbxc3 is possible, and White plays 1 d4 d5 2 c4 e6 3 tbc3
another at on f3 so that tbxe41 tbf6 4 iLg5 tbbd7 5 e3 c6 6 a3:

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
The hard part for Black is
reaching the tabia. The 1983
Kasparov-Smyslov candidates
match tested various orders and
showed that 9e5?! is bad in view
He can get a reasonable version of 10 a3! it.xc3 11 bxc3 'iixa3
of Rubinstein's Orthodox line, 12 e4 with a dangerous attack.
6... it.e7 7 tLln 0-0 8 'iVc2 and ~d1.
That kind of sacrifice is a major
But instead, Black can transpose Black concern. Alekhine said
to a good Lasker with 7...tLle4 or Black's best order to reach the tabia
7... h6 8 it.h4 tLle4. Then a2-a3 was 9..0-0 10 it.d3 e5 and then
plays little role. 11 0-0 h6 12 it.h4 ~e8. But this was
The strength of the main line based on his claim that 11 a3!?
it.xc3 12 bxc3 'ii'xa3 is unsound.
Cambridge Springs, 6 tLln 'iVa5,
rests today on whether White gets If a Cambridge Springs player
an edge from 7 cxd5 tLlxd5 8 'iVd2 has doubts about that, he should
it.b4 9 ~c1. Black has to decide consider the slightly different
when to play ... e5, ... h6 and ... 0-0 or 9 0-0 10 it.d3 h6.
he will be distinctly worse. The
crucial tabia is:

Then 11 it.h4 .l:.e8 12 0-0 e5

arrives at the tabia and 12 a3 it.f8 is
This has performed well for also safe. Black avoids 11 ... e5
Black, e.g. 13 e4 tLlf4 14 it.c4 tLlg6 12 a3!?, which may be a good
15 a3 tLlxh4! was an Alekhine- version of the gambit.

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 'Do you really know as much
11 i.d3 e5 reaches the tabia. But about the Botvinnik as I do?,' Black
White might improve with is asking. 'Or about the Meran?'
11 i.c4!? as in an Alekhine game
He means about 5 i.g5 dxc4 or
that went 11...ttJxc3 12 bxc3 i.a3
5 e3 ttJbd7 6 i.d3 dxc4. If White
13 l:tbl e5 14 i.g3.
doesn't like his answers to
Therefore Black should pick the those questions, he'll also be
order that avoids what he wants to disappointed when he looks for
avoid. If he doesn't mind 9... 0-0 alternatives and sees that 5 cxd5
10 i.d3 h6 11 i.f4!?, that's his best exd5! is an inoffensive Exchange
bet. But ifhe's willing to accept one Variation (6 i.g5 i.e7 and ... i.f5!).
of the gambits, other orders are And even ifhe knew Black's intent,
superior. White would be reluctant to play
cxd5 at move 3 or 4 because he'd
QGD: BOTVINNIK end up in another super-quiet line,
If you look up Mikhail the Exchange Slav.
Botvinnik's variation in books you So what a Botvinnik player is
may find it under 1 d4 d5 2 c4 e6 3 really saying in this order is, 'You
ttJc3 ttJf6 4 ttJf3 c6 5 i.g5 dxc4. don't have much choice. It's the
But the little secret of Botvinnik Botvinnik for you or quick equality
specialists is very few of them use forme.'
that order. It grants White too many When White goes into 5 i.g5
promising alternatives, including dxc4 he justifies the lost pawn with
4 i.g5 and 4 cxd5. 6 e4 b5 7 e5 h6 8 i.h4 g5 9 ttJxg5
It is other orders, such as 1 d4 d5 hxg5 10 i.xg5, with mind-boggling
2 c4 c6 3 ttJf3 ttJf6 4 ttJc3 e6, that complications. One of the less
make the Botvinnik the weapon it analyzed positions occurs when he
IS. retreats instead with 9 i.g3.

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 Persson- Antique theory claimed White is
Hector, Gothenburg 2006, won by much better (11...'i'xc4+ 12 'iitgl
White. 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
the diagram via other orders. For One way is 7 iLxc4 cxd4 8 4:Jxd4
instance after 1 d4 dS 2 c4 e6 3 4:Jc3 'i'a5 after which 9 iLxf6 iLxc3+
c6 4 4:Jf3 4:Jf6 S iLgS a Black who 10 bxc3 'i'xc3+ 11 'iitn gets there.
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
However, 6 iLh4 dxc4 7 e4 is for dark square defense. That's
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
include 9 eS, reaching the last Then 10 ... 4:Jbd7 11 iLxf6 'i'xc3+
diagram. 12 'iitfl gxf6 13 h4! and l:.h3 offers
good chances.
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.

Double QP Openings

The problem with 8 ... 'iVa5, The correct response is S...l2Jc6!.

according to authorities going back Then 6 ..tgS loses its punch to
to the 1930s, is '9 ..td2! and then 6.....te7!.
9... e5 10 l2Jc2.'
Black can also fall into the same
But 9 ..td2 is so passive that trap via the Semi-Tarrasch, 1 d4 d5
9... 0-0 makes White's pieces look 2 c4 e6 3 l2Jc3 l2Jf6 4 l2Jf3 c5, if he
misplaced (10 'iVe2 .:f.d8 11 l2Jf3 plays 5 cxd5 exd5? 6 ..tg5! instead
l2Jc6 12 a3 ..txc3 13 ..txc3 'iVh5, of 5 ... l2Jxd5!.
Summerscale-Wells, Millfield
2000). Instead, White may have to
risk 10 l2Jc2 ..txc3 11 ..txc3 'iVg5! SLAV DEFENSE
12 'iVe2 'iVxg2 13 0-0-0.
The early moves of the Slav
TARRASCHISEMI-TARRASCH conceal quite a bit of subtlety
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 Semi-
6 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 Sl2Jf6?!. Meran, with success going to the
better transposer.

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,
His punishment is 6 ..tgS!. He as in 1 d4 d5 2 c4 c6 3 l2Jf3 l2Jf6
has transposed into 1 d4 d5 2 c4 e6 4 e3 e6 5 ..td3 l2Jbd7 6 0-0.
3 l2Jf3 l2Jf6 4 ..tg5 when Black
plays the sloppy 4 ... c5?! and allows Or suppose 1 d4 dS 2 c4 c6 and
5 cxd5 exd5 6 l2Jc3. now 3 e3:

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.

White can initiate the Exchange
Why would White play such a Variation at move three (as most
quiet move? The reason is that an players do today), at move four (as
early e2-e3 is a good way of Botvinnik preferred) or later. The
avoiding the high-maintenance rationale for delay is that Black may
Open Variation (1 d4 d5 2 c4 c6 prematurely commit his pieces
3 tZJc3 tZJf6 4 tZJf3 dxc4). To get an before he knows that cxd5 is
advantage White has to play 5 a4 coming, e.g. 1 d4 d5 2 c4 c6 3 tZJf3
.if5 6 e3 or 6 tZJe5, or risk the tZJf6 4 ttJc3 .if5? 5 cxd5 cxd5
5 e4!? gambit. Either way he has to 6 ~3! favors White.
memorize a huge amount of book.
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 The same goes for 4 ... ttJbd7
Indian Defense, e.g. 4 tZJc3 .ig7 when 5 cxd5! cxd5 6 .if4 leaves his
5 tZJf3 0-0 6 .id3 d6! 70-0 tZJbd7 8 QN misplaced.
'ilc2 e5 9 l:tdl 'ile7 10 .in e4 11
White has the better pieces after
tZJd2 l:te8.
6... e6 7 e3 .ie7 8 h3 0-0 9 .id3,
There are two other drawbacks to e.g. 9... tZJb6 10 0-0 .id7 11 ttJe5
e2-e3 in any Slav order. One is that l:tc8 12 'ilb3! .ie8 13 l:tfc1 tZJfd7
... .if5! is safer to play when 14 a4!, Reshevsky-Bernstein, U.S.
White's QB is blocked in. Also, Championship 1951.

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 But in practice Black often plays
Caro-Kann (6 exd5 lbxd5 7 lbf3 3 lbc3 dxc4 with no intention of
iLg4 8 'iVb3 iLxf3 9 gxf3 etc.). ... b5. He knows all about the 3lbc3
lbf6 4 e3 trick to avoid the Open
But 6 e5 lbe4 7 iLd3 is much Slav. 'But I'm the one doing the
more promising, e.g. 7 ... lbxc3 tricking, ' he thinks to himself,
8 bxc3 e6 9 lbf3 h6 10 0-0 lba5 'because I'll get there after 3... dxc4
11 .l::tbl iLd7 12 lbd2 and 'i'g4, 4 lbf3 lbf6!.'
Gaprindashvili-Murshed, Polanica
Zdroj 1986. White can avoid the Open
with 4 e4!? But then 4... b5 has
SLAV: 3lbc3
more justification (5 a4 b4 6 lba2
This was once preferred to 3lbf3 lbf6!).
(after 1 d4 d5 2 c4 c6) because it
does a better job of discouraging Instead he often prefers 4 e3 .
... iLf5, as in 3 lbc3 lbf6 4 e3 iLf5?! There follows 4 ... b5 5 a4 b4 6lba2
5 cxd5 cxd5 6 'iVb3. e6 7 iLxc4lbf6 and since he has no
e4-pawn under attack he can
Today when White chooses continue 8lbf3.
3 lbc3 it's usually for a different
reason - to reach the Semi-Slav,
What happened: This often
after 3...lbf6 4 e3 e6 5lbf3, without
comes about after 1 d4 d5 2 c4 c6
allowing the Open Variation, 4 lbc3
3 lbf3 lbf6 4lbc3 dxc4 when White
does not try for advantage with 5 a4
Books say the downside to or 5 e4!? but plays the innocuous
3 lbc3 is 3.. dxc4 because ... b5-b4 5 e3.
will attack the knight.

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
The thematic idea is 6 axbS cxbS
Then S.. bS 6 a4 b4 7 ttJa2 e6
7 b3. But again Black should attack
8 .txc4 reaches the diagram, e.g.
e4, e.g. 7 ... ttJf6! 8 eS ttJe4 9 bxc4
8... ttJbd7 9 0-0 .tb7 10 .td2 as
.tb4+ 10 ttJbd2 ttJc6 11 .td3 ttJcS!,
11 ttJc1 .td6 12 ttJb3 cS 13 'i'e2
with good chances for Black in
0-0 141Hdl 'i'b6. Black has the last
Belyavsky-Kharlov, Ljubljana
laugh. He not only got into the
Open Slav that White tried to avoid
but even reached a quite equal More common after 1 d4 dS 2 c4
version of it. 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 compen-
sation after 6 ... cxb3 7 axb3 e6
8 .td2, e.g. 8... aS 9 .td3 i..b7
10 0-0 .te7 11 .tc3 (Nakamura-
Hansen, Malmo 200S).
There's no major drawback to
4 'i'c2. But 4.. g6 is a better-than-
usual version of the Schlechter Slav
because ... iHS will gain a te~po.

Double QP Openings

If White passes up the various SLAV: THREE PAWNS

evasions of the Open Slav and
When Black uses his first three
continues 4 ttJc3, Black has an
moves to put pawns at d5, c6 and e6
alternative to 4... dxc4 5 a4 it.f5 in
he sacrifices development for
the form of 5... e6!?
transpositions. He can drag a
reluctant White into the Botvinnik,
Noteboom, and other sharp
variations or even a better-than-
usual 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
The natural temptation is to 2 ... c6 is more likely to prompt
punish such a passive move with 3 ttJf3. In this way he won't have to
6 e4. But 6 7 e5 ttJd5 8 it.d2 face the Marshall gambit, 3 ttJc3 e6
b5 9 axb5 transposes to a line of the 4 e4. But another Black may prefer
Slav Gambit that's not at all bad for 2 ... e6 3 ttJc3 because it rules out the
Black (5 e4 b5 6 e5 ttJd5 7 a4 e6 ttJbd2 option in the Semi-Slav.
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 6c5 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 The 3 ttJc3 order deprives White
gets some control of d5 by of 4 it.g5 and discourages 4 cxd5
because 4 ... exd5! and ... .tf5
occupying the hole at b4 with his
knight, e.g. 10 'ii'e2 0-0 II ':'dl
ttJb4. Black seems to stand well , Black risks Marshall's 4 e4 but
e.g. 10 it.g5 0-0 11 ':'e 1 it.d7, can try to bail out with 4 .. it.b4 and
Navarra-Svidler, Turin 2006. then 5 exd5 cxd5 6 ttJf3 into a

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 Gelashvili-
Izoria, 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 This is White's way of saying,
of the Three Pawns Variation is 'I'm guessing you don't want a
illustrated by Black's options after Botvinnik because if you did you
4 tt:JO: would have played 4 ... tt:Jf6 and then
5 i.g5 dxc4. But now you've got a
(a) 4... fS creates a Stonewall. problem if you want to avoid the
Black may feel happy because he Botvinnik. '
won't face the tt:Jh3-f4 and tt:Jbd2
He means S..f6 when 6 i.d2 cS
maneuvers that are dangerous in
7 e3! is a dangerous gambit based
normal Dutch orders. But White
on the weakened e6, e.g, 7... tt:Jc6 8
hasn't ruled out 5 i.f4!, which
i.xc4! cxd4 9 exd4 tt:Jxd4 10 O-O!.
favors him.
For example, 1O ... tt:Jxf3+ 11 'i'xf3
(b) invites a Semi-Slav 'i'xd2 12 l:tadl 'i'xb2 13 tt:Jb5 or
(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
Noteboom is so theory-heavy that
only the brave and the booked up He is hinting that on S tt:JO he'll
will play 4 tt:Jf3 confidently build an excellent Stonewall
unless ... (S .. fS!). Then 6 i.e2 tt:Jgf6 7 0-0
i.d6 is the kind of aggressive
Unless White intends to meet position Black can only dream
4 tt:JO dxc4 with S i..gS!?: about in a Dutch Defense move

Double QP Openings

order, e.g. 8 a3 a5 9 l::tb1 fie7 White has extra options, such as

10 i.d2 0-0 11 b4 ttJe4 12 bxa5 8 .lif3 and ttJge2/a2-a4 in that line,
':f6! 13 i.e1 l1h6 13 g3 g5 with a or 7 .lid3 .lib7 8 ttJf3 a6 9 0-0 c5
fine game, Polak-Stangl, Brunn 10 .lie4 but they aren't very
1993. impressive.

White can counter with the His best in the diagram may be
finesse S i.d3, so that S... fS 6 cxdS! simply to transpose to a Reynolds
exdS? 7 i.xfS hangs a pawn and Variation of the Meran with 7 .lid3
6... cxd5 is an Exchange Slav in .lib7 8 ttJf3 a69 e4! and then 9... c5
which Black's QN is misplaced. 10 d5 c4 11 .lic2 ttJgf6!' But the
As a result, Black usually plays super-sharp Reynolds isn't for
5 ... ttJgf6 instead and 6 ttJf3 everyone.
gets White to the Semi-Slav he Clearly the Three Pawns holds a
wanted. huge number of feints and
Alexey Dreev and Yevgeny sidesteps. The other version occurs
Sveshnikov have met 5 .lid3 with a when White's KN comes out first,
counter-finesse, S... dxc4 6 i.xc4 1 d4 dS 2 c4 c6 3 ttJf3 e6.
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 Jelen-
Meran (10 d5 exd5 11 ttJxd5 ttJgf6 Sveshnikov, Nova Gorica 1998.
12 e4!? ttJxd5 13 exd5 i.d6,
Scherbakov-Sveshnikov, Rostov The major extra option White
1993). enjoys in this order is 4 .ligS.

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, Litinskaya-
Arakhamia, Womens Candidates
1988, is excellent.

Then 4...liJf6 is a QGD. Black Second, he doesn't allow Black's

could prefer 4.. ~e7 5 ~xe7 'ifxe7. best version of Mikhail Tchigorin's
system, 6... ~d6 7 e4 dxc4 (8 ..ixc4
Trading off the good bishop seems
eSt) because 8liJxc4! stops ... eS.
dubious but some might be happy to
face 6 e3 liJf6 7 liJc3 0-0 8 :c 1 Alekhine thought this trick was
liJbd7 9 .te2 dxc4 10 .txc4 eS. so strong that Black had to meet an
early S liJbd2 with S... cS!? That's
SEMI-SLAV DEFENSE why White often inverts the order
today - S ~d3 and if S... liJbd7 then
The Semi-Slav arises via the 6 liJbd2. His reasoning is that
QGD, Slav or Three Pawns orders. Black's QN is misplaced at d7 in
White's first finesse is developing case of a later ... cS.
his QN on d2 instead of c3, as in
1 d4 d5 2 c4 c6 3 liJf3 liJf6 4 e3 e6 Then Black's best may be to fall
and now 5 liJbd2. back on the QGA in which White is
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 anti-
6 ~d3 .td6 7 e4!. Meran order, 5 ~d3liJbd7 6 0-0.

Double QP Openings

This employs psychology: Meran 13 gxh8('i'). His only good

players usually don't want a QGA alternative is a highly complex
position, such as 6 ... dxc4 7 .Jixc4 variation (10 tDa4 c5 11 e5 tDd5)
c5, even if it's their best option. that may require him to sack at least
Instead, they will be attracted to the a pawn to have any hope of
Meran-like 7... b5 8 .Jid3 a6 and advantage.
take their chances after 9 a4.
For that reason many a White
But this trickery backfires if will avoid the issue by playing
Black adopts the Tchigorin plan of 9 0-0. Then 9 ... b4 10 tDe4 gets all
6...Jid6!. Then 7 tDbd2 0-0 8 e4 e5 the attention in books and the
is equal. verdict is mixed. But 9. a6!? is the
trickster's choice.
And on 7 tDc3 0-0 we get a
position that occurs more
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 What has happened is Black
respond 7 ... dxc4 8 .Jixc4 e5? reached a normal Meran (8 ... a6) in
which the second-best 9 0-0 has
9 dxe5 tDxe5 10 tDxe5 .Jixe5 in
been played and met by 9 ... .Jib7.
view of 11 'i'xd8+ and .Jixfl with
advantage. This is a good weapon against a
White who doesn't like the
In the main Meran line, 1 d4 d5
Reynolds Variation (10 e4 c5
2 c4 c6 3 tDf3 tDf6 4 e3 e6 5 tDc3 11 d5). The point is that 10 e4 c5
tDbd7 6 .Jid3 dxc4 7 .Jixc4 b5 11 e5 is like the old main Meran
8 .Jid3, there is a deceptive order line but with 0-0 and ....Jib7 added.
beginning with 8....Jib7 instead of
8... a6. The natural 9 e4, which is the This is a fairly new position and
right move against 8 ... a6, is so far chances appear mixed after
answered by 9 ...b4!. 11...cxd4 12 tDxb5 .JixfJ or
11...tDd5 12 a4 b4 13 tDe4 cxd4
Then White can enter a four-'ii 14 .Jig 5 'i'b8 (Dambacher-
middlegame (!) after 10 e5 bxc3 Stellwagen, Dutch Team Champ-
11 exf6 cxb2 12 fxg7 bxal('ii) ionship 2006).

Chapter Six:

The most popular answer to 1 d4 makes sense for a White who only
today is 1...tbf6 and the most likes the Tromp when he's allowed
popular reply is 2 c4, opening the to double pawns, i.xf6, or to build
door to the Hypermodern and a center with f2-f3, after 2... tbe4.
related defenses we call Indians. We For the Tromp player who has no
tend to overlook 2 tbf3 because it
interest in transposing to another
usually 'just transposes.'
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 Theory claims Black can
but it becomes a significantly equalize with 4.. :iVb6. If that's true,
stronger weapon in the hands of a how can White avoid this and how
flexible thinker who can shift into a can Black obtain it?
Torre Attack or a QGD when
One route is 1 d4 tbf6 2 ..tg5 c5
and then 3 d5 tbe4 4 i.f4. But
For example, after 2 ... e6 White Black has to be concerned about
can back out of the normal 3 i.xf6 and 3 tbc3!? Both are
Trompowsky (3 e3 or 3 e4) and dangerous and have their own body
enter the Torre with 3 tbf3. This of must-memorize book.


More precise is 1 d4liJf6 2 i.g5 5 dxc5!? Then 5.. :iia5+ 6 'i'd2!?

liJe4 and then 3 i.f4 c5 4 d5. 'i'xc5 7liJc3 favors him, e.g. 7... d5
Black's chances of reaching the last 8 0-0-0 i.e6 9 e4 or 7... d6 8 e4 a6
diagram are much greater because 9 0-0-0, Berg-Ilincic, Budapest
the B retreats to h4less often in this 2006.
But Black has a course correction
Moving on, let's suppose in 4'i'a5+! 5 c3liJf6. Then 6 dxc5
4.. :ilVb6 is met by 5 i.el!. (Tromp 'i'xc5 7 e4 e5! is fine for him.
players do things like that, they White's best is 6 d5, after which
really do.) White is hoping to build 6 ... 'i'b6! 7 i.el (not 7 'i'd2liJxd5!)
a center with f2-f3 and e2-e4 that e6 8 c4 exd5 9 cxd5 c4 leaves us
justifies his loss of time. But after back at the diagram.
5 . e6 he either has to play a
One last point about the
doubtful gambit, 6 c4 "ifb4+, or go
Trompowsky: After 1 d4 liJf6
into 6 f3 liJf6. The latter often leads
2 i.g5 liJe4 3 i.f4 d5, rather than
to 7 c4 exd5 8 cxd5 c4:
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 It becomes downright respect-
by means of 1 d4 liJf6 2 i.g5 liJe4 able. Even Garry Kasparov played
3 i.f4 c5 and now 4 f3 liJf6 this (in a simul) - 7... i.g4


8 h3 i.xf3 9 'iixf3 c6 10 0-0-0 e6 3 f4 h6 4 i.xf6 'iVxf6 5 lLlc3 d6,

11 i.c4 lLlbd7? 12 d5! cxd5 White gets his best play from f2-f4,
13 lLlxd5!, with a terrific attack in e.g. 6 'iVd2 c6 7 f4!. This means the
Kasparov-Cameiro, Sao Paolo Torre doesn't transpose well into a
2004. Tromp and that shift should be
reserved for use against a Black
TORRE ATTACK who knows only the 1 d4 lLlf6 2
When Black plays ... d5 in the i.g5 lLle4 and 2 ... c5 lines of the
Tromp or Torre, White usually has Tromp.
the option of shifting to a QGD with
To minimize the danger of ... h6
c2-c4, and that's often good
White can try the delayed Torre,
psychology against a 1...lLlf6
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.
Petrosian won one of his greatest White waits for Black to commit
games. himself to a system of development
(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 gave White an aggressive setup in
Tromp version, 1 d4lLlf6 2 i.g5 e6 Timman-Tiviakov, Moscow 1993.


BENKO GAMBIT can ignore 6 ... c5. Instead of 7 d5

The ... b5 sacrifice in a Benoni b5!? White should castle since
7 0-0 cxd4?! lands Black in an
pawn formation existed long before
inferior version of the Maroczy
the Benko Gambit, and there are
some players who offer it only after
starting with a King's Indian Another idea is to play ... a6
Defense or Benoni. In this way they before ... c5, as in 1 d4 l2Jf6 2 c4
benefit by knowing White's setup a6!? and 3 l2Jc3 cS. Then 4 dS bS!
before risking a pawn. For example, S cxbS g6 is a book Benko.
1 d4 l2Jf6 2 c4 g6 3 l2Jc3 JL.g7 4 e4
d6 S h3 0-0 6 l2Jf3 and 6... cS 7 dS. 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

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, Gheorghiu-
Nemet, 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
The English that arises after
11 ... l2Jb4.
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.


However, there's always a dxc5 to have a bishop or knight on a6.

problem when ... a6 is played before That's important when, for
... c5. White gets the upper hand example, White fianchettos his QB,
after 1 d4 tiJf6 2 c4 a6 3 tiJc3 c5 5 ....ltxa6 6 g3 d6 7 .ltg2 g6 8 b3!.
4 dxc5! and then 4 ... 'i!Va5 5 a3 'i!Vxc5 He has good chances of making the
6 e4 d6 7 .lte3 or 4 ... e6 5 e4 .ltxc5 extra pawn count, e.g. 8 ... .ltg7
6 e5. 9 .ltb2 lLlbd7 10 lLlh3 0-0 11 0-0
tiJb6 12 a4.
A slightly different version is
1 d4 tiJf6 2 tiJf3 a6, used by Miles But after 5 ... g6! White can't play
and Lev Alburt. Then 3 c4 c5 4 d5 6 g3 d6 7 .ltg2 .ltg7 8 b3?? lLlfd7.
b5 is a gambit-less Benko. This is Fedorowicz pointed out that even
mainly a psychological weapon, if White puts his queenside house in
designed to upset a White who order first, 6 b3 .ltg7 7 .ltb2 0-0
played 2 tiJf3 specifically to avoid 8 g3, Black can use the flexibility of
the Benko. his fifth move with 8 lLlxa6!
9 .ltg2 .ltb7 and then 10 lLlh3 e6! to
BENKO PROPER undermine d5. Then 11 tiJf4 tiJxd5
'Benko doesn't know how to 12 .ltxg7 cJi;xg7 13 tiJxd5 .ltxd5
play the Benko,' said another expert 14 .ltxd5 exd5 15 'i!Vxd5?? "iff6
costs material.
on the opening, John Fedorowicz.
He was referring to Pal Benko's The trickster's answer to the
reliance on the original move order Benko begins with 4 tiJf3. He
(1 d4 tiJf6 2 c4 c5 3 d5 b5 4 cxb5 knows that a devoted Benko-ite
a6 5 bxa6 and now 5....ltxa6) long doesn't want the queenside tension
after the merits of 5 .. g6! were to be resolved by 4 ... bxc4 or 4 ... h4.
recognized. 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
The point of 4 tiJf3 is revealed
when White accepts the gambit on
Black can retake the pawn later, the next move, 4 . g6 5 cxb5! a6
when he knows whether it's better and 6 b6!?


12 ttJxe5 ltxb7 13 \i'a4+! ttJbd7

14 ttJc6 or 13 ... ttJfd7 14 ttJc6 'ifb6
15 i.f4!, Milov-De Vreugt, Ohrid
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
Then 6...'iWxb6 7 ttJc3 d6 8 e4
Black found a good antidote in an
i.g7 is promising after 9 i.e2 i.g4
attack on the d-pawn, 5... axb5
10 e5! or 9... 0-0 10 ttJd2 ttJbd7
6 i.xb5 \i'a5+ 7 ttJc3 i.b7!.
11 ttJc4 'iic7 12 0-0 ttJb6 13 ltbl,
Burmakin-Ostertag, Bad Woris- But in the 4 ttJf3 g6 5 cxb5 a6
hofen 2006. 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.

Another benefit is that Black Black gets little out of 6... axb5
often replies 6... a5?!. He does this 7 i.xb5 \i'a5+ 8 ttJc3 i.b7?
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
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 ?


This transposes to 1 d4 liJf6 2 liJf3 And then 3 c4 exd5 4 cxd5 liJf6.

a6 3 c4 cS 4 dS bS. Then on 5 e3 The benefit of 1 d4 cS is this time
two thematic ideas of the Benko, 2 liJf3 allows Black to seize center
S... b4 and S... dxc4 make little sense ground with 2 ... cxd4 3 liJxd4 dS!,
because he's played ... a6. as Alexander Alekhine showed.

For example, S... bxc4 6 liJc3 d6 But this order has a major minus
7 e4 g6 8 .ltxc4. Black has an extra since White hasn't committed
. tempo compared with 4 liJf3 bxc4 himself to c2-c4. He can play a
S liJc3 d6 6 e4 g6 7 .ltxc4. But the more open game with 3 liJc3. Then
extra move is ... a6, which is more 3...liJf64 e4 exd5? 5 e5! favors him
harmful than useful. considerably and 4 ... d6 S .ltbS+ and
6 dxe6 less so.

THE BENONI FAMILY 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.

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


tiJd7 13 'iVd2 :b8 14 a4 a6 15 :abl such as 1 d4 tiJf6 2 tiJo c5 3 d5 d6

f5, Skembris-Hamdouchi, Bastia 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 Black often continues 3... g6
an answer to 2 e4. 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.


KID/BENONI That leaves 9 ttJd2, which

defends e4 and can reach a Modem
This hybrid is similar to the
Benoni tabia (9 ... exd5 10 cxd5).
Schmid Benoni except that White is
Black may feel happy that in this
committed to c2-c4. This is usually
way he avoids the ~c2 line. That is,
due to the order 1 d4 ttJf6 2 c4 c5
3 d5 d6. The finesses generally 1 d4 ttJf6 2 c4 c5 3 d5 e6 4 ttJc3
concern what Black does with his exd5 5 cxd5 d6 6 e4 g6 7 ttJf3 ..tg7
8 ..te2 0-0 9 0-0 lte8 and now
e-pawn. For example, 4 ttJc3 g6
10 ~c2 rather than 10 ttJd2.
S e4 ..tg7 and 6 ttJf3 0-0 7 ..te2 e6
80-0. But Black can delay further,
Now 8... exd5 9 cxd5 transposes 9 ttJd2 ttJa6!. He can spend the next
into a very familiar Modem Benoni few moves preparing ... b5 with
... ttJc7, ... a6, ... l1b8 while White
position that is rated somewhere
doesn't have equally useful moves.
between equal and plus-over-equal.
But Black can delay further with Once again 10 dxe6? is dubious.
8...:e8!? 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, Indon-
esia 1983) favors Black.
The delayed Modem Benoni is
still largely unexplored and the
indications are that it's worth
This is more than a waiting move
because he is preparing 9... exd5
10 cxd5? ttJxe4. On quiet moves The Modem Benoni comes about
such as 9 h3, Black gets a fine game from two main orders, 1 d4 ttJf6
with 9... exdS 9 exdS ..tfS. 2 c4 c5 3 d5 e6 followed by
... exd5/... d6 and ... g6, and 2 ... e6 and
And if White tries to exploit the
3... c5 (4 d5 exd5 5 cxd5 d6).
delay in ... exd5, by means of9 dxe6
..txe6 10 ..tf4, Black has adequate In both cases White can shift into
compensation for the lost pawn the English Opening with ttJf3
after 10...ttJc6! 11 ..txd6 ttJd4. Or instead of pushing d4-d5. A virtue
so experience indicates. of 1 d4 ttJf6 2 c4 cS is that it grants


Black the extra option of meeting i.a6 15 'iVb3! exf4 16 i.xf4 i.xc3
3 liJf3 with 3... cxd4 4 liJxd5 e5!?, 17 'i!Vxc3 'i!Vxe2 18 ltae1 with a
one of the most enterprising anti- powerful attack.
English lines.
Objectively, a good solution to
To avoid that White may prefer White's order is 6... d5!, a better-
3 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 Nimzo-
threat (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!,
drifts toward another poor Maroczy If White tries 1 d4 liJf6 2 c4 e6
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 Bisguier-
Kavalek, Tel Aviv 1964 after 11...f5
12 exf5 gxf5 13 f4 ~e7 14 i.e3


This is a tacit invitation to a 7... iLd7 8 'Yib3 is nice for him and
Semi-Tarrasch QGD (4 ... d5). A true 7 ... tiJbd7 8 tiJfd2! imperils the
Benoni player will prefer Mihai g4-bishop, e.g. 8... a6 9 h3 b5
Suba's 4... a6 5 tiJc3 "iic7 followed 10 "iic2 iLh5 11 a4! b4 12 tiJc4 and
by ... d6/ ... tiJbd7 and eventually iLf4. Or 11...iLg6 12 axb5 tiJxd5
... g6/ ... iLg7. 13 bxa6 and White won in
All of those moves are useful to Dokhoian-Lukov, Plovdiv 1988.
Black in a Benoni. That means he'll Of course, Black has an easier
be in great shape when White time engineering ... b5 in this order.
recognizes that his only chance for However, that push doesn't solve
an edge lies in d4-d5, e.g. 6 iLe2 d6 his problems in 6.. g6 7 h3 iLg7
70-0 tiJbd7 8 d5 exd5 9 cxd5 b5. 8 iLd3 0-0 9 0-0, e.g. 9 ... b5 10 :el!
a6 11 a4 b4 12 tiJbd2 or 10 .. J~e8
Or 8 a4 g6 9 d5 exd5 10 cxd5
11 iLxb5 :'xe4 12 tiJc3 :'xe1+
iLg7 11 e4 0-0 (Tukmakov-Suba,
13 "iixe 1 a6 14 iLa4 and iLf4, as in
Erevan 1980) - when Black has
I.Sokolov - Papaioannou, Turin
actually gained a move on a normal
There's a simple counter to a
White can try to exploit the
delayed tiJc3. After 1 d4 tiJf6 2 c4
Nirnzo-threat order, 1 d4 tiJf6 2 c4
e6 3 tiJf3 c5 4 d5 Black should play
e6 3 tiJf3 c5 4 d5 exd5 5 cxd5 d6,
4... d6! rather than 4 ... exd5.
with 6 e4!?

This is more than a one-move White can't play 5 e4? tiJxe4 this
trap (6 ...tiJxe4?? 7 'iVa4+). White is time and 5 tiJc3 allows Black to
holding his QN back until he knows transpose, with 5... exd5 6 cxd5, to
whether it belongs on c3, a3 or d2. normal lines.
He also benefits from tactics This also gives White reason to
such as 6...iLg4 7 'iVa4+!. Then think about 6 tiJxd5 tiJxd5 7 'iVxd5.


But the backward d-pawn is not the 8 tiJd2) in which Black has been
weakling it seems after 7..tiJc6, e.g. finessed into a ... tiJbd7 line, rather
8 e4 i.e6 9 "ii'dl i.e7 or 8 tiJg5 than one of the more popular ... tiJa6
'ike7 8 'iVe4 h6 as in the forgotten or ... i.g4 variations. There's no
game Bandal-Fischer, Meralco downside to White's order - or at
1967. least we haven't discovered it yet.
In the Modem Benoni's main
OlD-in to-KID
line, 1 d4 tiJf6 2 c4 cS 3 dS e6
4 tiJc3 exdS S cxdS d6 6 tiJf3 g6, Some people actually like the
the move 7 tiJd2 was considered a Old Indian. After 1 d4 tiJf6 2 c4
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
10 i.f4!. White from sharper lines, such as
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
tiJb6 is a good response. Some
annotators gave '7 tiJd2?!' and This finesse is a century old.
'7 ... tiJbd7!' when those moves Janowsky-Tchigorin, Ostende 1907
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:

There followed S..i.e7 6 i.d3

and Black soon got a bad game
White can play 8 e4!. Then he's (6 ... c6 7 0-0 'ikc7 8 i.d2 0-0 9 l:tc1
reached the Classical Variation l:te8 10 tiJg5 tiJf8?! 11 'iVc2 h6
(usually occurring via 7 e4 i.g7 12 tiJge4 i.e6 13 d5! i.c8 14 f4!).


But when the diagram arose in For instance, 7... exd4 8 tiJxd4
another Ostende 1907 game, Black 0-0 9 'iYd2 and O-O-O! or 7... h6
(Aron Nimzovich) deviated with 8 .te3 0-0 9 'i'id2 'it>h7 10 d5! a5
5... g6!. He tricked White into one of 11 g4! tiJg8 12 h4 tiJc5? 13 h5 g5
the most innocuous of KID 14 tiJg3 with a big positional
variations and equalized after plus (Kohlmeyer-Lymar, Balaguer
6 .te2 .tg7 7 0-0 0-0 8 b3 :e8. 2006).
White can insist on playing a A sly move like 5 tiJge2 deserves
Samisch setup, after say 4 e4 e5 a sly reply, 5... c6. Then 6 f3 .te7!
5 d5 tiJc5 6 fJ. But the point of this transposes to a pretty good Old
order is revealed when Black Indian, since 7 .te3 0-0 8 'iYd2
changes his mind about the KID. allows 8... d5! (9 cxd5 cxd5 10 exd5
After 6 a5 7 .te3 .te7! was played exd4 11 .txd4 tiJb6). Or 8 tiJg3 d5!
in the 1954 USSR Championship, 9 cxd5 cxd5 10 exd5? .tb4 with an
the tournament book said this was edge.
'an openmg accuracy even
The traditional counter to 1 d4
M.l. Tchigorin knew. '
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 better- Then if White continues in OlD
than-usual Samisch. mode (5 .tg5 .tg7 6 e3) he finds


himself in an offbeat KID, 6. h6

7 i.h4 g5 S i.g3 ttJh5, that is fairly

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 counter-
OlD 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.

Black has to be careful about

S i.gS because of the temptation in
After the traditional KID order, KIDs to kick the bishop as soon as
1 d4 ttJf6 2 c4 g6 3 ttJc3 i.g7 4 e4, it lands on gS. If he had played
Black is not required to play 4... d6. 4 ... d6 instead of castling, S... h6!
With 4... 0-0 he tries to lure White (6 i.e3 ttJg4) is recommended
into overextending himself, 5 e5 before White can stop it with
ttJeS. The pawn push hasn't 6 'iWd2.
recovered from Letelier-Fischer,
However on 4... 0-0 5 i.g5 h6?!
Leipzig 1960 (6 f4 d6 7 i.e3 cS
6 i.e3 Black lacks 6... ttJg4. The
8 dxcS ttJc6!).
result is an inferior Samisch, e.g.
Aside from provocation, 4... 0-0 6... d6 7 D! eS 8 dS c6 9 'iWd2 cxdS
serves Black by creating a gambit, 10 cxdS ttJe8 11 0-0-0 ~h7 12 g4!,
5 f3 c5!? and then 6 dxc5 b6. Ushenina-Siti, Kochin 2004.


On the opposite side of the coin, 11 0-0, e.g.ll ... a5 12 'ilVc2 "ikc7
Black can delay rather than 13 l'itadl tiJc5 14 %:lfel tiJfd7 15 f4
accelerate castling. This gives him tiJe6 16 "ikd2 tiJxd4 17 .i.xd4,
an extra tempo for development and Rogozenko - Dragomirescu, Timis-
denies White a kingside target, e.g. oara 2006.
1 d4 tiJf6 2 c4 g6 3 tiJc3 .i.g7 4 e4
d6 S f3 tiJc6!? and 6 .i.e3 a6 What has happened is White
7 tiJge2 %:lbS. 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
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.
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 After 1 d4 tiJf6 2 c4 g6 3 tiJc3
13 .i.xd4 %:lxb2. After 9 tiJb3 Black .i.g7 4 e4 d6 S f3 0-0 Black usually
can transpose into normal lines with chooses between breaking in the
9... exd4 10 tiJxd4 0-0. center with ... c5 or with ... e5. The
first seemed harder to achieve until
White can also profit from
it was discovered in the 1980s that
delayed castling. In the Fianchetto
6 .i.e3 c5 is a sound gambit. That
Variation, 1 d4 tiJf6 2 c4 g6 3 g3
told White he should consider other
i.g7 4 .i.g2 0-0 sllJf3 d6 he can try
sixth moves, such as 6 tiJge2.
Viktor Korchnoi's favorite 6 tiJc3
(instead of 6 0-0). It's a good waiting move since
there are only a few Samisch lines
Then 6...tiJbd7 7 e4 eS S h3 in which White develops the knight
offers to transpose into the elsewhere but there are several in
conventional ... c6 lines, such as which .i.g5 serves better than .i.e3,
8... c6 9 .i.e3 exd4 10 tiJxd4 %:le8 e.g. 6 tiJge2 eS?! 7.i.gS!.


On 7... c6 8 'iVd2 tiJbd7 9 dS! Now 7 ..td3 can be answered by

Black cannot easily play the 7 ... cS!. Then 8 dxc5 dxc5 9 ..txc5
thematic ... ~e8/... f5, e.g. 9 ... cxd5 ~c6 is an improved version of the
10 ~xd5 or 9 ... ~b6 10 ~g3 'iVc7 6 ..te3 c5 gambit and 8 d5 e6! is a
11 is awful (Dreev-Mork, good Benoni.
Linares 1999). Black's main idea is that 7 'iVd2
can be met by 7 ... c6!. ECO
Well, then, how good is 6 ~ge2 considered 8 0-0-0 b5 and 8 ..th6 b5
cS 7 dS ? Black can seek a double- to be equal - and among the few
edged Modern Benoni, 7 ... e6 8..tg5 Byrne Variation lines that don't
exd5 9 cxd5. favor White.
A similar idea was used by Tal in
However, 8 ~g3 retains more the 1961 world championship when
options for White, since on 8 ... exd5 he was trailing badly. He wanted to
9 cxd5 his bishop may belong on f4 play an ... e5/... c6 line but was afraid
(e.g. 9 ... ~a6 10 ..te2 ~c7 of 5 ... 0-0 6 ..te3 e5 7 dxe5! dxe5
11 0-0..td7 12 a4 a6 13 ..tf4! ~fe8 8 'iVxd8!, when Black has the worst
14 'ilVd2, with advantage in of a likely draw.
Khismatullin- N echepurenko, Tal refined the order with 6... c6,
Serpukhov 2004). 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
that via S f3 0-0 6 ..te3 a6. S... eS 6 dS tiJa6.


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 The younger - and wiser -

h5 and 8... h4. But 7 f3! and sibling of 6... ttJbd7 is 6...ttJa6! and
8 .tg5 reaches a better-than-usual 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, Smed-
erevska Palanka 1999).


The most common Classical One database shows only eight

Variation order is 1 d4 ttJf6 2 c4 g6 examples of this played before
3 ttJc3 .tg7 4 e4 d6 5 ttJf3 0-0 1988. In most cases Black will be
6 .te2. Then 6... e5 allows White to able to transpose into 6 ... e5 7 0-0
force a riskless, slightly favorable ttJbd7 positions that are quite good
endgame, 7 dxe5 dxe5 8 'iixd8 for him, e.g. 8 l:te 1 exd4 9 ttJxd4
l:.xd8 9 .tg5. 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
order also denies White the The benefits of having a knight at
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
(9 .tg5 h6! 10 .th4? g5). the e-pawn and because the


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 That was the case in Langeweg-
White to transpose into normal Vasiukov, Wijk aan Zee 1973:
... tbbd7 or ... tba6 lines - having 6... tbbd7 7 e4 e5 8 a4 exd4 9 tbxd4
dodged the ... tbc6 mega-theory. l:te8 10 a5! tbc5 11 l:tel tbg4?!


12 h3 4JeS 13 f44Jed7 14 .lte3 4Jf8 comes about via 1 d4 4Jf6 2 c4 g6

IS 'ti'd2 a6 16 b4! 4Jcd7 17 4Jb3 3 g3 .ltg7 4 iLg2 0-0 5 4Jf3 d6
with a clear edge. 6 0-0 4Jbd7 7 4Jc3 eS 8 e4 c6
9 h3 'it'aS. The major difference is
But Black can hide his intentions
that Black has avoided a heavily
with 5... c6!? Then on 6 4Jc3 d5! he
analyzed alternative in the second
should equalize (7 b3 .ltfS 8 iLb2
order, 9 .lte3!? 4Jg4 10 .ltgS.
4Je4!). The ... dS idea is naturally
stronger when Black hasn't spent a
This order also enables Black,
tempo on S... d6.
after 7 4Jc3 'ii'aS, to attack the
If White has advanced his c- c-pawn (8 dS 'ii'b4). In recent years
pawn and not the d-pawn, as in 1 c4 8 h3 'ii'a6 has had some vogue. Joel
4Jf62 4Jf3 g6 3 g3 .ltg7 4 .ltg2 0-0 Benjamin proposed a more accurate
5 0-0, then 5.. c6 is what Tal called move order beginning with
'a slight psychological diversion.' 7...'iib6.
Black gets a good version of the
Neo-Gruenfe1d with 6 4Jc3 dS!, This transposes after 8 d5 'iVb4 or
threatening ... d4, and then 7 cxdS 8 h3 'it'a6. But 8 e4, which is a
cxdS 8 d4 4Je4!. central line in the Kavalek order, is
easier for Black to meet here with
White can try for a reversed 8..iLg4!. Then 9 h3 .ltxf3 10 .ltxf3
version of that with 6 d4 so that his 4Jfd7 11 .lte3, which is played
knight is the one in the center after against the Kavalek, would hang
6 ... dS 7 cxdS cxdS 8 4JeS!. But then the b-pawn. And 9 b3 cS! equalizes,
Black can go back to the KID with e.g. 10 dxcS dxcS 11 .lte3 4Jc6
6d6, transposing into a position 12 h3 :ad8, Tregubov-Izoria,
pioneered by Lubos Kavalek. 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

But others wait for 4Jc3, such as

1 d4 4Jf6 2 c4 g6 3 g3 iLg7 4 .ltg2
Then 7 4Jc3 'it'a5 and 8 h3 e5 0-0 5 4Jf3 d6 6 4Jc3 and now 6.. c5
9 e4 4Jbd7 is a main line that also 7 d5 e6!.


White can try to punish Black's Here 7 ... b5!? 8 cxb5 a6

last move with 8 dxe6 because his transposes into the Fianchetto
pawns are loose after 8 ... fxe6 and Variation of the Benko. That's a
because 8 ... .1i.xe6 is a sacrifice solid line for White but no threat to
(9 l2Jg5 .1i.xc4 10 iLxb7). refute the Benko.

But after Korchnoi and Rafael KID: .1i.g5

Vaganian lost celebrated games to
the sacrifice, theory said White In Yuri Averbakh's variation,
should avoid this and merely 1 d4 l2Jf6 2 c4 g6 3 l2Jc3 .1i.g7 4 e4
transpose to a quiet Benoni (8 0-0 d6 5 .1i.e2 0-0 6 .1i.g5, Black cannot
exd5 9 cxd5). play 6 ... e5? 7 dxe5 dxe5 8 'i'xd8
and 9 l2Jd5. His main choice is
However, if White has played between 6 ... c5 7 d5 h6 8 .1i.e3 e6
6 0-0 instead of 6 l2Jc3, then 6 ... c5 and 6 ... h6 7 .1i.e3 c5 8 d5 e6, which
7 d5 e6 8 dxe6! is strong because transpose.
8... iLxe6 9 l2Jg5 iLxc4 10 .1i.xb7
Both orders have merit. Svetozar
l2Jbd7 is rudely surprised by
Gligoric preferred 6 ... h6 because
11 l2Ja3!. This gave Korchnoi two
after 7 .1i.h4 c5 8 d5 Black has a
easy victories when he introduced it
strong alternative in 8... a6 9 a4
(11...!:.b8 12 l2Jxc4 llxb7 13 l2Jxd6
'i'a5, since .1i.d2 is not possible. On
with a clear extra pawn). This gives 7 .1i.e3 e5 (or 7 ... l2Jbd7 8l2JO e5) he
White a major reason for delaying
reaches a Classical Variation
position in which the extra tempo,
The downside is allowing Black ... h6, is more asset than liability.
a good version of the Benko Other authorities prefer 6 ... c5
Gambit, e.g. 1 d4 l2Jf6 2 c4 g6 3 g3 7 d5 h6. But to play it Black needs
iLg7 4 iLg2 0-0 5 l2Jf3 d6 6 0-0 c5 to know a huge amount of theory
7 d5: about the pawn sacks that arise after


8 i.f4 e6 9 dxe6. In contrast, 6... h6 KID: FOUR PAWNS

7 i.f4 can be met by 7...t2Jc6 (8 12Jf3
In the Four Pawns variation (1 d4
12Jxd4! and 8 d5 e5!). That makes
12Jf6 2 c4 g6 3 12Jc3 i.g7 4 e4 d6
6... h6 the mental hygiene order.
S f4) Black usually challenges the
Another i.g5 system is 5 h3 0-0 center with ... c5 and White
6 i.gS. 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:

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
Black can get dynamic play with
10 d5 12Ja6 11 g4 with advantage,
8 .. l2Jg4!, e.g. 9 12Jxc6 bxc6
Paunovic-Illic, Belgrade 2004.
10 i.g1 e5! or 9 i.g1 12Jxd4
Theory used to regard 6 cS as 10 i.xd4 e5.
good because 7 dS e6 8 i.d3 exdS
9 exdS is a somewhat lifeless This is better than 8...0-0 9 i.e2!
Benoni. But 9 cxdS! transposes into which stops ... l2Jg4. White gets a
a good one, e.g. 9... l:.e8 10 12Jf3 a6 somewhat favorable middlegame or
11 a4 'iYa5 12 0-0 12Jbd7 13 :e1 a good endgame like 9 ... e5 10 12Jxc6
'iYb4 14 'iYc2 c4 15 i.fl, Bareev- bxc6 11 fxe5 dxe5 12 i.c5 and
Predescu, Kitchener 2006. 'iYxd8 and the related 9 ... l2Jxd4
10 i.xd4 e5 11 fxe5 dxe5 12 i.c5.
Note that since ... h6 and i.h4
haven't been played, all lines with The reason this is significant is
... 'iYa5 can favorably be met by that if Black plays 5... 0-0 instead of
i.d2, e.g. 6... c5 7 d5 a6 8 a4 'iYa5 5... c5, White can safely play 6 12Jf3
9 i.d2! e6 10 i.d3 or 7... b5 8 cxb5 c5 7 i.e2!? cxd4 8 12Jxd4, forcing
a6 9 a4 'iYa5 10 i.d2 with his way into the good line of the last
advantage. paragraph.


Therefore 5 ... c5 is more precise. But 8 ttJrJ! transposes into a

White can avoid it with a counter- favorable line of the related
finesse, S Jie2!. Then if Black ttJf31h2-h3 variation. For instance,
assumes he's headed for a Classical 8 ... cxd5 9 cxd5 ttJa6 8 Jie2 ttJc5
Variation, 5 ... 0-0 6 ttJf3, he'll be 9 ttJd2 and 9 ... ttJe8 10 ttJc4 a6?!
surprised by 6 f4 c5 7 ttJf3, slipping 11 b4 ttJd7 12 Qd2 as in
into the 5 f4 0-0 6 ttJf3 c5 7 Jie2 Aleksandrov-Zhelnin, Krasnodar
line. 1997.

There is one further finesse. But if White tries to reach this

Black can meet 5 Jie2 with S... cS, through the ttJf31h3 order, S ttJrJ
so that 6 ttJf3 cxd4 7 ttJxd4 ttJc6 0-0 6 h3, he can be stopped after
6.. eS 7 dS, when Black favorably
thwarts White.
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.
And now 8... exdS 9 exd5 ::'e8+
KID: OTHER VARIATIONS 10 Jie3 Jih6 or 10 ... ttJh5 are
considered unclear. But 9 cxdS!
An obscure variation that reaches a Benoni that many
deserves more attention, if only authorities say favors White. He is
because of its transpo possibilities, being rewarded for misdirection: if
is 1 d4 ttJf6 2 c4 g6 3 ttJc3 Jig7 White tries to get to the 9 cxd5
4 e4 d6 S h3 0-0 6 Jie3. Some KID position through a Benoni order
experts say 6... eS is best and give there are various ways for Black to
7 dS c6. Then 8 Jid3 b5!? leads to thwart him, such as I d4 ttJf6 2 c4
equal chances, they say. e6 3 ttJf3 c5 4 d5 exd5 5 cxd5 d6


6 ttJc3 g6 7 h3 and now 7... a6 8 a4 More common is 5 ttJf3, which

'fie7. can become a Classical KID
following 5 ... ttJd7, as Tigran
MODERN DEFENSE Petrosian, among others, played.
The Modem Defense after 1 d4 But if Black doesn't want to playa
g6 2 c4 i.g7 becomes a King's Classical he can exploit the 1 d4 g6
Indian after ... ttJf6. Black can use 2 c4 i.g7 3 ttJc3 order with 3 c5!?
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.

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
1996. option in 3 . ttJc6.

If, instead, 5 d5 in the diagram, The attack on the d-pawn turns

Black gets a normal KID after out well if Black gets to secure his
5... ttJf6. But another benefit of this knight on d4, 4 d5 ttJd4 5 i.e3 with
order is 5..ttJd7 6 i.e3 ttJe7! and 5 ... c5 and then 6 dxc6 dxc6. This
.f5, which theory smiles on. recapture isn't possible after ... d6.


But 3... tbc6 can tum out very But today we know ltJf3 fits in
badly after 4 tbf3! instead of 4 dS. quite well with the center exchange
Then a typical continuation such as because after, say, 7... cS 8 :blltJc6
4 d6 5 d5! tbb8 6 h3 heads for a White can favorably allow 9 dS!
Benoni (6 ... cS 7 i.d3) in which i.xc3+ 10 i.d2. And 4 ... dS no
White enjoys two extra tempi, longer deserves an exclamation
thanks to ... tbc6-b8. Life isn't easy point.
for the Modem player.
A Black who can play both the
KID and Gruenfeld may prefer the
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
pinning ... i.g4, he felt. That meant In that order, 1 d4 ltJf6 2 c4 g6
after 1 d4 tbf6 2 tbf3 g6 3 c4 i.g7 3 tbc3 d5, the most fertile area for
4 tbc3: traps lies in the Russian ('itb3)
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
Black could play 4.. d5! since defense. That seems trivial today
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
center with a timely ... i.g4. One of championship match when
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!?


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
But suppose you'd be happy in time for 9 ... 0-0 10 h3! .tc8 (or
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 A potential drawback to this
Lilienthal's. order is Black's extra option of
8... 0-09 i.e3 a5!? and ... a4. If you
This is both a ' something-to-
don't mind that - and you want to
think-about' order and a good way
dodge the Prins or Hungarian -
to limit Black's options. White
6 'iia4+!? suits you.
wants his opponent to think about
6... ltJc6 and worry about the A bit dicier is Wolfgang
consequences of7 e3 and 8 i.b5, as Uhlmann's similar 4 ltJf3 .tg7
well as of 6... c6 7 e4!, e.g. 7... ltJb6 5 'iVa4+. We again approach a
8 'iic2 .txd4 9 i.h6 and :dl or Russian tabia, 5....td7 6 "iVb3 dxc4
8... 0-0 9 i.e3 .tg4 10 ltJe5!. 7 'iVxc4! 0-0 8 e4.


tZJxc5 as in Goletiani-Neubauer,
Port Erin 2005. Only tricksters
should rely on 5 'ifa4+.


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 Black has played 6... i.g7:
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 White can meet 7...i.d7 with
he can try to punish White's order 8 .ite2 and transpose to an ttJf3
with 8...ttJa6!? Then on 9 e5 ttJg4 Exchange Variation in which
10 i.f4 he gets to put his extra Black's extra move, ... i.d7, blunts
... .itd7 to good use with 10...c5 his pressure on d4, e.g. 8... c5 9 ttJf3
11 d5 b5! as in Gerzhoy-Tiomin, ttJc6 10 0-0 0-0 11 l:.b 1 cxd4
Tel Aviv 2002. 12 cxd4 b6 13 d5! with advantage
(Almasi-Balusz, Gyula 1993).
Another added option is 8... b5!,
e.g. 9 ttJxb5 ttJxe4 10 'ifxc7 ttJc6 Instead, Black should meet
with good play. Or 9 ~3 c5! 7 .itb5+ with 7... c6! after which
10 dxc5 ttJa6 11 e5 ttJg4 12 .itf4 8 .ita4 begins a controversial


line championed by Alexander Then Lev Polugayevsky found a

Belyavsky. finesse, 10 :tel, instead of 10 0-0.
His first point was that 10..:i!ic7
However if he had played 6... c5
transposes into the Shamkovich
then on 7 i.b5+ he has nothing
plan after 11 O-O!.
better than 7.. i.d7, allowing White
to transpose to the Exchange line His second point was that the
with 8 i.e2!. 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


idea of meeting 9 j"e3 with 9 ... cxd4 But with the c-pawns on the
10 cxd4 'iia5+ 11 j"d2 'it'd8!. board, 12 j"d5 j"d7 13 ':bl 'it'c8
Experience with 12 j"c3 liJc6 is threatens the bishop with ... c4! and
inconclusive. If White doesn't have ... e6.
anything better than 12 j"e3 'iia5+ Another problem with ... cxd4
he is headed towards a draw by lies in the line made famous in the
repetition before the game is five 1987 world championship match,
minutes old. So far no downside to 13 j"xf7+ ':xf7 14 fxg4 ':xfl+ 15
9 0-0 has been found. ~xfl 'iib6 16 ~gl! 'it'e6.
Black delays ... cxd4. The old Then 17 "ii'd3 'iixg4 18 ':fl is
main line of the Exchange regarded as a slight edge for White.
Gruenfeld was 4 cxd5 liJxd5 5 e4 But with the c-pawns on the board
liJxc3 6 bxc3 j"g7 7 i.c4 c5 8 liJe2 we reach a slightly different
0-0 9 j"e3 liJc6 10 0-0 and then
position after ~g1/ .. :ife6.
1O ... cxd4 11 cxd4 j"g4 12 f3 liJa5.
But in the late 1970s Jan Timman Then 16 'it'd3 'it'c4! gives Black
demonstrated the value of delaying more vigorous queenside play, e.g.
an exchange of c-pawns, with 17 'it'd2 'it'e6! 18 liJg3 liJc4 or
10 ... i.g4. Play often transposes 17 'it'xc4+ liJxc4 18 j"f2 cxd4
following 11 f3 liJa5 and a later 19 cxd4 e5. As a result, the more
... cxd4. accurate 10 ... j"g4 is replacing
What's the point of 10... i.g4 1O... cxd4.
11 f3 liJa5 ? Two stand out.

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
After 10 ... cxd4 11 cxd4 j"g4 Black.
12 f3liJa5 theory says White can try
13 j"d5 and get a nice game More common is an order such
following 13 ... i.d7 14 l:tbl 'it'c8 as 1 d4 liJf6 2 c4 g6 3 g3 j"g7
15 'iVd3 and :t.fc 1. 4 j"g2.


7 ttJge2 eS 8 .te3? ttJg4 or 8 d5

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.
Note that 8 ttJf3 d6 transposes to
a Black-friendly KID that usually
Black often delays ... dS because arises from 1 d4 ttJf6 2 c4 g6 3 g3
he doesn't like 4 ... dS S cxdS ttJxd5 .tg7 4 .tg2 0-0 S ttJf3 d6 6 ttJc3
6 e4 ttJb6 7 ttJe2!. Instead, he waits, ttJc6 7 d5 ttJa5 when White plays
with 4... 0-0, for the easier-to-handle 8 'iid3?! cS 9 0-0 instead of8 ttJd2!.
S ttJf3 dS! 6 cxdS ttJxdS.
White's counter-finesse is S ttJc3, QUEEN'S INDIAN DEFENSE
to rule out S... d5. Black can
The Queen's Indian and Nimzo-
continue in KID mode (5 ... d6 or
Indian are opposite sides of the
S... c5). But there is also S...ttJc6!?
same coin, one designed if not
It says, 'You tried to avoid the minted by Aron Nimzovich. Before
Neo-Gruenfeld. But if you continue settling on his formula - 1 d4 ttJf6
6 ttJf3 now, I can transpose with 2 c4 e6 followed by 3 ttJf3 b6 or
6... dS.' 3 ttJc3 .tb4 - he experimented with
an order that didn't depend on
White's third move, 1 d4 ttJf6 2 c4

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 The benefit of this order is Black
were discredited in the 19S0s, e.g. has sharper options after 3 ttJf3


.Jib7 4 g3 such as 4 ... e5!?, 4 ....Jixf3 If White can conquer his Nimzo-
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
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
3 ..b64 e3 .Jib7 5 .Jid3 .Jib4+!? gambit again.

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 The critical move is 7 tDc3,
equality. His Nimzo-fear has hurt preparing to seize the center with
him since 6 tDc3! transposes into a 8 'ifc2 and 9 e4. Black usually
line that runs 1 d4 tiJf6 2 c4 e6 occupies e4, 7... tDe4, to mechanic-
3 tDc3 i..b4 4 e3 b6 5 .Jid3 .Jib7 ally stop e2-e4. But if ... .tb4
6 tDf3 and offers excellent chances, equalized earlier, why not 7..Jib4,
even in the 6 ... tDe4 7 0-0 gambit. a tempo down?


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 It's as if the game began 1 d4 d5
after 8 i.d2 but not by a great deal 2 c4 e6 3 liJf3 liJf6 and then 4 e3
(8 ... i.xc3 9 i.xc3 liJe4). i.e7 5 liJbd2?! and White got an
Or 8... d6 9 :tel liJbd7 10 d5 extra a2-a3 as compensation. It isn't
liJc5 as in W.Schmidt-Gofshtein, enough and 6 e3 0-0 7 i.d3 b6
Debrecen 1992 (11 liJh4 a5 12 'ifc2 8 0-0 i.b7 9 b3 liJbd7 10 i.b2 c5 is
:e8 13 :fe 1 i.xc3 14 i.xc3 exd5 easy equality.
15 i.xf6 'i'xf6 16 cxd5 g6 with
The main stem of the Bogo is
4 i.d2, not 4 liJbd2. Black usually
BOGO-INDIAN DEFENSE protects the bishop (4 .. :Yi'e7, 4 ... a5,
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 Nimzo-
women'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.


His point is that White would see But Black can avoid trans-
that S... b6 6 ii.g2 ii.b7 is a posing into Botvinnik-Capablanca
favorable version of the QID. But (7 ... llJf6?!) with 7 ... llJc6 8 .td3
after S g3 Black has a good fallback llJge7! and ... i.fS. For example,
- 5... d5! transposes into a good 9 'iVc2? c4 10 .te2 .tfS and 9 llJe2
version of the Closed Catalan, one c4 10 .tfS 11 a4 0-0 12 i.a3
that usually arises from 1 d4 llJf6 l:.e8 with sufficient counterplay as
2 c4 e6 3 g3 dS 4llJf3 .tb4+ S .td2 in Khurtsidze-Tsereteli, Tbilisi
.te7 - and one a 3 llJf3 player may 2006 .
know little about. His KN can also be put to better
use if White tries 4 "ika4+. This
NIMZO-INDIAN DEFENSE would transpose into a Ragozin
The Nirnzo-Indian is related to QGD after 4 ... llJc6 S llJf3 llJf6
an opening without a name, 1 d4 d5 which is double edged after 6llJeS.
But Black renders llJeS harmless
2 c4 e6 3llJc3 .tb4!?
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 Alexander Alekhine said 'the
have avoided the Nirnzo with 3 llJf3 most logical answer' to 1 d4 llJf6
or 3 g3 or even 2 .tgS. But now 2 c4 e6 3 llJc3 .tb4 was 4 .tgS
I've gotten into a Nirnzo after all.' because 4 ... dS is 'a rather
But this is not a normal Nirnzo unattractive variation' of the QGD.
because if Black doubles the That is, 1 d4 dS 2 c4 e6 3 llJc3 llJf6
c-pawns, White can undouble with 4 .tgS ii.b4?!, which favors White
after S 'ii'a4+ or S e3.
cxdS, as we saw in Botvinnik-
Capablanca in the Introduction. But the real test of 4 i.gS, the
That means the litmus test should Leningrad Variation, is ... cS!. A
be 4 a3 .txc3+ 5 bxc3. For main line is 4 .tg5 h6 5 .th4 c5
instance, 5... c5 6 cxd5 exd5 7 e3. 6 d5 .txc3+ 7 bxc3.



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 Now 6 a3 i.xc3+? 7 'ifxc3 favors
... i.xc3+. This rules out d5-d6 and White. He's transposed into 4 'ifc2
reaches the desirable blockade liJc6?! 5 liJf3 d5 6 a3 i.xc3+. In
following 7 e3 i.xc3+! 8 bxc3 eS, that order, 6... dxc4?? would lose a
with equal chances. piece. But with the queen on b3,
6... dxc4! 7 'iixc4 i.d6 is excellent
For example, 9 'iVc2 'ife7 10 liJf3 for Black.
g5 11 i.g3 liJh5 12 i.d3 liJd7
13 liJd2 liJdf6 14 0-0-0 ~d8!? and A similar case arises after 4 'ifb3
... <3;c7, as in Guseinov-Ibragimov, c5 and 5 dxc5. Then 5... liJa6 6 a3
Baku 2006. i.xc5 is an easy road to equality.


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
evaluate 5 a3 .te7!? avoid the messy issue entirely.

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.

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
White is better but only
Championship 2006.
somewhat, e.g. 6 cxd5 exd5 7 .ltg5
c6 8 e3 .ltg4!? 9 tiJge2 tiJbd7 10 h3 There is, however, a complic-
.lth5 11 tiJg3 .ltg6 12 tiJf5 .ltxf5! ation in 6 i.g5 'i'xd5!?


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
with the new gambit is to answer 'I can't figure this all out so I'll
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 What theory? The considerable
tries 6...exd5. Then on 7 ~g5 we amount of book that's been built up
transpose into a position that comes around 4 'i'c2 0-0 5 a3 i.xc3+
from 4 'ifc2 d5 5 cxd5 exd5 6 ~g5 6 'fixc3, into which he's
when Black tries the ambitious unwittingly transposed. This is a
6... c5 and White fails to find 7 a3!. problem if Black normally meets
Instead he's played 7 dxc5, 4 'ifc2 with something other than
giving Black a fine game after 4 ... 0-0.
7 h6! 8 ~h4 g5 9 ~g3 ttJe4 and Similarly, ifhe is unfamiliar with
then 10 .txb8 'iff6! 11 .tg3 ttJxc3 4 'ifc2 d5 5 a3 ~xc3+, because he
as in the sensational game plays some other fourth move as
I.Sokolov-Aronian, Turin 2006. Black, he will be a stranger in a
Finally, few books even mention strange land after 4 'iVd3 d5 5 a3!
the third queen move, 4 'fid3. ~xc3+? 6 'fixc3.


NIMZO: SAMISCH line, 4 a3 i.xc3+ 5 bxc3 0-0 6 e3 c5

After Fritz Samisch introduced 7 i.d3 tbc6 8 tbe2, with a quick e3-
e4 and f2-f4-f5 attack.
4 a3 attempts were made to refine it
by delaying a2-a3 in favor of But there's a better transposition
Nikolai Riumin's 4 D and Akiba than Samisch-into-Riumin or
Rubinstein's 4 e3, which are Samisch-into-Rubinstein. It is
supposed to be more flexible. But a Rubinstein-into-Samisch, 4 e3
case can be made that 4 a3 i.xc3+ followed by a2-a3. It arises in
S bxc3 is the most flexible, because several orders including 4 e3 cS
it allows White to choose between S i.d3 0-0 and 6 a3 i.xc3+ 7 bxc3.
f2-D and e2-e3 once Black
commits himself.
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 Keres-
since 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
7 i.g5!). ferocious attack).
On the other hand, if Black Black often gets confused in the
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


10 0-0 lba5 11 e4 lbe8! with the commitments to pawn structure and

idea of ...lbd6. Reshevsky would can't be forced to trade off his
have been able to transpose into that bishop (as 4 ... c5 5 lbe2 and 6 a3
game after 8lbe2 .tb7 or 8 f3 .tb7. does). But 4 ... 0-0 does commit his
But White crossed him up with king which is important if White
8 e4!. shifts to the delayed Samisch, 5 a3
.txc3+ 6 bxc3.
Castling also costs Black a tempo
The best known Nimzo-Indian and prevents him from reaching
tabia arises three moves after 4 e3. 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 30-
35 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 Black succeeds after 5 "ifc2 f5
and 7... lbc6. Others prefer 5... d5 6lbe2 b6 or 6lbf3 b6 7 .td3 .tb7,
and 6 ... c5. Still others began with both excellent 4 ... b6 lines for him.
4 ... c5 and then 5 ... lbc6 and 6 ... d5. The main benefit in the 4 ... lbe4
And others have faith in 4 ... d5, order is that he doesn't have to deal
5 ... 0-0 and 6 ... c5. with the enormous amount of 4 e3
b6 theory, as well as some specific
Today 4 ... 0-0 has the most dangers, such as the 5 iL.d3 iL.b7
supporters because Black makes no 6lbf3 lbe4 7 O-O!? gambit.


White can also meet the threat to position is considered unclear

c3 in the last diagram with 5 tbe2 following 6 'iia4 tbd7! or 6 0-0
instead of 5 'iYc2. If Black doesn't l:tb8!?
like the looks of 5... 'iYh4 6 g3 'iif6
Moreover the Black QN is well
he can try to reach a good 4... b6line
placed ifhe manages to transpose to
once more with 5... f5 6 'iYc2 b6.
a Bogo-Indian with 4 ... i.b4+
5 i.d2 'iie7. But White should
TANGO avoid that with 5 tbbd2! and 6 a3.
This is a transposition from a
In a remarkably short time, 1 d4
normal Bogo, 1 d4 tbf6 2 c4 e6
tbf6 2 c4 tbc6 has spawned a lot
3 tbf3 i.b4+ when 4 tbbd2 has
of theory, especially in sharp
been met by the weak 4...tbc6?!
continuations beginning with 3 tbc3
5 g3 'iie7.
e5 and 3 d5 tbe5. But in most
games White flees towards the If, instead, White plays 4 tbc3 in
familiar with 3 tbf3. After 3... e6: 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?
White can try 4 g3 after which
4... d5 makes it a Catalan. Such Or did White get into a 4 'iic2
positions, with a blocked c-pawn, line in which Black has been denied
were once considered heretical the more popular 4... d5 or 4... c5 ?
but now have not-too-shabby Once again, both players may have
reputations. After 5 i..g2 dxc4 the been tricked.

Chapter Seven:
Dutch Defense and Flank Openings

Everyone knows the first finesse White hints he is amenable to a

of the Dutch Defense: Black can Tarrasch French (2 ... d5 3 e4). But
avoid the Staunton Gambit, 1 d4 f5 his real intent may be a Colle (2 ... d5
2 e4, by means of 1...e6 and 2 c4 f5. 3 liJgf3!) in which Black has
In the last few decades 1...e6 has blocked in his QB. Black can
also been used to reach a Nimzo- also find himself in an inferior
Indian or Queen's Indian (2 c4 liJf6) Stonewall Attack after 2... c5 3 e3
or a Benoni (2 c4 c5) without risk of
liJc6 4 c3 liJf6 5 i..d3 d5 and 6 f4!.
facing a Trompowsky.
Of course, there's another White can only dream of getting
hazard. After 2 e4 Black has to be that position after 1 d4 d5 2 e3 and
ready to defend a French (2 ... d5) or it's quite good, e.g. 6..i..d6 7 'iff3
the more dangerous version of the i..d7 8 liJh3 'iib6 9 liJf2 0-0-0
English Defense (2 ... b6) or try 10 0-0 ~b8 11 e4!, Marshall-
2 ... c5, as Bent Larsen did on Rubinstein, Vienna 1908.
occasion. And he has to know what
to do after 2 liJd2!? 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:

Dutch Defense and Flank Openings


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
Trompowsky-like ~xf6/e2-e3/
~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
Now 5 lLlc3?! is fine for Black,
9 lLlge2 lLld7 10 g4! fxg4 11 hxg4 e.g. 5... 0-0 6 lLlf3 lLle4 7 'iic2 d6
'it'e7? 12 l:txh7! as in Korchnoi- 8 0-0 .txc3! 9 bxc3 lLld7 10 l:tdl
Kostjoerin, Havana 1966. ~e8 11 lLld2 lLldf6!? as in Hort-
Short, Prague 1990.
The most common response to
2 lLlc3 is 2.. d5, when 3 e4!? Better is 5 ~d2! since an
transposes to a good version of the exchange of bishops tends to favor
Staunton after 3... fxe4? 4 f3 exf3 White. This is particularly true if
5 lLlxf3 but 4 'iWh5+! is better. Black adopts a Stonewall pawn
structure, 5... ~xd2+ 6 lLlxd2 d5?,
That means 3... dxe4 is best. But instead of 6... d6! and ... e5.
this often transposes (4 f3 lLlf6 5 Black has more success delaying
fxe4 fxe4) into 3... fxe4 4 f3 lLlf6 ... .tb4+ in the hopes that White will
5 fxe4 dxe4. Then 6 ~c4 e5!? voluntarily develop his knight on c3
7 ~g5! and 6... lLlc6 7lLlge2 e5 8 d5 and allow the pin. But that's only
are unclear. good when White is committed to
lLlf3 or ~g2. The pin is premature
Bottom line: If Black can also after 1 d4 f5 2 c4 e6 3 lLlc3 .tb4?!,
play the French, avoiding all this for example, because White hasn't
with I d4 e6 may be his best route ruled out Akiba Rubinstein's plan of
to the Dutch. Moving on to 1...f5: e2-e3/lLlge2/a2-a3, which he used

Dutch Defense and Flank Openings

just as effectively against the Dutch control of e5 with ltJd3. But after
as against the Nimzo-Indian. ... 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!.
Consequently, if Black commits
his d-pawn too early, White knows
what to do with his KN - 1 d4 fS
That means 4 'iic2 ltJf6 S e3! 0-0 2 g3 e6 3 iLg2 ltJf6 4 c4 dS S ltJh3!.
6 iLd3 followed by ltJe2 and a2-a3,
The same applies if White develops
favors White, e.g. 6... d6 7 ltJe2 ltJc6
his knight at e2 rather than h3. Two
8 0-0 iLxc3 9 'iixc3 a5 10 b3 e5
of the 1951 world championship
11 iLa3! J:te8 12 d5 ltJe7 13 f4!,
games reached this position.
Yakovich-Gleizerov, Stockholm
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.
4 ...iLe7 5 iLd3 0-06 ltJge2 d6! with Mikhail Botvinnik played
a good game. 6 ... d5?!, got the worst of7 ltJge2 c6
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.

Dutch Defense and Flank Openings

The situation becomes more Mark Taimanov, who praised

difficult when Black seeks a 6 ... c6, explained that the
Stonewall with ... ~d6. Then he Stonewall's defects are best
can't use ... ~e7 as a waiting move. exploited when White can play
But its place as a 'pass' can be taken ttJf3-e5-d3 and ttJbl-d2-f3. He
by ... c6. cited a game that went 6... d5,
After 1 d4 f5 2 g3 ttJf6 3 ~g2 e6 instead of 6 ... c6, and turned in
4 c4 d5 White stands well with White's favor following 7 ttJbd2!
5 ttJh3! c6 6 0-0 ~d6 7 ~f4. But 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 However, once White is
both 5 ttJh3 d6 and 5 ttJO d5. 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 For example, 7 ttJc3 d5 8 b3 ttJe4
bishop (8 i..a3). 9 ~b2 ttJd7 10 e3 ~d6 11 ttJd2
ttJdf6. If White avoids 7 ttJc3 by
If White insists on the trade, with
means of 7 b3 Black has a good
8 a4, Black gets adequate play from
reply in 7 ... a5. He can answer
8... a5! 9 ~a3 i..xa3 10 ttJxa3 0-0
8 ~a3 with 8... ~xa3 9 ttJxa3 d6!
11 ttJc2 b6 12 ttJce 1 i..b 7 13 ttJd3
ttJbd7 and ... c5 as in Tenikashvilli- or 9 ... ~e7 10 ttJc2 d6!. His bad
Radjabov, Moscow 1998. bishop is no longer so bad.

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 Black can set up a Dutch on his
and then 6... c6!. Black waits for side of the board after 1 c4, 1 ttJO
7 ttJc3, which makes 7 ... d5 more or 1 g3 by playing 1...f5. He can
appropriate. also make a good Dutch out of a

Dutch Defense and Flank Openings

doubtful English Defense, as in 1 c4 pioneered by Carlos Guimard, 1 d4

b6 2 d4 ~b7 3 liJc3 e6 4 a3 and d5 2 c4 e6 3 liJc3 ~e7 4 liJf3 and
now 4.. f5!' now 4 f5.

White's a-pawn plays a minor This sequence rules out the

role, e.g. 5 d5liJf6 6 g3liJa6 7 ~g2 Staunton Gambit as well as the 1 d4
ltJc5 8 liJh3 ~d6 9 ~e3 ~e5!, fS 2 liJc3 and 2 ~gS lines, and gets
Karpov-Morozevich, Cannes 2002. White to commit his knights, so that
the liJbd2, liJh3 and b2-b3/~a3
But after I c4 fS or 1 liJf3 fS options are lost.
White can play d2-d3 rather than
d2-d4. This takes e4 away from But White can make a favorable
course correction with 5 ~f4! liJf6
Black's KN and may prepare a
6 e3 c6 7 ~d3 0-0 8 'iic2. For
powerful push of the e-pawn, e.g.
example, 8... liJe4 9 g4! liJa6 10 a3
1 liJf3 fS 2 g3 liJf6 3 ~g2 e6 4 c4
'iVaS 11 ~e2 liJxc3+ 12 bxc3 liJc7
dS S 0-0 c6 and now 6 d3 ~d6
13 cS liJe8 14 l:thg 1 ~h8 IS l:tg2
7 liJc3 0-0 8 e4!. and White crashed through in Lima-
Giardelli, Sao Paolo 1993.
Black was victimized by his
insistence on 3... e6 and ... dS. Ifhe's
willing to handle a ... d6 Dutch, he ENGLISH DEFENSE
should play 3... d6, preparing 4... eS.
If White deters him with 4 d4 Black Most lines of the English
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:

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 But a more adventurous Black
5 ltJf3 i.b4 since White can no will try 3...f5!? 4 ltJc3 ltJf6. White
longer break it favorably with ltJe2 is better off here than he was in the
and a2-a3. Better for White is position considered earlier, I d4 e6
5ltJge2 ltJf6 6 d5 but after 6...ltJa6 2 c4 b6 3ltJc3 i.b7 4 e3 f5!. In this
this is one of the better lines of the case he can benefit from a
English Defense for Black. fianchetto, 5 g3 i.b7 6 d5!.
An irony is that many English However, there is room for
Defenders rarely - if ever - begin subterfuge after 5 g3 c6! 6 i.g2 d5:
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.
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. Each side has an extra move
Another aspect of the English compared with a normal Stonewall.
Defense is that White can avoid the But ... b6 is more useful than

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 Rogozenko-
Morozevich, Bundesliga 2000.

Note how late Black waited to

put his bishop on b7. He was
holding back because after S dS it This retains the options of ttJc3,
makes more sense to play S... .ta6!. c2-c3 and even c2-c4, as in S... dS
Then 6 e3 exdS 7 cxdS .txfl and 6 eS ttJfd7 7 0-0 cS 8 c3 and 7... .te7
7 ttJxdS .td6! are quite nice for 8 c4!, e.g. 8... dxc4 9 i.xc4 a6
him. 10 ttJc3 bS 11 i.d3 ttJb6 12 i.e4
with an edge in Khalifinan-Bauer,
If Black insists on the English internet 2004.
Defense against 1 e4, a key
question is how White defends the CATALAN OPENING
e-pawn after 1 e4 b6 2 d4 .tb7. On
3 ttJc3 e6 Black is ready for A Catalan is i.g2 attached to a
4....tb4. Experience with 4 ttJf3 Queen's Gambit. Curiously, the
.tb4 S .td3 ttJf6 has varied widely Catalan comes about more often
but the results are better for Black these days via 1 d4 ttJf6 than 1 d4
than other 1 e4 b6 lines. 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, Opocensky-
Kotov, 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
be 3 i.d3 ttJf6 4 'iie2! and then This transposes to a normal
4.. e6 5 ttJf3. Catalan after 2 ... e6 3 c4. This order

Dutch Defense and Flank Openings

allows Black to develop his QB superior form of the Tarrasch QGD,

outside his pawn center (2 ... d5 4... c5 5 cxd5 cxd4! 6 'iVxd4 tbc6
3 tbf3 c6 4 .lig2 .lif5). But that is no that worked well for Boris Spassky.
longer considered an easy road to
Note that White can limit enemy
equality after, say, 5 0-0 e6 6 b3
resources by holding back the c- or
h6 7 .lib2 tbbd7 8 c4 .lid6 9 tbbd2
d-pawn. For example, 1 d4 tbf6
0-0 10 tbe5 and 2-f4, S.Polgar-
2 tbf3 e6 3 g3 d5 4 .lig2 .lie7 5 0-0
Ushenina, Dresden 2006.
and a later c2-c4 rules out ... .lib4+
There are two common I d4 tbf6 and early ... dxc4 lines.
2 c4 e6 Catalan tabias and they
He can also delay d2-d4, as with
often appear after 3 g3 d5 4 .lig2
1 c4 e6 2 g3 d5 3 tbf3 and then
and 4 tbf3 (or 3 tbf3 d5 4 g3). There
3... tbf6 4 .lig2 .lie7 5 0-0 0-0 6 d4
are very minor differences between
is the closed Catalan. Of course,
the tWo. But the third order, 1 d4
there is less chance of reaching a
tbf6 2 c4 e6 3 tbc3 d5 4 g3, has
Catalan middlegame via these
plusses and minuses.
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.
In this way he avoids 'iVa4+ lines
But this order is imprecise such 1 d4 tbf6 2 c4 e6 3 g3 d5
because Black can transpose into a 4 .lig2 dxc4 5 'iWa4+. This makes

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.
Quite a different story is 1 c4 g6
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
3 b3? d4!, for example, is bad. Another family of finesses
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
3 b3. can meet 2 ... e6 with the well-

Dutch Defense and Flank Openings

regarded 3 e4. This also enables ENGLISH: 1... e5

him to dodge the Gruenfeld with
After 1 c4 e5 Black typically
2 ... g6 3 e4 and transpose into the
chooses between open systems
King's Indian after 3 ... d6 4 d4.
( ... l2Jf6/ ... d5) and closed ones
Books devote a lot of attention to ( ... l2Jc6 and possibly ....tb4).
2 12Jc3 d5. But perhaps the best
The finesses begin with 2 g3.
aspect of the 1 c4 12Jf6 2 12Jc3 order
This will transpose into main lines
for Black is he can transpose into a
after, say, 2 ... l2Jf6 3 .tg2 d5 4 cxd5
1...c5 line with 2 ... c5! and then
break in the center with 3 ... d5. As 12Jxd5 5 12Jc3 or 2 ... l2Jc6 3 .tg2 g6
noted earlier, if he had played the 4 12Jc3. But in this way White
more common order such as 1 c4 c5 avoids all the nasty ... .tb4 business.
White can break first with 2 12Jf3!? After Anatoly Karpov made good
and 3 d4. use of 2 12Jc3 12Jf6 3 12Jf3 12Jc6 4 g3
Another virtue of 1 c4l2Jf6 is that .tb4 as Black in his 1987 world
2 12Jc3 e5 gets Black to a common championship match, Vasser
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 c-
pawn with 4 'ilVc2 or 4 b3, which
allows Black to solve his opening
problems with 4 ... e5!, e.g. 4 'ii'c2 e5 But Seirawan's 3 d4! gave him
5 d3 h6! 6 12Jf3 .td6 7 0-0 0-0 either a distinct edge in space
8 12Jc3 d4, Socko-Belov, Hastings (3 ... exd4 4 'ii'xd4) or a superior
2004. endgame (3 ... d6 4 dxe5 dxe5

Dutch Defense and Flank Openings

5 'i'xd8+). Karpov chose the latter against the OlD. For example, 8 e4
and lost. :e8 9 h3 i.f8 10 l:tel a6 11 d5?!
The drawback to 2 g3 is 2...tbf6 cxd5 12 cxd5 b5! 13 a3 tbb6 14 b3
3 i.g2 c6 since 4 tbfJ e4 5 tbd4 d5 i.d7 15 i.e3 'i'b8 with a fine game
6 cxd5 can be met by Mikhail Tal's for Black (A.Sokolov-Sherwin,
6.'i'xd5!. By attacking the knight Swiss Team Championship 2001).
Black obtains excellent chances, In the main closed variation of
e.g. 7 tbc2 'iih5 and ... i.h3. the Reversed Sicilian, 1 c4 e5
The comparable 2 tbc3 tbf6 3 g3 2 tbc3 tbc6 3 g3 g6 4 i.g2 i.g7,
c6 line never returned to fashion 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

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 White has missed his chance for
after 5.te7 6 0-0 0-0 7 d4 tbbd7. d2-d4. Moreover, 4 i.g2 d6 5 tbfJ
The latter is a good choice because e5! puts him in a closed variation
the fianchetto is one of the least with tbf3 and d2-d3, which theory
ambitious weapons White has calls fairly even.

Dutch Defense and Flank Openings

ENGLISH: ..ii.b4 3 tLlf3 tLlc6 4 g3 ii.b4 lines. But it's

harmless after 1 c4 e5 2 tLlc3 ii.b4
The reversed version of 1 e4 c5
3 g3 ii.xc3 4 dxc3?! because Black
2 tLlf3 tLlc6 3 ii.b5 has more nuance
can make better use of his KN than
in the English because Black can
transposing with ... tLlf6, e.g. 4 d6
play ... ii.b4 at move two, three, four
5 ii.g2 tLle7! 6 tLlf3 tLlbc6 with
or five. An instructive example is good chances.
1 c4 e5 2 tLlc3 tLlc6. White can try
to punish Black's last move with There are other English lines in
3 e3 and d2-d4-d5. Then 3 ...tLlf6 is which bxc3 is preferable. But in this
an excellent waiting move. case, on 4 bxc3 Black takes
advantage of the absence of ... tLlf6
by playing ... f5 or ... f6.

Now 4 tLlf3 ii.b4! is a well-

guarded line for Black. He has For example, 4. d6 5 ii.g2 tLlc6
transposed into 1 c4 e5 2 tLlc3 tLlf6 6 d3 f5 or 6 d4 tLlge7 7 tLlf3 0-0
3 tLlf3 tLlc6 and then 4 e3 ii.b4. 8 0-0 f6! (9 ii.a3 ~h8 and now
Theory indicates that after 3 tLlf3 10 d5? tLla5 11 'ti'a4 b6 12 c5 tLlxd5
tLlc6 White's only serious chance 13 cxd6 cxd6 14 tLlxe5 ii.b7
for an edge lies instead in 4 g3!. favored Black in Akesson-Sisatto ,
Stockholm 2006).
That was a case of White tipping
The drawback to 1 c4 e5 2 tLlc3
his center plans with his third move.
ii.b4 is, of course, 3 tLld5 which
He does better with 1 c4 e5 2 tLlc3
both avoids ... ii.xc3 and gains time.
tLlf6 3 g3 or 3 tLlf3. But Black can
But if you believe recent experience
anticipate that with 2 ... .tb4. He
indicating that 3 ....td6 or 3 ....tc5
wants to know how White will
equalize, then 2 ... ii.b4! is an
retake after 3 ... .txc3. improved order for a Black who
That's important information likes the positions that arise after
because the conservative dxc3 is 1 c4 e5 2 tLlc3 tLlf6 3 g3 ii.b4 or
recommended in some 2 tLlc3 tLlf6 3 tLlf3 tLlc6 4 g3 ii.b4.

Dutch Defense and Flank Openings

It may also pay to accelerate 3 g3 d5 4 cxd5 tiJxd5. Once more

... i.xc3 in that 3 tiJf3 tiJc6 4 g3 White's choice between the two
i.b4 line. Play usually continues depends on what he most wants to
5 i.g2 0-0 6 0-0. Since 6... .l:te8 avoid.
7 tiJd5! favors White Black's task is The first order allows Black
to time ... i.xc3 and ... e4 accurately. alternatives such as 4 ... i.c5 and
Books consider 6... e4 at length 4 ... tiJd4 that are quite reasonable. In
and hardly mention 6..i.xc3!? 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

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!. This means he can get the upper
If White retakes with the d-pawn hand with 4 i.g2 tiJc6 5 e3!
Black may do best to avoid ... e4, followed by tiJge2 and d2-d4. For
which could transpose to a difficult example, 5... a6 6 tiJge2 d6 7 d4
6... e4 7 tiJel i.xc3 8 dxc3 line. exd4 8 exd4 i.a7 9 h3 0-0 10 g4,
Instead he should maintain his dark- as in Moskalenko-Diaz, Badalona
square pawn structure, e.g. 7 dxc3 2000, among others.
d6 and then 8 tiJel i.e6 9 b3 'iVd7 The main objection to the second
and ..i.h3. order is 3 ... i.b4, which some
English players regard as more
REVERSED DRAGON promising for Black than 4 ... i.h4 is
in the four-knights order.
The mirror image of the
Accelerated Dragon arises after When the Dragon occurs in the
1 c4 e5 2 tiJc3 tiJf6 3 tiJf3 tiJc6 4 g3 four-knights order, White decides
d5 5 cxd5 tiJxd5 as well as after whether to delay d2-d3, as in

Dutch Defense and Flank Openings

1 c4 e5 2 liJc3 liJf6 3 liJf3 liJc6 4 g3 ENGLISH: l ... c5

d5 5 cxd5liJxd5:
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 doub1e-
8 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 - can try 4 e3 and if 4... e6 then 5 d4
provided White wants to quickly d5 6 dxc5! or 5... cxd4 6 exd4liJe7
push his b-pawn, such as 6 ... liJb6 7 d5! and 6... d5 7 c5.
7 0-0 .te7 8 a3 or 8 .l:lbl. If he More common IS 4... liJc6
doesn't, 6 d3 is easier to handle. 5liJe2!:
Note also that 1 c4 e5 2liJc3liJc6
would transpose into the four-
knights 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.

Dutch Defense and Flank Openings

This enables White to transpose apparent in Seirawan-Dlugy, World

safely into normal lines (5 ... e5 Open 1989, which went 6 'ilVb3 tbc6
6 tbbc3 or 5...tbf6 6 tbbc3) as ifhe 7 .txc6+ bxc6 8 tbf3 f6 9 'i'a4.
had played 2 tbc3 and Black had Black was lost soon after 9 ... 'ilVd7
passed up both the symmetrical 10 d3 e5 11 i.e3 tbb5? tbd4
... e6/ ... d5 and the unbalancing 13 0-0 h5? 14 h4 .
... .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

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
This variation is almost always
4 cxdS tbxdS S tbc3 puts
treated as some offbeat Gruenfeld,
immediate pressure on d5.
with analysis of 4 g3 met by
If Black responds with the 4 ... tbxc3. However, Black also has
natural S... e6 he reaches his goal 4... cS! and then 5 i.g2 tbc7.
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 innocent-

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,
avoided it (by means of 1 c4 c5 One of the drawbacks to 2 tbc3 is
2 tbc3 tbf6 3 tbf3 d5 or 2 ...tbc6 that 2 ... g6!? allows Black to
3 tbf3 tbd4). concentrate firepower on d4. His
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.

The natural 5... g6? is faulty

because 6 d4! cxd4 7 "iVxd4 attacks The d4-knight is under attack
Black's KR. before White can smooth out his
development with 6 e4 and 7 i.e3.
Black is also worse after 6 ... i.g7
If the knight retreats to b3 or c2
7 e4 tbc7 8 d5. What has happened
then 6 ... i.xc3+! puts White's
is we've reached a Neo-Gruenfeld
queenside under severe pressure.
(1 d4 tbf6 2 c4 g6 3 g3 d5 4 i.g2
i.g7 5 tbf3) in which Black has
Trading that bishop was once
played 5... c5?! (and 6 cxd5 tbxd5
thought to be foolhardy but that
7 e4!) instead of the superior 5... 0-0
view was overthrown by examples
or 5... dxc4.
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.

Dutch Defense and Flank Openings

There is a possible drawback to prefer 6 ... liJa5, e.g. 7 e4 liJf6 8 h3

1 c4 c5 2 liJc3 g6 in the rarely 0-0 9 .td3 a6 10 0-0 b5 11 cxb5 c4
played 3 d4. Then 3.. cxd4 4 'iixd4 12 i.c2 axb5 13 liJxb5 liJb3!
attacks the KR. This leads to (lzoria-Savchenko, internet 2006).
4 ... liJf6 5 i.g5 liJc6 6 'iid2, which Despite that, and support for
is hard to evaluate, when compared 6 ... i.xc3+ 7 bxc3 liJa5 as well,
with vaguely similar positions, such theory doesn't trust Black's system.
as 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4
But if there is no knight on c6,
4 liJxd4 g6 5 c4 liJf6 6 liJc3 liJxd4
the pawn push is weak, as Bisguier-
7 'iixd4 d6 8 i.g5 i.g7 9 'iVd2.
Karpov, 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).
Fischer also met 1 c4 c5 2 liJc3 RETI OPENING
with 2..liJc6 and then 3 liJf3 g6.
Kramnik made 1 liJf3 famous as
This rules out a 'iixd4 recapture
a purely transpositional device.
after d2-d4/ ... cxd4. Black gets
After 1...d5 he reaches normal
where he wants to go following
Queen's Gambits with 2 d4 and
4 d4 cxd4 5 liJxd4 i.g7.
3 c4. But after 1...liJf6 he almost
But White can play 4 e3! and always replies 2 c4.
threaten 5 d4 and 6 d5, which is
In this way he avoids the Nimzo-
stronger when it attacks a knight on
Indian by meeting 2 ... e6 3 liJc3
c6. ?-. \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 Of all the major first moves,
4 ... i.g7 5 d4 d6 6 d5 liJe5. Others 1 liJf3 grants Black his best

Dutch Defense and Flank Openings

opportunity for 1... b6 since White - 1...f5 tries to exploit White's

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
b62 e4 ~b7 31LJc31LJf6!: 2 d3.

The attack on the e-pawn White's goal is to transpose into

interrupts White's center-building, a favorable version of the Lisitsyn
4 e5 lLJd5 5 lLJxd5 .txd5 6 d4 e6, (2 e4) Gambit after 2.lLJf6 3 e4
e.g. 7 c4 ~b4+ or 7 .te2 c5 8 c4 fxe4 4 fxe4 lLJxe4 5 ~d3 lLJf6
~xf3! 9 ~xf3 lLJc6 10 d5 lLJxe5 6 lLJg5 and lLJxh7! or ~xh7!. The
11 ~e4 f5 with a fine game original Lisitsyn runs 2 e4 fxe4
in Shchekachev-Bauer, Clermont- 3 lLJg5 lLJf6 4 d3 exd3 5 ~xd3 but
Ferrand 2003. Black gets the edge with 3... d5! in
Black's other options are more that version.
If Black is willing to playa ... d6
- l...c5 is a favorite of Dutch, he should reply 2 ... d6,
Sicilianistas (2 e4). Since 1 1LJf3 which avoids the gambit as well as
players generally prefer closed
2 ... lLJf6 3 ~g5. But if Black wants a
lines, the typical result is a
Stonewall, then 2 d3 is a major
Symmetrical English (2 c4) or a
problem for him.
King's Indian Reversed (2 g3).
- 1. .. d6 seems to say, 'I'm - 1. .. g6 keeps more Black
offering a KID this way (2 d4 lLJf6 options open. Against 2 e4 he can
and 3 ... g6) because I don't want to shift to a hyper accelerated Dragon
end up in a Pirc after 1...g6 2 e4.' If Sicilian with 2 ... c5 or head toward a
White tries to trick him into a Pirc, safe Pirc, with 2... d6 or 2 ... ~g7,
1... d6 2 e4, he can find himself in a because White has ruled out 2-f4
Sicilian,2 ... c5. and 2-f3 lines.

Dutch Defense and Flank Openings

After 2 d4 or 2 c4 Black retains More common after 1 tbf3 d5 is

the option of various kinds of 2 c4. The simplest defense is
Modem Defenses, Englishes and 2... dxc4. Then 3 e3 becomes a book
the Dutch Indian (2 ... f5). If he QGA in the vast majority of games
meets 2 d4 with 2... c6 the results (3 ... tbf6 4 i.xc4).
include a Neo-Gruenfeld (3 g3 d5),
a Modem (3 e4 d5), or some Can White do better? Richard
nondescript Indian (3 c4 d6 Reti liked 3 tba3. But 3 ... c5 4 tbxc4
4 tbc3 i.g7 or 4 ... i.g4). tbc6 and ... e5 transposes into a kind
of English (1 c4 c5 2 tbs tbf6 3 g3
- 1... e6 is, next to 1. .. tbf6, the d5 4 cxd5 tbxd5 5 i.g2 tbc6 6 0-0
most elastic response to 1 tbS. It's
e5 7 d3 i.e7) in which White
often used by a Dutch loyalist (2 d4
has chosen a suspect knight
f5) who is trying to trick White into
development, 8 tbbd2 and tbc4.
a French (2 e4 d5!) or a Sicilian
(2 e4 c5). After 3 tba3 fell out of fashion,
White tried 3 'i'a4+. The simplest
RETI: 1...d5
response to that is 3...tbd7 and
After 1...d5 the Hyperrnodemist ... a6/ ... b5, seeking a Catalan.
will play 2 c4 or 2 g3. But the
transpo-minded will look at 2 e3.

For example, 4 'i'xc4 e6 5 g3 a6 6

i.g2 b5. And on 4 g3 a6 5 tbc3 e6
Taimanov used this to get into a
6 i.g2 ltb8!.
Slav (2 ... c6 3 c4 tbf6 4 tbc3 e6
5 d4) or QGA (2 ... tbf6 3 c4 dxc4 The threat of 7 ... b5 leads to
4 i.xc4 e6 5 d4, dodging ... i.g4 7 'i'xc4 b5 8 'i'b3 i.b7. Then White
lines). White enjoys many of the has nothing better than to transpose
same opportunities as he would into an equal Catalan with d2-d4.
after 2 c4 c6 3 d4 tbf6 but without That is, 1 d4 tbf6 2 c4 e6 3 g3 d5
having to find an answer to 2 c4 4 i.g2 dxc4 5 'i'a4+ tbbd7 6 tbS
d4!? a6 7 tbc3 ':b8 8 'i'xc4 b5 etc.

Dutch Defense and Flank Openings

If Black doesn't like these 4lbf3 e4 5lbd4 i.c5! 6lbxc6 dxc6

positions he should consider 7 e3 i.f5 8 'ifc2 'ife7 9 i.e2 0-0-0
meeting 2 c4 with 2c6 since the and White was virtually lost after
Slav Defense (3 d4) is holding up 10 f4?
well these days. White can remain
Theory concluded that White
true to the spirit of Reti with 3 b3.
Today's theory says Black can should play lbe2 and f2-f4 instead.
equalize whether he adopts But another point of view says
Lasker's defense ( ... i.f5) or Larsen's real error was playing c2-
Capablanca's ( ... i.g4). A major c4 before lbf3. Instead of 3 e3 and
reason is that 3 b3 rules out 'ifb3, 3 c4 lbf6 4 e3, which are played
the attack on b7 that is so most often today, White may prefer
troublesome to Black in ... d5/ ... i.f5 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/e2-
e4. The KIR enjoys a pretty good
reputation against ... e6.
If, however, Black brings out his Black defends the e-pawn with
QB, with 3i.f5. White can switch ... f6 in some 1 b3 lines but here
from the KIR back to the Reti with 3. f6?! 4 e4! and 5 i.c4 can't be
4 c4. In this way he avoids 2 c4 d4 good for him. Moreover, 3... d6,
and 2 ... dxc4. which is fine following 3 e3, allows
White to punch at e5 with 4 d4, e.g.
NIMZO-LARSEN OPENING 4e4 5 d5! exf3 6 dxc6 bxc6
Bent Larsen helped make 1 b3 e5 7 gxf3 lbf6 8 'ifd4! with advantage.
2 i.b2 lbc6 famous by attacking For example, 8... i.f5? 9 e4 i.g6
the e-pawn with lbf3. Then after 10 lbd2 i.e7 11 'ifa4! 'ifd7
... e4 he occupied the central dark 12 i.h3! resigns, Hansen-Bergfalk,
squares with pieces and assaulted correspondence 1992, or 8... i.e7
... d5 with c2-c4. 9 l:tgl c5 10 'ife3 0-0 11 lbc3 g6
12 0-0-0 l:te8 13 'iff4 and 14 e4.
But this strategy was almost
retired in 1970 by a Larsen-Spassky Note also that in similar positions
miniature, which went 3 c4 lbf6 ... i.d6 defends the e-pawn well -

Dutch Defense and Flank Openings

1 b3 e5 2 ~b2 ~c6 3 e3 ~f6 4 ~b5 without reaching the Spassky

~d6!' But it fails here, 3 ~f3 ~d6 position (5 ... .i.xc5?? 6 ~xc6 and
4 ~a3! and ~c4 or ~b5. ~xg7).

Black most natural response to If White likes such positions he

3 ~f3 is 3 .e4 and then 4 ~d4 ~f6. should be happy (after 1 b3 e5
The absence of c2-c4 plays a role in 2 i.b2 ~c6) with 3 e3 d6 because
S e3 ~cS 6 ~xc6 dxc6: of 4 ~f3, rather than the book-
endorsed 4 .i.b5.

This is better for White than

Larsen-Spassky after 7 d4!? ~e7
Now 4... e4 will bring about the
8 c4 and perhaps after 7... exd3
desired position, a tempo ahead,
8 ~xd3 as well. If instead S...~xd4
after S ~d4 ~xd4 6 .i.xd4 when
6 ~xd4 dS then 7 c4 is timed well
Black inevitably plays ... dS.
and transposes into positions
Larsen had shown to be good. For Instead 4 ... ~f6 S d4! favors
instance, 7 ... ~f5 8 ~c3 dxc4 White after 5... exd4 6 ~xd4 i.e7
9 ~xc4 ~d6 10 'ifc2 0-0 11 f4! c6 7 ~b5, e.g. 7... i.d7 8 .i.xc6 bxc6
12 0-0 l:e8 13 ~dl 'ife7 14 ~f2, 9 'iif3! d5 10 ~f5 as in Hodgson-
Welling-Minasian, Capelle la Percerias, Breda 1989. White has
Grande 1996. transposed into a good 3 e3 d6
4 ~e2 .i.e7 5 d4 exd4?! 6 ~xd4
Nevertheless books say that after
1 b3 eS 2 ~b2 ~c6 the best move line.
is 3 e3. They often give 3... dS And if White is happier facing
followed by 4 ~b5 ~d6. Yet here 1 b3 d5 than 1...e5, he should
again ~f3 is an attractive idea - consider 1 ~f3 so that 1.. .d5 2 b3!
4 ~f3 e4 S ~d4 ~xd4 6 ~xd4 gets him closer to the middlegame
transposes into the above line, again he wants.

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

Breyer Gambit 14

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

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

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

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

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

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