James Schuyler

the
dark knight
system
a repertoire with 1 Nc6
...

EVERYMAN CHESS
Gloucester Publishers plc www.everymanchess.com
First published in 2013 by Gloucester Publishers Limited, Northburgh House,
10 Northburgh Stree t, London EClV OAT

Copyright © 2013 James Schuyle r

The righ t of Jame s Schuyle r to be identified as the auth or of this work has been
asserted in accordance with the Copyrights, Designs and Patents Act 1988.

An righ ts reserved. No part of this publication may be reproduced, stored in a

re trieval system or transmitted in any form or by any mean s, electronic,
ele ctrostatic, magnetic tape, photocopying, recording or otherwise, without prior
permission of the publishe r.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN: 978 1 8 5744 995 2

Distributed in North Ame rica by The Globe Pequot Press, P.O Box 480 ,
246 Goose Lane, Guilford, CT 06437-0 480.

All other sale s enqUiries should be directed to Everyman Chess, Northburgh House,
10 Northburgh Street, London EC1V OAT
tel: 020 725 3 7887 fax: 020 7490 3708
email: info@everymanche ss.comj website : www.everymanchess.com

Everyman is the re gistered trade mark of Random House Ine. and is used in this
work under licence from Random House Inc.

Everyman Chess Series

Chief advisor: Byron Jacobs
Commissioning editor: John Emms
Assistant editor: Richard Pallise r

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteve rde.
Printed and bound in Great Britain by Clays, Bungay, Suffolk
 Abo ut t h e Aut h o r

James Schuyler is a FIDE Master. He was Nevada State Champion in 2007 and won
the Virginia State Championship in both 2011 and 2012. He has been teaching
chess for over 25 years.
 Co nte nts

Introduction 7
Weak Colour Complex 13

Section One: 1 d4t'tJc6 15

1 2ttJf3 18
2 2 c4 34
3 2 dS 43

Section Two: 1 e4t'tJc6 55

4 2 d4 57
5 2ttJf3 92
6 2 ttJc3 114

Section Three: Others 119

7 1 c4 ttJc6 120
8 1 ttJf3 ttJc6 127
9 Others 129
10 Miscellaneous Topics 132

Illustrative Games 134

Index of Variations 213
Index of Games 222
 I nt roduction
1 ... ttJc6 a n d the Kevitz System
Why another repertoire book on 1 ... llJc6
. ? Didn't one come out just a few years
ag o? As it turns out, the subject matter
here is completely different, as are the
types of positions reached. 1 ... llJc6 isn't
really an opening yet, but the starting
point for many openings. This book
generally concerns itself with Black's
plans for using 1 ... llJc6 to force ... e7-es,
as played by Tony Miles, for instance.
Thi s idea is properly known as the
Kevitz System. (Wisnewski's repertoire
book on ... llJc6 was all about the
Nimzowitsch and Chig orin Defences, in
which Black plays 1 .. dS or 2.. dS.)
. .
The Dark Knight System?
For reasons I will lay out, I will b e rec­
ommending a fianchetto of the king's
bishop if White resists the ...e7-eS ad­
vance. These are in fact the most com­
mon positions, and they have a much
different feel from a typical Kevitz Sys­
tem. Furthermore, I believe that the fi­
anchetto is a substantial improvement
over the commonly played moves, and
therefore a new name is in order. Since
it is a black knight venturing out from a
dark square to initiate a strategy of
dark-square control... need I go on ?
Does the Dark
Knight System work?
It works wonderfully, and in two ways.
Firstly, it can throw White on his own
devices as early as move one ! When
forced to improvise, even titled players
can play shockingly weak moves or ex­
pen d their time and energy in the
opening. Secon dly, the opening is fully
sound and playable against all calibre
of opposition . Black is fundamentally
okay so there will be no need to aban ­
don the repertoire just because your
opponents are no longer surprised.
Furth ermore, this book is intended to
leave you a step or two ahead of even
very well-prepared opponents.
I would like to take the opportunity
to say that, in general, the quality of
play in Dark Knight variations has been
low for both colours. In many common
positions, as early as moves six, five,
four, and three (!) the unquestionably
best moves have been rarely or never
7
The Dark Knig ht System
played! To some extent this is under­
standable in an "unorthodox" opening,
particularly for White, who has more
important things to worry about. It is
less understandable for BlacK, who can
hardly be surprised by his own open­
ing. However, from BlacK's standpoint,
this is highly correctable - and what
better opening to use than one in
which there is a long history of incom­
petence b y the opponent, even at the
GM level? I suggest that BlacK's practi­
cal results (which are by no means bad)
can be substantially improved.
Who plays it?
As far as I KnOW, nobody plays the DarK
Knight in exactly the fashion I will be
recommending, but many strong play­
ers use large parts of the repertoire,
and most of the bits and pieces have
been tested in high-level encounters.
In spite of the relative obscurity of
the Kevitz System, it should be noted
that it was one of Miles's regular
weapons against both 1 e4 and 1 d4,
and IM Zvonimir Mestrovic plays it fre­
quently - they each have hundreds of
1 ...l2Jc6 (with the idea of ... e7-es) games
to their credit.
It has also seen use by GMs Bogol­
jubow, MiKenas, Short, Hoi, Lazic, Veli­
mirovic, Benjamin, Gause1, Svidler,
Huang Thong Tu, Hort, A.5oKolov, Sul­
sKis, Godena, Izeta Txabarri, GulKo,
Klinger, Rogers, Olafsson, Tolnai,
Art.Minasian, Ubilava, Sadler, Anand,
Dizdarevic, Gonzales, Speelman, Mohr,
8
Bachmann, ZarnicKi, Gelashvl1i, Le KO,
Johansen, ShKuro, Rohde, Karpatchev,
ErmenKov, Bezgodov, and de facto GM
NiKolaevsKY - not to mention IMs Kjeld­
sen, CvetKovic, PrzewozniK, Vlassov,
Danailov, Tarlev, Barle (frequently),
Z.NiKolic, Vujadinovic, Wohl, Sommer­
bauer, Mascaro, MatiKozian, O'Donnell,
Eid, Ambrus, Kos, Bus, and presumably
many others that I have missed.
So, as we can see, not only are
strong players willing to play these po­
sitions (and against other strong play­
ers) they do so over and over, in some
cases without any expectation of sur­
prising their opponents. This says a lot
about the hidden consensus as to the
merits of the opening among those in
the Know.
Coverage
This is a repertoire booK, but I am not
adhering slavishly to the concept. Side­
lines for BlacK are presented if they are
useful or enlightening. One situation
that sometimes comes up is that a
main line, while objectively fine for
BlacK, offers very few winning chances.
In this case, I will try to offer an alter­
native which maKes it more practical to
play for a win, normaTIy with substan­
tial additional riSK (otherwise it would
have been chosen as the main line).
Transpositions to other openings
are obviously frequent, but I will not
abandon the reader just because we
have reached a position that happens
to be Known by a different name. I win

mention transpositions when available
and cover the transpositions that I rec­
ommend.
That being said. it is not simple to
fit a whole Black repertoire into one
volume. and decisions needed to be
made about what to devote space to.
Except for here. I win not waste space
expressing the wish that I had more
space. However. if certain positions
receive light treatment. this is gener­
ally the reason. When deciding what to
focus on. I weighed both frequency and
danger. only intentionally ignoring
White moves that are both rare and
weak. Besides. space aside. I see no
point bogging down the reader with
information he won't need.
Transpositions? Aargh!
Why would anyone want t o learn inde­
pendent Dark Knight and Kevitz posi­
tions when they are just going to have
to learn regular (transpositional) open­
ings on top of it? One part of the an­
swer is that a player may greatly enjoy
the non-transpositional positions. and
these are reached frequently. Another
important part is that White normaTIy
has to give up valuable options in order
to enter the transposition. For instance.
in the Pirc reached through the Dark
Knight System. White can only play the
Classical Variation which. though fairly
popular. is just not very challenging for
Black. Admittedly. Black's k night
reaches the slightly unusual square c6.
and does so unusually early. but I will
Introduction
demonstrate that this is not a problem.
With White's options limited and Black
committed to this sideline. the study
material is relatively small.
To continue. a player who plays 1 e4
e5 must typically learn the Ruy Lopez.
Two Knights. Scotch. King's Gambit.
Vienna. and other sidelines. Compared
to this. the Scotch reached via the Dark
Knight is a light workload. not particu­
larly dangerous. and not a popular
choice for White. Therefore. play the
Dark Knight System still.
Oh, the humanity!
I am admittedly human and. further­
more. fallible. but I will refrain from
continuously hedging in the tex t (e.g.
"If my analysis holds up, it seems to me
that perhaps Black may indeed have
the better practical chances. though
this idea is untried and further investi­
gation is needed"). If there are particu­
lar doubts about conclusions. the nor­
mal solution is not to express them.
but to rectify them.
Hopefully I am far less fallible with
the help of chess engines. especially
Houdini (whom I sometimes refer to
affectionately as "Mr. H"). Everything
presented is computer-checked. which
offers the reader substantial protection
when relying on the analysis. However.
I have only used long computer­
generated variations when absolutely
necessary; i.e. there are no relevant
human games to draw from. and the
positions aren't settling down into
9
The Dark Knight System
something that can be understood and
assessed. In other words, fairly often.
I am inevitably prone to error when I
quote statistics, or when I say that a
move is new. These statements are nec­
essarily based on games I have access to.
I will try to avoid saying, "according to
my database" every time, since that
should be taKen as a given. And I apolo­
gize in advance to the true originators
for such errors in attribution.
Untested? (*gasp*!)
In opening booKs, untested - or lightly
tested - moves are typically treated liKe
embarrassing relatives, introduced
quicKly for propriety's saKe and then
shuffled off to somewhere they won't
bother anyone. Admittedly it is far eas­
ier to discuss and analyse moves that
have been played repeatedly by GMs,
but ultimately moves need to stand on
their own merits, and we should not
shy away from a little worK in order to
play better chess. Besides, isn't it good
to catch our opponents unprepared?
As for enemy novelties, it is also
sensible to be ready, especially if it is a
computer novelty. After all, if "my"
Houdini says a move is best, my oppo­
nent's will too, and I will soon be fadng
this move at the board.
Who?
"I" is me, James Martin Schuyler. "You"
is you, the reader. "We" is not the royal
we - it is me and you, the reader. "Our"
opening is the DarK Knight System. I
10
am nobody i n particular. My qualifica­
tion for writing this booK is the fact
that I wrote the excellent booK you are
now holding in your hands.
Assessments
Chess writers wm often tell you that
your understanding of a position is
more important than the objective as­
sessment. No doubt this is true, but this
is not a good reason to be unconcerned
with assessments. An objectively poor
position will require a great deal of
preparation and understanding in order
to be worth playing. Also, what if your
opponent happens to understand it
too?! Wouldn't it be better to taKe the
time to understand a sound position
instead of a questionable one?
I will try to convey as much of my
understanding as pOSSible, but I am
also extremely concerned with the ob­
jective quality of the position (to the
extent that it is possible to determine
it). I do not want to place us one or two
inaccuracies away from an extremely
difficult position, nor do I want our op­
ponent to have the luxury of one or
two inaccuracies and still retain
chances for an advantage.
If you are not concerned with as­
sessments, simply ignore them, or cross
out the words and write in crayon,
"BlaCK is oKay". I do not find this useful,
but it is sufficient for many and true as
far as it goes - if the position were not
extremely playable, it would not be in
the book

Houd;n; is not the final arbiter of
anything - especially since it is people
who must play the positions - but he is
a far stronger player than I am, and he is
nothing if not objective, so when look­
ing for what passes for the truth, his
assessments carry considerable weight.
When his opinions have not made sense
to me, I have looked deeper. Typically, I
have become convinced, but sometimes
I a m able to convince him - rarely do we
continue to disagree.
Assessments in this book are in­
tended to apply to narrow ranges.
"Equal" corresponds to an advantage
for one player of no more than 0.09
pawns. "Comfortably equal " is the
more pleasant half of that range. IITiny
advantage", "tiny edge", or Ilslightly
better" is an advantage of 0.10 to 0.17
pawns, while linearly equal" would be a
similar disadvantag e. In most chess
works, such positions are simply la­
belled as equal, but I believe that there
is far too big a difference between
+0. 15 and -0.15 (two to three inaccura­
cies or even two to three tempi in many
positions) to let it go without mention.
A n "edg e" or liSmall advantage" is be­
tween 0.18 and 0.25 pawns. In other
works, such positions are often called
"approximately equal" or "tl=. I under­
stand that the style of assessment I am
using implies a degree of precision that
is difficult to attain, but I would rather
strive for precision and risk falling
short than strive for vagueness in the
hopes of evading criticism.
In troduction
I have not found it nec essary to in­
clude in the repertoire positions worse
than a quarter pawn disadvantage, but
they are not uncommon in the notes. I
have not tried to be as precise in my
descriptions of theoretically unimpor­
tant positions, but the unadorned
words "advantage" and Ilbetter" mean
approximately 0.26 to 0.39 pawns,
while 0.4 to 0.6 is a "comfortable ad­
vantage" and more would be Ilclearly
better" or some such, while more than
one pawn would be "nearly winning".
Personal history (with 1 .. .lLJc6)
My love affair with 1 ... ltJc6 goes back to
the late 180s, and my trusty old Batsford
Chess Openings. I had owned it for
some time before I ca me across a sing le
line by Bogoljubow concerning the
amazing 1 d4 ltJc6 ! ? Should White
"take the bait" and try to play a kind of
mirrored Alekhine's, a wonderfully in­
teresting position may be reached: 2 d5
ltJe5 3 f4 tbg6 4 e4 e5 5 f5 (n) 'Wh 4 + 6
�d2 �xe4(?) 7 fxg6 'li'xd5+ 8 �el
'it'xdl+ 9 �xdl hxg6.
11
Th e Dark Kn ight System
Although labelled as unclear, Black's
compensation seemed tremendous to
me, with three premium pawns and
the half-open h-file for a small knight.
My shoddy but practical analysis con­
firmed this: after 10 tLlc3 ? ! c6 11 tLlf3
f6 ! 12 �d3? ! tLle7 13 �d2 dS 14 �e2
tLlfS
12
White will be lucky t o survive, even
should he find a defence to lS ... e4 and
16 ...tLlg 3 17 :gl �c S. Thi s is, in fact,
what many players tend to do as
White. Importantly, after the correct 10
C4! I still preferred Black.
Two of Bogo's opponents were kind
enough to allow S .. .'.h4+! . Alas, after
26 years of 1 ... tLlc6, I have yet to bring
this variation to the board during a
tournament game. (And now I never
will. Even if White plays into it, I will be
obligated to correct Black's sixth move.
More on this in Chapter Three.) On th e
plus side, I have yet to encounter any
real opening difficulties against any
calibre of opposition.
Therefore, play the Dark Knight Sys­
tem !
 Weak Co l o u r Co m p lex
There is no way to play chess well while
adhering to a single idea, or even two
or three ideas - the g ame is far too
complicated - but I have noticed that,
in the Dark Knight System, on e concept
assumes far greater than normal im­
portance, and that is the notion of the
weak colour complex. In many of the
high -level games that Black wins, it is
by taking advantage of White's weak­
nesses on the dark squares. I would
assume that most readers are familiar
with the idea of a weak colour complex,
but since it is especially important in
the D KS, I will prattle on about it any­
way.
In some positions, a player is more
likely to have problems because of
weaknesses on a single colour. The con­
ditions:
1 Most of the player's pawns are
on a single colour (at least in a
certain area of the board). Natu­
rally, the weak colour complex
will occur on the opposite colour.
1 The player is missing the bishop
that he would need to guard the
weak colour.
1 The player's king is in the vicinity
of the weak squares.
1 The player's opponent still has
the bishop that can infiltrate on
the weak colour.
A player will usually experienc e
problems if three of the conditions are
met. Furthermore, if you notice two
conditions in your opponent's position,
it is worth seeing if you can ag gravate
his situation.
A few more observations:
1 Obviously, if a player's opponent
has no access to the /tweak"
squares, there is no weak colour
complex, regardless of what
other conditions are present.
1 A knight is a handy piece for the
invader to have, because it al­
lows him to extend the attack to
the other colour. (A knight sit­
ting on a weak dark square at­
tacks light squares).
1 A space advantage is no protec­
tion again st a weak colour com­
plex - it can even be a vulner­
ability.
13
The Dark Knight System
t Weak cobur complexes are com­
mo-n i-n fia-nchetto ope-ni-ngs - for
the oppo-ne-nt of the player that
fia-nchettos, that is. This is be­
cause of h ow each player is likely
to set up their paWTls, a-nd also
because the fia-nchettoed bishop
(t he o-ne that is likely to be i-nfil­
trati-ng the e-nemy weak-nesses)
14
has extra protectio-n agai-nst be­
i-ng traded off for a k-night.
Th e weak colour c omplex win -not
come up ofte-n m the theoretical sec­
tio-n because it does -not usually appear
m a full-bloWTl form u-ntil the middle­
game, but th e idea permeates the
g ames sectio-n.
 Sectio n On e
1 d4 ctJc6
The Dark Knight first caught my at­
tention as a defence to 1 d4. Although
playing it against 1 e4 requires very
little additional knowledge, Black may
find it useful that his 1 d4 opponents
are unlikely to know the ins and outs of
the Scotch and the Pirc. Besides, since it
is not so easy (for most of us) to meet 1
d4, it is espec ially nic e to find an effec­
tive defenc e. It even neutralizes the
London, Trompowsky, and other white
"easy" systems.
Apart from 2 e4 which is covered via
1 e4 in Section Two, White has three
main moves:
2 tLlf3 - Chapter One
2 c 4 - Chapter Two
2 dS - Chapter Three
Others:
a) 2 e3 eS (or 2 ... d6 and 3 ... g6, to keep
things interesting - White's pawn on
e3 makes a poor impression in this
King's Indian type of position) 3 tLlf3
sees White try to play a French with an
extra move, but this move order gives
Black a few good options. 3 ... e4 4 tLlfd2
fS S c4 tLlf6 6 tLlc 3 .i.e7 ! transposes to
line B2 in Ch apter Seven. Also possible
is 3 ... exd4 4 exd4 dS (an Exchange
French) to dry up the game. Instead, 3
c4 transposes to A2; whereas 3 dS is
schizophrenic nonsense: 3 ...tLlc e7 4 c4
d6 S tLlc3 fS 6 .i.d3 tLlf6 7 .i.c2 g6 gave
Black an extremely comfortable version
of a King's Indian in l Paasikangas
Tella-T.Lindqvist, Finnish Team Cham­
pionship 1996 (see G ame 1).
b) 2 tLlC3 eS will soon transpose, af­
ter 3 d5 tLlce7 4 e4 or 3 dxeS tLlxe s 4 e4,
15
Th e Dark Kn ight System
into positions considered in Chapter
Four (see 4 lbc3 in lines A an d B respec­
tively).
c) 2 c3 eS 3 e4 - see 3 c3 at the be­
ginning of Chapter Four.
d) 2 .tg S ? ! looks like an attempt to
outdo Black in the weird department,
though it does prevent ... e7-eS for now.
Black must not shy away from the bi­
zarre: 2 .. .f6 ! (breaking the pin an d chal­
lenging White to prove that the bishop
is well placed on the rim) 3 .th4 dS !
(this is not normal for us, but White
has given us space and time and cen­
tre, so we do not wish to be disturbed
by a belated d4-dS) 4 e3 (4 C4? ! es! S
dxeS .tb4+ 6 lbC3 d4 7 a3 .te7 s lbdS
fxes 9 lbxe7 lbgxe7 is just good for
Black) 4 ...lbh6! S lbc3 (S c4 eS or S .td3
es! 6 dxes lbxe s 7 lbc 3 c6 S lbf3 lbxd3+
9 'ii'x d3 lbfS) s ...lbfS 6 .tg 3 and we
have a few sound choices (6 ... e6, 6 ... g6),
but the point of our play is to pursue
the bishop, so let's g o!
6 ...h S ! ? 7 .te2 g6 S lbf3 .te6 (time to
take a break; S ... h4 9 .tf4 gs 10 lbxg s
fxg s 11 .th S+ Wd7 12 .txg s is pretty
16
dangerous) 9 .tf4 .tf7 10 h 4 a6 1 1 0-0
Vi'd7 an d 12 ... 0-0-0 is equal.
e) 2 f4 ! ? is another radical way to
prevent 2 ... eS. In the very limited prac­
tice, Black has done well with 2 ... dS 3
lbf3 .tg4 4 e3 f6 !?; e.g. S c4 es! 6 cxdS
'ii'x ds 7 lbc 3 .tb4 and White was al­
ready worse in C.Depasquale-A. Ker,
New Zealand Championship 2001;
while after S .tbS 'ifd6 6 0-0 a6 7 �xc6
�xc6 S c3 lbh 6 9 lbbd2 0-0-0 10 'ii'e l
.tfS 11 �e2 e6 12 :rel .te4 13 c4 .tb4
14 cxdS exds lS a3 .txf3 16 'ii'xf3 .txd2
17 .txd2 fS, Black h ad continuously
maintained his grip on e4 and went on
to win with his g ood knight versus bad
bishop in J.Vialatte-F.Giroux, Pari s 2006
(see Game 2). Of course, White also has
3 e3 .tfS 4 .td3 e6 ! S lbf3 lbf6 6 0-0,
and although White has no advantage,
he h as reached a Stonewall position
with which he is presumably comfort­
able.
Nothing wrong with any of that, but
taking the opponent out of his comfort
zone is one of the things this book i s
about, and for this purpose the brand­
spanking-new, I-just-made-it-up,
2 .. .fS ! ? fits the bill. No sane person
would play the Stonewall against the
Dutch - White gives up the e4-square
without getting the eS-square in re­
turn - so only 3 dS can be critical. But
will the Stonewall player be happy in
the resulting positions ? He may not
play 3 dS at all, but here's what h ap­
pens if he does: 3 dS lbb4! 4 a3 lba6 S
lbC3 lbf6 6 lbf3 lbc S ! 7 e3 (or 7 b4 lbce4
 1 d4 tLJc6

S i..b 2 as 9 b5 e6 10 dxe6 i.. C 5 ! 11

tLJxe4 tLJxe4 12 i..d4 d 6 13 e3 i.. xe6 14
i..d3 'ike7 with a bizarre position that
slightly favours Black) 7 ... e6 S dxe6
tLJxe6 (S ... d5 ! ?) 9 i.. d3 g6 10 0-0 d5 and
while Mr. H calls it equal, I would be far
more comfortable sitting behind the
black pieces.
Seeing how the black army con­
verges upon the e4-square, White may
rethink his decision to evict the tLJb4,
but leaving it there is not convenient Mr. H likes White here, but what do
either, since White must fortify the d5- you think? Without further ado, the
pawn and he can no longer play i.. d 3; not sane Stonewall Attack vs. Dark
e.g. 4 c4 tLJf6 5 tLJc 3 g 6 6 g 3 i.. g 7 7 i.. g 2 Knight Dutch, which I can't wait to see
0-0 S tLJf3 e6 is equal. happen in real life: 1 d4 tLJc6 2 f4 f5 !
Alternatively, White may try to ad­ (I'm giving this move an upgrade as of
dress the weak e4-square by placing a now) 3 tLJf3 e6 4 e3 tLJf6 5 i.. d3 b6! 6 0-0
black pawn there: 4 e4! ? fxe4 5 a3 (oth­ i.. b7 7 a3 (if White doesn't play this
erwise 5 ...c 6!) 5 ...tLJa6 6 tLJc 3 tLJf6 7 i.. e3 soon, ...tLJb4 is going to be extremely
c6 ! ? (7 ... g6 leads to more " normal" po­ annoying) 7 ... tLJe7 (7 ... g 6 !?) S c4 tLJcS !
sitions) S i.. xa6 bxa6 9 dxc6 d5 10 with ... i.. e7 ... 0-0 . . .tLJd6 coming, and a
ttJge2 e6. wonderfully fun equal position.

17
 C h a pte r One
1 d4 tZJc6 2 tZJf3
This is the most common reply, and a
very logical one for a 1 d4 player. White
stops Black's planned 2 ... e5, while try­
ing to maintain a familiar position
(unlike 2 d5).
2 d6
•••
There may be nothing wrong with
2 .. d5, but with 2 ... d6 Black continues to
. 3 ... e5 4 e4, which is now some sort of
fight for the e5-square. Also, after 2 ... d6
Black's !bc 6 will usually be presented
with an excuse to move soon, freeing
the c-pawn to join in the battle for the
centre. While the knight is not exactly
glued to the board in Chigorin-type
positions (i.e. after 2 ... d5), it may not
18
find a convenient opportunity to relo­
c ate for some time, leaving the c-pawn
out of play.
White h as:
A: 3 c419
B:3 dS 22
c: 3 i..f426
D: 3 g3 30
E:3�gS 33
Instead:
a) 3 e4 is covered via 1 e4 - see
Chapter Five.
b) 3 !bc 3 !bf6 4 e4 also reaches
Chapter Five.
c) 3 e3 can be met by 3 ... g6.
d) 3 h 3! ? may be a trick to induce
Philidor. 4 ... exd4 5 !bxd4 g 6 ! ? is not
really bad, but 3 ...!bf6 4 !bc 3 g 6 5 e4
will transpose into lines we are more
familiar with - see line Cl in Chapter
Five again.
e) 3 c3 was used to g ood effect in
B. Kurajica-Z.Mestrovic, Bosnian Team

Championship 2003, continuing 3 ... e5
4 e4 liJf6 5 i.. d 3 (5 i..b 5 ! ?) 5 ... i.. e 7 6
8bd2 and White went on to win. In­
stead, 3 ... liJf6 4 e4! g6 transposes to 4
c3 g6 at th e beginning of Ch apter Five.
A:. 3 c4 g6!
Until I started researching this book,
always played 3 ... e 5 ? ! here, but if
White follows up correctly, 4 d5! will
lead to unpleasant positions for Black
(though this is not widely known). The
text move is in keeping with the idea of
provoking d4-d5 while leaving the al­
hS diagonal open, a common theme in
the Dark Knight.
4 dS!
White should play this now, while
Black is mid-fianchetto, or it will be less
effective.
4 liJc3 i.. g 7 5 e4 i..g4 6 i..e3 e5 7 d5
�d4 S i..e2 i..xf3 9 i..xf3 cS is a com­
fortable variation of the Modern D e­
fence in which Black has outscored
White, though the game should be
equal after 10 dxc6 bxc6 11 0-0 liJe7 12
c S ! 0-0. Another option is 9 ... liJe7 ! ?, and
1 d4 liJc6 2 liJj3
if 10 i..xd4? ! exd4 ll liJe2 c S 12 dxc 6? !
i..xc6 13 liJxd4?! then 13 . . :�a5+ 1 4 Wfl
"c S regains the pawn with a clear ad­
vantage.
On 5 d5 Black could transpose to
our main line with 5 ... liJbS, though
5 ... liJe5 ! ? 6 liJxe 5 i..xe5 7 e4 liJf6 ! S i..d3
0-0 9 liJe2 ? ! (9 0-0 c 6! 10 h 3 ! improves,
when White retains an edge) 9 ... liJd7 i s
more fun . This was actually played in
M.Tratar-M.5rebrnic, Slovenian Cham­
pionship 2010, and A.lpatov­
R.Antoniewski, G erman League 2011.
Black h as equalized, reached a fascinat­
ing new position and, furthermore,
went on to win both times (see Games
3 and 4).
4 liJb8 S liJ c3 i..g 7 6 e4 liJf6 7 i..e 2 0-0
..•
This is now an obscure King's Indian
variation that can arise via 1 d4 liJf6 2
c4 g6 3 ltJc3 i.. g 7 4 e4 d6 5 liJf3 0-0 6
i..e2 liJc6 ! ? (6 ... es is "normal") 7 d5
liJbS. It is also .. .
Position One
White's position certainly is large
and, indeed, he has the advantage, but
19
The Dark Kn igh t System

Bl ack has counterplay on the dark We will need to look carefully at:
squares. One important thing to realize
is that both ... e7-eS and ...c 7-cS are poor Ai: 8 h320
at this stage in the game (though they A2: 80-021
are frequently played). Both moves take
squares away from Black that he will Ai: 8 h3 �bd7
enjoy using, while neither m ove puts
any pressure on White's centre, or
anywhere else for that matter. If that's
not clear enough, notice that ... e7-eS or
...c 7-cS will plac e Black two tempi down
in a main line King's Indian ( Petrosian
System) or Benoni. Meanwhile, ...c7-c6
or ... e7-e6 actually does pressure
White's centre while reserving the cS­
and eS-squares for Black's piec es. Even
so, there's no rush, sinc e White has no
convenient pawn break. Because Black With h2-h 3 included, it becomes
has g ood control over when and how possible to start considering ... e7-es,
much the board opens up, it makes though the immediate 8 ... eS 9 .i.gs has
sense for him to wait for a particularly scored 100% for White.
g ood opportunity. 9 .i.e3
Sinc e Black's play is on the dark Logically, White takes measures
squares, White has a space advantage, against 9 .. �CS.
.

and the game is not open, the trade 9...�c5!

... .i.g4xf3 suggests itself - but in spite
of the favourable factors, it is still no
bargain to part with the bishop pair. In
any case, Black is not the one in control
of the trade, since White could have
played h 2-h3 at any point going back to
move five. Indeed, GM Neverov and IM
Bonin did choose 5 h3 when confronted
with this situation. Personally, that
would please me, as it costs White a
tempo and saves me from a difficult
decision. If you prefer ... .i.g4, play it as But we play it anyway! White's dark
soon as White plays .i.e2. bishop is way too valuable to trade.

20

10 "c2
In Y.Balashov-G. Kuzmin, USSR
::: h ampionship, Vilnius 1980, White
:Jlayed 10 e5 tLlfd7 ( 10 ... tLlfe4! ?) 11 exd6
exd6 12 .i.d4 tLlf6 and the game soon
oetered out to a draw (see G ame 5).
10 .i.xC 5 ? ! dxc5 11 e5?! (11 0-0 e5!)
's no good because of 11...tLld7 12 e6
-=Xe6 13 dxe6 tLle5 14 'ii'b 3 tLlc6 15 0-0
�xe6 16 'ifxb7 tLld4 with the better
�ame for Black.
10 as 11 0-0 tLlfd7!
...
Black's main idea i s to play 12 . . .e 5 .
Th i s resembles Yates's plan in line D of
�his chapter. Black improves his pieces
and clamps down on the dark squares
�fore engaging in pawn play. White
should also be concerned about the
!)Ositional damage he could suffer after
for instance) 12 l:tadl .i.xc3 ! ? 13 bXC3,
:hough this is obviously a double­
edged sword.
Now if 12 tLld4 e5 13 tLldb 5 f5 with
good play, or 12 .i.d4 .i.xd4 13 tLlxd4 e5
14 dxe6 fxe6 15 :adl e5 16 tLlf3 b6 17
:fel .i.b7 18 .i.fl wg 7 19 l:te3 tLle6
with equality. Black may consider
1 d4 tLlc6 2 tLlf3
...lIxf3 at some point. As in many simi­
lar cases, White hurts his own position
by trading off Black's fianchettoed
bishop.
A2: 8 0-0 as
8 ....i.g4! ? is a logical alternative.
9 tLld4
9 tLlel is also played, when 9 ... tLla6
10 tLld3 b6 11 .i.e3 tLlC5 or 11 ... tLld7 is
similar to the main line. 9 h 3 is seen as
well, with a likely transposition to Al;
e.g. after 9 ...tLlfd7 10 .i.e3 tLla6 and
11. ..tLlac5.
9 tLla6
...
9 ... e5?! is more common, but this
move of Stefano Rosselli del Turco's is
the most accurate. Piece play before
pawn play in this variation !
10 .i.e3 tLlcS 11 f3 eS ! 12 tLlb3 tLlfd7
The two-time Italian Ch ampion's
12 ...b6 13 -.d2 tLle8 ?! 14 1!ael f5 1 5
exf5 tLlxb3 1 6 axb3 gxf5 17 f4 was not
very successful in S.Flohr-S.Rosselli del
Turco, Zurich 1934.
13 'ii'd 2 b6
White may be slightly better be-
21
Th e Dark Kn ight System
cause of his extra space, but he has no
pawn play.
Meanwhile, we can advance ...f7-fS
at any time; we are also threatening
14 ...ltJxb3 15 axb3 ltJc s, which will force
White to play an awkward m ove to de­
fend the b-pawn (or else give up his
good bishop).
B: 3 d5
The time t o play this was move two.
Now White will h ave great difficulty
finding an advantage.
3 ...ltJe5
The most accurate reply, though there
is a tendency for the positions to dry up
22
early. I f Black must play for a win,
3 ... ltJb8 is better, probably transposing
to line A above.
4 ltJxe5
If White delays this capture, pre­
sumably with 4 e4, Black plays 4 ...ltJf6 5
ltJc3 ltJxf3 6 'ifxf3 g6 and the fianchet­
toed bishop will enjoy its open diago­
nal, while White's queen will soon need
to move again (but to no particular
effect); as for example in R.Fi scher­
lSchuyler, Richmond 2008 (see Game
6).
Sometimes in blitz White tries to
avoid trading my "problem" knight by
playing 4 ltJd4, but White's knight is no
better off after 4...cS. If 5 dxc6 ltJxc6 6
e4, we are in a Sicilian with Black hav­
ing slipped in an extra move while
White's attention was diverted. Focus
on the board, White!
4...dxe5 5 e4
This is nearly automatic, but not
obligatory - in L.Altounian-1Schuyler,
Las Vegas 2008, White played 5 c4 e6 6
ltJC3 ltJf6 7 g 3 ! ? (see Game 7).
5 ...ltJf6!
 1 d4 tbc6 2 tbf3

The move Black wants and needs to White is doing well with 7 f4! .
play i s ... e7-e6, but after s . . .e 6 ? ? Black i s
already lost! Then 6 .i.bs+! .i.d7 7 dxe6 !
,j.xbs 8 'ii'h s ! was brutal i n lBonin­
J.5chuyler, New York 1988 (see Game
8). I shouldn't feel too badly, I suppose
- Mestrovic has made this blunder, and
Miles made it twice ! ! (Since Miles was
�iles, he actually lost neither game.)
White now has:

81: 6 ttJc323
82: 6 .i.bS+24 7 .i.bS+ .i.d7 8 dxe6 ..txbS!
Previously, both Jonathan Speelman
6 .i.d3 is less common and less logi­ and Emmanuel Bricard have tried
cal. H.Keskar-J.5chuyler, Norfolk 2008, 8 .. .fxe6, each securing a draw - though
continued 6 ... e6 7 c4 ..tc s 8 0-0 0-0 9 if you see the games, you may not be so
&3 exds 10 cxds, when I set about eager to repeat their methods.
blockading the d-pawn with 10 ... tbe8 Instead, after 8 ... ..txbs, White has
1 1 .i.e3 ..txe3 12 fxe3 ..td7 13 "f3 c S 14 several paths to the endgame, but al­
"g 3 f6 15 l:tadl tbd6 and Black is a bit most no chance of extracting anything
better already. The game concluded 16 from the position.
b3 bs 17 'ifh4 'ifas 18 :Cl b4 19 tba4
:ac8 20 l:tf3 c4 21 ..tfl cxb3 22 ':xc8
:Xc8 23 axb3 .i.xa4 24 bxa4 b3 25 'iVg4
b2 26 'iWe6+ �h8 27 :xf6 'iVd8 0-1.
Three years later, H.Keskar-
J.5chuyler, Hampton 2011, went 7 dxe6
l.xe6 with comfortable equality, albeit
with a long struggle for a win (see
Game 9).

81: 6 tbc3 e6!

This is still the move Black needs to
play - and now he actually can, as my
computer explained to me a few years
ago. This is of great theoretic al impor­
tance because following 6 ... a6 or 6 ... g6,
9 tbxbs
White should capture neither the
queen nor the f-pawn, though this is
not simple for him to figure out.
a) 9 'ii'xd8+ :xd8 10 tbxbs has been

23
Th e Dark Kn ight System
played four times, but it is slightly
weaker than the text. Then 9 .. .fxe6 ! 11
liJxC7+ wf7 12 liJbs :c8 ! (better than
12 ... liJxe4? !, as in I,Kreitner-H.5tenzel,
Long Island 1997) 13 liJC3 i..b4 14 i.. d2
i.. xc3 15 i.. xc3 liJxe4 16 i.. xe 5 ::tXC2 is
equal .
b) 9 exf7+ costs a move (as opposed
to waiting for .. .f7xe6) an d improves
Black's king position, so it is not a good
idea. 9 ... wxf7 10 liJxbs (or 10 'ii'xd8+
l:lxd8 11 liJxbs liJxe4 12 f3 a6 13 fxes
axbs and Black has slightly the better
of the probable draw) 10 .. .'iVxdl+ 11
wxdl liJxe4 12 We2 c6 13 liJc3 liJxc3 1 4
bxc3 is roughly equal again.
9 'ii'xd1+ 10 Wxd1 0-0-0+ 11 �e2 a6
•••
12 liJc3 i.. b4 13 liJdS liJxdS 14 exdS
:XdS 1S exf7
White has a nominal edge because
of the isolated e-pawn, but the game is
all but drawn. Still, let's be careful and
avoid any mishaps by taking the a7-gl
diagonal immediately before White's
bishop entrenches itself on e3: i.e.
l s ...i..c s 16 i.. e 3 i.. x e3 17 Wxe3 J:lf8 18
lIadl, and now let's centralize the king:
24
18 .. ':d6 19 <iit e4 ::txf7 20 f3 wd 7 - al­
though White can't win, he can try to
lose if he likes with 2 1 wxes l:[e7+ 22
Wf4 l::t e 2.
82: 6 i.. b S+ i.. d 7
6 ... liJd7?, as in E. Bukic -Z.Mestrovic,
Belgrade 1978, is an experiment that
should not be repeated. Had White
seen 7 'ifhs ! our hero would have
found himself a pawn down for noth­
ing.
7 'it'e2
7 'ii'd3 is rarely played: 7 ... a6 8 i.. xd7
'ii'xd7 9 0-0 ( D .Haessel-J .5chuyler, Paw­
tucket 2008, continued 9 liJc3 e6 10
i.. g s i..b4 11 0-0-0 0-0-0 12 f3 fle7 13
-'C4 h 6 with equality, though there
was still some play, and I went on to
win - see G ame 10) 9 ... e6 10 c4 i.. e 7 1 1
liJC3 i s obviously similar t o the main
line. White's queen has some extra op­
tions, but his d-pawn is pinned. These
differenc es are important enough to
change Black's best method of coun­
terplay: l1 ... bs! 12 �dl b4 13 liJe2 0-0
14 i..g s as with a slight edge for White.

One important point of Black's
queenside expansion is that it secures
the cs -square for his bishop, ensuring
that it won't get shut out of play (as it
would if White were allowed to seize
space on the queenside with a2-a3 and
b2-b4). White's knight has also been
taken out of contact with the impor­
tant ds-square. Notice that if White
ever plays dsxe6, and for some reason
Bl ack doesn't feel like recapturing with
a piece, ...f7xe6 is positionally sound
because the f-file is valuable and the
e6-pawn controls critical squares.
Hold on ! Couldn't 12 ... b4 have been
prevented? Indeed, 12 a3 is possible for
White and not a bad move, but 1 1 ...bS
was not played with only 12 ...b4 in
mind. After 12 a3 0-0 13 l:tdl exds 14
cxdS lIfd8, Black's idea is to pl ay lS ...c S
and, i f White doesn't capture, 16. . .ttJe8
and 17 ... ttJd6. White's edge is tiny.
7 a6
.•.
I'm not crazy about 7 ... g 6 ! ? with
that silly pawn sitting on es, but it does
avoid spending a tempo on 7 ... a6, and
Mestrovic is 2-0 with it, which suggests
it is worth a try in a must-win game.
Check out D . Rasic-Z.Mestrovic, Croatian
Team Championship 2001, and J.Barle­
Z.Mestrovic, Slovenian Championship
1997, in the games section (Games 11
an d 12).
8 i.xd7 'i¥xd7 9 0-0
Somehow White has done well with
9 i.g s, though it is not a move that
should cause problems. In p.staniszew­
ski-H.Kaulfuss, D armstadt 1996, Black
1 d4 ttJc6 2 ttJj3
certainly had the right idea and was a
bit better following 9 ... e6 10 i.xf6 gxf6
11 dxe6 fxe6 (11 .. .'iVxe6 is also good) 12
O-O ?! (12 'ifh S+ 1ff7) 12 ... 0-0-0 13 �C4
itg8 14 ttJc3, even if he went on to lose
after the passive 14 ...ite8 ? ! (here
14 ...�b8 15 l:tfdl i.d6 16 b4 fS was bet­
ter).
For White, the most accurate con­
tinuation may be 10 ttJC3 i.b4 11 0-0
i.xc3 12 bXC3 exds 13 i.xf6 gxf6 14
itfdl VibS 15 'ii'f3 0-0-0, though he has
nothing to show for it - instead he
should have been accurate on move
three!
Alternatively, the untried 9 ... h6!? is
playable, although in this case 10 i.xf6
exf6 11 0-0 fS 12 exfs 'ii'xfs 13 ttJC3
i.d6 14 ttJe4 0-0 15 c4 is a tiny edge for
White.
9 e6 10 c4 i.e7!
...
Allowing Black to castle short. In­
stead 10 ... i.c S ? ! 11 i.g 5 spells trouble.
11 ttJc3 0-0 12 ':d1 exdS 13 cxdS
White's ds-pawn is currently an as­
set, and Black has two possible ways to
neutralize it. First, he can attack it di-
25
Th e Dark Kn ight System
rectly with ...c7-c6. This is definitely
worth considering in some similar po­
sitions, but it doesn't work so well here;
e.g. 13 ... c6?! 14 ..tg s ! �fd8 15 ..txf6
..txf6 16 dxc6 .-xc6 and the eternal
knight sets in with 17 ttJdS.
If Black tries to prepare thi s with
... h7-h6 (in fact not a bad move at all) it
is unlikely he will be fully ready for
14 ... c6 anyway; e.g. 13 ...h 6 14 a3 (14
..te3 ttJg4!) 14 ... c6 15 ..te3 cxds 16
ttJxds ttJxds 17 :txds 'ii'e6 18 l:tadl and
while Black's position is playable, it is
very dull, and White has a small but
clear advantage.
The second plan is very appropriate
here, which is a timely ... ttJe8 and
...ttJd6. This idea is useful in many
variations of the Dark Knight, but espe­
cially the ones starting with 3 ds. The
point of the knight transfer is fourfold:
the knight is safe, as White's own d­
pawn shields it from attack; the knight
i s active - centralized and controlling
the important e4-, fs-, c4-, and b7-
squares; the knight blockades the
strong d-pawn, so Black does not need
to worry about an eventual ds-d6 by
White; and finally, the knight has
cleared itself from the f-file, so it is now
possible (and usually desirable) for
Black to play -.f7-fs. Coincidentally, the
.. .f7-f5 break also does (at least) four
things: frees Black's rook(s), isolates
White's d-pawn (or pressures White's
e-pawn), clears the second rank for
easy defence of the c7- and g 7-pawns,
and gains space.
26
Therefore, 13 .. �ad8 1 4 ..tg 5 ttJe8 1 5
..txe7 'ii'x e7 1 6 llacl ttJd6 1 7 b 3 fs and
with all Black has accomplished, per­
haps White should resign? Unfortu­
nately, chess is not quite that simple,
but Black can now start fighting on
equal terms.
c: 3 ..tf4
White plays the London System,
which stops the ... e7-e5 break for now,
but the bishop bites a granite pawn on
d6.
3 ttJf6!
...
To be honest, I've always played
3 ... ..tg4?! here, intending to force
through ... e7-e5 one way or another (4
c4 e 5 ! ; 4 e3 e 5 ! ; or 4 ttJbd2 ttJxd4! 5
ttJxd4 es ). However, in researching this
book, I found that 4 d5 ! ttJb8 gives
White a large advantag e. Although this
h as only been played once in my data­
base and never against me, I believe it
is a bad idea to play moves one knows
to be poor, however unlikely it might
be to encounter the refutation. This
kind of "hope chess" is bad for one's
 1 d4 liJc6 2 liJf3

confidence and psychology - one is no 5 .i.e2 .i. g7

longer in control of the game.
As for 3 ....i.g4?!, it is not actually sur­
prising that this move is questionable -
there are virtually no cases in the Dark
Knight where the bishop goes to g4
early. With Black's pawns set up on dark
squares, it is costly to trade the light­
squared bishop for a knight, and the
bishop rarely has anywhere decent to
retreat to. Should you wish to ignore my
advice and take your chances, I will
point out that 4 dS ! is White's only good White has:
move, and it goes so far against the
grain for a typical London System player Cl: 6 0-0 2 7
that you, too, may never encounter it. C2: 6 h 3 28
4 e3
If 4 dS?!, then 4 ... e S ! 5 dxc6 exf4 6 Cl: 6 0-0 0-0 7 h3
cxb7 .i.xb7 7 'iid4 dS ! 8 'iixf4 liJe4! 9 c3
l.d6 10 'ii'Cl 0-0 11 e3 cS and, with
12 ... d4! coming, White's tiny material
advantage does not make up for all of
his pathetic grovelling. Instead, S .i.g s
�7 6 c4 (or 6 liJC3 h 6 7 .i.xf6 gxf6 8 e4
a6 9 .i.d3 fS) 6 ... liJe4! 7 .i.d2 (7 .i.h4?!
c6!) 7 ... g6 is equal.
4 g6
•••

4 ... liJh S has done well in practice;

for instance, S .i.g s h6 6 .i.h4 gs 7 .i.g3
�xg 3 8 hxg 3 .i.g7 reaches a position
Black is normally happy to have even
with a tempo less. However, White
once again h as the annoying novelty 5
dS! liJxf4 6 exf4 liJb8 7 liJC3 and Black's
knig ht excursions have placed him too
far behind in development, a situation
he will have trouble fixing; e.g. 7 ... g6 8
'ii'd4! .
Black is finally ready to play 7 ... liJh S,
so White prepares a retreat, but this is
still a slow move. Instead:
a) 7 c3 liJh S 8 .i.g s h6 9 .i.h4 gs 10
liJel liJf6 11 .i.g3 liJe4 12 liJd2 liJxg 3 13
hxg 3 eS 14 liJd3 .i.e6 15 .i. f3 dS is
equal. Of course, White wasn't playing
for an advantage anyway, but this i sn 't
the position he wanted either.

27
Th e Dark Kn ight System

b) 7 c4 es! S ..tg s (or S dxes lLJe4 9 sponse, and Black i s comfortably equal.
h 3 dxes 10 .i.h2 .i.fS with an edge) 9 l:e8
••.

S ... h6 9 ..th4 gS 10 .i.g 3 exd4 11 lLJxd4

lLJxd4 12 exd4 lLJe4 13 lLJC3 J:teS and
Black is a little better.
7 es!
...

After this unplayed novelty, White is

t he one who must be careful to keep
the balance. Furthermore, in doing so,
he may not be able to maintain his
typically comfortable London structure.
8 .i.h2
S dxes gives up more of the centre
for no gain. Black's point is S ... lLJe4! and Simple chess! Black has no problems
White's extra e-pawn is pinned to the at all. Moves worth considering in the
b2-pawn. 9 'ii'dS ? ! dxes 10 'iVxe4 exf4 near future are ....i.fS ( ... .i.e6), ... lLJe4
11 lLJc3 fxe3 just makes matters worse ( ...lLJdS), ... .i.h6, ... 'iWd7, ...:tbS, ...b7-bS. If
for White. lLJbd2, then ... a7-aS is a good idea,
8 exd4!
••• while if lLJC3, then ... a7-a6 and ... b7-bS
A second annoyance for th e London is effective.
System player, who is hoping for ... eS­
e4 at some point, after which White C2: 6 h3 0-0 7 c3
h as excellent long-term prospects be­
cause of the h2-bishop, which makes it
h ard for Black to generate meaningful
play on the kingside, whereas White
has no difficulties making progress on
the queen side, due in large part to that
same piece. It is not h ard to see why
that bishop appeals to so many players!
However, we will be sure to maintain
control over it.
g exd4
9 lLJxd4 may be slightly preferable This slows White's queenside play
(because the f3-square will be useful to a crawl, but it does avoid problems
for the e2-bishop), but White players on the long diagonal.
do not seem to consider such m oves. In 7 c4? ! is a bad idea. Black was al­
any case, 9 . .lIeS is still the correct re-
. ready better following 7 ... eS! S .i.h2

28

exd4 9 exd4 lbe4! in K.Gunasekaran­
K.Akshayraj, Dhaka 2005, and went on
to win after 10 lbc3 ? (but if 10 .i.f4!
:e8 11 0-0 g S ! 12 .i.e3 g4! 13 hxg4
8g 3! 14 :el lbxe2 15 l:txe2 .i.xg4 with
a big edge to Black) 10 ... lbg S ! 11 0-0
�3+ 12 .i.xf3 lbxd4 with a free pawn.
7 tbd7
..•
I shall soon try the untried 7 ... as ! ? I
do like the idea of avoiding the ... e7-eS
break for the time being, as it makes
contact with the enemy where he is
already fortified. Black's plan is 8 0-0 a4
9 tbbd2 a3 10 b4 lbdS 11 'ii'c l lbxf4 12
exf4 - equal according to my com­
puter, but this does not even vaguely
resemble the position White was hop­
ing to play. Notice how we have rid
ourselves of the London bishop, while
starting to soften the long diagonal for
our own dark-squared bishop.
Naturally, White could stop the a­
pawn with 8 a4, in which case we go to
plan B: 8 ... lbdS 9 .i.h2 eS 10 0-0 exd4
(the mini-operation succeeds - White
must give up the b4- or f4-square) 11
exd4 .i.h6 ! 12 I:tel :e8 13 lbbd2 lbf4 14
1 d4 lbc6 2 lbf3
.i.xf4 .i.xf4 1 5 'itb3 dS 16 .i.d3 :xdl 17
:xdl lbe7 and although White is n ot
worse, he is once again without his fa­
vourite bishop. I've played the London
System quite a lot, so I know exactly
how annoying these ideas can be.
8 0-0 e5 9 .i.h2 'ii'e 7
This is the main line - which shows
that every once in a while a main line is
actually good! Interestingly enough,
Houdini prefers White, but extensive
practice shows the opposite. Black
normally continues with 10 .. .fS,
l1 ..sPhS, and then looks for a g ood op­
portunity to shut out White's London
bishop with .. .fS-f4!. Perhaps Mr. H is
underestimating the problem of the
h2-bishop. Sometimes Black prepares
.. .fS-f4 with ... eS-e4 and ... g6-g s (or just
... g6-g S). Meanwhile, White shoves the
a- and b-pawns. Sooner or later the c6-
knight gets kicked and usually re­
routes itself to the f7-square via dS.
These ideas come to life in
P.B. Pedersen-D.Bekker Jensen, Danish
Team Championship 2008, and
R.Valenti-V.Tkachiev, Corsica (rapid)
29
Th e Dark Kn ight System
1997 (see Games 13 and 14); whereas
V.Golod-E.5utovsky, Netanya (rapid)
2009 (Game is) shows Black, a strong
GM, fail utterly to contain th e London
bishop.
0: 3 g3
A subtle (read: boring) move.
Mestrovic has responded 3 ... e s ! ? here
five times, drawing all corners from
expert to GM. This is a good bet for
Black theoretically, since the endgame
after 4 dxeS ttJxes S ttJxes dxeS 6 'ii'xdS
�xdS gives only a tiny edge for White.
Perhaps your opponent will turn away
from this Mutually Assured Dullness,
but I would as soon not give him the
opportunity. Thus, the usual solution:
3 g611
•••
When we must look at:
01: 4 dS 30
02: 4 i..g2 31
01: 4 dS
If White is going to kick the knight,
he should do it now - though, accord-
30
i n g to my database, 4 dS has never ac­
tually been played.
4...ttJb8
I have selected this retreat over
other options in many variations, and
the more I think about it, the more I
like it. Like General MacArthur, the
knight sh all return, likely settling on
the newly soft cs-square. Incidentally,
if ... ttJb4 is played, it is with a similar
idea: ... a7-as, ... ttJa6 and ...ttJcS.
S i..g 2 i.. g7 6 0-0 ttJf6
I also like 6 ...e S ! ? 7 dxe6 (7 e4 ttJd7 S
c4 ttJe7 is likely to transpose to D2; e.g.
9 ttJC3 0-0) 7 .. .fxe6 S e4 ttJh 6 ! 9 c4 ttJf7.
7 c4 0-0 8 ttJc3
This line was topical in the 1920s (!)
with Frederick Yates seen frequently
behind the black pieces against the
best players of his day, while Alekhine
and Grunfeld championed White. Rich­
ard Reti played both colours. Yates was
able to defeat RHi, Kmoch, and
Alekhine, the last of these games win­
ning a brilliancy prize at Carlsbad 1923.
8...ttJbd7
The most common move, S ... es?!,
 1 d 4 llJc6 2 llJ/3

didn't work in the 1920s and still

doesn't work in the 21st century. Even
with the centre closed, Black will lose
too many tempi with his knights in
order to play .. .f7-fS. Another common
move, 8 ... cS?!, is equally illogical and
unsuccessful.

02: 4 .i.g2 .i.g7 5 0-0

The board resembles Position One,

with White's fianchetto not particu­
larly helpful to him. Black should de­
velop and establish his pieces on the
dark squares before initiating pawn
play.
g h3
Prophylaxis against ...llJg4 and
... ttJges.
g... a5 10 .i.e3
Or 10 e4llJcs 11 'iVC2 and now, with
the knight anchored on the cs-square,
Black is ready for 11 ... eS (12 dxe6 .i.xe6
= ) 12 ... llJfd7 and 13 .. .fS .
10 ..llJc5
. 5 ...llJf6 6 c4 0-0 7llJC3
Black intends 11 ...llJfe4 with near
equality. 11 llJd4 .i.d7 transposes to
H . Kmoch -F.Yates, Hastings 1927/28, a
beautiful demonstration by Black of
how to build an attack (see Game 16).
S dS is playable now and at any
point, but it was most forcing on move
four, when Black was obligated to play
...llJb8. Now Black also has s ...llJes,
S ... llJb4! ?, and s ...llJas ( ! ) as in
A.Galliamova-M. Krasenkow, Koszalin
1997 (see Game 17), though transpos­
ing to Dl with S ...llJb8 is obviously sim­
plest.
By transposition we have reached
the fi anchetto variation of the King's
Indian Defence. There is a n obscure but
logical sideline for Black that has been
played successfully by strong players.

31
Th e Dark Knig h t System
7 ...th d7!?
This prepares both ...e7-eS and ...f7-
fS, and holds up White's c4-cS break.
The immediate 7 ... eS is frequently
played, but White has done very well
with 8 dS the7 9 CS!.
8 e4
8 dS thas (or 8 ...thces 9 thxes thxes
10 'ii'b 3 and now 10 ... CS, 10 ... aS or
10 ...b6 is level) 9 'ifa4 c6 10 .i. g s thcs 11
'iVb4 tha6 12 'ifa4 thcS is a draw.
8 ...e5
9 d5
These other moves are almost as
common:
a) 9 .i.gs .i.f6! ? (9 ...f6) 10 .i.xf6 thxf6
32
11 h3 (11 dS thb8!) 11 ...l:te8 12 a3 (12 dS
thb8!) was A.Kotov-A.Lein, USSR Team
Championship 1962; then 12 . . exd4 12
.
thxd4 a6 is a tiny edge for White.
b) 9 .i.e3 ? ! exd4 10 thxd4 thdes 1 1
thxc6 bxc6 1 2 'ii'a4 c S and Black is bet­
ter; e.g. 13 thds c6 14 thC3 :tb8 ! 15
%ladl (15 b3 thd3 !) lS .. J�b4 16 "'C2
thxC4 with a big advantage.
9 th e7 10 th el f5 11 th d3 h6 12 f4
•.•
This is not a very good move, but it's
what White has been playing for. Black
must try to prove that White's position
is overextended.
12 exf4 13 thxf4
.•.
13 .i.xf4 thb6 ! 14 'ii'b 3 fxe4 1 5 thxe4
thfS is equal.
13 thes 14 'ifb3 fxe4 15 th xe4
...
15 .i.xe4 g s 16 the6 :xfl+ 17 Wxfl
.i.xe6 18 dxe6 c6 gives Black the advan­
tage, since if 19 'it'xb7?! thxC4 20 .i.xc6 ?
Itb8 2 1 'iVa6 'iff8+ 22 Wg2 .i.d4, White's
exposed king loses him the game.
15 ... g5 16 th h 5 lhfl+ 17 .i.xfl .i.h8
Things are still complicated, but
Black is slightly better due t o White's
looser position.

E: 3 i..g 5?!
Preventing ... e7-e5 for now, this
move has given Black plenty of practi­
cal problems, but objectively it is poor
to allow Black to play ... h 7-h 6 and ... g7-
g 5 for free with White's knight already
committed to f3.
3 h6 4 i.. h4
...
4 i..f4 g5 5 i.. g 3 comes to the same
thing; 5 i.. d 2 g4 wins the d-pawn;
while after 5 i.. C l White may die of
shame. As a matter of fact, my en gine
recommends both 5 i.. cl and 4 i.. c l,
which certainly makes clear what it
thinks of this whole 3 i.. g 5 fiasco.
4 . g5 5 ..tg3 g4!
..
Dare I say it? This logical move is
another strong, unplayed novelty
which fully turns the tables. White has
1 d 4 ttJc6 2 ttJj3
deliberately provoked these moves,
only to find that his knight has no­
where to go. He will also find his i.. g 3
awkwardly placed.
6 ttJg1!
Limiting the damage. Others are
clearly worse.
a) 6 d5 gxf3 7 dxc6 bxc6 8 exf3 �b8 !
with some advantage t o Black i n a
complicated position.
b) 6 ttJh4?! e6! threatens 7 ...i.. e 7,
trapping the knight, which explains the
following contortions: 7 h 3 h 5 8 e4 i.. e 7
9 d5 i..x h4 10 i..x h4 1fxh4 11 dxc6 g 3
1 2 f3 bxc6 and Black pockets a pawn.
6 ... i..g 7 7 e3 h5 8 h3 h4 9 i.. h 2 ttJh6 10
hxg4 ttJxg4 11 ttJC3 e5
Black has an edge in this bizarre po-
sition.
33
 Cha pte r Two
1 d4 tLJc6 2 c4
T o my mind, this is already a lax move.
White puts up no resistance to Black's
logical follow-up. Presumably some
players are hoping for a Chig orin (after
2 ... d5).
2 eS
•••
3 dS
Instead:
a) The lame 3 e3 offers Black a few
methods: 3 ... exd4 4 exd4 dS 5 lbf3 lbf6
is an equal Exchange French position.
3 ... d5 ! ? is a kind of Chigorin/Albin
which has been played repeatedly by
many GMs - with poor results, how­
ever.
34
Personally, I like 3 ... i..b4+ 4 lbc3
i.. x c3 5 bXC3 d6, playing a kind of
Nimzo-Indian with ... e7-e5 in one go.
After 6 i.. d 3 f5 ! or 6 lbf3 e4 7 lbd2 f5,
the game can also be thought of as a
reversed Grand Prix Attack (cf 1 e4 cS 2
lbc3 lbc6 3 f4 e6 4 lbf3 d5 5 i..b 5). Of
course, 4 i..d2 is also possible: 4... exd4 5
i..xb4 (5 exd4? 'ii'e 7+! wins the d-pawn)
5 ... lbxb4 6 exd4 lbf6 (not 6 ... 'ii'e 7+ 7
i.. e2 'iVe4? ! 8 �fl ! and Black's queen is
worse than White's king) 7 lbf3 d5 is at
least equal - the exchange of bishops
will help Black in the coming isolated
queen pawn position. If instead 7 a3
lbc6 8 d5 ?!, then 8 . .'ife7+! 9 i.. e2 lbe5
.
and Black is better, as White must fig­
ure out some way to develop an d
guard the c4-pawn.
b) 3 dxe5 lbxe5 (3 ... d6 ! ?) cannot be
dang erous either. Black's position re­
sembles a Budapest Gambit, but with­
out any of the inconvenience normally
associated with recovering the e-pawn;
e.g. 4 e3 lbf6 5 lbC3 i..b 4 6 i..d2 0-0 7
i.. e2 c6 ! ? was fine and worked out well

for Black in R.Aghasaryan-A.Chibukh­
chian, Kajaran 2011 (see Game 18). And
4 e4? ! is an especially bad idea: after
4... �cS White is already worse and
must be very careful; e.g. S �e2 'ifh4!
or S 4.Jf3 4.Jg4 6 4.Jd4 dS ! ? 7 cxdS 'iWf6 8
.1e3 4.Jxe3 9 fxe3 'ifh4+ and White will
be needing both his chess resources
and his sense of humour as he plays 10
�d2.
c) 3 4.Jf3 transposes to 1 c4 4.Jc6 2
!2lf3 es 3 d4, covered at the beginning
of Chapter Seven .
3 �b4+!
•• •
3 ...4.Jce7 is played 90% of the time,
and with excellent results, but if White
really understands what's going on,
Black will be forced to play positions I
cannot recommend; e.g. 4 4.Jf3 ! 4.Jg6 S
h4! �b4+ 6 4.Jbd2 hS 7 93 4.Jf6 8 4.Jg S ! ?
and Black doesn't have much to look
forward to - White has all the squares
and will soon have the bishop pair unless
Black makes a pathetic retreat Black also
needs to worry (after S 4.JC3 4.Jf6) about
an of the "dynamic" tries mentioned in
Richard Palliser's 200S book on the
Tango, some of which pose questions to
which Black has yet to find answers.
(Whenever Black reaches the Tango via
the Dark Knight, White is already com­
mitted with a pawn on e4, which takes
away all " dynamic" tries, leaving "classi­
cal" tries, which Palliser rightly considers
to be pleasant for Black.)
Attempts to transpose to a King's
Indian with 4 4.Jf3 d6 do not bring hap­
piness either, because (to make a long
1 d4 4.Jc6 2 c4
story short) White gets to break on the
queen side early with 6 cS or 7 cS.
Instead, with 3 ... �b4+, Black devel­
ops his bishop before it gets obstructed
by ... 4.Je7 or ... d7-d6, incidentally solv­
ing his space issues an d defusing
White's h2-h4-h S ideas. What could be
more logical ?
White blocks with :
A: 4 .i. d 2 36
B: 4 4.Jd2 40
4 4.JC3 ?! is legal and it doesn't lose
material. So much for its positive
points. 4 ... 4.Jce7 S 'Wli'C2 (otherwise
S ... �xC3+) S ... aS 6 a3 �xc3+ 7 'iWxC3 d6
8 e4 fS gives Black a comfortable posi­
tion typical of this chapter. Our devel­
opment is simple and we can consider
clamping down on the queenside with
... as-a4 when we feel we can spare the
time. 9 Vig 3 4.Jg6 10 exfS .i.xfS 11 h4
Vie7 12 4.Je2 4.Jf6 13 hS 4.Jf8 14 4.JC3
4.J8d7 lS Vig4 �c2 ! might make White
feel like a tough guy, but it does not
lead to an advantage.
35
Th e Dark Knight System
A: 4 .i.d2 .i.xd2+
5 'ii'x d2
Altern atively, s llJxd2 llJce7 6 d6 ! ? (or
6 e4 d6 7 i.. d3 llJf6 ! - in this particular
position it is too costly to play 7 .. .fS?!,
activating both White's i..d3 and his
ridiculous llJd2; instead, Black plans
...0-0, ...llJg6, ... 'fIe7, ... llJh S, and/or ... a7-
as, ... b7-b6, ... .i.d7) 6 ... cxd6 7 llJe4 'Was+
8 'iWd2 'ii' xd2 9 �xd2 llJfS 10 g4 llJh4 11
llJxd6+ cj;e7 12 cS b6 13 b4 llJf6 14 llJf3
llJxf3+ 15 exf3 llJe8 is equal.
5 llJ ce7
.••
A1: 6 d61? 36
A2: 6 llJC3 38
36
Here 6 'fig s llJg6 7 'ii'xd8 �xd8 8
llJc3 d6 is just equal, though not with­
out play.
A1: 6 d6!?
6 cxd6
•..
A.Hoffman-A.Fernandez, Mar del
Plata 1996, went 6 ... llJc6 ? ! 7 llJc3 cxd6 8
llJbs llJf6 9 llJxd6+, and gives a good
example of what Black must avoid (see
G ame 19).
7 'ii'x d6
7 llJC3 llJf6 (or 7 ... dS 8 llJxdS llJxdS 9
'fIxds 'ii'e 7!) 8 llJf3 dS 9 cxdS (or 9 llJxes
d6 10 llJf3 .i.e6) 9 ... d6 10 e4 0-0 is no
problem for Black.
7 llJf61 S llJc3
. .•
8 'ii'x es frees Black's d-pawn and
does nothing to address White's devel­
opment - he is still four ( ! ) moves away
from castling kingside and the queen­
side is not a safe place: 8 ...0-0 9 llJc3 dS !
10 cxdS (10 e3 .i.e6 11 llJf3 llJg6 12
'ii'd4 dXC4 13 'fIxd8 l:axd8 is equal; or
10 llJf3 llJc6 11 'Wf4 'Was 12 cxdS llJxds
13 'ii'd2 llJxC3 ! 14 'ii'x C3 llJb4 15 llJd4
l:d8 with considerable pressure; or 10

0-0-0 �d7 ! 11 ltJf3 :c8 12 cxdS? :e8 13
'i'd4 ltJexds 14 �bl ltJxC3+ 15 bXC3
!!Je4, winning) 10 ...ltJexds 11 ltJxdS
!!JxdS 12 a3 (this sad move is necessary
to prevent 12 .. .'it'as+; castling just loses
after 12 O-O-On �e6, with 13 ... llc8+ 14
�bl ltJc3+ coming) 12 ...11e8 13 'ii'd4
(after thirteen moves, White's kingside
is hilarious - the most plausible expla­
nation is that he rolled very bad dice in
his Chaturanga game) 13 ... b S ! 14 ltJf3
lob7 15 e3 'ii'C7 16 �e2 llad8 gives
Black plenty for the pawn.
8 0-0
•••
Which brings us to:
Position Two
This position has never occurred,
but it is the obvious way to disturb
Black's easy play after 3 ... �b4+, so I
think it will be contested many times in
the future. Furthermore, Black's meth­
ods of counterplay are tricky and there­
fore require special attention.
The first time I saw this position was
on an analysis board about 25 years
ago, when my teacher was trying to ex-
1 d4 ltJc6 2 c4
plain to me why people didn't play
3 ... �b4+. It seemed obvious at the time
that Black's position was bad - after all,
the d-pawn is hopelessly backward. How
can Black possibly evict White's queen
and achieve the ... d7-dS advance?
Often we can't, but there are other
options and, depending on how White
continues, it is actually possible to
"play around" White's queen and leave
the d7-pawn alone for the foreseeable
future. After all, it is securely guarded
and Black's pieces do have oth er ways
to develop. This idea is demonstrated
by the variation 9 e4 ltJc6 10 ltJf3 'ifas !
11 0-0-0 ( 1 1 �d3 :e8 12 �C2 b6 13 a3
'iic s or 12 0-0 :e6 13 'ii'a3 'ii' xa3 14
bxa3 and White can stop bragging
about his superior structure) 11 ...a6 12
�bl lle8 13 a3 bS 14 cxbS axbs 15
�xbS �a6.
Black's whole army is activated and
the d-pawn is neither an obstruction to
Black nor a target for White's counter­
play. Black's compensation is more than
sufficient. Notice the ... a7-a6, ... b7-bS
idea, which develops Black's bishop,
37
Th e Dark Kn ight System
weakens White's control over dS, and
opens lines against White's king. It is
powerful enough that it can sometimes
be used even when White has not cas­
tled queen side (and ... b7-bS can occa­
sionally be played without ... a7-a6).
Was White's play too co-operative?
Presumably 9 e4 is th e culprit, provid­
ing a target for Black's ... tDf6 and
.. :�as. It also leaves a hole on the d4-
square which could turn into a long­
term problem, although it did stop
... tDfs, a useful move for Black. There­
fore 9 e3 :te8 10 tDf3 tDfs 11 'ifd2 d6 !
(ll ...b S ! ? is a great try, but 12 tDxbs
i..b7 13 i.. e 2 tDe4 14 'ifc2 'ii'a s+ 15 tDc3
tDxc3 16 bxc3 tDd6 17 Itxd7 Aab8 is a
little better for White) 12 i.. e 2 (12 l:ldl
'iVb6 13 .i.e2 i.. e 6 14 0-0 h6 or
14 ... :ad8! ?; or 12 e4 tDe7 13 0-0-0 'Wb6
14 'ii'xd6 'ii'xf2 15 tDxes and lS ... i..fS,
lS ...i.. g 4!?, ls ...i.. e 6 ! ?, or lS ... tDc6 ! ?
with equal chances i n all cases) 12 ...b6
13 e4 tDe7 14 0-0 i.. b 7 15 'ii'd 3.
By playing for ... d6-dS (with the
queenside fianchetto) Black has man­
ag ed to provoke e3-e4 again, which is
38
once again a target, and again leaves
the dark squares weak. On the other
hand, if Black doesn't find a plan, he
may find himself statically worse with
his backward d-pawn, despite White's
"bad" bishop. It is Black's idle tDe7 that
will save the day by repositioning to
target those soft dark squares: lS ...tDg 6
16 g 3 tDf8 ! 17 :adl tDe6 ! 18 'ii'xd6 tDd4
(or 18 .. :i"xd6 19 lIxd6 tDcS) 19 'ii'xd8
tDxe2+ 20 tDxe2 l:Iaxd8, which is at
least equal for Black.
Can White save a tempo by keeping
his e-pawn flexible? Not unless he
wants an e-pawn shoved u p his king's
file: 9 tDf3 :te8 (or the wild 9 ...e4 10
tDd4 e3 11 fxe3 bS! 12 tDdxbS a6 13
tDC7 lta7 14 tD7ds tDexdS 15 cxdS l:tb7
with excellent play) 10 :Idl tDfs 11 'ifa3
(11 'iid2 e4 12 'ifC2 dS! with advantage)
11. .. e4 12 tDd4 tDxd4 13 :xd4 e3 and
although Black's d-pawn remains,
White now has a matching one on the
lovely e2-square, and has at least as
much to worry about.
A2: 6 tDc3 d6

7 e4
7 tt)f3 fS 8 g 3 was V. Rao-J.S chuyler,
New York 1986 (see below).
7 fS
••.
7 ... tt)f6 is surely playable, hoping for
a better opportunity to play .. .f7-f5, but
I prefer this active move in spite of a
few downsides; i.e. weakening the e6-
square, and opening the game for
White's lousy bishop. Black's rook will
soon be enjoying the f-file, and the tt)e7
will gain access to d4 via the f5-square.
After 7 .. .fS, there is only one game
in my database, H.Titz-C.Barlocco,
Dresden 2004:
8 exfs i.. x fS 9 i.. d3 tt)f6 10 tt)ge2!
1 d4 tt)c6 2 c4
This knight covers the soft f4-
square. Instead, tt)f3 is vulnerable to
... i..g 4 and will be loose if White finds it
necessary to play g2-g3.
10 0-0 11 0-0 i..x d3 12 'i¥xd3 tiJhs
• .•
This typical move usually provokes
White into playing g2-g3, a long-term
weakness.
13 g 3 'iVd7 14 f3 a6 15 :ad1 :ae8 16
tt)e4 h6
An important move: tt)g s-e6 must
be prevented.
17 cS tiJf6 18 tiJ2c3
There is nothing wrong with this,
but knights that control e4 are not
permitted to control d4 as well !
18 tiJfS ! 19 b4 tiJxe4 20 tiJxe4 gs
••.
This is intended to discourag e
White from playing f3-f4, which would
undermine Black's knight as it arrives
on d4.
21 cat>g2 tt)d4
The position has been equal since
move six, but somehow Black contrived
to win in 75 moves (see Game 20).
It is now my great displeasure to
39
The Dark Kn ight System
show V.Rao-J.Schuyler, New York 1986. I
had not yet started to use 1 ... tLlc6 regu­
larly, but it seemed like the perfect
choice against the straight-laced,
booked-up senior master. We pick up
after 1 d4 tLlc6 2 c4 e5 3 d5 .i.b4+ 4 .i.d2
.i.xd2+ 5 'ii'xd2 tLlce7 6 tLlc3 d6 7 tLlf3 f5 .
8 g31? tLlf6 9 e4 h6
Generally a useful move, preventing
tLlg 5 and making ... g7-g 5 possible. Here
Black prepares to play 10 ... fxe4.
10 exf5 .i.xf5 11 .i.g2 0-0 12 0-0
I was justifiably h appy with my po­
sition. Houd;n; prefers Black and sug­
gests ... 'iVd7, ... .i.h7, .. .:f7, .. .:af8. How­
ever, I was a 15-year-old expert, there­
fore ...
12 g5??! 13 h4 g4 14 tLlh2 �h7 15 f3
.••
gxf3 16 :txf3 .i.g6?! 17 :tafl tLleg8 18
g4 'ile7 19 h5 .i.e8 20 'iVd3+ �g7 21
tLle4?!
I have been barely hanging on, but
there is a light at the end of the tunnel
since I will be in good shape if I can
reach an endgame. Also, it seems that
White's attack is not simple to play.
21 ...:f7 22 tLlg3 .i.d7 23 'iVe3 b6 24 g5?1
40
hxg5 25 'iYxg5+ Wh8 2 6 tLlf5 .i.xf5 2 7
:xf5 .:laf8 28 h 6 "e8
29 'iVg6??
White, short on time and frustrated
about being unable to break through,
commits a horrible blunder, allowing
his queen and rook to be forked. In­
stead, after 29 l:lxf6 ! ? tLlxf6 30 l%xf6
'iIIe 7 3 1 lbf7 'iYxg 5 32 :xf8+ �h 7 a
crazy endgame arises - presumably
White has some advantage.
29 tLlh7???
.•.
From completely winning to com­
pletely losing in one move, as we ap­
proach the time control. In my haste I
both overlooked the fork and the fact
that 29 'iIIg 6 attacks f7. White's posi­
tion would have disintegrated com­
pletely after 29 ...tLle7. If you think I'm
over this after just 26 years, you would
be wrong. I resigned shortly.
B: 4 tLld2
Considering the pawn structure and
the closed position, it will be fine for us
to trade off our dark-squared bishop
for a knight.

Besides, White's knight would have
controlled the important e4-square, a
job for which White's dark-squared
bishop is uniquely unsuited. I beat fu­
ture-GM Jesse Kraai (then a senior
master) in this variation, though the
game doesn't survive. Apparently I am
not one of those awesome people who
remember every game they have ever
played.
4 liJce7
...
5 a3
Just about everyone plays this, but
White has more challenging moves:
a) 5 liJf3 and now Black can't play
S ... d6?? 6 'iVa4+, so S ... i.. X d2+ is nearly
1 d4 4Jc6 2 c4
forced: 6 i.xd2 (don't worry about 6
�xd2 d6 7 "g s liJg6 S �xdS+ <it>xdS 9
e4 h 6 ! 10 i.d3 liJ Se7 11 0-0 fS with
equal chances) 6 ... d6 7 e4 fS S exfS liJf6
9 i.. e 2 0-0 10 0-0 liJxfS ! with interest in
... eS-e4, or ... a7-aS, ...liJd7, ...liJcS, ( ...b7-
b6), while ...h 7-h 6 is generally very use-
ful as well. Houdini claims that all
roads lead to equality, but the fact that
he likes 11 i.cl a whole bunch seems
like a bad omen for White.
b) 5 'iVa4! ? was played in B.Avrukh­
I.Ben Menachem, Israeli Team Champi­
onship 1999 - an annoying move be­
cause it pins the d-pawn, complicating
the defence of eS (if Black has to play
... liJg 6 in order to guard the pawn, it
defeats the purpose of 3 ...i.b4+). The
game continuation s ... cs 6 a3 (6 liJf3 ! ?)
6 ...i.xd2+ 7 i.xd2 liJf6 S liJf3 e4
(S ... O-O ! ?; S ... liJe4 ! ?) 9 liJh4 0-0 10 d6
was pleasant for White.
Although Black has ways to try to
make s ... cs work, I prefer S ... aS 6 a3
(a fter 6 liJf3 e4 7 a3 i.. x d2+ S liJxd2 fS 9
g3 liJf6 10 liJb3 0-0 11 i.h 3 d6 12 liJd4
i.d7 1 3 '6'C2 a4 14 0-0 �eS White's
bishops barely register as an asset;
lS ... bS ! ? is likely to fonow) 6 ... i.xd2+ 7
i..xd2 liJf6 S liJf3 liJe4! 9 i.. e 3 0-0 10
"C2 (10 liJxes liJfs 11 i.f4 d6 12 liJf3 g s
1 3 i.cl 'ii'f6 1 4 h 3 h S or 1 4... liJh4 gives
Black considerable pressure for the
pawn) 10 ...liJf6 1 1 liJxes d6 12 liJf3 (or
12 liJd3 :eS 13 i.d2 i.fS 14 e3 c6 15
dxc6 liJxc6 and 16 .. .d S ) 12 ... c6 ! 13 dxc6
i.fS 14 '6'a4 bxc6 with full compensa­
tion ; e.g. 15 :dl :eS 16 i..d4 liJe4 17 e3
41
The Dark Knig h t System
cs 18 iLc3 lbxC3 19 bxc3 lbg6 20 iLe2
: b8 21 0-0 lbf4 or 21 ...iLe 4!?, or 15
iLd4 lbe 4 16 g4! ? iL g6 17 h4 h6 with
just a big me ss.
F or the cowardly it is 11 0t strictly
l1ecessary to sacrifice a paWl1: 8... e4
(il1stead of 8 ...lbe 4!) 9 lbd4 0-0 is pl ay­
able, though White has a small adval1-
tage after 10 Vi'a (or 10 e3 d6 11 iLe2
iLg4 12 f3, or 10 g3 CS 1 1 lbbS lb fS 12
e3 b6 il1tel1 dil1 g 13 ...iLa6, 14 ...iLxbS
al1 d l s...lbd6) 10 ...d6 11 g3 c6 12 dxc6
lbxc6 13 lbxc6 bxc6 1 4 iLg2 dS 15 cxdS
cxdS 16 iLc3 .
c) 5 'iWc2!? fS 6 lbf3 hd2 7 lbxd2 d6
8 e 4 lbf6 9 iLd3 was tried m SAriste Cas­
tal1o-J.Salgado GOl1zale z, Saragossa
1998, at which poil1t BlaCK pal1 icKed al1 d
played 9-.f4?, a decisiol1 wh ich led to
difficulties sil1ce he lacKed coul1te rplay
agail1st White's big queel1side (10 CS! is
best). 111stead, 9_.0-o! 10 exfs 'iWe8 is ab­
solutely fil1e for BlaCK; e.g. 11 f3 'iWhs 12
g4?! 'iWh 4+ 13 Wdl hS 14 gs 'iWxgs 15
lbe4 'iWxfs 16 lbgs e 4!, or 11 h3 Whs 12
g4?! lbxg4 13 iLe2hfS 1 4 'iWdl 'iWh 4 1S
iLxg4 iLxg4 16 'iWxg4 'iWxf2+ 17 Wdl lbfS
18 4Je4 Vi'd4+ al1 d the sufferil1 g of
White's Kil1g � far greater thal1 our
small material il1 vestmel1t.
5 iLxd2+ 6 iLxd2
•••
6 Vi'xd2 d6 7 'iWgs is pOil1 tless be­
cause of 7 ...h6! 8 'iWxg7?! lb g6 9 lbf3
(White shoul d probabl y prefe r 9 h 4!
lbf6 10 hS �h7 11 'iWxh7 lbxh7 12 hx g6
fxg6 1 3 :xh6 lbf8 1 4 lbf3, whel1 1 4 ... c6
42
o r 1 4. ..'iWd7 give s BlaCK slightly the be t­
te r chal1 ces) 9 ...lb f6 10 ttJgs fie7 11 g4
lbxg4 12 lbe 4 iLfS 13 ttJg3 .i.d7 1 4 .i.h3
0 -0 -0 15 iLxg4 al1 d 110W m D.J usto­
P.Germal1 , Buel10s Aires 1995, BlaCK
spoiled some fil1e worK with
l S ...iLxg4?! 16 lbe4 iLfS 17 lbf6 wh ich
was approximately equal, whe reas
lS ...:dg8! 16 iLxd7+ �xd7 17 lbfS
i!xg7 18 lbxe7 lbxe7 would h ave left
him with chal1 ces to COl1vert his adval1 -
tage .
6... d6 7 e4 f5 8 exf5
Sil1ce the mail1 lil1e offers White
110thil1 g, it maKes sel1 se that il1
M.Gurevich-D.loler, Al1twerp 1998,
White looKed for al1 d foul1d somethil1 g
else - 8 'ii'h S+!? - al1 d was immediately
rewarded by 8 ...�8? ! , after which
Bl aCK'S static Kil1g provided a l1ice tar­
ge t for Gurevich's Kil1 gside storm (see
Game 21). 111 stead, 8 ... g6 9 'iWh4 fxe4 10
f3 exf3 11 lbxf3 lb fS 1 2 'iWxd8 �xd8 13
iLd3 lbf6 leaves White with el10ugh for
the paWl1 , but 110 more thal1 that; e .g.
14 0 -0 Irf8 15 l:[ael We7 (ls...�e8 ! ? ,
l s...as! ?) al1 d 11 0W Mr. H sees 110thil1g
be tte r thal1 16 .i.as �d7 17 iLd2 we7,
re peatil1 g positiol1 .
8 _iLxf5 g lLJe2 lbf6 10 lbg3
Either 10 ....i. g6, as il1 E.Arlal1 di­
M.Lal1 zal1 i, Sal1 Maril10 1998 (see Game
22), or 10 ...0 -0, as il1 G.G rigore­
P.Brochet, Creol1 1999 (see G ame 23), is
good el1 0ugh for equality, though il1
the latter game BlaCK cOl1 trived to lose.
 C h a pter Th ree

1 d4 ctJc6 2 dS

White takes the bull by the horns - a
surprisingly uncommon reaction. It is
worth noting th at this is how Miles
played against his own specialty when
he faced Zvonimir Mestrov1c.
2 .lhes
••
White normally chooses between:
A: 3 e4 43
B: 3 f4 46
Others:
a) After 3 ttJf3 Black may wish to try
3 ... ttJxf3 4 exf3 eS (or 4... e6 or 4 ... g6),
but I don't care for it - White has de-
velopment, space, an d open lines in
exch ange for his anti-position al cap­
ture. Simply 3 ... d6 transposes to 1 d4
ttJc6 2 ttJf3 d6 3 dS ttJes in Chapter One.
b) 3 i..f4 ttJg6 4 i..g 3 - does the
bishop dominate the knight, or does
the knight domin ate the bishop? In the
blitz game E.Bacrot-C.Bauer, Ajaccio
2007, it looked like the latter after
4...ttJf6 5 ttJc3 eS 6 e4? ! i..b4 7 i..d3 d6 8
f3 ttJh 5 9 i..f2 c6! 10 dxc6 bxc6 11 g 3
i.. as 12 f4?! ttJf6 13 fS ttJe7 1 4 ttJge2
ttJg4, when Black has a huge advantage
(though he went on to lose). Bauer's
method of meeting 3 i..f4 should be
copied exactly. 6 dxe6 is an improve­
ment for White, but Black has nothing
to fear after 6 .. .fxe6 and 7 ...i..b 4.
c) 3 ttJC3 e6 will soon transpose to
other lines in this chapter; e.g. 4 dxe6
fxe6 5 e4 (see line A) or 4 f4 ttJg6 5 dxe6
fxe6 6 e4 (see 6 ttJc3 ! ? in line B).
A: 3 e4 e6
Mestrovic played 3 ... d6! ? 4 f4 ttJd7 5
ttJc3 c6 and went on to draw with GM

43
Th e Dark Kn ight System
Drasko (Game 24). Black has also scored
well with 3 ... lhg6?!, which can transpose
into lines considered below, but his po­
sition is highly suspect after 4 h4!.
4 dxe6!
4 f4? ! is seen here most often and
seemed fine th e first 5 5 times it was
used, but on the 56th, the G erman
master Lutz Diebl played 4 ... exd5 ! and
drew with GM Gyimesi in the
Bundesliga. Bravo! If White takes the
knight, Black draws by perpetual: 5
fxe5 'iWh4+ 6 �e2 1fh 5 + 7 �d2 'ii'h 6+ 8
�c3 'Wc6+ 9 'itd3 'ifa6+ etc.
If 6 g 3 ?, as Gyimesi actually played,
then 6 .. :�xe4+ 7 'iWe2 'Wxh l 8 lhf3 b6 9
lhC3 and now, rather than Diebl's pre­
mature 9 ... i.a6?, the preliminary 9 ... c6!
maintains Black's nearly winning ad­
vantage (10 ... i.a6 is still coming to ex­
tricate the queen).
Meanwhile, White doesn't need to
go in for Gyimesi's contortions (Gyim­
n astics?) to avoid a forced draw; he can
play 5 exd5 or 5 'ii'x d5, although he
cannot hope for an opening advantage
after such concessions. The best re-
44
sponse i s 5 lhc3 ! lhg 6 6 'Wxd5 lhf6 7
�d3 i.C5 8 i.e3 'ii'e7 9 0-0-0 0-0 10 lhf3
with a small advantag e for White.
Diebls's 4 ... exd5 was tested again in
V. Erdos-R.Rapport, Hungarian Team
Championship 2012, to produce an­
other entertaining draw (see G ame 25).
If it is Black who is keen to avoid the
draw, he needs to forego 4 ... exd5 and
play 4 ... lhg 6, transposing to positions
considered in line B below.
4 fxe6
•••
This recapture is certainly dynamic,
although the somewhat exposed posi­
tion of Black's king requires careful
treatment.
Many players prefer 4 ... dxe6 5
'iVxd8+ �xd8, but Black's results have
been poor - pretty much draws and
losses. Nonetheless, Short made it look
easy to defend in S.Gordon -N.5hort,
British Championship 2011, so his
method could certainly be tried (see
G ame 26); and M.Gurevich-M.Rohde,
Philadelphia (blitz) 1989, shows that it
is possible for Black to win if White
overextends (see Game 27).

s ltJ c3
This flexible move makes it hard for
Black to activate his f8-bishop, since
5 .. .i.cS ? ? loses to 6 �S+ and S .....tb4?
.
is met by 6 'iVd4!, targeting Black's
l oose bits on b4, es, and g7.
Alternatively:
a) S f4 l2Jg6 is line B below. S ... l2Jf7 ! ?
and S . . .l2Jc6 ! ? are fi n e too, but no better
than the text.
b) S l2Jf3 l2Jxf3 6 'ii'xf3 'ii'f6 is already
equal according to Kalinin, but 7 '6'g 3
will gain some advantag e - White's
queen is active while Black's is mainly
awkward Therefore, just S ...l2Jf7 which
is similar to the main line (and trans­
poses after 6 l2JC3 b6).
c) S .i.f4 is untried but should lead
White to a normal plus; e.g. S ... l2Jg6
(S ...l2Jf7 ! ?) 6 .i.g 3 .i.cS 7 12Jc3 a6 8 12Jf3
ltJh 6 9 'ii'd2 0-0 10 0-0-0 d6 11 h4 bS.
d) S .i.e3 ? ! 12Jf6 6 12Jc3 .i.b4 7 .i.d4?!
12Jc6 8 a3 ..tas 9 e s l2Jxd4 10 "xd4 l2JdS
is a pleasant position for Black.
5 ... b6!
The fastest way for Black to mobi­
lize. The fianchettoed bishop is active
1 d4 12Jc6 2 d5
and occupies a diagonal which may
soon be lengthened by White's e4-eS .
S . . .l2Jc6 ! ?, a s recommended by Rybka 3,
is also possible - after all, the knight
retreats sooner or later, and c6 is not a
bad square.
6 ltJf3
V.Burmakin-J.Ulko, Moscow 1995,
continued 6 .i.f4 12Jg 6 ? ! 7 ..tg3 .i.b7 8
h4 h S ? ! , and after the simple 9 12Jf3
White would have been much better.
Instead, 6 ...l2Jf7 7 12Jf3 .i.b4! 8 .i.d3
.i.xc3 9 bXC3 .i.b7 10 0-0 12Jf6 11 litel
0-0 12 es l2Jds 13 .i.d2 cS 14 � e2 C4! lS
.i.xc4 :c8 16 ..tb3 'ii'C 7 is only a bit bet­
ter for White.
6 ... l2Jf7!
It is a bad idea to activate White's
queen with 6 ...l2Jxf3+, as tried by
B.5avchenko.
7 .i.c4?1
Instead:
a) 7 .i.f4 transposes to 6 .i.f4 l2Jf7 7
12Jf3 above.
b) 7 .i.d3 is stronger - at this stage
White is more likely to play eS than
Black is. After 7 ... .i.b7 8 0-0 12Jf6 9 'ii'e 2
45
Th e Dark Kn ig h t System
i.b4 10 Itel ttJg4! 11 h3 ttJge5 12 ttJxe 5
ttJxe5 13 'ifh5+ ttJf7 14 i.f4 .i.xc3 1 5
bXC3 g6 16 'ii'g 4 'iIIe 7 White has a small
advantage, though he has long-term
concerns about his pawn structure.
Black still has a useful choice as to
where he should castle. 17 ... 'lWC5 is
usually a good move.
7 ... i. b7 8 'it'e2 a6
This useful little move prevents ttJb5
and .i.a6, while preparing ...b6-b5-b4.
9 i.b3 .i. b4
Here 9 ... ttJf6 ? ! was played in C .Crouch­
A.Karpatchev, Cappelle la Grande,
1993, starting complications which
objectively favour White. The straight-
46
forward 9 ... .i.b4 can also lead to com­
plications - e.g. 10 0-0 ttJf6 11 Ildl 'fie7
12 e5 .i.xc3 13 bXC3 ttJe4 14 �d4 ttJxc3
15 'ii'd3 ttJd5 - but here Black is fine.
B: 3 f4 ttJg6
4 e4
Other moves:
a) 4 h4 is an interesting attempt to
take advantage of Black's inflexible
knig ht. In the expert section of the
1985 New York Open, I fell for White's
trap and played 4 ... e5? 5 h 5 ! ttJxf4!? 6
e3, though I went on to win an ugly
miniature with 6 ... 'ifg 5 ! 7 'it'f3 ttJxd5 8
-.xd5 -'g 3+ 9 �dl d6 10 .i.b5+ �d8 1 1
'ifxf7 ttJf6 1 2 ttJf3?? "'xg 2 13 IU l .i.g4
14 i.e2 .xfl+! 15 .i.xfl .i.xf3+ 16 i.e2
.i.dS (trapping White's queen) 0-1. Hi­
larious !
Instead, 4 ... e6! 5 h 5 ttJ6e7 6 C4! ? (6
dxe6 fxe6 7 e4 d5 transposes to 6 h4 d5
7 h5 ttJ6e7 in the notes to Position
Three below) 6 ... ttJf6 7 ttJc3 ttJf5 8 'ifd3
(if 8 dxe6 fxe6 9 g4, then 9 ... ttJxg4! ? 10
e4 ttJfe3 or 10 ... ttJfh 6 is possible, but
relatively simplest is 9 ... ttJg3 10 l:th3

lLlxfl ll g5 lLlxh 5 12 l:txh 5 lLlg 3 13 :h 3
lLlf5 14 e4 lLld6 15 lLlf3 lLlxC4 16 lLlh4
11g 8 17 "' 5+ g6 18 "xh 7 1:[g7 19 'i¥h6
:g 8 with a draw) 8 ... i.. c 5 ! 9 e4 lLlg4 is
fine for Black, according to Mr. H,
though there are some crazy variations
to consider:
al) 10 exf5 lLlf2 11 'ii'g 3 lLlxh l 12
'ifxg7 �4+ 13 �d2 (or 13 �dl 'i¥g4+ !
14 'ii'xg4 lLlf2+) 13 ... I1f8 14 lLlge2 lLlf2
15 lLlb5 i..b6 16 d6 'i¥xh 5 17 fxe6 lLle4+
18 �C2 lLlxd6 is in Black's favour.
a2) 10 dxe6 lLlf2 11 'ii'd5 (not 11
exf7+? �f8 12 'ild5 lLlxh l) 11...d6 12
'ii'xf5 i..x e6 13 'ii'xC5 dxc5 14 <iitxf2
�xC4 with an unbalanced but roughly
equal position .
a3) 10 lLldl lLlfh6 11 dxe6 fS! 12 i.. e2
'ii'e7! with excellent play for the pawn.
b) 4 f5 overextends: 4... lLle5 5 i..f4
d6 6 e4 g 6 ! (in practice, Black has
played 6 ... lLlf6 ? ! or 6 ...e6? but it is best
to challenge White's space immedi­
ately) 7 lLlf3 i..g 7 8 i..b 5+ i.. d 7 9 i..xe5
if.. xe 5 10 i..x d7+ 'ii'xd7 1 1 lLlxe5 dxe5 12
0-0 lLlf6 with equality; e.g. 13 'i¥d3 c6
14 c4 cxd5 15 cxd5 :c8 16 lLlc3 0-0.
1 d4 lLlc6 2 dS
c) 4 lLlf3 is legal and was in fact the
move order for Onischuk-Shkuro men­
tioned below (see note 'e' to Position
Three). 4 ... e6 5 dxe6 fxe6 6 e4 trans­
poses to the main line, while 5 c4 i.. c 5
will not bring White any happiness.
4 e5
•••
4... e6 may transpose after 5 fxe6, or
it may turn into a kind of mirrored
Alekhine, which usually works badly for
White because of the weak a7-g 1 di­
agonal; e.g. 5 c4? ! exd5 6 cxd5 i.. C 5 or 5
lLlf3 exd5 6 exd5 ? ! i..c 5. However, S
lLlc3 ! exd5 6 'ilxd5! lLlf6 7 'ii'd3 i.. c 5 8
i..e3 'ile7 9 0-0-0 0-0 10 lLlf3 is better
for White (though Black is still okay).
5 dxe6!
White's only good move.
a) 5 f5? could lead to the position
mentioned in the introduction to this
book after 5 .. :iVh4+ 6 �d2 'ii'xe4 7 fxg6
'ii'xd5+ 8 �el 'ii'x dl+ 9 �xdl hxg6, as­
sessed as unclear by Bogoljubow. The
endg ame is favourable to Black, but
that is a moot point because 6 ...lLlf6 ! is
even stronger - as one of my students,
Matthew Shih, was kind enough to
47
The Dark Kn ight System
point out to me last year.
al) 7 fxg 6 ? tbxe4+ 8 �e2 iVf2+ 9
�d3 tbcs+ 10 WC3 ttJa4+ 11 wb3 'ifb6+
12 �C4 iVa6+! 13 �b3 tbcs+ 14 �c3
tbe4+ 15 wb3 'i¥b6+ 16 Wc4 'i'b4+ 17
�d3 tbf2+ wins White's queen.
a2) 7 tbC3 �b4 S fxg6 tbxe4+ 9 We2
iVf2+ 10 �d3 fs (10 ... �xc3 11 Wxe4!
�as 12 c3 hxg6 13 �d3 d6 14 '1itC4 �fs
1 5 -.e2 ""6 may be better, but it's far
m ore complicated, as Black will still be
down material for some time) 11 tbxe4
-.d4+ 12 We2 'ii' xe4+ 13 �e3 f4 14 -.d3
'ii'xe3+ also leaves Black much better.
a3) 7 '6'f3 tbxe4+ S '1ite2 tbf4+ 9
�xf4 iVxf4 10 iVxf4 exf4 11 �f3 tbf6 12
tbc3 �b4 13 tbge2 (13 :el+ '1itfS 14
i.. C 4 b5! 15 i..xbs .i.b7) 13 ... d6 14 wxf4
h 5 ! is relatively best, but still very good
for Black.
b) 5 tbf3 ? is apparently tempting (it
h as been played several times), but af­
ter 5 ... exf4 6 tbc3 �c5 7 �d3 'ii'e 7! S
iVe2 d6 9 tba4 (9 �d2 ? ! a6 ! is even
worse) 9 ... �b6 10 �d2 tbf6 White has
far too little for the pawn.
c) 5 iVf3 exf4 6 tbc3 (6 �xf4 tbxf4 7
48
-.xf4 'ii'f6 S 'ii'xf6 tbxf6 leaves Black
with a superior pawn structure and
White with an acute shortage of dark­
squared bishops) 6 ... i.. c5 7 i..xf4 tbxf4 8
iVxf4 _f6 9 'iVg 3 d6 10 �bs+ wfS 1 1
tbf3 'ii'g 6 and Black can look forward to
a long and pleasant endgame.
d) 5 tbe2 (or 5 tbh 3) 5 ...exf4 6 tbxf4
�d6 ! already puts the enemy kingside
under pressure: 7 tbxg6?! hxg6 S 'ii'f3
'Wh4+ 9 �dl tbf6 10 �d3 i.. e5 is obvi­
ously not satisfactory for White, but 7
'iff3 tbf6 S tbC3 0-0 9 i..d3 leaves Black
with several good ideas, the simplest
being 9 .. .i.es (9 ... c6 ! ? 9 ... tbe5 ! ?) 10 0-0
.
d6 11 h3 cS, when Black's activity and
strong e5-point give him the advantage.
S fxe6
•••
The endgame after s ... dxe6 6 'ii' xdS+
'1itxdS is playable in theory, but with
Black's slightly misplaced g6-knight, it
is less appealing than the similar end­
ing in line A (without 3 f4 tbg 6). In
practice, Black's results are quite poor.
Instead, 5 ...fxe6 brings us to:
Position Th ree

Not to put too fine a point on it, but
the players who have reached this posi­
tion with the black pieces need collec­
tively to have their heads examined.
White's main asset is his powerful
pawn duo on e4 and f4. Black's main
asset is his central pawn majority. One
need notice only one of these two
things in order to come up with the
correct plan (or at least the correct
sixth move) for Black. In fact, I'm not
even going to insult my readers by say­
ing it out loud, so if you still don't
know, see Wood-Penrose below and
then read Pawn Power ;n Chess by Hans
Kmoch .
Instead of taking the opportunity to
strike in the centre, Black has generally
been seduced by 6 ... .i.cS, occupying the
a7-g1 diagonal, presumably to stop
White from castling. Naturally this is
less important than the central battle
and, what's worse, it doesn't even
work. If White finds the bishop trouble­
some, he can trade it off with a timely
ttJa4 or 'ii'e 2 and .i.e3. To add insult to
injury, White is usually better off cas­
tled long anyway. Sometimes 7 ... .i.cS is
a good idea (or 7 ... .i.b4+ S c3 .i.cS), but
if there is a knight on c3, it is much bet­
ter to put pressure on White's centre
with 7 ... .i.b4, which prepares ... lLlf6-e4.
Black's e4-knight can be a very an­
noying piece. As we see in the following
analysis, White's light squares are usu­
ally too weak (because he has had to
play g2- g3) to allow him to eliminate
the knight comfortably with .i.d3 and
1 d4 lLlc6 2 dS
.i.xe4, even if this wins a pawn .
Let's get to the analysis.
6 lLlf3
Other moves:
a) 6 .i.d3 is less popular and less
successful. Then Black has done fine
with 6 .. ics, but 6 ... dS! is more accu­
rate, as in B.Wood-J. Penrose, Southend
19S7(!),
which continued 7 eS lLlh 6 S .i.e 3?!,
and now not the game's s . . .lLlfS ? ! 9
.i.xfS exfs 10 lLlf3 .i.e7 ? ! (10 ... CS!) 11 c4
which is good for White, but instead
S ... lLlh4! 9 'ii'e 2 CS! and Black is better.
No better is 7 lLlC3 .i.b4 S .i.d2 lLlh 6
9 lLlf3 0-0 10 g3, when Black gently
plays 10 ... eS ! ! and lets White try to
work out the details with his king in
the centre. The correct solution is 11
lLlxdS .i.xd2+ 12 'ii'xd2 c6 13 lLle3 exf4
14 .i.c4+ �hS 1 5 'ii'xdS :xdS 16 gxf4
ltJxf4 17 :gl bS with equality. Accord­
ing to Houd;n;, 11 fS dxe4 12 lLlxe4
.i.xd2 13 'ii'xd2 .i.xfS 14 0-0-0 'ii' e7 is
also equal, but to me it looks a lot like
White is down a pawn for nothing. For­
tunately, it's not really our problem.
49
Th e Dark Kn igh t System
b) 6 i.. e3 is seldom played - al­
though it prevents ... i.. c s, White's im­
portant dark-squared bishop is vulner­
able:
6 ... ds (of course) 7 g3 (7 l2Jc3 i..b4 8
�d3 i.. xc3 9 "xc3 l2Jf6 10 0-0-0 0-0 11
exds exds 12 ttJf3 i.. g 4 is equal) 7 ... Cs! 8
ttJd2 "b6 9 :bl (sad, but there's really
nothing better: 9 fs exfs 10 exds ttJf6
11 "e2 i.. e7 12 ttJC4 �a6 13 d6 b s ! or
13 i.. xcs 0-0 14 d6 i.. d8 15 d7 i.. xd7 16
i..xf8 ttJxf8 gives Black more than
enough for a small exchange; while 9
i..d3 i.. e 7 10 ttJgf3 ttJh 6 11 "e2 ttJg4 12
exds exds 13 i..g l 0-0 14 0-0-0 i.. d 6 15
h 3 ttJf6 is fine for Black) 9 ...i.. d 7 10
l2Jgf3 l2Jf6 11 es l2Jg4 12 i..g l 'ii c 7 13
C4! d4 14 i.. d 3 i.. c 6 15 h3 ttJh 6 16 i..f2
i.. e 7 17 0-0 0-0 18 b4 b6 19 l2Je4.
White's play makes an excellent im­
pression, but here Black has 19 ...l2Jxf4!
20 gxf4 :xf4 2 1 �e2 ttJfs 22 :b3 :f8,
which reverses the initiative at the cost
of a small material investment. Hou­
d;n; calls it equal, but if 1 h ad the
choice, I'd sit behind the black pieces.
c) 6 g3 ds 7 ttJf3 transposes to 6 l2Jf3
50
ds 7 g 3 below.
d) 6 h4! ? has never been played, but
it's a venomous move. The tactical jus­
tification is 6 ... ttJxh4? 7 "g4! i.. e7
(7 ... ttJg 6 8 l:xh7) 8 'ii'x g7 with a large
advantage for White. Correct is the
anti-shocker 6 ... ds even though 7 hs
ttJ6e7 blocks the f8-bishop. After 8 ttJf3
l2Jc6 9 l2Jc3 i..b4 10 i..d2 l2Jf6 11 es
i.. xc3 12 i.. xc3 l2Je4 13 i.. d 3 ttJxC3 14
bXC3 Black gets out of Dodge with
14 ... �e7 15 ttJg s i..d7 16 i..xh 7 0-0-0,
when White has space an d a pawn, but
is badly overextended with nowhere
for his king. 1 will utter a naughty word:
unclear.
e) The rare 6 ttJc3 ! ? is logical, fight­
ing for th e ds-square, but Black forces
... d7-ds anyway with 6 ... i.. b4! 7 l2Je2
(other moves, such as 7 ttJf3 and 7 i.. d 3,
transpose elsewhere) 7 ... ds 8 �d3 c6 9
i..d2 l2Jf6 (after 9 ... dxe4 10 'iixd8 �xd8
11 a3 i.. as White is a little better in the
endgame) 10 es ttJg4 11 h3 ttJh 6 12
0-0-0 0-0 13 g4 bs 14 �bl i.. cs 1 5 i.. g 2
as and although White had a head
start in the race, it is difficult for him to
advance further. Then 16 ttJd4!? �b6
17 ttJxc6 i..b 7 18 ttJxds exds 19 ttJxas
l2Jxf4 20 i..xf4 "xas 21 i.. xds+ �h8 2 2
i..xh 6 gxh6 is certainly complicated,
but not unfavourable to Black.
The text move, 6 ttJf3, is by far the
most common - apparently with g ood
reason since White h as won the last
five games in a row from this position,
most notably A.Onischuk-I.Shkuro,
Ukrainian Team Championship 2009,

which is a perfect example of what
Black must avoid (see Game 28). Clearly
we need some new and improved
ideas.
6 d5!
•.•
This move is part of my original
analysis that dates back to 1986, and
inspired the above variations. In
twenty-five opportunities, Black has
somehow failed to try thi s, so we ana­
lyse in a vacuum. For most players, the
following lines need hardly be memo­
rized, but offer an excellent opportu­
nity to get acquainted with the wide
variety of plans for both sides.
7 liJc3
a) The first thing I realized 2 6 years
ago was that 7 fS ?! is not a problem:
7 ... dxe4 8 'iVxd8+ �xd8 9 liJg s exfs 10
liJf7+ �e8 11 liJxh 8 liJxh8 12 liJc3 c6 13
�c4 (13 g4! ?) 13 . ..tbf6 14 �e3 �d6 i s a
bit better for Black. As it turns out,
7 ... exfs is also okay: 8 exdS liJf6 9 liJc3
�b4 10 'it'e2+ �f7 ! 11 liJg S+ �g8 12
�d2 �xc3 13 �xC3 liJxdS 14 �d2 h6 1 5
liJf3 �h 7 16 0-0-0 l:te8 and i t is White
who has the hard job of proving full
1 d4 liJc6 2 d5
compensation.
Notice that Black isn't actually
threatening 7 ... dxe4, so White has an
array of options:
b) 7 eS may not be best, but it is cer­
tainly critical.
White plans liJg s and liJxh7, an ar­
gument he will try to enhance with h4-
hS, �d3, or 'ii'h S . This plan is indeed
dangerous, especially if Black is castled
on that side of the board. However,
Black does not skip his turns, and as
long as White is pursuing this plan he
is neither developing quickly nor at­
tending to his own king. For instance,
7 ... liJh6 ! 8 g3 (or 8 �e3 lbg4 9 'ird2
lbxe3 10 'iVxe3 'ire7! 11 lbc3 'ifb4) 8 ... cS
9 liJg s �e7 10 liJxh7 (after 10 �bS+
�8 White has to worry about both
11 ... c4 and 11 ... lbxes) 10 ...liJfS 1 1 "' 5
�f7, when Black h as good compensa­
tion after White's failed attack; e.g. 12
liJc3 �g8 13 lbf6+ �xf6 14 -'xg6 �e7
15 �g 2 c4 16 0-0 l:th6 1 7 '6'g4 'ifb6+ 18
�hl �d 7 and White is having trouble
with development, the centre, and his
king, which doesn't leave much to be
51
Th e Dark Kn ight System
happy about. There is also 8 h4!? .i.cs 9
h 5 (or 9 .i.d3 tLJg4 when, according to
Houdini, White has nothing better than
10 tLJd4 tLJh6 11 tLJf3, with a repetition)
9 ...tLJe7 10 tLJgs tLJhfs 11 'ii'd3 h6 12 g4
tLJd4 13 c3 tLJdc6 14 tLJf3 as ! with equal
chances. Black intends ...b7-b6, ... i.a6
( ... .i.b7), ... �d7 (or ... d5 -d4, ... 'ifds), and
will have the choice of which side to
castle. White may enjoy his space, but
may also find himself overextended.
Hold on: what's so great about
7 ... tLJh 6 - ?
I used to have problems in this posi­
tion in blitz games because I didn't
know the proper arrangement for the
pieces - especially whether to play
7 ... .i.cs or 7 ... cs, and also whether to
play ...tLJh6 or ...tLJ8e7. As it turns out, it
is not yet clear whether it is the bishop
or the pawn that belongs on cs, so it
makes sense to wait on that decision,
but the knight is just about always best
on the h 6-square. All of ... tLJg4, ...tLJfs,
and ...tLJf7 are useful options from
there, and it can hold up White's g- and
f-pawn s. Just as important, developing
52
the knight to h 6 avoids a traffic jam on
the e7-square, which may be needed
for Black's other knight, not to mention
the bishop and queen. (You may now
resume your normal programming.)
c) The immediate 7 h4! ? is also pos­
sible, but Black is already fine after
7 ... i.b4+! 8 c3 (not 8 i.. d 2? dxe4!)
8.. .i. cs 9 h s tLJ6e7; e.g. 10 b 4 i..b 6 11
C4! ? dXC4 12 'ifxd8+ �xd8 13 i.b2 (not
13 tLJg s ? i..d4!) 13 ... exf3 14 .i.xg 7 tLJfs
15 .i.xh 8 f2+ 16 �d2 tLJg 3 17 cS tLJxhl
18 cxb6 axb6 19 tLJc3 tLJe7, which is cer­
tainly no worse for Black.
d) 7 i..d3 allows us to surprise White
with 7 ... dxe4! 8 .i.xe4 '6'xdl+ 9 �xdl
tLJf6 10 .i.xg6 hxg6.
Although our pawns are vile, our
bishop pair and overall activity are
quite enough, particularly since the
enemy king is a bit loose; e.g. 11 tLJbd2
.i.d6 12 tLJC4 bs 13 tLJces i..b 7 14 �e2
.i.xes 15 fxes tLJd7 and Black is more
comfortable because of White's bad
bishop; or 11 i.. e 3?! .i.d6 12 tLJc3 tLJg4
13 �e2 O-O ! 14 g 3 ? ! (14 .i.cl!) 14 ...b6!
and suddenly White is in big trouble; or

11 llJC3 .i.d7 12 llJe5 (12 We2 .i.d6! 13
llJe5 .i.xe5 14 fxe5 llJ g 4 nets a pawn)
12 ... 0-0-0 1 1 13 llJf7 .i.c6+ 14 llJxdS WxdS
15 h4 .i.xg 2 16 l:th2 .i.f3+ 17 �el .i.c5
and how is White going to untangle
himself without shedding any mate­
ri al ?
e) 7 C4! ? does force 7 ... dxe4, but
weakens White's position as well: S
·"xdS �xdS 9 llJg 5 �eS 10 llJxe4 llJf6
11 .i.d3 b6 12 llJbC3 .i.b7 13 0-0 :dS
and Black is comfortably equal.
f) 7 g 3
7. . ..i.c5 (7 . . ..i.b4+ S c3 .i.c5 9 'iWe2
·fle7 10 llJbd2 llJh6 11 llJg 5 0 - 0 12 h4
:eS is okay, too) S 'ii'e 2 (S llJc3 llJf6 ! 9
e5 llJg4 10 llJd4 0-0 11 'iVxg4 .i.xd4 12
l.. d2 cS! 13 0-0-0 'ifb6 with equal
chances) S ...llJf6 9 es llJe4 10 .i.e3 .i.xe3
11 'Wxe3 llJe7! 12 .i.d3 .i.d7 13 llJbd2
�xd2 14 llJxd2 (14 'ii'xd2 CS) 14 ... 0-0 1 5
0-0 llJf5 and with 16. . .'iWe7 (or 16 . . .b6)
and 17 ... c5 coming, Black h as sufficient
cou nterplay.
g) 7 exd5?! exd5 S .i.d3 makes no
sense - it surrenders the centre and ac­
tivates Black's problem piece, the cS-
1 d4 llJc6 2 d5
bishop. Unsurprisingly Black has many
ways to play: S ... .i.g4 9 0-0 .i.C5+ 10 �hl
llJSe7 11 h3 bf3 12 'iWxf3 0-0 seems
simplest, or Black can enter an equal
endgame with S ....i.b4+ and 9H.'it'e7+.
7 .i.b4
..•
8 g3
White usually finds it necessary to
play this sooner or later. Otherwise:
a) S .i.d3 .i.xc3 9 bXC3 dxe4 10 .i.xe4
'iWxdl+ 11 �xdl llJf6 12 .i.d3 0-0 13
:el llJg4 14 .i.xg6 hxg6 15 h 3 llJf6 16
llJe5 llJh 5 17 llJxg 6 l:tf6 lS llJe7+ wf7 19
llJxcS llJxf4! with equal chances.
b) S e5 llJSe7 (S ... C5 isn't bad either)
9 g3 0-0 10 .i.d3 cS and Black's good
centre and rapid deployment ensure
that he will not be rolled up on the
kingside and that his chances are not
worse.
c) S 'iWd3 ! ? .i.xc3+! 9 "xC3 llJf6 10 e5
llJe4 1 1 'ii' a 3 'iVe7 12 'iVxe7 llJxe7 13
.i.d3 b6 14 .i.xe4 dxe4 15 llJg 5 h6 16
llJxe4 .i.b7 17 llJc3 llJf5 lS 0-0 0-0-0 and
Black has the d-file, while White's cl­
bishop is a huge problem, a situation
which fully compensates for the pawn.
53
The Dark Kn ight System
8 .lt�f61 9 i.. d2 0-0 10 eS i.. x c3 11 i..x c3
..
lbe4 12 i.. d 3 b6!
White's C3-bishop is a silly piece - at
the moment Black is far better off
keeping the e4-knight and blocking the
other bishop.
a) 13 i..xe4 dxe4 14 lbg s is pointless
54
because o f 1 4... h6! 15 lbxe4?! i.. b 7 16
'ii'e 2 'iWdS 17 lbf2 lbxf4! 18 gxf4 lIxf4 19
Ilf1 (not 19 Ild1?! Ilxf2 !) 19 ... Ilaf8 20
i.. d2 (still not 20 :d1? ! 'ii'xa2 ! 21 l:td7
iVa4 22 l%xC7 i.. g 2 and White is toast)
20 .J:f3 21 a4 l:[8f5 22 0-0-0 :xes 23
iVxf3 iVxf3 24 lbd3 'iWds 2 5 lbxe5 iVxe5
26 i.. C 3 'ife3+ 27 <;t>b1 i.. d S 28 h4 gS
with some advantage to Black.
b) 13 'ife2 i..b7 14 0-0-0 'iWe8 15
i.. xe4 dxe4 16 lbg 5 e3 17 l:[he 1 h6 1 8
lbf3 lbe7 1 9 'ii'x e3 lbdS with enough
play for the pawn; e.g. 20 iVd3 as 2 1
lbd2 lbb4! o r 2 1 lbd4 a4 2 2 a 3 c S 2 3
lbe2 b5.
c) 13 0-0 lbe7 14 'ife2 i.. b7
(14 ... lbc5 ? 15 i..xh7+ is too strong) 15
l:Iad1 'ife8 is equal.
 Section Two
1 e4 lbc6
One might argue that it makes
more sense to play the Dark Knight
against 1 e4 th an 1 d4, because th e
dangerous move 2 dS is n o lon g er
available. On the other hand, White, as
a 1 e4 player, is more likely to be fa­
miliar with th e possible transpositions
to th e Scotch an d the Pirc. Personally, I
find th e second argument to be more
compelling than the first (for a decade
I played 1 d4 Nc6 with no plan for de­
fending the Scotch), but th e reader
can make up his own mind, or better
yet j ust play 1 ...ltJc6 against every­
thing !
2 d4 (including The Scotch) -
Chapter Four
2 ltJf3 (the Dark Knight Pirc) -
Chapter Five
2 ltJc 3 - Chapter Six
Other moves:
a) 2 .i.C4 commits the bishop way
too early: 2 ... ltJf6 3 ltJC3 e6! 4 d3 .i.b4 5
.i.d2 dS 6 exds exds 7 .i.bS 0-0 8 ltJge2
ltJe s! 9 a3 .i.d6 10 .i.g s c6! was P.Cruz­
A.Kogan, Lisbon 2000. Black has a big
advantage, largely due to having shut
White's light-squared bishop out of
play.
b) 2 f4 is not as bad as it looks.
2 ... dS ! 3 exds 'ii'x ds 4 ltJc3 leads to a
bizarre kind of Scandinavian. Fortu­
nately, 4 .. :ii'e6+! is more awkward for
White than it is for Black; e.g. 5 .i.e2
ltJd4! or 5 'ii'e 2 ltJb4! or 5 ltJge2 ltJd4! 6
d3 ltJf6 7 ltJe4 ltJfs 8 c3 g 6 (8 ... ltJxe4 9
'ii'a4+) with some advantag e to Black in
an cases, according to Houd;n;.
3 es is a good Nimzowitsch Defence
55
The Dark Knight System
for Black since White's pawn does not
belong on f4 (at least until _.f7-f6 is
played). In R.Barkman-S.Lejlic, Karls­
krona 1 997, Black used this to good
effect: 3 ... lLl h6 4 d4 .tg4 5 .te2 .txe2 6
lLlxe2 lLlf5 7 c3 e6 8 'ii'd 3 h 5 (8 _ .'Wh4+!)
9 lLld2 ? ! fih 4+ with advantage due to
White's light-square problem s - White
decided to pitch a pawn with 10 lLlg3
'ii'xf4 11 lLlxf5 'ii'xf5 12 'ii'xf5 exf5, but
he had no compensation.
Alternatively, Black can throw a clog
in White's machinery with 3 ... g 5 ! ? 4 d4
gxf4 5 .txf4 .tf5 6 c3 'iWd7 7 lLlf3 h5 8
..te2 0-0-0 9 lLlh4 ..tg4 10 lLld2 ..th6 as
in H.Gohlke-S.Wiezer, G orlitz 1985,
56
with a small advantage according to
New In Chess. Actually, I don't believe
Black is better, but the position is in­
teresting and Black has the making s of
a kin gside attack Wisnewski's recom­
men dation is 3 ... d4! ? - one idea is to
clear the d 5-square for Blafk's queen,
but I am n ot keen on relinquishing con­
trol over e4.
For those who are interested, 2 ... e 5
3 lLlf3 f5 ! ? i s the Adelaide Counter­
Gambit, a wild and fascinating defence
to the King ' s Gambit, but truly outside
the scope of this book Besides, there is
also 3 lLlc3, with wildness that White is
presumably prepared for.
 C h a pte r Fo u r
1 e4 tDc6 2 d4 eS
Now White has:
A: 3 dS 57
B: 3 dxeS 73
c: 3 ttJf3 79 The Scotch
- A2: 4 c411 62
On 3 C3, Black can transpose to a
Ponziani with 3 ...ttJf6 4 ttJf3; but 3 ... dS 4
dxes dxe4 S 'ii'x d8 ttJxd8 is simpler and
equal, while for a more interesting
endgame Black can play 4....te6 ! ? S
exdS 'ii'x dS 6 'ii'xdS i.. x dS with fun
compensation; e.g. 7 f4 f6 8 exf6 ttJxf6
9 c4 .te6 10 ttJf3 O-O�O 1 1 ttJc3 ttJb4
with a large advantage to Black.
A: 3 dS ttJce7
White normally chooses between:
A1: 4 ttJf3 58
Others:
a) 4 f4? is a fairly common mistake at
amateur level. The apparently hard-to­
spot 4 ... ttJg6! transposes to 1 d4 lbc6 2
dS ttJes 3 f4 ttJg6 4 e4 eS with the differ­
ence that White's only good move, S
dxe6 en passant, is not permitted by the
rules. How unfortunate! White does not
have my sympathy though, since
57
Th e Dark Kn ight System
4... exf4?! 5 i.xf4 lLlg6 is also lousy.
b) Miles had a wonderful answer to
4 i.e3 in 4 .. .f5 ! ?, with which he drew
with Beliavsky and beat Campora (see
Games 29 and 30). The point is to take
advantag e of the newly moved bishop,
which is vulnerable to both .. .f5-f4 and
5 exf5 lLlxf5. If Black does not play 4 .. .f5,
White may be able to stop the f8-dark
bishop from developing comfortably,
e.g. 4 ... lLlg6 5 a3 ! ?, though 5 ... b6 is sat­
isfactory for Black.
c) 4 lLlc3 lLlg6 5 i.e3 lLlf6 (5 ... i.b4!?)
6 a3 is very lightly tested. Then 6 ...b6 7
lLlf3 i.c5 is fine, when 8 i.xc5 bXC5 9
d6?! 0-0 and 10 ... i.b7 is at least equal.
d) 4 d6 has been played a few times
and Black usually just takes, which is fine
- but I prefer 4... lLlg6, accelerating devel­
opment after 5 dxc7 "xc7; e.g. 6 lLlC3
i.b4 7 lLle2 (after 7 i.d2?! i..xc3 ! 8 i.xc3
lLlf6 9 i..d3 0-0 10 lLle2 :d8 and 11 ... d5,
Black is better) 7 ...lLlf6 8 a3 i..c5 9 b4 i..b6
and Black is fine, since 10 lLlb5 ? ! runs
into 10 ...i..xf2+ 11 Wxf2 �6+.
A1: 4 lLlf3 lLlg6
58
4. . . d6 is recommended by De Fir­
mian in MC014 as a way to keep the
play obscure. In my opinion, the reverse
is true: 4 ... d6 will transpose to Pirc or
King's I ndian-type positions, whereas
4 ... lLlg6 maintains a unique "Kevitzian"
flavour. Furthermore, 4... d6? ! 5 C4! lLlf6
6 lLlC3 g 6 runs into 7 c S ! which is in­
deed a certain kind of King's Indian -
the bad kind.
After the text White has two impor­
tant moves:
All: 5 h4! 5 9
A12: 5 .te3 61
After other moves, Black is already a
statistical favourite, reaching positions
similar to line A2.
a) 5 lLlc3 lLlf6 6 i..g 5 ? ! (this is wrong,
as usual; it was not too late for 6 h4!
h 5, though this is rarely played here)
6 ... h 6 ! ( Black can wait, but putting the
question immediately places the most
pressure on White to follow through
with this ill-conceived exchange) 7
i..xf6 'ii'xf6 8 g3 i.. c5 9 i..h 3 a6 and
Black is already a bit better due to his
powerful dark-squared bishop - and if
White doesn't prepare th e exchange of
light-squared bishops Black has a com­
fortable advantage.
b) 5 i.d3 i.c5 6 0-0 lLlf6 7 lLlc3 c6?! 8
lLla4 i.. e 7 was Y.Dembo-R.Goldin, Petah
Tiqwa 1996, which Black went on to
win - even so he should have preserved
the bishop on the diagonal with 7 ... a6
or 7 ... a5.

All: S h4!
This is the way to start posing prob­
lems for Black and his knight on g6.
S hS
..•
Black is forced to weaken his king ­
side or accept an extremely cramped
position. But let's not forget that White
has also weakened his kingside.
6 i.. g s
Alternatively:
a) 6 ttJc3 ttJf6 7 ttJg S ! ? (7 i.. g s trans­
poses to the main line) 7 ...i.. b4! 8 i.. e 2
d6 9 'ifd3 i..d7 10 a3, and instead of
10 ... i..x c3+ as in M. Kravtsiv-K.Tarlev,
Evpatoria 2007 (see Game 3 1), I prefer
10 ... i.. c S, retaining the bishop for now;
e.g. 11 g3 0-0 12 0-0 ttJg4 with a tiny
edge to White. Black intends ... 'ife7 and
.. .f7-f6 to evict the knight, and probably
... a7-a6 and ...b7-bS.
Riskier is 7 ...i.. c S ! ? 8 d6 ! cxd6 9 i.C4
0-0 10 0-0 bS! 11 ttJxbS i.a6 12 a4 :c8
13 i.e2 dS 14 exdS ttJxh4 15 i.xh S e4 -
Black has solved the problem of his
weak pawns by sacrificing them all and
now has reasonable play for the pawn
minus, but White had many other tries.
1 e4 lbc6 2 d4 e5
b) 6 g3 i.cS 7 i.g s f6 ! ? (7 ... ttJf6 8
i..h 3 d6 9 i.. xc8 'ifxc8 10 i..xf6 gxf6 is
simpler - Black will play .. .f6-fS at his
convenience, with near equality) 8 i.. d2
d6 9 ttJC3 i..d7 10 i.. e 2 ttJ6e7 was a little
better for White in H.Meissner-A.Miles,
European Cup, Slough 1997, though
Miles methodically ground out the win
in an instructive ending (see Game 32).
c) 6 i.. e 2 ttJf6 7 ttJ C3 a6 8 i.. g s i.cS 9
0-0 d6 is already equal since Black has
quickly established his bishop in the
ideal position . Black will normally fol­
low with 10 .. . 'iVd7, breaking the pin
and intending ... lbg4 or ...'ifg4.
6 ...ttJf6 7 ttJc3
7 ... i.. b4
7 ... i.. c S is a possible alternative.
Then 8 ttJa4?! i..b4+ 9 c3 i..e7 10 i..xf6
bf6 11 d6 was A.lvanov-J. Benjamin,
US Championship, Parsippany 1996,
which Ivanov won, but Black's inaccu­
racies are yet to come. Indeed, after
11 ... cxd6 12 g 3 ! ? dS ? ! 13 'ii' xds d6? 14
i..b S+ �f8 White was much better, but
Benjamin could h ave played 12 ... b6 13
c4 (13 1Wxd6 ? ! i..b 7 14 'iWd3 i.. c 6 15 b3
59
The Dark Knight System
dS is great for Black) 13 ....i.b7 14 lLlc3
.i.e7, followed by ...lLlf8-e6 - Black's
extra pawn may not be much of an as­
set, but he is not suffering for having it.
However, the correct response to
7 ... .i.cS, namely 8 lLld2 !, puts Black un­
der some pressure.
s lLld21
Instead:
a) 8 a3 .i.xc3+ 9 bxc3 c6 10 c4 d6 11
lLld2 Vi'as was L.Christiansen-1Ben­
jamin, US Championship 2000 (see
Game 3 3). Black h ad nothing to com­
plain about and went on to win .
b ) 8 .i. e 2 allows u s t o preserve the
bishop with 8 ... d6 9 0-0 .i.d7 10 a3 .i.cS
11 �d3 a6 12 g3 .i.h 3 13 l:Hdl and
break the annoying pin by 13 ... Vi'c8 (or
13 ... 'ifb8! ?), when we're just about
equal. (This queen manoeuvre should
be kept in mind throughout this sec­
tion .) 14 ... lLlg4 is likely, targetin g the f2-
square.
The text move is more challenging.
S d6!
•••
In E.Mortensen-C.Hoi, Ostrava 1992,
Black played 8 ... c6 9 .i.e2 (9 g 3 ! ?)
60
9 ....i.xc3 10 bxc3 cxdS 11 .i.xh S ( 11
exds ! ?) 11 ...lLlf4 12 .i.f3 lLle6 13 .i.xf6
'iVxf6 14 exdS lLlcs 15 g 3 (15 h S ! ?)
lS ... d6 with only a tiny disadvantage,
but White missed some good chances
along the way.
After the text, we're in uncharted
territory ag ain, which is just how we
like it. Right?! Right.
a) 9 .i.bS+ is certainly not a big deal
after 9 ... .i.d7 10 .i.xd7+ Vi'xd7. Sure, we
liked that bishop, but White has solved
our space problem and helped us break
the annoying pin on our f6-knight.
Then 11 .i.xf6 gxf6 12 "'f3 'if g4
(12 ....i.xc3 13 'ii' x C3 fS 14 g3 0-0-0 is
okay, too) 13 'irxg4 (or 13 'ii'xf6 .i.xc3 14
bXC3 lLlf4! ls lLlf3 "'xg 2 16 lLlg s l:f8 17
0-0-0 'ir'xf2 18 lLlh7 lLle2+ 19 �d2 'ir'xf6
20 lLlxf6+ rtt e 7 2 1 lLlxh S :h8 22 �xe2
lixh S 23 lIdg l with a draw, presuma­
bly, though only Black can pretend to
play for a win) 13 ... hxg4 14 g3 .i.xc3 1 5
bxc3 lLle7 1 6 c 4 fS i s level.
b) 9 .i.e2 lLlf4! (as in many similar
positions, White does not want to take
this knight, especially since his bishop

is strong on the g 5-square) 10 .i.b5+
M8 (this time 10 ....i.d7 doesn't work
well - White will gain the very useful
g2-g3 with tempo) 11 0-0 .i.g4 12 f3 (or
12 .-el .i.d7 ! - the situation has
ch anged again already; the bishop ex­
change now brings equality: 13 .i.xd7
·.xd7 and Black will follow with
14 ...'6'g4 if possible, otherwise 14 ... c6,
or similarly 13 .i.d3 ! ? c6) 12 ... c6! 13
...t a 4 .i.xc3 14 bXC3 .i.d7 15 dxc6 .i.xc6
16 .i.xc6 'iVb6+ 17 <ii?h l 'ii' x c6 18 .i.xf4
exf4 19 ltJb3 d5 20 es ltJd7 2 1 :el 1:e8
22 'ii'd4 l:th 6 is roughly equal.
There are many other ways the
game could go, and White does have
chances for an advantag e, but the pre­
ceding lines give a good indication of
Black's resources.
A12: S .i.e3
White doesn't actually play this very
often, but Black's attacking ideas in
this variation are too important and
widely applicable to relegate to a foot­
note. Study carefully. Don't worry, it's
fun and easy.
1 e4 ltJc6 2 d4 e5
S ltJf6 6 .i.d3
•••
Not 6 ltJbd2 c6! 7 C4? ! (this is n o
good, but the alternative is to give up
the centre) 7 ... ltJg4! 8 .i.g 5 ? 'ifb6 9
.-e2 ? ! .i.c5 and White needs (PR.
6 ltJg4! 7 .i.d2
••.
Or 7 .i.g 5 .i.e7 8 .i.d2 .i.c5, transpos­
ing.
7 .i.cS 8 0-0 as 9 h3 ltJf6
•••
White has g otten h2-h 3 for free, but
free is still way too expensive - he must
be extremely careful because of his
weakened kingside:
a) 10 �el d6 11 .i.xa5? ! .i.xh 3 ! 12
gxh3 'ifd7 13 <ith 2 ltJf4 14 ltJgl ltJg4+!
15 �h l (not 15 hxg4?? .xg4 and .... g2
mate) 15 ... ltJxf2+ 16 :xf2 .i.xf2 17 'ifxf2
l:xa5 18 'ii'f3 0-0 19 ltJc3 f5 20 exf5
ltJxd3 2 1 .xd3 :xf5 and Black is better
because of White's weak king and
awkward knights.
b) 10 a4 0-0 11 ltJa3 d6 12 �el (or
12 ltJC4 c6! 13 dxc6 d5 ! 14 exd5?! e4! )
1 2...ltJh 5 13 .i.e3 ltJgf4! 14 .i.xc5 �f6 ! !
1 5 .i.e3 .i.xh 3 1 6 .i.xf4 ltJxf4 1 7 ltJh2
.i.xg 2 18 .-e3 .i.xfl and even though
White survives, Black maintains some
61
Th e Dark Kn ight System
material and positional pluses.
Note that in many similar positions
White voluntarily spends a tempo on
h2-h 3(?) - a move which contributes
greatly to his downfall. Note, too, that
Black's attack with ... tLJh 5, ... tLJgf4, .....f6
is so strong that he can sacrifice a full
piece to accelerate it. One free tempo
and the game would have been over.
A2: 4 C411
This weak move is almost as com­
mon as 4 tLJf3 . It wastes a tempo in a
critical position, obstructs the fl­
bishop, and leaves a big hole on the d4-
square. In exchange, White has forti­
fied dS, but the price is too high - he
has already squandered his first move
advantage, and if he plays at all indif­
ferently he will soon stand worse.
White imagines he will eventually be
playing c4-c5, but this is difficult to
achieve.
4 tLJg6
•.•
4 ...tLJf6 is similar, when 5 tLJC3 tLJg6
transposes to a position often reached
through the Tango. However, 6 h4! is a
62
dangerous move here, and Black has
both objective and practical problems.
The text move order makes it difficult
for White to play an early h2-h4 and so
avoids this troublesome line.
White has:
A2l: 5 tDc3 62
A22: 5 iLe3 64
A2 3: S ct:Jf3 6 7
A24: 5 iLd3 68
A2 S: 5 g3 70
A26: 5 a3 71
By the way, I hate all these nested
letters and numbers which make the
book look like the awful, tedious out­
lines we had to write in seventh grade
Social Studies. It makes my eyes glaze
over just looking at it, and probably
yours, too, so just read about 5 tLJc3
and 5 .i.e3 and then skip to line B. The
rest doesn't much happen and it's
pretty obvious anyway. Maybe some­
day, when you need to look something
up, you'll thank me that the material is
so well organized. I take cash and
cheques.
A21: 5 tLJc3 .i. C5 !
If S ... tLJf6 then 6 h4! and Black is
back on less solid ground.
6 tLJf3 tLJf6 7 .i.e2
By now 7 h4? is not possible be­
cause it does not address the threat of
7 ... tLJg4!.
For 7 .i.d3, see A24, n ote 'b' to White's
seventh move.
 1 e4 lbc6 2 d4 e5

White tried 9 a3 as 10 i. d2 fie7 11

l:bl .i d7 12 b4? ! axb4 13 axb4 i. d4!,
but she was worse after the queenside
had opened and it was Black who was
successful in the end (see Game 3 5).
9 ... i.. d7
This stops lba4 for now.
Instead, D. Baramidze-E.Griezne,
Baunatal 1999, continued 9 ... a6 10 a3
'iie7 11 b4 i.. a7 12 lbdl, when Black g ot
to demonstrate the typical manoeuvre
7 0-0
••• 12 ...lbh S 13 lbe3 lbgf4 which he used

Black will be playing ... d7-d6 and to good effect (see Game 36).
usually ... a7-aS soon, but this move or­ 10 a3 as
der is the most accurate since 7...d6 ? ! is
met by S b4 (S ... i..xb4?? 9 "a4+).
7 ... as is sometimes played, but
White is actuany quite far from being
able to play lba4 because the e4-pawn
would hang. Holding off on ...a7-aS
makes sense because:
1. Black may change his mind and
play ... a7-a6.
2. Sometimes Black can prevent
!tJa4 with ... i.. d7 instead.
3. It is possible to allow the bishop Black is slightly better.
to be traded off under some circum­ If 11 l:bl, we will certainly fix
stances - I have recently noticed that White's queen side pawns with 11 ... a4.
tempi sometimes matter in chess. If 11 b3, we have available the slow
8 0-0 d6 plan 1 1 .....e7, 12 ... h6, 13 ...lbh S ,
S ... as 9 lbel d6 (which could just as 1 4. . .lbgf4 , ls . . :ii'f6, etc, which I provide
easily have been S .. d6 9 lbel as) 10
. because it's often necessary in similar
lbd3 i..d4! was W.Weisser-L.Trumpp, positions - but here we just blast out
German League 2003, in which White 11 ... lbh S ! 12 lbxes (12 g 3 is safer, but it
quickly reached a difficult position and does create a weakness, and Black is
got abused tactically (see G ame 34). better after the simple 12 ... lbf6, menac­
9 '6'C2 ing ... i..h 3 and/or ... lbg4) 12 ... lbxes 13
In B.Perrusset-I.Moullier, Paris 2005, i..xh S 'tWh4 14 i.. e 2 fS 15 g 3 '6'h3 16

63
Th e Dark Kn ig h t System
.i.f4 :ae8 (or 16 ...g S ! ? 17 .i.xes dxes 18
exfs :a6 with more than enough for
the pawn) 17 b4 (not 17 .i.xe s ? :xes 18
.i.f3 f4 19 .i.g 2 'ii'h S 20 b4 .i.h3 21 bxcS
.i.xg 2, winning) 17 ... axb4 18 axb4
.i.xb4 with clearly the better game for
Black.
A22: 5 .i.e3 lhf6
6 lhc3
Other moves:
a) 6 .i.d3 b6! (6 ... lhh4! ? 7 �fl b6 8
g 3 lhg6 9 �g 2 .i.cS is similar, but White
is certainly not suffering from his in­
ability to castle) 7 lhc3 .i.cS (in V.Vilkov­
A.Provotorov, Kaluga 1996, the only
time this position has been reached,
Black played the highly inconsistent
7 ... .i.b4?! and went on to lose) 8 lhf3
0-0 9 0-0 and there are many roads for
both players, but they all lead to equal
positions - the imminent and posi­
tionally favourable bishop trade bal­
ances White's space advantage; e.g.
9 ... aS 10 �d2 'fIIe 7 11 :fel d6 12 .i.xcS
bxcS 13 g3 a4 and the game is still bal­
anced.
64
b ) 6 f3 .i.b4+ ( 6 ... b 6 i s still interest­
ing, but after 7 lhC3 .i.cS 8 .i.xcS bxcS 9
d6 ! White has balanced the chances) 7
lhd2 �e7 (7 ... b6!?). Miles scored 4-0
from this position, beating GM Kaida­
nov and IMs Shirazi and Langeweg ; e.g.
8 g 3 0-0 9 .i.h 3 c6 (or 9 ... aS or 9 ....i.cS)
and Black is better, going on to win in
G .Kaidanov-A.Miles, Palma de Mallorca
1989 (see G ame 37).
c) 6 .i.g S ? ! is ridiculous here and in
all positions where Black can play ...h7-
h6. The bishop is way too valuable to
trade, and even when it's possible to
retreat, ... h 7-h 6 is always a useful move
for Black. So 6 ...h6! 7 .i.xf6 'fIIxf6 and
White has even more dark-square
problems than usual. Black's queen is
also very happy on the f6-square.
6 i.b4 7 f3
...
Against 7 "c2 or 7 .i.d3 Black plays
exactly the same way.
7 i..x c3+
•••
Black's results are excellent (+21 -12
=8) with this move, so there is little
need to worry about alternatives.
Sometimes Black plays 7 ...'ile7 with the
idea of trading the dark-squared bish­
ops, but this is time-consuming com­
pared to the text, and hasn't worked
well in real life.
Quoting Joel Benjamin: "This is im­
portant: do not hang your bishop with
7 ... d6 8 'iVa4+." {Actually, according to
my database, 7 ... d6? ? has been played
three times and White has yet to play 8
'iVa4+, even in the game where White
was a GM. I won't name the GM. Can

this possibly be right? ! Can you explain
yourself, Pablo?)
8 bxc3 d6 9 "'d21?
In fact, just about anything is better
than the lemon 9 CS?! that White
squeezes out most of the time, which
leads to:
Position Fou r
Apparently i t i s just too tempting to
liquidate the weak c-pawn, but the
lines that open are all going to be
seized by Black; and while White's
weakness on c4 is no more, he h as an
even bigger problem on C3, which is
directly in the line of fire, sitting on
Black's half-open c-file. The games all
go the same way: 9 ...0-0 10 i.. d3 tLld7 !
11 cxd6 cxd6 12 tLle2 .-as ! (or 12 .. :.C7)
13 0-0 tLlcs, and ... i.. d7/ ... :ac8/ ...tLla4,
or ... b7-b6/ ... i.. a 6/ ...:tac8 on the queen-
side, and after White is pacified there,
.. .f7-fS on the kingside. S.Brudno­
J.Benjamin, Boston 2001, is a classic
example (see Game 38), whereas in
R.Mitchell-J.Schuyler, Bloomington
1991 (Game 39) Black focuses entirely
1 e 4 tLlc6 2 d4 e5
on the queen side, a strategy which is
far less effective, as w e will see.
Can things really be that simple for
Black? Not quite, but if White wants to
get any play, he pretty much needs to
go crazy like a Grandmaster (in lJelen­
E.Dizdarevic, Ljubljana 1992 - see
Game 40 - White went crazy like an I M,
which did not work), as in E. Rozentalis­
M.Lazic, Genoa 2004, which went 9 cS
0-0 10 cxd6 cxd6 11 i..d3
11 ...'ii'c7 12 tLle2 tLld7 13 h4!? tLlcs
14 hS tLle7 15 g4! an d while White may
not h ave been better, he certainly was
scary, an d he did go on to win after
lS ... tLlxd3 ? ! 16 'ii'xd3 bS 17 tLlg 3 as 18
�2 i.. a 6 19 'ii'd2 f6 20 tLlfS l::tf7 21
:hgl :c8 22 :acl �8 2 3 'Wb2 tLlxfs 24
gxfs "'c4 25 'ii' d 2 h6 26 i..b 6 'ii'a4 27
wg 3 :b7 28 i.. e 3 :bC7 ?? 29 i..x h6 b4
30 Wh4 bxc3 31 i.. x g7+ �e8 32 "g 2
1-0 .
Notice that White's pawn storm had
less to do with blowing Black open than
controlling th e fs-square, preparing it
for his knight. The square is especially
important since, had Black been able to
65
The Dark Kn ight System
break with .. .f7-fS, White's strategy
would have failed. Notice, too, that
Black was never losing - in fact, he was
never significantly worse - until
28 ... l:lbc7?? Instead 2 8 ... :tf7 ! (preparing
the escape 29 ... 'it;e7 if necessary) would
have pre-empted White's sacrificial
attack and held the balance. White's
ideal position wasn't so dan gerous af­
ter all !
Let's go back to move 11, set up the
proper defence without fear, and try to
play a little more quickly and actively
on th e queen side: l1 ..:ii'as ! (or
11 ... �d7 first) 12 �e2 �d7 13 h4 �e7
14 hS h6 lS g4 f6 16 1i'd2 �cS 17 .i.c2
b6 18 �g 3 (18 C4! VWa6 is equal)
18 ... .i.a6 19 �fS �xfS 20 gxfs l:lf7 2 1
:g l �f8, followed by 22 . . ': c 8 with
pressure on the c3 -pawn.
We now return to our regular pro­
gramming .
9 b6
.••
Palliser concludes that there is no
point delaying castling, but I am find­
ing that 9 ... 0-0 10 h4! (intending
10 ...�h S 11 .i.f2 or 11 �e2) is very
66
dangerous for Black, who must be ex­
tremely resourceful in a dizzying array
of variations in order to avoid disaster.
(See for yourself - I won't bore you with
the details. Frankly, I wouldn't even
know where to start.) After the text
move, Black can bail out to the queen­
side if things get too hairy too quickly -
a surprising but valuable option !
10 h4
Not forced, obviously, but if this is
what White is about, he is likely to get
right to it.
If 10 .i.d3 (or 10 �e2, for that mat­
ter), Black could castle, having lost
nothing - but I would prefer to stay
flexible with 10 ...�d7; for example, 10
�e2 �d7 11 h4 h6 12 h S �gf8 ! (in
Z. Koczka-Zsu.5imon, Hung arian Team
Champion ship 2003, 12 ... �e7? was 1-0
in 30 moves) and 13 ... �h7 will be like
the main variation. If 10 .i.d3 �d7 11
�e2, Black can still delay ...0-0 in favour
of 1 1 ...�CS ! . Can White keep waiting
before committing to either castling or
h 2-h4 - ?
10 h6
•••
This stops White's pawn from going
to h6, which would leave Black with a
big cramp and weaknesses on the dark
squares.
11 g3
Someone will try 11 hS �f8! (I just
love this move - the knight heads for
the open cs-square, whereas a knight
on e7 would be going nowhere for a
long time) 12 g4 �6h 7 ! (preventing 13
g s and preparing an eventual ... �g S !,
 1 e4 lbc6 2 d4 e5

while ......, 4+ is also sometimes useful)

13 i.d3 ltJd7 (Black is at least equal in
spite of White's space - his knights are
happy pieces, unlike White's sad bish­
ops) 14 �f2 ltJcS 15 �g 2 as 16 ltJe2
ltJg 5 etc. It is even safe to castle king ­
side now.
11 ... ltJd7
This transposes to A.Karpov-D.Chev­
allier, France 1993 (see Game 41). Ac­
cording to Palliser, this game was very
influential and popularized the varia­ We no longer have to worry about a
tion with i.e3 for White. If so, the positional squeeze by White, real or
game's true theoretical significance imagined. White's knight will soon be
was greatly overestimated - Black was established on the e4-square, but it has
fine well into the middlegame, and no forward movement and does not

even better at one point. His only real
problem was that he was playing Kar­
pov!
12 ltJh3 ltJcs 13 ltJf2
control important squares. One likely
plan for us is ... lbe7-cS-b6 to close the
b-file and pressure White's weakness
on c4. We can also consider ... l:tabS-b6
and ... J:fbS. Black is clearly no worse
and I believe White has to be careful
that he is not punished for his overex­
tended pawns.

A23 : 5 ltJf3 ltJf6 6 ltJc3 i.b4

So far we have followed the famous

game, and Black could certainly con­
tinue more or less like Chevallier (who
tried 13 .. .'ii'd7 h ere), but I would rather
play actively with 13 .. .f5 ! ?; e.g. 14 exfs
i.xfS 15 i.xc5 bxcS 16 i.d3 0-0 17 0-0
'ii'd 7. This is a difficult decision, especially

67
The Dark Knight System

sin ce we will need to know the posi­
tions after 6 ... �cs anyway because of
the different move orders White can
use (see line A2 l for this). Although
Black's bishop is very strong on the a7-
This is not good, but i t is always
played, if it hasn't been played already.
Even Houd;n; likes it ! Black's attacking
ideas should be familiar by now:
10 �h5
•••

gl diagonal, the doubled c-pawns are

also a big problem for White, and one
that cannot possibly be fixed. Not only
are they targets in the late middlegame
and endgame, they immobilize White's
queenside, makin g it difficult for him
to gen erate meaningful play on that
side of the board. Furthermore, White
is forced to play an awkward m ove like
7 .i.d3 or 7 .-c2 to defend his e-pawn.
Because it is possible to inflict the posi­
tional damage and still attack, I prefer Black intends ... �hf4 and ...�h4, or
to double the pawns if possible, though ... �gf4 and ...'iif6. On the queen side, it
this is largely a matter of taste. will probably be necessary to play ... b7-
7 i..d 3 b6 soon, while a2-a4 should be met by
7 �d2 ? ! keeps the pawns from be­ ... a7-aS to eliminate counterplay. In
ing doubled, but it is bad nonetheless: theory White should be okay, but in
7 ... .i.xc3 8 .i.xc3 �xe4 9 �xes �xes 10 reality Black has won every game.
�xes 'ii'f6 11 �d3 0-0 12 .i.e2 d6 13 0-0 W.M.Buehl-J. Benjamin, Reno 1999 (see
.i.fS and Black has a significant advan­ G ame 42) is typical.
tage in activity.
A24: 5 i.. d3 �c5

7 �xc3+ 8 bxc3 d6 9 0-0 0-0 10 h31!

...

68

6 ttJc3
With 6 ttJe 2 ! ? ttJf6 7 0-0 White gets
his knight to e2 without spending a
tempo on h 2-h3 (as compared to the
main line), but he will still find that he
has not found a route to an advantage:
7 ...0-0 (7 ... d6? ! S b4!) S ttJbc3 a6 9 ttJg3
d6 10 ttJfs ttJf4! 11 .i.e3 (not 11 .i.xf4?!
exf4 12 'ir'f3 ? ! g 6 ! 13 ttJh6+ �g7 14
'i'xf4 ttJh s 15 'iWd2 'iVh4! 16 b4 '6'f4! 17
'i'xf4 ttJxf4 winning, because the ttJh6
is still trapped) 11 . . .ttJd7! 12 .i.C2 g 6
with equal play.
6... ttJf6
7 h3
Given an exclamation mark by some
sources, which to me seems a little ex­
treme - the move is more like an ar­
guably necessary evil.
a) Admittedly 7 h3 is far superior to
7 ttJge211 ttJg4 S O-O? 'ifh4, winning.
b) 7 ttJf3 isn't so great either. 7 ...0-0 S
0-0 as (S ... a6 is also reasonable, but the
text move is sound and Black is 5-0 with
it) 9 a3 d6 10 l:tbl ttJhs with a danger­
ous initiative in practice. In A.Hahn­
J.Bonin, New York (rapid) 2003, White
1 e4 ttJc6 2 d4 e5
managed to block Black's favourite
bishop with 11 b4 axb4 12 axb4 .i.b6 13
ttJa4 .i.a 7 14 .i.c2 ttJhf4 15 cS, but Black's
attack was still strong, and he won a
topsy-turvy battle (see Game 43).
c) 7 .i.e3 b6! transposes to A22
above (the first note). 7 ... .i.xe3 isn't the
worst move ever, but it loses th e f4-
square for Black's knights, which love
to settle there and checkmate White.
7 0-0 8 ttJge2
•••
8... a6
Not the only good move. S ... as i s
also worth considering - i t slows down
White's b2-b4, though it does leave
Black's queenside less flexible and
weakens the bs -square.
K.Ellmauer-D.Huber, Schwarzach
2001, went S ... d6 (S ... as is a better
move order) 9 0-0 as 10 �h l?! (10 ttJg 3 )
10. . .ttJh s 11 g411 Wh4?! (11 ...ttJhf4!) 12
�g2 .i.xg4 13 hxg4 1Wxg4+ 14 ttJg3 ? (14
�h2) 14 ... ttJgf4+ 1 5 .i.xf4 ttJxf4+ 16
�g l 'tixg 3+ 0-1. Short and sweet.
lJelen-Z.Mestrovic, Slovenian Team
Championship 1996, saw S ...ttJhS ! ? 9
g3?! (9 ttJa4!), which is incorrectly as-
69
The Dark Kn ight System
sessed by De Firmian as better for White
- a case of annotating by result. The
game continued preposterously with
9 ...'ii'f6?! 10 :h2 d6?! (10 .. . 'iVd8! is equal)
11 lba4 .i.b4+?? 12 lbec3! and Black's
h s-knight is toast.
Going back to move nine, it is true
that White has more space, but his po­
sition is very loose. Simply 9 ... a6! fol­
lowed by ... d7-d6 and ... c7-c6, and what
moves does White imagine that he is
going to be playing ? For example, 9 ... a6
10 0-0 d6 11 �h 2 lbf6 12 :bl c6 13 b4
"d7 14 lbgl .i.d4 15 lbce2 .i.a7 and
White's position is slowly degrading.
9 0-0 d6 10 lbg3!
Otherwise, the nasty 10 ...lbh S! is
coming; e.g. 10 :tbl lbh S ! 11 b4 .i.a7 12
<iith l lbhf4 and ...Vh4 or ...lbh4 with
threats ag ainst g2, h3, and f2 .
(9 ... lbh S ? ! was no good because of 10
lba4! J..a7 11 d6 ! with a small edge and
a large disruption of Black's plans.)
The position after 10 lbg 3 has never
been reached, but we have many rea­
sonable and sensible moves such as
10 ....i.d7, 10.. J:te8, or 10 ...h6 (with equal-
70
ity), and ...c7-c6 may be on the cards.
One interesting idea is ...lbf4. One
thing we absolutely do not need to
worry about is .i.xf4 exf4 - White's
good bishop is too valuable to squan­
der in such a fashion, and our pawn on
f4 is strong, not weak. We will very
much enjoy the use of the es-square
for our knight, bishop, rook, or queen.
Even if ... g7-gS is necessary to defend
f4, Black's kingside will remain safe.
Another idea is ... .i.d4 and ... C7-CS,
anchoring the bishop. In this closed
position, we will not mind if our bishop
gets traded off, especially since we will
wind up with a protected passed pawn
on the d4-square (if the bishop is taken,
either recapture can be considered,
though ... e5xd4 is the typical answer) .
Naturally, White will not be eager to
play dsxc6 (en passant) because of the
loss of space and centre. 10 ... J..d4 is not
the most accurate, but 10 ... h6 11 lba4
can definitely be met by 11 ... .i.d4 and
12 ... c5.
A25: 5 g3

Taking the f4-square away from
Black's knight is sensible, but this move
is slow, and the white bishop is not go­
ing to be active on g2.
5 ttJf6
...
6 i..g2
If 6 ttJc3, we have that choice again
- to take the strong diagonal or to
double White's c-pawns. I prefer to
have a queenside target, particularly
since White has taken measures
against our king side play. Also, White's
fianchetto leaves the c4-pawn without
protection. So 6 ... i..b4 7 i..g 2 i..xc3 8
bXC3 d6 9 ttJe2 (or 9 h4 h6 with a com­
fortably equal position very similar to
A.Karpov- D.Chevallier in A22 - if 10
h s ? ! then 10 ...ttJf8 ! with advantag e)
9 ... 0-0 and Black has won every game.
White's plan to play f2-f4 is far too
weakening - after ... eSxf4 g3xf4, White
has problems on C4, e4, f4, g4, and h4.
While Black is waiting for f2-f4, he can
play ... b7-b6, ... ttJd7, ... ttJcs, ...f7-fs and
perhaps ... a7-as and ... i.. a 6. E.5chien­
dorfer-D.Recuero Guerra, Herceg Novi
2006, is a typical disaster for White (see
1 e4 ttJc6 2 d4 e5
Game 44).
6 i.. c 5 7 ttJC3
...
The popular 7 ttJe2 allows Black to
take the initiative immediately with
7 ... h s ! , when 8 ... h4 cannot be stopped
because 8 h4? ttJg4 9 O-O? (or 9 lIfl
.f6 !) 9 ... ttJxh4! is awful for White, as is
8 i..g s ? i..xf2+.
7 0-0
•.•
7 ... d6, 7 ... as and 7 ... hs are also in­
teresting, but castling is flexible and
leads into the amusing miniature
H.Titz-C.Rossi, Austrian Team Champi­
onship 2001: 8 ttJf3 d6 9 0-0 a6 10 .d3
bs 11 b3 i..d7 12 a4 b4 13 ttJdl 'ii' c8 14
i..e 3 ttJxe4 15 ttJxes ttJxes 16 .xe4 i..fs
17 'it'h4 ttJg6 0-1. White resigned be­
cause 18 'i'hs i.. g 4 19 -g s h6 traps the
queen.
As seen in this game, ... a7-a6 is of­
ten the best way to preserve the bishop
in this variation, because with White's
bishop fianchettoed, the advance ... b7-
bs is easy to achieve and likely to be
effective.
A26: 5 a3 ttJf6 6 ttJc3 i.. c 5
71
Th e Dark Knight System
7 i.d3
Other moves:
a) 7 tiJf3 ? tiJg4! or 7 tiJge2? tiJg4! is
terrible. Have you noticed that White
needs to be careful on the dark
squares? Eh?
b) I've seen White play 7 h3 here, af­
ter which he should have his right to
play White permanently revoked - but
the fact that he always loses here has
as much to do with his obviously pas­
sive attitude as the defects of his posi­
tion. 7 ... d6 8 tiJf3 as 9 g3 0-0 10 i.g2 c6
was C.Baluta-A.Cioara, Bucharest 1996
(see Game 45), in which White, an FM,
got man handled while Black demon­
strated all the available ideas ( ... c7-c6,
... b7-bS, ... tiJg4, .. .f7-fS, etc).
c) 7 b4 certainly looks stupid and
has lost all the games so far, but in
analysis things are not so clear: 7 ...i.d4
8 tiJge2 tiJxe4 (8 ... c s ! ? looks okay) 9
tiJxe4 (not 9 tiJxd4?? tiJxC3 10 'ii'd3 exd4
11 'it'xd4 'iVe7+, winning) 9 ... i.xal 10
d6 ! (stronger than 10 i.e3 ? ! i.d4 11
tiJxd4 exd4, though after 12 'iWxd4! 0-0
13 d6 White still has some play) 10 ... b6
72
11 h4 i.b7 12 dxC7 "xC7 13 tiJd6+ with
full compensation an d more than one
way to pursue th e attack. Ag ainst a
dangerous an d prepared (GM) oppo­
nent, Black would probably do well to
avoid the whole mess with 8 ... cs, but
otherwise it makes sense to pocket the
material and let White try to figure out
how to prove compensation .
7 ...aS
7 . . a 6 has done well in practice, but I
can't see the logic of allowing White to
accelerate his queenside play.
S l:lb1
In stead:
a) 8 tiJge2?? is still terrible due to
8 ... tiJg4 9 o-o ?! -.h4.
b) 8 tiJa4 is interesting, but not es­
pecially strong: 8 ... i.a7 9 cS d6 10 i.e3
was R.5taudte-M. Roth, Chemnitz 1998,
when 10 ...tiJh4! (instead of 10 ... tiJg4?! )
11 �fl tiJg4! 1 2 i.d2 dxcs already
leaves White in poor condition. 9 d6 is
a little better, but 9 ... cxd6 and 10 ... ds
will be sufficient for easy equality.
S ... d6 9 h3 0-0 10 b4 axb4 11 axb4 i.a7
Black was already slightly more com-

foTtable in Bu Xiangzhi- L.Christiansen,
Deizisau 2000 (see Game 46), and Bu's
12 g 3 ?! should have worsened White's
situation after 12 ... c6 or 12 ... .i.d7 (as
opposed to Christiansen's 12 ...ltJe8?!).
Black, who was better throughout most
of the opening and early middlegame,
.. .lost.
B: 3 dxes ltJxes
White has:
B 1 : 4 ttJf3 73
B2: 4 f4 76
Instead, 4 ltJc3 .i.c5 5 ltJf3 ? (5 f4 ltJc6
transposes to line B22 below) 5 ... ltJg4! 6
ltJd4 is like a Two Knights Defence (1 e4
e5 2 ltJf3 ltJc6 3 .i.c4 ltJf6 4 ltJg 5 d5 5
exd5 ltJxd5) with reversed colours, ex­
cept that Black's king's knight is at
home instead of his queen's knight. This
makes it harder to pressure the ltJd4, so
the standard idea of 6 ... ltJxf2 is less ef­
fective; e.g. 7 Wxf2 'ii'f6+ 8 �e3 ltJe7 9
ltJb5 �e5 10 c3 f5 11 Wd2 fxe4 12 �C2
0-0 13 g3 d5 14 .i.f4 does not give Black
1 e4 ltJc6 2 d4 e5
enough compensation. However, Black
still has 6 ... d5 !?; e.g. 7 ltJxd5 c6 8 f3 ltJe5
9 b4 .i.d6 10 ltJc3 .i.xb4 11 i-b2 1i'h4+
12 g3 'iV e7 and White is very loose. Al­
ternatively, 6 .. :ifh4 is less ambitious but
much simpler: 7 g3 '6'f6 8 iVxg4 .i.xd4 9
ltJdl d5 ! (since the c8-bishop is de­
fended!) 10 _e2 dxe4 11 iVxe4+ ltJe7
and Black is a bit better.
B1: 4 ltJf3 ltJxf3 5 'ii'xf3 ltJf61
The usual move is 5 ...'i¥f6, but then
White can either allow or avoid the
queen trade as he pleases, with a
pleasant position in either case. Al­
though the text move has had no seri­
ous trials, analysis demonstrates its
viability an d none of the ideas previ­
ously tried have proven themselves
reliable. In m ost cases, we intend to
play like a Ph ilidor, counting on the
knight trade to ease our space disad­
vantage, while trying to show that
White's queen is awkward on f3; and
there are other possibilities depending
on how White reacts. It is not possible
to analyse exhaustively, but the follow-
73
The Dark Kn ight System

ing information should leave us a few

steps ahead of the opponent, who has
never even considered this position .

811: 6 l2JC3 74
812: 6 �C4 74
813: 6 e 5 75
814: 6 �d3 75
815: 6 �e2 75
816: 6 �g5 76

811: 6 ttJC3 .i.b4 7 .i.d3 .i.xc3+ 8 bxc3 6 ... .i.d6

0-0 9 0-0 d6 10 :e1 6 ... d6 ! ? is not bad, and keeps things
very Philidor-esque, but I prefer the
bishop to be more active if possible.
7 ttJC3 0-0 8 .i.gs
Or 8 0-0 'fie7 9 .i.f4 (9 .i.g S ? ! 'iVes !
1 0 "g 3 'lVcs 1 1 'i'h4! �es ! i s level)
9 ....hf4 10 .xf4 d6 11 :adl :e8, in­
tending ...�e6 or ... 'iVes . The exchanges
have eased Black's space problem, and
White's edge is tiny.
8 ... c6 9 i..xf6 'ii'xf6 10 "xf6 gxf6 11
0-0-0 .i.cS 12 f3 bS 13 .i.d3 d6
10 �g s h 6 11 �xf6 'ii'xf6 12 'Wxf6
gxf6 is an equal endgame.
After 10 :el, White is a little better
in spite of his bad pawns. It is time for
Black to reposition the knight with
lo ...ttJd7, which steps out of the poten­
tial pin, prepares •f6, and eventually
...

... ttJcS or ...ttJes, where it should have

g ood prospects. Black further intends
... b7-b6, ... �b7, and ... :te8, targeting
White's e-pawn, and may eventually
try for .. .f7-fS. Black's powerful dark-squared
bishop is (nearly) enough to balance his
812: 6 �C4 ugly pawn structure.

74
 1 e4 lbc6 2 d4 e5

813: 6 eS 'ife7 On the surface, this looks bad for

Black, who is still four or more moves
away from completing his develop­
ment - but he controls the e6-square,
and White lacks the means to orches­
trate a breakthrough, so White's ad­
vantage is small; e.g. 11 'iitb l .i.fs
(ll ...h S ! ?) 12 lbb3 lbxb3 13 'iWxb3 0-0-0
14 .i.d3 -'.e6! with no real problems.

814: 6 .i.d3 d6 7 0-0 .i.e7 S lbc3 0-0 9

.i.f4
7 -'.f4 9 lbds lbxds 10 exds .i.f6 11 c3 :e8
7 -.e2 lbds 8 c4 'iVb4+ 9 lbd2 lbf4 10 gives White very little.
..w e4 lbe6 is equal. 9....i.g4 10 'ii'g 3 -'.e6
7 ... c6 s lbd2
Other moves lead to equality:
a) 8 ..-e2 lbds 9 -'.g 3 fs ! 10 c4 lbC7
11 lbc3 lbe6 12 f4 d6 ! 13 0-0-0 filc7 14
..-c2 g6 15 'iitb l dxes 16 fxes -'.g7 (or
16 ... -'.e7 or 16 .. .f4 !?).
b) 8 lbc3 d6! 9 0-0-0 dxes 10 l::t e l
lbd7 11 -'.c4 f6 and Black keeps his ex­
tra pawn, although White has the ap­
propriate compensation. Black intends
... lbb6 (or ... lbcs), ... .i.d7 and ...0-0-0.
S ... d5 9 0-0-0 lbd7 10 filg3 lbcs The threat to trade off White's good
bishop with 11 ...lbh s is annoying. As­
suming White defends against that,
Black should consider the candidate
moves ... c7-c6, ... :e8, ... lbd7 (or ...lbg4),
... lbes, ... lbxd3, ... -'.h4, ... -'.f6, and ... ..wC7
(or ... 'iWe7). White is only slightly better.

815: 6 -'.e2 dS 7 eS ltJe4 S lbC3 lbxc3 9

'ii'x c3 c6 10 .i.e3 .i.e7 11 f4 0-0
And with 12 .. .f6! coming, Black has
equalized.

75
Th e Dark Kn ight System
816: 6 i.. g 5 d6 7 lbc 3
As usual 7 i..xf6 ? ! 'Wxf6 8 'ifxf6 gxf6
is equal.
7 ... i.. e7
In principle, we can already notice
that, compared to a regular Philidor,
White's i..g s and 'tWf3 are awkwardly
placed, so White cannot expect much
from the opening; e.g. 8 h 3 0-0 9 0-0-0
c6 10 "'g 3 (10 eS lbe8 doesn't do any­
thing, while after 10 i..e2 "as 11 i.. d2
'iII C7 12 '6'g 3 bS 13 i..h 6 lbe8 Black is
certainly no worse) 10 ..... aS 11 f4 .ie6
leads to a wild race with approximately
even chances.
76
82: 4 f4
This is far less popular than 4 lbf3
among the strongest players, but gives
excellent results. Fortunately, it's easy
to see where Black has been going
wrong.
4... lbc6
White has:
821: 5 i..c4 77
822: 5 tLlf3 78
Alternatively:
a) S lbC3 i..cS 6 tLlf3 transposes to
line B22.
b) S i.. e 3 prevents S ...i.. c S, but Black
is okay after S ... i..b4+ 6 c3 i.. a s
(R.Hilbner-V.Hort, German League
1984, see G ame 47). Even better is the
unplayed novelty S ... d S ! which brings
equality; e.g. 6 exds lbb4 7 i..b S+ .id 7 8
i.. x d7+ 'irxd7 9 lbf3 lbxdS 10 'ire2 0-0-0
- come to think of it, White is m uch
worse here, so 8 '6'e2 lbXC2+ 9 '6'xC2
.ixbS 10 lbc3 i..a 6 11 lbf3 i.. d6 12 0-0-0
lbf6 13 l:[hel 0-0, when Black's bishop
pair balances White's space advantage.
 1 e 4 liJ c 6 2 d4 e 5

821: S i.C4 liJd5+ 2 3 �c4 liJe3+ 2 4 �c3 with a draw.

Preventing S ...i.cS because of 6 6 i.b4 7 liJe2
..•

i.xf7+ �xf7 7 'iVdS+. Here 7 e5 d5 8 exf6 dxc4 9 'ii'e 2+?!

S llJf61
... �8 ! ? (9 ...i.e6 is also good) 10 i-e3
'ii'xf6 11 0-0-0 i.e6 was better for Black,
who went on to win in M.Orso­
G .Bordas, Budapest 2000 (see Game
48). G o Bordas! 9 'ii'xd8+ improves for
White, when 9 ... ciif xd8 10 fxg 7 :g8 1 1
i-e3 i.xc3 1 2 bXC3 l1xg7 13 cM2 i..f5 1 4
:dl+ �c8 15 :d2 b 6 i s equal.
7 ...liJx e41
7 ... d6 is equal; Godena finds more.
S i-xf7+ �xf7 9 �dS+ �fS 10 'iVxe4 dS

This strong move, from the IM-GM

clash L.Shytaj-M.Godena, Italian Cham­
pionship, Marta Franca 2008, leads to
very sh arp play, not unfavourable to
Black. Before this, S ...i.b4+ was the
standard reply, as played by E.Pedersen,
Przewoznik, Miles and Hort, who scored
one draw between them ! (It was Miles.)
6 liJc3
Shytaj's choice.
Only 6 eS is critical, but Black has A sharp reversal of fortune has oc­
sufficient resources: 6 ... dS! 7 i.b3 liJg4 8 curred in the centre, and this is far
i.xd5 i.b4+ 9 liJc3 (or 9 c3 i-cS 10 more important than any minor incon­
i.xc6+ bxc6 11 'ii'xd8+ �xd8 12 liJe2 venience suffered by Black's king .
i.f2+ 13 Wfl i.b6 14 liJd4 f6 15 exf6 11 .f3 'ii' h4+! 12 g3 i.g4!
i.a6+ 16 �gl l:te8 17 i-d2 :e4 18 fxg7 After seeing this, Godena is my new
We7 19 g 3 i.d3 with dangerous play - hero.
Black's light-squared bishop is an abso­ 13 'ifd3?
lute monster) 9 ... 0-0! 10 i-xc6 'ii'xdl+ 11 White, who is suffering badly, may
<ii? xdl bxc6 12 �e2 f6 ! 13 h 3 liJh6 14 as wen take a pawn for his troubles. 13

liJf3 liJf5 15 liJe4 fxes 16 fxeS i-e6 17 'ii'xds is not clearly losing, whereas the
i.d2 i.xd2 18 �xd2 lIad8+ 19 WC3 liJe3 text move is.
20 liJd4 i.dS 21 liJcs i-xg2 22 lIhgl 13 .. :ifhS 14 0-0

77
The Dark Kn ig h t System
14... d4?1
14 ... :e8! allows less counterplay: 15
.i.e3 d4! 16 tt:)xd4 J::t x e3 17 'iiVx e3 tt:)xd4
18 'iVd3 tt:)f3+ 19 :xf3 .i.xf3 .
15 f51 dxc3 16 tt:)f41 'iVf7 17 tt:)e6+ �g8
18 bxc3 i.d6 19 'iiVe4 tt:)e5 20 'ii'x b7 :e8
21 tt:)d4 cS 22 'ii'a 6 cxd4 23 'ii'x d6 tt:)f3+
24 �g2 'irb7
White has had enough. Did I men­
tion that G odena is my new hero?
822: 5 tt:)f3
Rare at a high level, this lazy-looking
move is White's best, apparently doing
nothing to contest the a7-g1 diagonal.
5 ... i.c5 6 tt:)c3
78
6 ...d6
Instead:
a) 6 ... a6? ! 7 "e2 ! d6 8 .i.e3 .i.xe3 9
'ii'xe3 tt:)f6 10 0-0-0 is pleasant for
White.
b) 6 ... tt:)f6 ! ? leads to wild complica­
tions and is fully playable if you enjoy
such positions. Here are some sample
variations: 7 e 5 ! tt:)g4 8 .i.C4 (or 8 tt:)e4!
.i.b6 9 i.c4 d5 10 'it'xd5 'ire 7
{10 ...'ii'xd5 ! ?} 11 h 3 i.e6 12 'W'b5 a6 13
'ii'a4 i.xc4 14 'ii'xc4 tt:)e3 15 i.xe3 i.xe3
16 g3 0-0-0, when Black has compensa­
tion for most of a pawn) 8 ... d6 9 tt:)g5
(or 9 tt:)e4 i.e3 10 exd6 0-0 {10...i.XC1 !?}
11 i.xe3 tt:)xe3 12 'iWe2 tt:)xc4 13 'ii' xC4
cxd6 14 0-0-0 i.e6 15 'i'b 5 d5 16 fs a6
17 'Wcs i.xfs 18 :xds 'We7 19 tt:)d6 i.e6
20 J::t d 2 i.xa2 ! 21 b3 'ii'f6 22 tt:)d4 as)
9 ...0-0 10 h3 tt:)h6 11 tt:)a4 i.b6 12 tt:)xb6
axb6 13 0-0 dxes 14 'ii' xd8 :xd8 1 5
fxes tt:)xes 16 i.b3 cS 17 i.f4 c4 with
equality.
7 tt:)a4! i..b6 8 tt:)xb6 axb6 9 i..d 3 tt:)f6
9 ... ds transposes to s. Fedorchuk­
A.Miles, Ohrid 2001, which Miles won
(see Game 49), but the idea does not

merit the exclamation mark bestowed
by Kalinin - the simple text is best.
10 0-0 0-0 11 lItel
So far we have followed M.Heyne­
R.Vogel, Passau 1999, which actually
started as an Englund G ambit! (1 d4
es?!). Instead of l1 .. . :eS (which could
be met by 12 es !), Black should have
played 11 .. .liJb4, an annoying attack on
White's bishop pair, which he can no
longer preserve (12 i.. c4 i..e 6!). After 12
i..d2 liJxd3 13 cxd3 Black's plan is to
mobilize his queenside majority with
13 ... cS, ... b6-bS -b4 and ... b7-bS (again!),
develop the bishop with ... i.. d 7-c6, and
restrain White's centre by ... :leS. It is
also important to prevent i.. g s , so if
White plays f4-fS, then ... h 7-h6 is the
normal response. I cannot tell a lie -
White has a tiny edg e. Notice that
13 ... cS 14 a4 b S ! is tactically possible
(and desirable) because 15 axbs l:Ixal
16 Vixal i.. d7 recovers the pawn and
activates the bishop.
If White prevents ...liJ b4 with 11
i..d2, the answer is 11 ... dS! 12 eS liJe4
13 i..e l liJcS 14 fS liJxd3 15 'iVxd3 Vie7
1 e4 liJc6 2 d4 e5
16 i.. c 3 'ifcS+ 17 �h l 'ii'C4 with equal­
ity.
White has also tried 11 b3 ! ? i.. d7 12
i..b 2 liJb4! 13 'iVd2 liJxd3 14 cxd3 cS 15
fS i.. c 6 16 '6'g s h 6 17 i..xf6 'ii'xf6 lS
'ii'xf6 gxf6, when Black had equalized in
lR.Capablanca(!)-M.H.McGuire, New
Orlean s (simul) 1911. Black went on to
win, outmanoeuvring Capablanca with
his better minor and queenside pawn
majority. Outrag eous !
C: 3 liJf3 exd4
Now White has:
Cl: 4 lDxd4 80 the Scotch G ame
-
C2: 4 i.. c4 88 the Scotch Gambit
-
If 4 c3 (the Goring Gambit), we de­
cline with 4 ... dS ! S exds .-xdS 6 cxd4
i..g 4 7 i..e 2 i..b4+ S liJC3 i.. xf3 9 i..xf3
�C4 and White, who can't yet castle,
has done terribly after 10 'i'b3 .-xb3 11
axb3 with vile pawns, or 10 i.. x c6+
bxc6 ! 11 ii'e2+ 1i'xe2+ 12 �xe2 0-0-0 1 3
i..e 3 liJe7, when Black's king is safe and
guards his weaknesses. White should
79
Th e Dark Kn ight System
really be okay, but has had serious
problems in practice. One good idea for
Black is ... .ta5-b6 to pressure White's
d4-pawn and shore up the queenside.
Cl: 4 ltJ xd4 .t cS!
We are allowed to play the normal
move on occasion . There's always time
for weirdness later.
White's main moves are:
Cl1: 5 .ie3 81
C12: s ltJxc6 85
First let's take out the trash:
a) 5 ltJb3 ? ! leaves our bishop uncon­
tested on a strong diagonal: 5 ... i.b6 6
a4 a6 7 ltJc3.
The position is highly reminiscent of
a Caro-Kann Classical Variation (in mir­
ror image), but with a few advantages
for us. We have saved a move on ... c7-
c6 (i.e . .. .f7-f6 here), it is hard for White
to trade our strong dark-squared
bishop, and we will find castling short
to be far more efficient and effective
than castling long in the Caro (where
80
.. .c�b8, ...:c8, ... c6-C5 i s time consum­
ing) .
If anyone tells you to let White play
8 ltJd5, don't believe them: 7 ...ltJf6 !
(7 ... ltJge7 is good too, transposing to 6
ltJc3 ltJge7 below, but not the insipid
main variation with 7 ... d6?) 8 .tg 5 (8
i.e2 is preferable) 8 ... h6 9 i.h4 d6 and
Black is better already.
Apparently Magnus Carlsen has h ad
some good results with 6 'ife2 d6 7 i.e3
but, given the above comparison, it is
clear that this should not be danger­
ous. Black simply needs to keep in mind
the .. .f7-f5 break: 7 ...ltJge7! 8 ltJC3 0-0 9
0-0-0 f5 ! - a few people have noticed
this, and Black is +7 -4 =2 in this posi­
tion (most of the games reach ed by
transposition) . G oh Wei Ming-F. Bellini,
Turin Olympiad 2006, is a nice win by
Black (see Game 50).
A cagier move order which is some­
times used is 6 ltJC3, when it is not yet
clear whether White will be playing a2-
a4 or .-e2 . It is more important to be
prepared for the fashionable 'iVe2, so
6 ... ltJge7 7 'iWe2 d6 8 i.e3 transposes to

the previous paragraph, while 7 a4 a6 S
i..g 5 f6 9 i..h 4 0-0 is absolutely fine for
Black, who intends ... d7-d6, ...ltJg6 and
.. .f6-f5, or else ... d7-d6, ... ..te6, ... 'iVd7 (or
...'iVeS) an d .. .f6-f5 .
b) 5 ltJf5? ! performs well if Black
doesn't know his stuff (he usually
doesn't) . Surprise the surpriser by
knowing this short variation : 5 ... d 5 ! 6
ltJxg 7+ wfS 7 ltJh 5 'ilh4 S ltJg3 ltJf6 ! 9
i.. e 2 dxe4 and Black has a small advan­
tage with which he almost always
wins.
c) 5 C3?! indicates that White has
forgotten the move order. Don't lazily
transpose with 5 .....f6? ! 6 ..te3, but
play 5 ... ltJf6! instead, with advantage.
(11: 5 i..e 3 'iff6
6 c3
White can try 6 ltJb5 here, but after
6 ...i.. x e3 7 fxe3 'ilfh4+ (this intermezzo
forces a concession) S g3 "dS 9 "g4
g 5 ! 10 ltJlc3 ltJe5 11 'ii'e 2 d6 12 h 3 c6
13 ltJd4 ltJf6 14 0-0-0 'ile7 15 'ii'f2 ..te6
16 ..te2 0-0-0 Black was obviously fine
and went on to win a marathon game
1 e4 ltJc6 2 d4 e5
in P.Bontempi-OJovanic, Nova Gorica
200S (see Game 5 1).
6 ...'ifg61?
From here the queen threatens e4,
pressures g2, and clears the f6-square
for the knight. Also, we steer clear of
the most heavily analyzed continua­
tions. Black has done very well with this
move!
There are two main replies, and a
bunch of minor ones.
(111: 7 t2Jd2 82
(112: 7 tLJb5 84
Others:
a) The unlikely-looking 7 'ile2 has
also been popular. White hopes for
7 ... 'iVxe4?? S ltJxc6 ..txe3 9 ltJd4, win­
ning (this has yet to work, but hope
springs eternal). Instead, 7 ... ltJxd4! S
cxd4 (or S i..xd4 i..xd4 9 cxd4 ltJe7 10
ltJC3 0-0 11 0-0-0 c6 and with 12 ... d5
coming, Black is slightly better)
S ... ..tb4+ 9 ..td2 ..txd2+ 10 ltJxd2 ltJe7
11 g 3 ! (to discourage 11 ... d5) 11 ...0-0 12
..tg 2 'i'b6 and White will have compen-
81
Th e Dark Kn igh t System
sation for the pawn he is losing, but no
more than that.
b) 7 f3 a6 ! ? (this is a novelty - it's
time to put a stop to tLlbs once and for
all) 8 tLld2 d6 9 'ifc2 tLlge7 10 0-0-0 i..a7
11 �bl 0-0 is equal; or 8 'ifd2 tLlge7,
transposing to S.Vajda-S.5kembris,
Naujac 1999, which continued 9 tLlC2
i..x e3 10 tLlxe3 d6 11 i.. e2 and Black
found plenty of activity with 11 .. .fS,
going on to win (see Game 52).
c) Dembo and Palliser (D&P) like 7
'ii'f3 pretty well,
but after 7 ... d6 8 tLlbS i.. g 4 9 'iff4
i.. x e3 10 'ii'x e3 :c8 I don't agree that
White is better; e.g. 11 f4 tLlf6 12 i..d3
0-0 13 0-0 l:tfe8 14 tLld2 i.. d7 and Black
has no particular problems. 15 tLlxa7 ? !
is met by 15 . . .tLlg4 16 fS 'iVh 5 17 'iVg 3
tLlxa7 18 :f4! 'Wxh 2+ ( 1 8. . .tLle5 19 :h4
'ifxh4 20 'ii'x h4 tLlxd3 21 f6 is less clear)
19 'ii'xh2 tLlxh2 20 �xh 2 tLlc6 with a
small but persistent endgame advan­
tage for Black due to White's backward
e-pawn, weak e5-square, and defensive
bishop.
d) D&P also like 7 i.. e 2, and it is in-
82
deed dangerous to try to win a pawn,
but in E.5tavropoulou-M. lkonomo­
poulou, Athens 2003, Black found an
excellent alternative: 7 ...tLlf6 ! 8 tLld2 (8
0-0 d6 9 It>hl i.. d7 10 tLld2 0-0 is equal)
8 ... dS ! 9 exds tLlxds 10 i..f3 tLlxe3 11
fxe3 tLle5 ! with a clear advantage.
(111: 7 tLld2
This is pathetic, but it's played most
of the time, so I guess that makes it the
main line! I can't imagine why White
would play the Scotch if this is his plan
for dealing with 6 ... 'i¥g6. The following
draws significantly on Dembo and Pal­
liser.
7 tLlf6
•••
8 .f3
Other moves:
a) 8 f3 is more common, when Black
should strike in the centre with 8 ... d5! 9
i..b 5 ! (stronger than 9 .c2 ? ! i..xd4 10
cxd4 0-0 11 0-0-0 dxe4 12 fxe4 :e8 13
a3 tLlxe4 14 i.. d3 i..f5 15 dS tLld6 16
i..xfS 'ii'xf5 and Black keeps an extra
pawn; or 9 tLlbS ? ! i..x e3 10 tLlxC7 It>f8
11 tLlxa8 dxe4 12 'ii'e 2 'ii'h 6 13 tLlxe4

liJxe4 14 fxe4 liJes and White needs
both to survive and to extricate the
knight, which is more hard than easy)
9 .. . 'ii'x g2 10 :g l 'i¥xh 2 11 :txg7. Up un­
til now we've been following E.Berg ­
I ,Morovic Fernandez, European Cham­
pionship, Saint Vincent 2000, which
continued 11 ... �d7 12 'ib3 1lfh4+ 1 3
�dl and O& P says favours Black, but
Houd;n; calls even, and Black did go on
to lose. Instead, I am recommending
11 ... �f8 ! ? 12 :g l liJxd4 13 cxd4 i.. e 7 14
'ifb3 c6 15 es! (after 15 i.. d3 dxe4 15
fxe4 Wh4+ 16 .tf2 "f4 17 Ilfl i.. e 6,
White's compensation is insufficient)
lS ... cxbS 16 exf6 .-h4+ 17 �e2 'ii'xf6 18
"xdS Vi'fs 19 'ii'e4 'ii' x e4 20 fxe4 hS 2 1
liJf3 i.. g 4 with approximate equality i n
an unbalanced endgame.
b) 8 i.. e 2 dS ! transposes to Stavro­
poulou-Ikonomopoulou a few para­
graphs above. I must say it's interesting
that Black so rarely captures on e4 or
g2. However, 6 ...Vi'g 6 has other points
to it, and White can't leave those
pawns hanging forever.
c) 8 "C2 ? ! liJg4! 9 0-0-0 liJxe3 10
fxe3 0-0 11 liJ2f3 d6 12 liJfS :te8 and
Black is on top O&P.-
d) 8 'i¥e2 liJg4 9 liJc2 liJxe3 10 liJxe3
liJe7 11 liJb3 i..b 6 12 liJfs "f6 13 liJxe7
"xe7 14 g3 0-0 15 .tg 2 d6 with a small
advantag e in R.Roszkowski-A.Leniart,
Grodzisk Mazowiecki 2007 O& P. -
e) 8 liJfs (or 8 liJc2) 8 ... i.. x e3 9 liJxe3
0-0 10 i..d3 dS 11 exdS 'i¥xd3 12 dxc6
:e8 13 liJb3 iVbs 14 a4 occurred in
J.Hoogendoorn-A.Van de Oudeweeter-
1 e4 liJc6 2 d4 e5
ing, Leeuwarden 2001, which was
equal after 14 ... 'iVxc6 15 0-0 (O&P), but
Black can keep some pressure with
14 ..... a6 !, stubbornly denying White's
castling rights. After 15 'i¥e2 'i¥xc6 16
0-0 liJg4 17 liJd4 'iVd6 White will soon
have an isolated e-pawn and a some­
wha t worse position.
f) 8 f4 has worked very wen for
White in practice, so be careful ! The
careful response is 8 ... liJxd4 9 cxd4 i..b4
10 fS "g4 11 "xg4 liJxg4 12 i..f4 dS 13
h3 liJf6 14 eS liJe4 15 g4 g6 16 fxg 6
fxg 6, which is slightly better for Black
because of his imminent f-file control,
though it was not enough to win in
A.Motylev-S.Gligoric, Yugoslav Team
Championship 2000 (see Game 53).
g) 8 h4 hS ! is not helpful to White in
any variation.
8'HliJg4! 9 liJfs
The unplayed 9 liJbS ! ? is a better try,
though after 9 ...liJxe3 10 fxe3 �d8, the
position is balanced. And imbalanced.
Go figure.
9 liJxe3 10 liJxe3 0-0 11 liJds liJes 12
...
Vi'g3
83
Th e Dark Kn ight System
So far this is B.Kharashkina­
O.5tjazhkina, St. Petersburg 2001, and
now instead of 12 ... tiJg4?!, Black should
have preferred 12 ... 'ifxg 3 13 hxg 3 c6,
which D&P call "at least equal", though
I think is safe to call a small advantag e
for Black.
(112: 7 tiJ bS!?
This is the only critical move, but it
is rarely played.
7 ... .i.xe3 8 tiJxc7+
Just as often White wimps out with
8 fxe3 �d8, when White's structural
problems are far more serious than
Black's king position . Black win s virtu­
ally every game; e.g. 9 tiJd2 tiJf6 10 'it'f3
d6 1 1 h3 lie8 12 0-0-0 .i.d7 and Black
can capture the e-pawn at his conven­
ience. IM Rathnakaran somehow got
caught behind the white pieces in this
variation and tried 10 'ifb3 tiJg4!? 11
O-O-O? tiJf2 12 .i.e2 tiJxhl (K.Rathna­
karan-P.Negi, Indian Ch ampionship,
Mangalore 2008), but the game is far
too silly to put in my book.
8... �d8 9 tiJxa8 .i.f4! 10 'it'f3 .i.h6!
84
White's queen is lousy on the f3-
square - it can't reach the queenside,
ifs vulnerable to ... .i.g4, and makes it
impossible for White to guard his e­
pawn securely with f2-f3 . The position
is still extremely complicated, but ap­
proximately equal.
Some more analysis follows, though
it is hardly practical to be exhaustive,
so here are some bits to keep in mind:
1. The material imbalan ce (two mi­
nor pieces for a rook, once the a8-
knight drops) is inherently the most
useful for Black in the middlegame -
the minor pieces are well suited for
attack. This means that Black should
not be eager for trades even though his
king is a bit loose.
2. If an endgame is reached, it is far
better for Black to retain his single rook
than to trade it for one of White's two
rooks.
3. Pawn exchanges favour White,
whose rooks have the most to gain
from an open board.
4. Naturally, Black will try to keep
his bishop pair if possible. Even deep in
 1 e4 ttJc6 2 d4 e5

an endgame, two bishops are normally shows that in the endgames that are
equal to a rook and two pawns. normally reached, White sometimes
Black h ad a chance to show the wins and just about never loses. Yuck!
power of points #2 and #4 in Don't worry, it's all taken care of.
P.Hromada-L.Ostrowski, Moravian
Team Championship 2002 (see Game
54).
In D.Campora-V.Tkachiev, Biel 1995,
White tried 11 'it'f5 �xf5 12 exfS b6 13
�a3 i.b7. As I just mentioned, a queen
trade generally favours White, but the
price was too high: a tempo, a crippled
pawn majority, and a weakened centre
position. Black had no problems after
pocketing th e knight, and had winning
chances, although the game ended in a 5 - 'i'f6 6 -.d2
draw (see Game 5 5). Apparently, 6 �f3 ! ? is topical at the
Dembo and Palliser recommend 11 moment: 6 .....xf3 7 gxf3 bxc6 8 i.e3
.1e2 ttJf6 12 0-0 ttJxe4 13 i.d3 (end of i.xe3 9 fxe3
analysis), but Houd;n; greatly prefers
Black's position after 13 .. .fS . Naturally,
Black will play ...b7-b6 and ... i.b7 at his
earliest convenience; e.g. 14 ttJ a3 ttJes !
(14 ... ttJd2 1 5 'iYdS i s not worth it) 1 5
jfh3 (not 15 'i'e2 ? i.f4! 1 6 g 3 ? ! ttJg S ! 1 7
f3 1:te8 with a brutal attack) lS ... b6 15
:adl i.b7 and it's still complicated,
but Black is trapping the kni ght, acti­
vating his pieces, generating threats
against the enemy king, and not being
checkmated, which adds up to a n ex­ At first glance the position looks
cellent position. completely equal, which goes to show
you that sometimes fi rst glances are
(12: 5 ttJxc6 dead on. In practice White has a nag ­
This move is popular among Ex­ g i n g edge (+13 -5 = 2 1), but that's only
tremely Boring GMs and people with because Black h as not found the cor­
no idea what's going on . (I'm thinking rect plan until this very moment: 9 ... d6 !
of taking it up myself.) Experience 10 ttJc3 (or 10 :gl g6) 10 ...ttJf6 ! (the

85
Th e Dark Kn ight System
novelty) 11 �f2 (or 11 0-0-0) 11 ...0-0 12
l:tdl l:te8.
The point is to immobilize White's
centre pawns and prepare ... nes, from
which the rook has quite a pleasant
view! One obvious idea is .. .l::t h S, while
under some circumstances ... d6-dS can
be played (especially if White has cas­
tled long). Black's activity is quite suffi­
cient to neutralize the practical advan­
tage White has been getting after
9 . ..l�Je7?!. This idea was inspired by
I .Grynfeld-A.Bisguier, Helsinki Olympiad
19S 2 ! (see Game S6), though Bisguier in
fact missed his chance to play ... l:tes.
6 ...dxc6
86
7 lbc3
Or:
a) 7 -.f4 -'.e6 8 lbc3 0-0-0 9 -'.d3
h 6 ! ? - kingside expansion with 10 ... g S
will be useful whether the queen s are
traded or not. In any case, Black has
activity in exchange for his crippled
queenside majority. For the adventur­
ous, 9 ... h S ! ? is also possible: 10 0-0 h4
11 h3 lbe7 12 'irxf6 gxf6 13 lbe2 :hg 8
with more weaknesses for more activ­
ity.
b) 7 -'.d3 .te6 and (to make a long
story short) Black gets to castle long,
with excellent development.
7 ... -'.d4!
White intended 8 -.f4 to enter an
annoying endgame. The text move
stops this insidious and somnolent
plan and equalizes, according to
Dembo and Palliser. In practical play,
Black seems to do even better.
8 -'.d3 lbe7 9 0-0 lbg6
Black's knight is taking the scenic
route to the g4-square, where it will be
dangerous to the white king. If Black
had tried to take a shortcut with
 1 e4 ltJc6 2 d4 e5

8.JtJh 6, White would have put a stop

to it immediately with 9 h 3 ! .
10 wh1
Since the main variation offers noth­
ing, White sometimes tries 10 ltJe2 ! ?
which frees h i s position after 10...i..b 6.
So far, taking the pawn has performed
badly for Black, but it is the critical
move, and I do not believe in letting
White off so easily: 10 ...i..xb2 11 i..xb2
"xb2 12 f4 'li'a3 13 f5 ltJe5 14 "g5 i.. d7
is equal, as in J.Smeets-A.Beliavsky, Mar­ 11 i.. e 2 ltJg4 12 ltJd1 .d6
ibor (rapid) 2004 (see G ame 57). Even 12 ... i.. e 6 is fine too, intending to

better is 14.....f8 ! followed by 15 .. .f6, castle long.

when White does not have quite
enough for the pawn. If 15 f6 ? ! "c5+ 16
'it>hl g6, the threat of 17 ... ltJxd3 gives
Black time for ...i..d7 and ...0-0-0.
1 3 i..xg4
After 13 g 3 h5 14 c3 i..b 6 15 'ii'g 5
'ife5 1 6 f3 Vi'xg 5 1 7 i..xg 5 ltJe5, Black
had the initiative in D.Pavasovic­
N.V.Pedersen, Bled Olympiad 2002; or if
13 f4, Black has the pleasant choice
between 13 ...ltJxh2, with another per­
petual, and 13 ...h 5 ! ?, trying for more
(Dembo and Palliser).
13 ... i..xg4

10 ...ltJ e5
The knight is strong, but if White
tries to dislodge it with 11 f4, then
11 ... ltJg4 12 'irel (12 ltJdl?? ltJxh 2)
12 ...ltJxh2 (in B.5ultimov-N.Pokazanjev,
Russia 2007, Black tried for and got
more with the risky 12 ... i.. d7 ! ? - see 14 c3 i..f6 15 'ii' xd6 cxd6 is about
Game 58) 13 '1t>xh2 'it'h6 + 14 <ltg3 "g6+ equal, while the common 14 .g 5 'ii'g 6
is a draw by perpetual. is a little better for Black.

87
The Dark Knig h t System

C2: 4 .i.c4 compensation for the sacrificed mate­

The Scotch G ambit. rial.
4... lLlf6! 9 Jie7 10 lLlxd4
•••

The Two Knights Defence. I have 10... fSI 11 :f4

chosen sound sidelines for Black to
simplify the study material.
White has:
C21: 5 0-0 88
C22: 5 eS 8 9
Or S lLlg s dS! 6 exds 'ii'e 7+ 7 �fl
lLles 8 'ifxd4 lLlxC4 9 'ifxc4 h 6 10 lLlf3
"cS 11 'iVe2+ Jie7 12 c4 lLlxdS
(12 ... b S ! ?) with an edge for Black.
Or 11 Jih 6 ! ?, when Black can take a
draw with 1 1 .. .fxe4 12 Jixg 7 l:tf8 13
'ifhS+ lU7 14 l::t d l 'ifd6 l S lLldbS 'ii'f4 1 6
lLldS 'iWxf2+ 17 �hl Jid 7 18 lLlf6+ Jixf6
19 lLlxC7+ �e7 20 lLldS+ �e6 2 1 lLlc7+,
or play on with 11 ... 0-0 12 lLlxc6 bxc6
13 l:td4 "e8 14 Jif4 .i.f6 with equal
chances.
11 ... 0-0 12 lLlxc6 'i¥xd1+ 13 lLlxd1 bxc6
14 :c4 cS 1S .i.e3 l:te8

C21: S 0-0 lLlxe4 6 �e1 dS 7 JixdS 'ii'xd s

8 lLlc3 "d81?
This is rare but sound. The idea is to
make it more difficult for White to re­
cover the d-pawn.
9 l:txe4+
Just as common is 9 lLlxe4 ? ! Jie7 10
Jig s f6 ! and White, down a pawn and
lacking his light-squared bishop usu­
ally panics with 11 lLlxf6+ gxf6 12 Jixf6 Black's bishops are strong enough
0-0 13 Jixe7 lLlxe7, but there is no that he is not worse.

88

(22: 5 e5 ttJg41?
At first this seems insane because of 6
.i.xf7+ �xf7 7 ttJgS+, but as in so many
similar positions, Black's king position is
fine since White has traded off his most
powerful attacking piece (the light­
squared bishop). So 7 ...�gS ! S '-xg4 (af­
ter S 'iff3 .i.b4+! 9 c3 'fIe7 10 'ii'd S+ eMs
n 0-0 ttJgxes 12 cxb4 ttJxb4 13 iVb3 h6
White has nowhere near enough for two
pawns) S ...h6! 9 ttJf3 d6 10 'it'e4 (or 10 e6
'-f6) 10 ... dxes 11 ttJxes and now 11.Ji'f6
12 ttJxc6 bxc6 13 0-0 .i.fS is simplest,
though Houd;n; likes 11 .. :iVe7 12 f4
'Wh4+ 13 g3 'ii'f6 14 0-0 .i.fS 15 'ii'd S+
�h7 with a large advantage because
White has a stupid position - or, to be
more specific, due to Black's bishop pair
and White's weak light squares.
White also has:
(221: 6 'ife2 89
(222: 6 0-0 90
(221: 6 'ii'e 2
Most often played, but White strug­
gles to equalize in the main lines.
1 e4 ttJc6 2 d4 e5
6.....e7 7 .i.f4 f6! 8 exf6 'i'xe2+!
This is the most forcing.
9 .i.xe2
9 �xe2?! is more common, when
the brand new, never been opened, still
in its original packaging 9 ...gxf6 ! gives
Black the advantage; e.g. 10 :el ttJges
11 �1 d 6 12 ttJbd2 .i.d7 13 ttJe4 .i.g7,
and although Black's position is far
from perfect, White is h aving a hard
time recovering the pawn. Or 10 :dl
ttJges 11 .i.dS d3+ 12 cxd3 ttJxf3 13
bf3 ttJd4+ 14 �d2 ttJxf3+ 15 g xf3 c6,
when White has recovered the pawn,
but wishes he hadn't.
9... ttJxf6
89
Th e Dark Knig h t System
There is still chess to be played, but
the position is equal after (amon g oth­
ers) 10 tbbd2 d6 11 tbb3 d3 12 .i.xd3
tbb4 13 0-0-0 tbxd3 14 llxd3 h 6 and
1 5 ... �f7, though Black has man aged to
stir things up a bit. Worse for White is
10 .i.xC7 ? ! d6 ! 11 iob5 .i.d7 12 .i.xc6
.i.xc6 13 tbxd4 .i.xg 2 14 :g 1, as in
R.Stranz-K.Neumeier, Austrian Team
Ch ampionship 2004, when Black
should have continued 14 ... .i.h 3 with
an edge.
(222: 6 0-0 .i. e7
7 l:te1
Or 7 .i.f4 g 5 ! S tbxg 5 (S .i.g 3 h 5 ! 9 h 3
h4!, D.Von Wantoch Rekowski-J. Peric,
Tivat 2001, was a strong case for Black
- see G ame 59) S ... d5! 9 e6 (9 exd6? !
.i.xg 5 1 0 dXC7 'iff6 11 'it'e2+ wfS
doesn't give White enough for th e
piece) 9 ....i.xg 5 10 'ifxg4 .i.xe6 1 1 .i.xg 5
.i.xg4 12 .i.xdS l:txdS and Black keeps a
pawn, for which White has not nearly
enough.
7 d6 8 exd6 cxd6!
...
Only S .....xd6 has been played, but
90
th en 9 b3 ! 0-0 10 i.a3 .-dS 11 .i.xe7
tbxe7 12 -.xd4 'iWxd4 13 tbxd4 gives
White a good endgame.
Instead, the text move is very dy­
n amic - a main idea is to use the g4-
knight to harass White on the soft f2 -
square. Let's analyze: 9 tbxd4 (not 9
h 3 ? ! tbge 5 and Black is better already)
9 ... 0-0 and now:
a) 10 h3 d5! 11 tbxc6 (or 11 .i.b3 ? !
.i.C5 ! 12 c3 tbxf2 ! 13 �xf2 tbxd4 14
cxd4 "'4+ 15 �fl .i.xd4 16 .i.e3 'it'f6+
17 �g l .i.xb2 lS tbd2 .i.xal 19 'irxal
.-xal 20 :xal l:teS with advantage -
quite a small one actually; the material
balance very strongly suggests trades
for Black, so if 21 �f2 .i.e6 22 tbf3 ? !
then 22 . . .d4! ! makes progress) 1 1. . .bxc6
12 hxg4 (12 .i.d3 tbf6 is equal) 12 ... dxc4
13 'ii'f3 .i.e6! 14 'ifxc6 .i.xg4! 15 'wxc4
ioe6 16 -.e2 lleS 17 tbc3 .i.f6 and the
board is so open that the bishop pair
offsets the missing pawn. White also
lags in development.
b) 10 tbc3 .i.h4! 11 g3 .i.f6 12 tbdb5
a6 13 tbxd6 tbxf2 !
Wow! It's complicated, but at least

you won't be seeing the position for
th e first time, unlike your poor oppo­
nent. As it turns out, White has several
acceptable routes to a draw, but no
good way to play for a win:
b l } 1 4 tiJxf7 'ii'xdl 15 tiJxdl tiJxdl
and White can repeat moves now with
16 tiJd6+ �h8 17 llJf7+, or play 17 l:lxdl
�d4+ 18 :txd4 {18 �h l �g4! is trouble
for White, as is 18 �g 2 J:.f2+ 19 �h l
�g4} 18 ... tiJxd4 and repeat moves here
with 19 l1Jf7+ c;t>g 8 20 tiJg 5 + etc.
b2} 14 'ifd5 i..d4!, when White h as a
1 e4 lLlc6 2 d4 e5
couple draws, but 15 :te8 ? ! is not one
of them: 15 .. :ifxe8 16 lLlxe8 i.e6 ! 17
� 5 i.g4 18 tiJf6+ i.xf6 19 �d5 l:tad8
20 �xf2 l:txd5 21 tiJxd5 i..d4+ is better
for Black, who is more active, and
White's king is still a problem. Instead,
15 'ii'xf7+ l:txf7 16 llJxf7 lLld3+ 17 �h 1
llJf2+, and now i t i s Black who is best
advi sed to repeat moves; or 15 tiJxf7
tiJd3+ 16 �e3 'iWxd5 17 i.xd5 lLlxel 18
l:txel i.f5 19 �xd4 lLlxd4 20 l::te 7 i.xc2
21 l::txb7 llJf5 and the usual repetition
follows.
b3) 14 i.xf7+ �h 8 15 �xf2 �d4+ 16
�g 2 'ii'xd6 17 �f4 'iWc5 18 lLle4 'ii'f5 and
the complications are not over, but
White's king position is far too fragile
for him to have serious thoughts about
winning - losing is quite attainable
though !
b4} 14 �xf2 ? ! doesn't make sense.
14 ... i..d4+ an d 15 ... 'i¥xd6 is not terrible
for White, but he has nothing to com­
pensate for his loose king .
91
 C h a pter Five

1 e4 tDc6 2 tDf3

is unlikely to keep the bishop pair and

he won't be able to stop Black from
shifting his pawn structure to light
squares. Nonetheless, 3 ...ttJf6 4 ttJC3 g6
5 d4 a6 is simpler, especially since we
will need to know this position anyway.
3 ...ttJf6

With this move, White hopes to en­

ter his favourite double-king-pawn po­
sition, which he will be happy to do
after 2 ... e5. We are obligated to disap­
point him, and in the resulting Pirc po­
sitions White may wish he h ad not
committed his knight so soon - the
Austrian Attack (f2-f4) and the Arg en­ 4 ttJ c3
tinean Attack (f2-f3) are no longer Other moves:
available to him, nor are several other a) 4 c3 is possible and does not ac­
aggressive variations. tually drop a pawn : 4... ll)xe4?? 5 d5
2 d6 3 d4
••• ttJb8 6 '6'a4+ picks up the e4-knight.
After the premature 3 �b5 we have However, 4... g6 i s fine for Black. Miles
an extremely rare situation where it is downed GMs Becerra Rivero and Zelcic
fine to play 3 ...ll)f6 4 ttJc3 �g4 - White starting with this move (see Games 60

92

and 61). Mestrovic is also 2-0 with 4 ... g 6
here. T o continue: 5 i.. d3 i..g 7 6 0-0 0-0
7 h 3 e5 S l:tel.
Although Black has tried many
things in hundreds of games, I am rec­
ommending S ... b6!, as played once in
I.Vasilevich-M.Allakhinova, Russian
Team Champion ship 2002. One point is
that, after 9 d5 tiJe7, Black's pawn
holds up White's queenside play; while
9 i.. e 3 exd4! 10 cxd4 tiJb4 picks up the
bishop pair: 11 tiJc3 tiJxd3 12 'ii'xd3 lieS
13 l:tadl (or 13 :acl CS) 13 ... i..b 7 14 d5
:tcS with only a sman advantage to
White; ...:cS helps prepare an eventual
... c7-c6, and pre-empts any White pres­
sure down the c-file.
The game continuation should not
have been too dangerous either: 9 'iia4
i..d7 10 i..b 5 "'eS 11 d5 tiJe7 (ll ... tiJbS!)
12 i..xd7 tiJxd7 (12 ...'iVxd7!) 13 -'C4 tiJc5
14 b4 b5 15 "e2 tiJa4 16 i..d2 and now,
instead of 16 ... c5?! as played, 16 _.f5 ! ! 17
tiJg 5 fxe4 lS tiJe6 'ii'f7 19 tiJxc7 l:tabS 20
tiJe6 llfcS 21 'ifxe4 i..f6! is good for Black,
who plans 22 ...tiJb6 and 23 ... tiJbxd5.
White tried S tiJa3 in L.Rozman-
1 e4 tiJc6 2 tiJf3
J.5chuyler, Washington 2012: S ... d5! 9
i..g 5 ? ! (9 exd5 tiJxd5 10 dxe5 tiJxe5 11
tiJxe5 i.. x e5 12 l:tel i.. g 7 is equal)
9 ... dxe4 10 i..xe4 exd4 11 i.. xc6 ? ! (11
cxd4 is still fairly level) 11 ...dxc 3! with
great complications favouring Black
(see G ame 62).
b) 4 tiJbd2 is similarly met by 4... g6.
Miles took down G Ms Zapata and
Nijboer with this move (see Games 6 3
and 64). Given his successes with this
fianchetto, it is surprising to me that
Miles was never willing to play 4... g 6
against 4 tiJc3, even when oth er moves
were giving him trouble.
c) 4 i.. d 3 doesn't look right because
Black can pick up the bishop pair with
4 ... tiJb4. My first inclination was that
this must equalize immediately, but
apparently tempi sometimes matter in
chess! After 5 0-0 tiJxd3 6 'ii'xd3 g6 7 c4
i.. g 7 8 tiJC3 it will be difficult for Black
to disrupt White's huge centre. There­
fore, 4 ... g6 here too! An three of White's
fourth move sidelines may transpose
into each other.
4 g61
•••
93
The Dark Kn ight System

A: 5 �bS 94
B: 5 d S 9 6
c: Others (without d4-dS) 102

A: 5 -'.b5 a6

4...�g4 is by far th e most common

move, ensuring that Black reaches po­
sitions unique to 1 ...l'bc6, but Black has
both objective and practical problems
after White's most accurate reply S
�e3. Then Miles always played s ... e6
and, when that stopped working, s ... a6, 6 �xc6+
which was worse, while Mestrovic has 6 �a4 doesn't make much sense,
repeatedly but unsuccessfully tried to bu t it has been played several times,
uphold S ... eS. After 6 �bS, his record most significantly in M.Kozakov­
speaks for itself (+2 -9 =3). A.Zajarnyi, Lvov 1998, which continued
On the other hand, our standard fi­ 6 ...bS 7 �b3 �g 7 8 h3 0-0 9 0-0 e6 10
anchetto with 4 ... g 6 is reliable, trans­ a3 �b7 11 :el l'bas 12 �a2 cS 13 dS
posing to a Classical Pirc (1 e4 d6 2 d4
l'bf6 3 l'bC3 g6 4 l'bf3) with the odd
4... l'bc6. There are many strong players
h appy to use this placement for th e
knight; for instance, GMs Smirin and
Adorjan h ave played (4...�g 7 S �e2 0-0
6 0-0) 6 ... l'bc6 twenty times between
them, with seven wins and one loss. Of
course, 6 ...l'bc6 is not precisely 4 ... l'bc6,
but I intend to show that the differ­
ences are not critical.
White has many moves h ere, so I and now after 13 ... e s ? ! 14 b4! cxb4
will divide the material into three lS axb4 'ii C7 16 bxaS "xc3 17 �d2
groups: "C7 18 C4! White eventually converted

94

his advantage. Instead, Black could
have kept the balance with 13 ... exd 5 !
14 ltJxd5 .i.xd5 15 .i.xd5 ltJxd5 16 'iVxd5
ltJC4. The important feature of the posi­
tion from Black's stan dpoint is the ac­
tive g7-bishop.
The simple S ... ltJa5, picking up the
bishop pair, also brings equality; e.g. 9
0-0 .i.b7 10 es dxe5 11 dxe5 ltJxb3 12
axb3 ltJe4 13 'iixdS+ :xdS 14 ltJxe4
.i.xe4 15 :xa6 .i.xf3 16 gxf3 .i.xe5 and
if you're still awake after all these
trades, you'll notice that White's dis­
gusting pawn structure is balanced by
some extra rook activity.
Based on the Kozakov game, 6 .i.a4
is recommended (sort of) by An drew
Greet in Beatjng Unusual Chess De­
fences: 1 e4, so it is a good idea to be
prepared.
6 ... bxc6 7 0-0 .i.g7
8 h3
White usually plays this move
sooner or later in the Dark Knight Pirc,
as he gets tired of worrying about
....i.g4. Others:
a) S e 5 ! ? ltJd5 g ltJe4 0-0 10 :el was
1 e4 ltJc6 2 ltJf3
equal in O. Biti-G.Belamaric, Portoroz
2005. White's lunge has pushed into
where Black is strong and opened up
squares for the cS-bishop. Black should
play 10 ... a5 11 h3 CS ! ? 12 exd6 cxd6 13
dxc5 dxc5 14 ltJxc5 'iVd6 1 5 ltJe4 �c6 16
c3 a4 and in this open position, the
bishop pair and queen side pressure are
fully worth the sacrificed pawn.
b) S 'iie 2 0-0 9 lIdl .i.g4 (g ... a5 !
equalizes) 10 h3 .i.xf3 11 'ii'xf3 ltJd7 12
.i.e3 e5 with a small edge for White in
E.5veshnikov-T.Gelashvili, Cappelle la
Grande 200g, albeit one he was unable
to convert (see Game 65).
8...0-0 9 l:Iel l:lb8 10 l1bl ltJd7
l1 .i.g 5?!
Th e strong White and Black players
to hold this position were focused on
the ... e7-e5 break when they should be
preparing for ... c6-c5.
11 .i.e3 is a bit better, though it still
doesn't stop 11 ... C 5 ! ; e.g. 12 dXC5 �xb 2 !
13 l:txb2 .i.xc3 14 l:tb3 .i.xel 15 -.xel
ltJxC5 16 .i.xc5 dXC5 equal, or 12 �d2
cxd4 13 .i.xd4 ltJf6 14 e5 ltJh 5 15 exd6
cxd6 16 .i.xg7 ltJxg7 with equality.
95
Th e Dark Kn ight System
11 C5!
•••
After 11 ...h6 12 i.h4 g 5 13 i.g3 e5?!
(13 ... c5 or 13 .. .f5 ! ? is better) 14 dxe5
dxe5, as in B.Vuckovic-Z.Petronijevic
and D.Nestorovic-N.Ristic (from the
2003 Serbian Team Championship),
White enjoys a large structural advan­
tage which is not compensated for by
Black's inactive bishop pair.
Whereas l1... c 5 ! 12 e5 :e8 13 exd6
cxd6 14 ttJd5 f6 15 dXC5 ttJxcS sees
Black starting to get the better of it. Of
course, White could swallow his pride
and play 12 i.e3, which is equal after
12 ... e6.
B: 5 d5 ttJb8
As I have mentioned before, one of
the significant benefits of the fi­
anchetto is that 5 d5 opens the diago­
nal for the g7-bishop. It would be a
shame to close it with s ...ttJe5 6 ttJxes
dxe5.
6 i.e2
Other moves:
a) 6 �C4 isn't as stupid as it looks -
after all, when Black plays ... c7-c6
96
White will enjoy having the option to
open th e a2-g 8 diagonal whenever he
wants. However, Black is allowed to
change his plans too: 6 ... i.g7 7 0-0 0-0
8 h3 (this is usually played by this
stage, on moves 5, 6, 7 or 8; otherwise
White has to worry about both ... �g4
an d ...ttJg4) 8 ... e 5 ! ? (this is a new move
inten ded to punish �C4 and h2-h3; if
White refuses to open the position, the
c4-bishop is atrocious - both passive
and hindering White's natural plan of
C2-c4-c5) 9 �e3 (or 9 dxe6 i.xe6 10
i.xe6 fxe6 11 e5 dxe5 12 'i¥e2 ttJc6, and
with the knights jumping to the d5-
and d4-squares, Black has full compen-
sation for his weakness on e6) 9 ... a6 10
b4 (if 10 a4 then 10 ... a5 and 11 ... ttJa6)
10 ... ttJh5 11 'ii'd2 ttJd7 12 ttJe2 'ike7 and
with _ .f7-f5 coming, Black has reached
a good King's Indian-type position .
Instead of 8. . .e 5 ! ?, N.Praznik-A. BeJi­
avsky, Bled 1999, continued 8 ... c6 9 a4
as 10 l:[el ttJfd7 11 �e3 ttJa6. Though
White was better, Black played effec­
tively on the dark squares throughout
the game and went on to win, making
this a must-see (see Game 66).
b) 6 i.g 5 is in fact the most often
pla yed. White intends long castling and
a quick attack, but despite the time
gained against Black's knight, this strat­
egy is questionable - Black's counter­
play with ... C7-c6 is very fast: 6 ... �g 7 7
'i¥d2 c6 8 �h6 (after 8 0-0-0, Black cou­
rageously castled in F.Lukez-S.Lejlic,
Rodeby 1998, and equalized with 8 .. 0-0 .
9 �h6 �g4 10 �xg7 wxg7 11 i.e2
 1 e4 ltJc6 2 ltJj3

ltJbd7 - White's "attack" is going no­ position to B2. Of course White could
where; e.g. 12 h3 �xf3 13 �xf3 '11> 6 14 have played 6 h3, insisting on Bl.
h4? ! ltJes, threatening ...ltJc4) 8 ... �xh6! 9 7 0-0 0-0
'ii'xh6 'ib6 10 0-0-0 �g4 is an interest­
ing position, not at all unfavourable for
Black; e.g. 11 l:[d2 �xf3 12 gXf3 ltJbd7 13
f4 cxdS 14 exds :c8 ( 14. ..0-0-0 ! ? and
15 ... �b8) with great interest in ... l:1xc3;
or 11 �e2 'ii'xf2 12 :hfl 'iVcs and while
White is certainly well developed, he has
nothing concrete for the pawn. In
N.5ulava-M.Muse, Croatian Team
Championship 2002, White changed the
course of the game with 11 eS?! dxes 12
d6, but had Black spotted 12 ...ltJbd7 13 White has:
dxe7 'i'b4, followed by 14.....xe7 and
15 ...0-0-0, White would have found him­ 81: 8 h3 97
self without sufficient compensation. 82: Others 99
c) 6 �e3 is very similar to 6 �g s,
into which it often transposes (i.e. after Or 8 a4 as ! - Black cannot allow
�e3-h6). One time that didn't happen himself to become further cramped.
was in K.Nemcova-F.Olafsson, Marian­ The insertion of the two a-pawn moves
ske Lazne 2008, which went 6 ... �g7 7 is helpful to Black though, since it helps
'ifd2 c6 (by transposition) 8 h 3 b S ? ! him to establish knight outposts on the
(8 ...0-0 9 � e 2 bS!, as i n J. Hjartarson­ cS- and b4-squares.
F.Olafsson, Reykjavik 1995, was a better
move order - see Game 68) 9 a3?! (9 81: 8 h3
dxc6! b4 10 �bS ! is unpleasant for
Black) 9 ... a6 10 dxc6 ltJxc6 1 1 �d3 0-0
12 0-0 �b7, resulting in an equal
Dragodorf type of position which Black
went on to win (see Game 67).
d) 6 h3! is the most accurate move,
reaching line Bl below after 6 ... �g7 7
�e2 0-0 8 0-0.
6 �g7
•••

There is something to be said for

6 ... �g4 (or 7 ...�g4), with a likely trans-

97
Th e Dark Knig h t System

This is played most often, spending a sition where his normal play (with c2-
tempo to stop 8 ...�g4. As we will see, c4-cS) is blocked by the c3-knight. This
8 ... �g4 was indeed Black's intention, problem is serious for White, more so
but a tempo is a tempo. There is some than Black's funny knight on b8 (which
disagreement about the merits of 8 h3 is n ormally on the e7-square). Black will
- Alburt and Chemin adorn i t with an play carefully for .. .f7-fS . Part of being
(I), while Nunn says it is "not really careful is considering ... h7-h6 to pre­
necessary". I think it is the best move in vent White's lbgs-e6.
the position . 9 ... �xe6
8 ...esl
Smirin and Gulko have each chosen
8 ... eS three times, with an even (total)
score, while Finkel, Urban, an d Gufeld
have four wins and three draws with
8 ... c6, but against this, Benjamin's plan
(9 a4 as 10 :el lba6 11 �xa6 ! - see the
note on 8 ... c6 in line B2) is very strong
for White.
For those interested in making ... c7-
c6 work, I suggest 8 ... a6 (or 8 ... b6) 9 a4
b6 10 �e3 �b7 11 'Wd2 c6 12 %ladl Black has caught up in development
'fIC7 13 �h6 lbbd7 as in A.Bachofner­ and now has only to worry about a
P.Hopman, Amsterdam 2006, when small space disadvantag e. If White is
White is only slightly better. not alert, Black will fix thi s with a quick
. .. d6-dS.
White has now only showed interest
in:

811: 10 ii.g5 98
812: 10 tDd4 99

811: 10 �gS!

Dubious according t o Nunn, in his

1 9 8 9 The Complete P;rc, this has been
9 dxe6 by far the more dangerous move in
If White does not play thi s move, he practice.
winds up in a King's Indian type of po- 10... h6

98

ll .i.e3
Instead:
a) 11 .i.h4 ttJc6 12 'ii'd 2 gS 13 �g 3
dS equalizes.
b) 11 .i.f4 ttJc6 12 'ii' d 2 gS 13 .i.h 2
:e8 14 :adl ttJd 7 offers White a tiny
edge, but typical dark-square play for
Black.
11 ... ttJc6 12 'ii'd 2 �h7 13 :adl :e8 14
:fel a6 15 a4
So far we have followed lPiket-
8.Gulko, Amsterdam 1989, which con­
tinued lS ....i.d7 16 .i.C4 .i.e6 17 .i.e2
'ii'e 7 18 ttJdS 'Wd8 19 ttJc3, indicating
th at nobody could come up with a plan
- this is already a threefold repetition,
1 e4 ttJc6 2 ttJj3
but somehow the game continued and
Gulko contrived to lose, even after a
later repetition.
Alternatively, l s .. :6'e7 is fine to test
the opponent for a draw (16 ttJdS "d8
17 ttJC3), but if one wants to continue,
there is lS ... ltJd7 ! ? 16 .tfl b6, intending
17 ... ttJcS, while ...ttJdes is also a possi­
bility, as is the manoeuvre .. :"c8-b7.
White is a little better.
Y.Gruenfeld-I.Smirin, Israeli Team
Championship 1997 (see G ame 69),
shows what happens if White does not
play a2-a4 to contain Black's queenside
expansion - Black's counterplay was
more than sufficient.
812: 10 ttJd4 .i.d7
The insertion of these moves makes
it easier for Black to pressure the e4-
pawn .
11 :el ttJc6
Now White either loses time re­
treating the knight or agrees to an ex­
change that will ease Black's position.
Similar is 11 ...:e8 12 .i.fl ttJc6 13 ttJb3
as 14 a3 a4 1 5 ttJd2, and instead of
99
The Dark Knight System
ls ... tZlas ? ! 16 tZlf3, which was g ood for
White in K.Hulak-S.Marangunic, Yugo­
slavia 1977, Black should play lS ... tZld4!
16 tZlf3 tZlxf3+ 17 'ii'xf3 i.. c6 with an
equal position.
12 tZlf3 :e8 13 i..c4 i..e6 14 tZld5 'ii'd 7
15 c3 i..x d5 16 exd5 :xe1+ 17 'ii'xe1
tZle5! 18 tZlxe5 dxe5
The position is equal, White's bish­
ops being offset by his poor develop­
ment and the inconvenience of guard­
ing the ds -pawn . If 19 'i'xes, then after
19 .. .l:te8 and 20 ...:el+, White will never
complete his development.
82: Others (besides8 h3)
1 00
If White is going to refrain from 8
h3, it makes very little difference which
move he chooses, but we'll take the
following as the main line:
8 i..e 3
Altern ative ly:
a) 8 i..g s should be treated simi­
larly: 8 ... i.. g 4! 9 h3 i..xf3 10 i..xf3 c6 11
'ii'd2 tZlbd7 and Black follows with
...'ifb6, ... a7-aS, ... tZlcS and ... tZlfd7, with
typical dark-square play, even if White
is still slightly better.
White can also try to do without h2-
h3. For instance, 9 'i'd2 c6 10 l:.fel
tZlbd7 transposes to L.Vajda-M.Marin,
Rumanian Championship, Bucharest
1998, and M. Kolosowski-Dan. Fraczek,
Legnica 2011, which both continued 11
dxc6 ! ? bxc6 12 tZld4 :c8? ! 13 f3 i.. e 6 14
tZlxe6 fxe6 with advantage to White.
According to Houd;n;, Black can equal­
ize in this line with 12 ... 'ifb6 ! 13 tZla4
'ikC7 14 i..xg4 tZlxg4 lS i..xe7 dS 16 f4
(16 tZlf3 ? ! tZldes! 17 i..xf8 tZlxf3+ 18
gxf3 'iWxh 2+ 19 �fl l:txf8 20 fxg4 'iVh l+
21 �e2 'iWxe4+ 22 �1 'i'xa4 is g ood for
Black) 16 ...l:tfe8 17 exdS cxdS 18 h 3
tZlges 1 9 dxes :xe7.
b) 8 :el i.. g 4 (of course) 9 h3 i..xf3
10 i..xf3 tZlfd7! 11 i.. e 3 c6 was
D.Rogozenco-V.Nevednichy, Rumanian
Team Championship 200S, where
White went wrong immediately with
12 i.. d4?!, initiating an exchange bene­
ficial to Black; e.g. 12 ...i.. x d4 13 'ikxd4
�6 is equal. Without this mistake
White is a little better, but Black's mi­
nor pieces are all good, and he has the

simple plan of ... a7-as, ... l2Ja6, ... l2Jacs,
... as -a4-a3 (if possible), and .....b6 or
.. :ikC7.
8 .i.g4
...
Black has done very well with S ... c6,
but Joel Benjamin (who played 1 ... l2Jc6
frequently in the 1990S) warned me
that Black's position is difficult after 9
a4! as 10 h 3 (oth erwise 10 ... .i.g4)
10 ... l2Ja6 11 .i.xa6 ! (an idea I have not
been able to find in any published
games - the point is to stop Black's ac­
tive 11 ... l2Jb4) 11 ... :xa6, when it is hard
to find a constructive plan for Black,
whereas White can still build; e.g. 12
"d2 l:taS 13 l:tadl :eS 14 .i.d4 '6' c7 lS
l::tf el. Black lacks space, development
and pawn play, while his only "asset",
the light-squared bishop, is more of a
problem than anything else. Indeed,
this type of position acts more closed
than open, in part because nobody
wants to relieve the tension between
the dS- and c6-pawn s - for White to
trade would assist Black greatly in the
central battle, while if Black trades, he
has accessible weaknesses on the bS­
and e7-squares.
A possible antidote is 8 ... aS ! ? 9 a4
l2Ja6, when White should be less eager
to snap off the knight. Notice that since
Black h as not yet played ... c7-c6, he can
later try ... e7-e6 or ... e7-eS instead.
However, there are other moves to
worry about besides 9 a4.
8 ... .i.g4 is simplest, transposing to a
favourite line of the great Pirc expert
Alexander Chernin, who used it with
1 e4 l2Jc6 2 l2Jj3
great success in the 1990s (three
draws, three wins, all against GMs). The
normal move order to reach this posi­
tion is 1 e4 d6 2 d4 l2Jf6 3 l2Jc3 g6 4 l2Jf3
.i.g 7 S .i.e2 0-0 6 0-0 .i.g4 7 .i.e3 l2Jc6 8
dS l2JbS. (S ... .i.xf3 is the traditional
main line, but it is not stron ger.)
9 h3
White must play this sooner or later
unless he is intending to allow the ex­
change of light-squared bishops. White
should at least retain the bishop pair if
he is hoping to keep an advantage:
a) 9 l2Jd2 .i.xe2 10 '6'xe2, and now
Chernin's recommendation 10 _ .l::teS 11
f4 e6 has been tested only once, i n
A.Czebe- N. Resika, Budapest 2000 -
Black, an FM, held the draw against the
GM.
b) 9 l2Jd4 .i.xe2 10 .-xe2 cS 11 l2Jf3
�6 12 l:tabl '6'a6 13 .-d2 l2Jg4 14 .i.g s
l:te8 with equality in B.Chatalbashev­
M. Popchev, Cacak 1991 (see G ame 70).
c) 9 l2Jg s .i.xe2 10 '6'xe2 c6 11 l:tadl
.-as 12 f4 .-a6 13 '6'f3 l2Jbd7 with
equality. Black, with more experience
in this type of position, went on to win
101
Th e Dark Kn igh t System
in R.Ziatdinov-A.Chemin, N ew York
Open 1998 (see Game 71).
d) 9 'iVd2 ("Here the 'threat' of i.. e 3-
h6 is a fiction because the exchange of
bishops is in Black's favour. But what
other plan can White try?" - Chemin)
9 ... c6 10 l:Iadl 'ii'a s 11 a3 :tc8 12 b4
'fic7 13 i..d4 lLlbd7 14 :fel as was
equal in P.Blatny-A.Chemin, Pardubice
1993, and ended as a draw.
9 i.xf3 10 i.xf3 c6 11 a4 a s
...
The G M t o hold this position as
White found nothing better than 12
lLlbl ? ! . To me, this is a good indication
of the health of Black's position.
12 ...lLlbd7 13 g3 was V.Arbakov-A.Cher­
nin, Bern 1995, where Black can equal-
ize with 13 ... lLle5 14 i.g 2 lLlc4 15 i..d4
lLlh S or 13 ... lLlc5 14 i.. xc5 dxcS 1 5 dxc6
bxc6, as well as Chemin's 13 ... lLlb6 14
i.g 2 'fiC7.
Instead, in D.Primel-A.Nowocien,
French Team Championship 2007,
White tried 12 '6'd2 fic7 13 :adl lLla6
14 l:.fel lLlb4 15 i..d4 lLld7 16 i.xg7
�xg 7, again with equality; while
R.Kashtanov-A.Lug ovoi, st. Petersburg
1 02
2000, saw 12 i.e2 "tIC7 13 f4 lLla6 14
i.c4 lLld7 15 'ii'e 2 lLlb4 (15 ... i.. xc3 ! ?) 16
dxc6 bxc6 17 l:tadl, and now Black took
a break from his dark-square strategy
to play 17 ... e6! 18 i.b3 dS!.
If there is a way t o an advantage i n
this variation, White h as yet t o find it.
C: Others (without d4-dS)
In these lines White chooses to play
a regular-ish Pirc instead of trying to
challenge the correctness of Black's
early ... lLlc6 with 5 d5 or 5 i..b 5.
There are a few choices:
C1: 5 h3 1 02
C2: 5 i.. e2 106
C3: 5 i.. e 3 1 08
C4: 5 i..c4 1 09
CS: 5 i.gs 112
C1: S h3
This is the most frequently played
move here. White does not want to be
bothered by ... i..g4 or ... lLlg4. Neverthe­
less, a tempo is a tempo, however well
motivated.

S i..g7 6 i..e 3
...
By far the most common, but also
seen are:
a) 6 i.g s 0-0 7 'iid 2, when both
7 ... a6 and the surprising 7 ... dS have
worked extremely well for Black.
al) 7 ... dS S exdS lbxdS 9 i.. h 6 (9
0-0-0 lbxc3 10 "'xc3 'iVd6 11 i.c4 i.fS
12 :hel :adS 13 'ii'e 3 lbas 14 i.d3
i.. e 6 gives White a tiny edge) 9 ... lbxC3
(9 ... i.e6 ! is equal) 10 i..x g7 �xg7 1 1
'ii'x c3 'ii'd S 12 0-0-0 i.. e 6 13 b3 i.fS and
Black converted White's advantage
(har!) in So.Polgar-1Fries Nielsen, Ri­
mavska Sobota 1991 (see Game 72).
a2) 7 ...a6 S 0-0-0 bS ( S ... dS ! ?) 9 a3 (9
i.. d 3 i.. b7 10 �bl :eS 1 1 :hel eS 12
dS lbe7 is slightly better for White)
9 ...l:lbS 10 i.h6 b4 11 axb4 lbxb4 12
i.xg7 �xg 7 13 eS lbfdS was equal in
M.Yilmazyerli-D.Arutinian, Istanbul
2007, though the strong er player
(Black) went on to win (see Game 73).
b) 6 i.. e 2 i s not very consistent with
5 h3 because the pin has been pre­
vented already - White is normally
hoping to retain the option of i..C4.
1 e4 lbc6 2 lbf3
Then 6 ... 0-0 7 0-0 (7 i.. e3 transposes to
7 i.. e 2 in the notes to the main line)
7 ... eS (7 ... a6 ! ? is more combative, and
was tried successfully in N.Ryba­
J.Schuyler, Washington 2012 - see
Game 74) and now:
bl) S dxes
Hold on ! Seriously, how do we de­
cide how to recapture in this position
(and similar positions) ?
First of all, i t is safe t o assume that
if White took once on eS he will take
again given the opportunity, so we will
wind up with the same pawn structure
in either case - the only difference be­
ing the c6- and f3-knights. Do we like
having them there or not? The main
factor if the knights are on is that
White can play lbdS without havin g to
worry about being evicted by ... C7-c6,
but sometimes Black's ... lbd4 is useful,
too. If the knights are off, White has the
possibility to pressure Black's centre
and activate his rook with f2-f4, but
this also frees Black's g7-bishop. In
general, the ability to play ... c7-c6 is the
most important factor, so ... lbxes is the
1 03
Tn e Dark Kn ig h t Sys tem
normal choice. The reader would be
well advised to take special note of any
exceptions.
Here 8 ...4:Jxes 9 ttJxes dxeS 10 .i.g s
c6 is equal, when 11 .i.C4 bs allows
Black free expansion on the queen side.
b2) 8 .i.e3 is a bad version of 7 .i.e2
eS in the main line - if White is castled
king side he has no attack to compen­
sate for his troubles; i.e. 8 ... exd4 9 ttJxd4
�e8 and White already lacks a comfort­
able way to defend his e-pawn . 10 ttJxc6
bxc6 11 .i.f3 is most common, when
Black plays ... .i.a6, ... ttJd7, and can con-
sider ... :tb8, ... �8, ... ttJes, and/or ... ttJb6.
White is equal according to Mr. H, but
Black wins most of the games. There are
no worthy examples because White
can't seem to hold onto his pieces.
b3) 8 dS ttJe7 9 .i.e3 is similar to the
m ain line - except that White is not
fast enough with his pressure on the d­
file, so 9 ... c6 is already equal, and after
10 dxc6 bxc6 11 ifd2 'WIC7 12 :adl dS
13 exds ttJexds 14 ttJxdS ?! cxdS Black
took over the centre in N .Jhunjh­
nuwal a-S.Gligoric, Lucerne Olympiad
1982 (see Game 75).
c) O n 6 .i.C4 Black could proceed
"normally" with 6 ... 0-0 (cf line (41 be­
low), but the immediate 6 ... ttJxe4! is an
equalizer: 7 .i.xf7+ �xf7 8 ttJxe4 dS 9
ttJcS 'WId6, threatening 10 ... ttJxd4 or
10 ... eS. White can sacrifice a pawn with
10 ttJd3, when 10 ... ttJxd4 11 J.f4 ttJxf3+
12 'ii'xf3 �e6+ 13 J.es+ �g8 14 'ii'g 3
.i.xes 15 ttJxes 'WIa6 leaves h i m with
enough compen sation but no more. Or
1 04
Black can decline the gift: 10 ...�f8 11 c3
(11 .i.f4 'ii'e 6+ 12 .i.es �g8 13 c3 ttJxes
14 ttJdxes 'WId6 is equal) 11 ...�g8 12 0-0
h6 13 l:tel g S, though White is slightly
better here.
d) 6 J.bS looks silly since Black can
castle out of the pin, but after 6 ... 0-0 7
0-0 Black has nothing better than 7 ... a6
8 .i.xc6 bxc6, transposing to line A
above.
6 0-0
...
7 'ii'd 2
Or 7 .i.e2 eS (7 ...a6 is more aggres­
sive, but riskier) 8 dxes (other moves
transpose elsewhere: 8 dS ttJe7 9 'i¥d2
is 9 J.e2 in the main line, while 9 0-0
and 8 0-0 are respectively notes 'b3'
and 'b2' above) 8 ...dxes (Didn't I just
say this was wron g ? - this is one of
those exceptions; actually, 8 ...ttJxes is
fine too, but the text move h as per­
formed much better, so why not?) 9 0-0
'WIe7 with equal chances. The point is
that 10 ttJdS is not dangerous because
of 10 ... ttJxdS 11 exdS ttJd4! 12 liel ttJfS !
(chasing the more dangerous bishop)
and Black is comfortably equal.

7-.eS 8 dS
Others:
a) S 0-0-0 exd4 9 ttJxd4 ':eS 10 f3 -
3lack wins nearly every game from this
�sition. Indeed, White's pawns on h3
and f3 make a ludicrous impression (he
7S essentially down a full tempo in a
Ph ili dor Defence Larsen Variation),
though White should not actually be
worse. lo ...ttJxd4 11 .i.xd4 i.. e6 12 g4 (12
..i.f2 a6 13 �bl bS 14 h4 CS! was about
equal in G. Bastrikov-E.Geller, Tashkent
1958 - see Game 76) 12 ... cS ! 13 .i.e3
'.as !? (1 3 ... dS ! equallzes) 14 iVxd6?! (14
l.h6! .i.xh6 15 'ii'xh6 .i.xa2 16 :xd6 lie6
is slightly better for White) 14...ttJxe4! 15
fxe4 .i.xc3 16 'ii'xcS ? .i.xb2+ 17 �xb2
'''xa2+ lS �Cl l:acS and White soon
called it quits in D.Bescos Cortes-S.Garza
Marco, San Jose 1995. 16 bXC3? 'iVa3+ 17
�d2 l:lad8 wouldn't have worked either;
instead after 16 i..d3, closing the d-file,
the game still continues, though Black is
clearly better.
b) 8 dxeS seems like it is headed for
dullness, but things could get interest­
ing if nobody trades those queens; e.g.
S ...ttJxes 9 ttJxes dxes 10 0-0-0 .i.e6 11
g4 c6 12 g S (12 'ii'xd8 l:tfxd8 13 :xd8
l:lxdS 14 .i.xa7 .i.h6+ 15 .i.e3 .i.xe3+ 16
fxe3 hS 17 g S ttJh 7 18 h4 f6 19 gxf6
ttJxf6 is equal despite the slight pawn
deficit: Black's activity and future
passed h-pawn are sufficient) 12 ... lDd7
13 h4 "as and the race is on . Black has
no reason to be pessimistic about his
prospects.
8 ttJ e7
•••
1 e4 ttJc6 2 ljjf3
S ... ttJbS is also quite reasonable. The
e7-knight supports the ... c7-c6 break,
but with White likely to castle long and
form a battery on the d-file, the useful­
ness of this option is reduced.
9 0-0-0
Instead:
a) 9 .i.h6 is statistically the most
dang erous, scoring 6/7 for White, but
Black was heavily outrated and his play
quite uninspired. Artashes Minasian
shows us the way in D.5aduakassova­
Art.Minasian, Dubai 2011: 9 ... .i.d7 ! 10
.i.d3 (10 o-o-o? ! bS!) 10 ... c6 (or 10 ... bS
11 a3 as 12 0-0 b4 intending ... c7-c6
with counterplay) 11 dxc6 .i.xc6 (taking
firm control over the ds-square; back­
ward pawn? what backward pawn?) 12
.i.xg 7 �xg 7 13 0-0 'fIC7 (13 ...bS ! ?) with
easy equality (see Game 77).
b) 9 .i.e2 .i.d7 (9 ...ttJxe4 ! ? 10 ttJxe4 fS
11 lDc3 f4 12 0-0-0 is slightly better for
White) 10 g4 (not 10 .i.h 6 ? ! c6 ! 11 dxc6
.i.xc6 with advantage) 10 ...b S ! 11 g s
( 1 1 .i.d3 ? ! b4 12 lDe2 c 6 13 dxc6 .i.xc6
14 lDg 3 lDxe4! is good for Black; while
11 a3 :b8 is equal, since 12 .i.xa7? runs
1 05
The Dark Knight System
into 12 ...l:laS! 13 i.e3 b4) 11 ...b4 12
gxf6 bXC3 13 �xc3 i.xf6 with a level
position.
c) 9 1:tdl tLlh s 10 94 tLlf4 (this ag­
gression is called for because White
can no longer tuck his king away on the
queenside) 11 i.xf4 exf4 12 'ii'xf4 fS 13
exfs gxfs 14 g s i.xc3+ 15 bXC3 �e8 16
i.e2 tLlg6 with compensation.
d) 9 g4 a6 should transpose to the
main line once White castles long.
9 a6 10 g4 b5 11 g5
•••
11 i.d3 i.b7 ? ! 12 tLle2 ? ! (better 12
a3 or 12 gs with an edge) 12 ... c6?! was
equal in K.Haznedaroglu-T.G elashvili,
Antalya 2009; but Black could have
blown it open with 11 .. b4 12 tLle2
.
tLlexds! 13 exds e4 14 i..xe4 tLlxe4 15
'Wxb4 l:teS, when White has a great
deal to worry about for his extra pawn .
In the actual game 12 ...tLlfxdS ! would
have been even stronger, giving Black a
comfortable advantage, but White
need not have allowed that.
11 tLlh5
•••
A race scenario is developing in
which Black's chances are n o worse -
1 06
White has a hard time evicting the hS­
knight, and will also fi n d it hard to
profit even if the h-file opens. One con­
tinuation of many: 12 Wbl i.d7 13
tLlh 2 tLlcS ! ? 14 tLlg4 tLlb6 15 b3 b4 16
tLle2 as, still with approximately even
chances.
C2: 5 i.e2 i.g7 6 0-0 0-0
Azmaiparashvili plays 6 ... i.g4 here,
making sure White can't change his
mind and play 7 dS and S h3. Black may
even have considered S ...i.g4! ? It is not
clear which is the most accurate. Be­
sides, why should White change his
mind?
7 i.. e 3
White tries to make do without h2-
h3. Is he inviting trouble or saving a
tempo? The alternative is 7 i.gs h6 and
then:
a) S i.h4 g S 9 i.g 3 tLlh S 10 dS tLlxg3
11 hxg3 tLles is equal; e.g. 12 tLld4 c6 13
tLlfs i.xfS 14 exfs 'ifb6 1 5 llbl tLld7 16
i.f3 lIfe8 17 g4 tLlf6.
b) S i.e3 tLlg4 9 i.d2 eS 10 dS tLle7
11 h3 tLlf6 12 i.e3 tLld7 13 "d2 �h 7 14

ltJel fs most closely resembles a King's
Indian, where White's attack will be
greatly delayed by his pawn stuck on
the c2-square.
c) 8 �f4 ltJg4! 9 h3 (if White does
not play h 2 -h 3, the game will transpose
to note 'b' just above; e.g. 9 ds es 10
�d2 ltJe7, or 9 �d2 es, or even 9 �Cl
es 10 ds etc) 9 ...es 10 dxes ltJgxes 11
ltJxe s ? ! ( 11 'ifd2 is better, retaining a
tiny edge after 1 1 ... <ith 7; Black can con-
sider ...ltJxf3+ and ...ltJd4, or ...i.e6, or
.. .f7-fs, or ... l:te8) 11 ... dxes was level in
A.5akh arov-A.Adorjan, 50chi 1976,
though Black went on to win a wild
game (see Game 78).
7 es!
•••
As usual, 7 ... a6 can be tried - the
main line is a bit drawish - but then 8
ds ltJb8 9 a4 is an excellent answer.
Alternatively, 7 ... �g4 8 ds ltJb8 trans­
poses to line B2 (8 ...�xf3 9 �xf3 ltJes
10 �e2 c6 is a main line Classical Pirc
which will not be covered).
8 dxes
Or:
a) 8 ds ltJe7 9 'ifd2 ltJg4 10 �g s h 6
11 �h4 g s 1 2 �g 3 fs 13 h 3 ltJf6 1 4 exfs
ltJxfs 15 �h 2 'ii'e 8 with a tiny edge as
Black considers a kingside attack based
on his space advantage there, or the
... es -e4-e3 lunge, or ...'ii'f7 and ... ltJe7,
building pressure on White's d-pawn.
White is not well situated to use his
asset - the e4-square.
b) 8 'ifd2 (as usual, it is a bad idea
for White to try to maintain the tension
- this only works if Black is not happy
1 e4 ltJc6 2 ltJj3
to take! ) 8 ... exd4 9 ltJxd4 l:te8 10 f3 (10
ltJxc6 bxc6 11 f3 ds transposes;
11 ... 'ii'e7 is also fine) 10 ... d s ! 11 ltJxc6
bxc6 12 l:tadl 'ii'e 7 13 �d4 dxe4 14
fxe4 ltJxe4 15 ltJxe4 'ii'xe4 16 �xg 7
<itxg7 17 �f3 'ife3+ 18 �xe3 l:txe3 19
�xc6 l:tb8 20 l:td8 l:tb6 with a level
endgame.
8...ltJxes
8 ... dxes is also playable. Z.Bratanov­
B.Chatalbashev, Bulgarian Champion­
s hip 2004, continued 9 'ii'xd8 l:txd8 10
�C4 h6 11 h3 b6 12 ltJds ltJas ! 13
ltJxf6+ �xf6 1 4 �e2 �b7 15 b4 ltJc6 16
c3 as 17 a3 ltJe7 18 ltJd2 �g s ! and
Black went on to win (see Game 79).
g ltJxes dxes 10 'iWxd8 :xd8
In practice this is equal, but White
can try to extract a little something:
a) By far the most common is 11
l:tfdl (or 11 Iladl - it doesn't much
matter) 11. .. �e6 12 ltJbs, when the
new move 12 ... :tdc8 ! keeps things level;
e.g. 13 ltJxa7 l:txa7 14 �xa7 b6 15 l:td3
l:a8 16 l:ta3 �f8 17 l:ta6 .i.c8 18 l:ta4
�d7 etc.
b) 11 ltJbs �d7 12 f3 looks scary,
107
T r; e 0 0 '1{ K ni9h t System
and both 12 ...b6 and 12 ....i.xbS give
White something to work with. The
novelty 12 ... a6 13 ti:JxC7 :rac8 14 i..b6
i..c6 15 ti:Jxa6 :d2 16 i..d3 bxa6 17 i.. e3
l:ixd3 18 cxd3 is a tiny edge for White,
but the imbalances should provide
Bl ack some winning chances as well.
c) 11 ..tc4 c6 12 :adl l:te8 13 a4 i..f8
14 f3 �g 7 15 :d2 ..tb4 16 lifdl l::te 7 17
�f2 :d7 and White is running out of
things to play for.
(3: 5 ..t e3 i..g 7 6 'iid 2
Other moves are covered elsewhere:
6 dS ti:Jb8, 6 h3 0-0, and 6 ..te2 0-0 7 0-0
were seen in lines B, Cl and C2 respec­
tively, while 6 .i.C4 is C43 below.
6 0-01
•••
6 ... ti:Jg4?! should not work here. 7
i..f4 es 8 ..tg s f6 9 i..h 4 0-0 10 i..C4+
�h8 11 dS ti:Je7 12 h3 ti:Jh6, intending
13 ... ti:Jhg 8, 14 ... h6 and lS .. .fS is satisfac­
tory for Black, but after 7 i..g s h6 8 ..th4
the bishop chase is n ot paying divi­
dends and White wins nearly every
game.
7 .i.h6
1 08
7 h 3 is Cl again, and 7 dS ti:Jb8 8 h 3
c 6 is i n the notes t o line B. Others:
a) 7 0-0-0 h as not scored well, but
White's set-up is challenging he h as -
man aged to omit h2-h 3 and stop ... e7-
es, while 7 ... ti:Jg4? ! 8 i..f4 eS 9 dxes
ti:Jgxes 10 ti:Jxes ti:Jxes 11 h4 is still not
convincing.
However, White has committed his
king, so: 7 ... a6 ! 8 i..h 6 bS 9 i.. xg 7 (9
i.. d 3 allows 9 ... es!) 9 ... �xg7 and the
annoying threat of 10 ...b4 already
forces some sort of concession (such as
the weakening 10 a3 or an awkward
defence of the e-pawn). White is only
slightly better.
b) 7 i.. e 2 eS (7 ... a6 8 dS! ti:Jb8 9 a4
was K.Wang-J.Schuyler, Washington
2012; I felt my a-pawn was misplaced,
and I did go on to lose, but I was not
without chances - see Game 80) 8 dxe s
(or 8 0-0-0 ti:Jg4 9 dxe s dxeS 10 'iVxd8
ti:Jxd8 11 ti:JdS ti:Je6 with an even pos i­
tion, while 10 'iie l ! ? ti:Jd4 11 �bl c6 12
.i.xd4 exd4 13 h3 'iif6 is unclear)
8 ... dxes is equal. With White's queen
on d2 and his passive bishops, this is

not the time for the exchange varia­
tion, and 9 0-0-0 'fIe7 (or 9 .. .ltJg4) 10
.i.g 5 .i.e6 11 ltJdS ?! .i.xdS 12 exds l:fd8
just makes matters worse for White.
c) 7 .i.c4 ltJg4 8 .i.g s h6 9 .i.h4 gs 10
.i.g 3 eS 11 dS ltJd4 is about equal, but
not a very rational position - the tac­
tics would take pag es. Instead,
7 ... .i.g4!? keeps things under control;
e.g. 8 dS .i.xf3 9 gxf3 ltJes 10 .i.e2 cS 11
0-0-0 'ii'as 12 �bl :tab8, or 8 0-0-0
ltJxe4 9 ltJxe4 dS 10 .i.d3 dxe4 11 .i.xe4
-.d7 12 dS .i.xf3 13 .i.xf3 ltJes 14 .i.e2
ltJg4 15 .i.d4 eS 16 dxe6 '6'xe6 17 .i.xg 7
�xg 7 with near equality.
7 eS!
•••
Centre play beats wing play.
S .i.xg7 �xg7 9 0-0-0
After 9 dS ltJe7 10 0-0-0 l:lb8 11 .i.d3
(if 11 h4 .i.g4, or 11 �bl bS) l1 . .. bS,
1 1 ...c6 or 11 ... .i.g4, Black has full coun­
terplay.
g .i.g4 10 dxes dxes
•••
This position has led to three draws,
but Black h as a slight advantage based
on White's weak bishop. For instance,
11 "'e3 ltJd4 12 ltJbS ? ! was F.5aez-
1 e4 ltJc6 2 ltJ/3
A.5afranska, G renoble 2003, when
12 ... cS 13 c3 'ifas ! 14 ltJa3 (14 cxd4?
cxd4 15 'ii'a 3 l:tac8+ 16 �bl 'ii'xa3 17
ltJxa3 ltJxe4 18 l:tel ltJxf2 19 :g l .i.xf3
20 gxf3 d3 21 %:[xe s d2 22 .i.e2 :fe8 2 3
ltJC4 b S i s nearly winning) 14. . ..i.xf3 1 5
gxf3 ltJe6 would h ave given Black a
large positional advantage.
(4: S .i.c4 .i.g7
And now:
(41: 6 0-0 1 1 0
(42: 6 'i'He2 1 1 0
(43: 6 .i. e 3 1 1 2
Instead:
a) 6 i.. g S ? ! ltJxe4! 7 ltJxe4 (or 7
.i.xf7+ �xf7 8 ltJxe4 dS 9 ltJcS l:i.f8)
7 ... dS 8 C3 ! ? dxC4 9 dS fS ! 10 dxc6
"xdl+ 11 :xdl fxe4 12 ltJd2 bxc6 13
ltJxe4 l:lb8 is good for Black.
b) 6 .i.f4 - As a novice, I remember
thinking that this was some kind of
ideal position for White. Nowadays, it
looks like White is begging for trouble,
as the bishops are vulnerable and do
1 09
The Dark Kn igh t System

not defend his centre. Indeed, my re­ only a little worse for White) 9 ... dxe4 10
search uncovered a large collection of �xe4 t'Llxd4 is clearly unsatisfactory for
games by some of the lowest-rated White.
players I have ever seen in any data­ 8 11xf7 9 t'Llxe4 dSI
•••

base. - 6 ... t'Llxe4 (6 ...�g4 is m ore com­

bative) 7 t'Llxe4 (7 �xf7+ �xf7 8 t'Llxe4
l::tf8 9 dS �g 8! 10 'ii'd2 'iVd7 ! ! 11 dxc6 ? !
"'g4! o r 1 0 .itg s t'Ll e s with a tiny edge)
7 ... dS 8 �d3 dxe4 9 .itxe4 t'Llxd4 10
t'Llxd4 "'xd4 11 'ifxd4 .itxd4 12 �xC7
.itxb2 13 llbl fS 14 �f3 .itd4 with a
level endgame.

(41: 6 0-0

White has four knight retreats and

they're all bad - he has reached an in­
ferior version of 5 h3 �g7 6 �C4 t'Llxe4!
etc in line Cl (see note 'c' to White's 6th
move). The most frequent is 10 t'LlcS ? !
and now, rather than 10. . ...d6 ? ! a s al­
ways played, 10 ...b6! 11 t'Llb3 'Wd6 12
t'Llg s llf8 13 l:tel eS gives Black a com­
fortable advantag e.

6 0-0
••• (42: 6 'We2 �g4!
6 ... .itg4 is also good; e.g. 7 �e3 0-0 8
h 3 �xf3 9 'iVxf3 eS 10 dxes (or 10 dS
t'Lle7) 10 ... t'Llxes 11 'i'e2 t'Llxc4 with
equality.
7 h3?1
This is no g ood, but it is the most
common, and other moves are either
met by ... �g4 or else are covered else­
where: 7 dS t'Llb8 is line B again, while 7
�e3 t'Llg4 is C43 below.
7 t'Llxe4 8 �xf7+
...

8 t'Llxe4?! dS 9 .itd3 (9 C3 ! dXC4 is 7 �e3

110

GM Robert Ze1cik has twice played 7
es and won (against much weaker op­
ponents). If Black takes the bull by the
horns with 7 ... i.xf3 8 gxf3 tLlxd4 9 exf6
tLlxe2 10 fxg 7 :g 8 11 �xe2 ! ? (or simi­
larly 11 i.xe2 - White is 2-0 here as
well) 11 ... lixg 7 we reach this mess:
Queen and two pawns is a lot for
three minor pieces, even three pieces
that include the bishop pair. Is there
really no way to use the material? The
key to the position is to make sure the
pawns stay mobile. It is not bad to sac­
rifice a pawn if it helps to open files,
trade rooks, disorganize the white
pieces, or expose the enemy king. A
close look at the games shows that one
of the candidate masters was crushing
the GM, so w e should n ot despair, and
instead resolve to emulate Black (up to
a point) in R.Ze1cik-M.Djurkovic, Pula
2001 (see Game 8 1).
Anyway, if this is not to your taste,
Black can chicken out without adverse
consequences; e.g. 7 ... dxes 8 dxes i.xf3
9 gxf3 tLlh s 10 f4 i.h 6 which is equal,
though still not simple.
1 e4 tLlc6 2 tLlf3
7 e5 8 dxe5
...
8 ds?! tLld4! 9 i.xd4 (or 9 'ifdl i.xf3
10 gxf3 0-0) 9 .. exd4 10 tLlbs (10 e s ?
.
O-O! 11 exf6 dXC3 wins) 10. . .0-0 11 0-0
:te8 was much better for Black in
M.5trubreiter- K.Rogetzer, Austrian
Team Championship 2004, and after 12
tLlbxd4? ! tLlxe4 13 �d3 tLlcs 14 'iVd2
Black should have cashed out: 14 ...i.xf3
15 tLlxf3 i.xb2, with a squeaky-clean
extra pawn to go with his positional
advantages.
8 tLlxe5 9 i.b3 0-0 10 0-0-0
•••
10 h3 i.h s 11 0-0-0 as simply
transposes.
10 a5
•••
This is a new move. Black has equal­
ized an d should start thinking about
how to exert the most pressure. White
will n ot enjoy playing a2-a4, but it is
necessary: 11 h3 i.h s ! 12 a4 (not 12
g4? tLlfxg4 13 hxg4 i.. x g4 and
14 ... i.xf3) 12 ... tLlxf3 13 gxf3 tLld7 14
l:Ihgl tLles 1 5 :g 3 c6 is still equal, Then
16 i.cs i.h6+ 17 i.e3 ! i.g7 repeats,
since 17 �bl? dxc s ! 18 l:txd8 l:taxd8
should terrify White.
111
The Dark Kn ight System

(43: 6 �e3 0-0 tural advantage with ....i.xc3) 11 ... �xh 6

12 ttJf3 c6 and White is only slightly
worse.
9 ... g5
This potentially king-weakening
move is acceptable because White has
also castled short.
10 .i. g3 ttJf6

7 0-0
7 'ii'd2 transposes to 7 .i.c4 in line
C 3 (see note 'c' to White's 7th move).
7 . ltJg4 8 .i. g5
. .

Instead:
a) 8 .i.f4?! ttJxd4! (fork tricks every­
where!) 9 ttJxd4 eS 10 �e3 ttJxe3 11
fxe3 exd4 12 exd4 .i. e 6 was H.Hughes­ Chances are equal; e.g. 11 h 3 ttJxe4,
K. Richardson, British League 2004. or 11 �b3 .i.g4 12 dS ttJas 13 h3 .i.hS
Black had the edge and went on to win 14 'iWd3 .i.g6.
(see G ame 82), although 12 ... c6! was
simpler with a comfortable advantage. (5: 5 .i.g5 .i. g7 6 -.d2 h6!
b) I don't know who would play it,
but Houd;n; likes 8 .i.cl, when 8 ... es
allows White a small advantage after 9
.i.g S ! .i.f6 10 .i.xf6 ttJxf6 11 dS ttJb8.
Instead, 8 ...ttJf6 returns to 6 ...0-0 in line
C41 and offers (or bluffs) a repetition.
8 h6 9 .i.h4
...

On 9 .i.f4?! ttJxd4! is best, as in the

previous note, even though White has
the extra possibility 10 ttJxd4 eS 11
.i.xh 6 ! (after 11 ttJe6! ? fxe6 12 'ii'x g4
exf4 13 �xe6+ Wh 7 14 .i.xc8 'i'xc8 This bishop hunt does work.
Black's active bishop allows him a small 7 .i.f4
plus - which he may convert to a struc- Others:

112

a) 7 i..h4 g5 8 i..g 3 tLlh 5 9 d5 tLlb8 (it
is good form to delay ...tLlxg 3 until the
last moment, even without any specific
idea in mind) 10 tLld4 cS 11 i..b 5+ ct>f8
was equal in A.Grilc-G .Mohr, Slovenian
Team Championship 2008, though the
better player (Black) went on to win
because White could not control the
dark squares (see Game 83).
b) 7 i.. e 3 tLlg4 (8 i..f4 tLlxd4! 9 tLlxd4
e5 equalizes; e.g. 10 h3 exd4 11 tLlb5
tLle5 12 tLlxd4 fie7 13 i..b 5+ i..d7 14
i.. x d7+ tLlxd7 15 0-0-0 0-0-0) 8 0-0-0
tLlxe3 9 'ii'x e3 0-0 10 h3 a6 11 g4 b5 was
equal in D.Janowski-F.Yates, Marienbad
1925(1), a wild game that ended in a
draw (see Game 84) .
7 dS!
•.•
Abruptly, Black aborts the bishop
chase in order to seize this opportunity
for an unexpected central thrust. This
is a new move.
8 exds
Or 8 i..b 5 tLlxe4 9 tLlxe4 dxe4 10 tLle5
i.. d 7 11 i.. x c6 i..xc6 12 tLlxc6 bxc6 13
0-0-0 'ii'd 5 14 �bl 0-0-0 1 5 c3 g5 16
1 e4 tLlc6 2 tLlf3
i..e3 f5 17 g 3 e5 18 'ife2 'iWb5 19 'ii'xb5
cxb5 20 dxe5 i.. x e5 2 1 i.. xa7 Wb7 with
some chances in the endgame for Black
due to his extra space, superior bishop,
and "queenside" (i .e. away from the
kings) pawn majority.
8 ... tLlxdS 9 tLlxdS 'ii'xds 10 c3
10 i.. x C7 is too greedy; e.g. 10... i.. g 4
11 c3 (11 i.. e2 :c8 12 i.. g 3 i..xf3 13
i..xf3 'iWe6+ 14 'ire3 tLlxd4 15 'iVxe6
tLlxf3+ 16 gxf3 fxe6 is equal) 11 ...i..xf3
12 gxf3 ':c8 13 i..f4 'ii'xf3 14 :g l e5 15
dxe5 i.. x e5 16 i.. xe5 tLlxe5 17 i..b 5+
�e7 and Black has slightly fewer king
problems than White.
10 i.. g4 11 i.. e2 0-0-0
•••
Black has reached a very satisfactory
cross between a Modern and a Scandi­
navian, and has excellent activity to
compensate for White's extra pawn
presence in the centre, with ... g6-g5
and/or ... e7-e5 to follow. Now 12 0-0
e5?! 13 tLlxe5 doesn't quite work be­
cause the g4-bishop is h anging, but
12 ...i.. h 5 ! threatens both 13 ... e5 and
13 ... g 5 14 i..g 3(?!) f5, with the initiative.
113
 C h a pte r S ix
1 e4 liJc6 2 liJC 3
To play the Dark Knight against
both 1 e4 and 1 d4 (as opposed to just
against 1 d4) requires some plan for
dealing with this relatively rare move.
As with 2 lLlf3, White h opes to re­
turn to familiar territory with 2 ... e5,
and while this is fine theoretically,
learning the Spanish Four Knights, Bel­
grade Gambit, Vienna Game, Vienna
Gambit, and Scotch Four Knights is not
a practical answer to an "uncommon"
variati on.
Another possibility is 2 ... e6 3 d4 d5,
transposing to a French sideline cov­
ered in Wisnewski's Play 1 ... lLlc6!.
114
Ideally we should find something
that resembles - and is likely to trans­
pose into - positions we already know,
but we need to be careful of our move
order because the Dark Knight Pirc is
not well suited for generating counter­
play against the Argentinean Attack.
(For instance, 2 ... g6 3 d4 i.. g 7 4 .lte3 d6
5 'ii'd 2 lLlf6 6 f3 e5 7 lLlge2 is tricky for
Black at best, though 7 ... a6 S 0-0-0 b5 is
a decent try.)
2 lLlf6 3 d4
•••
Others:
a) 3 f4 d5 4 e5 d4 5 exf6 (5 lLlce2?!
d3 ! 6 cxd3 lLld5 7 a 3 i.. g 4 S 1i'b3 lLlb6 is
a mess for White) 5 ...dxC3 6 fxg7 cxd2+
7 1i'xd2 'ii'x d2+ (or 7 ... i.. x g7 S 1i'xdS+
lLlxdS) S .ltxd2 .ltxg7 is just equal.
b) 3 lLlf3 d6 followed by ... g 7-g6 will
be a Dark Knight Pirc.
3 d6
.••
We intend to fianchetto but we are
not yet committed, so it is impossible
for White to start an Argentinean At­
tack. The most common reply by far is 4
lLlf3, transposing to Chapter Five.

Here, we will take a close look at:
A: 4 f4 1 15
B: 4 d 5 1 1 7
But first:
a) 4 i.. e 3 makes no sense because of
4 ... CtJg4! . If 5 i.. g 5 h6 6 i.. h4 9 5 7 i.. g 3
i.. g 7, we have gotten our fianchetto
with gain of time, reaching positions
similar to Ch apter Five. While thi s does
not actually give easy equality, it is cer­
tainly satisfactory - 4 i.. e 3 is almost
never played.
b) 4 f3 e5 5 CtJge2 exd4 6 CtJxd4 i.. e7
is a kind of Philidor where White's f­
pawn does not belong on f3. White
normally continues in "Argentinean"
style with 7 i.. e 3 0-0 8 'ifd2, when Black
has 8 ... CtJxd4 (8 ... d5 ! ? 9 0-0-0 dxe4 10
CtJxc6 'ifxd2+ 11 :xd2 bxc6 12 CtJxe4
CtJxe4 13 fxe4 offers a tiny endgame
advantage for White) 9 i..xd4 c6 10
0-0-0 and now 10...b5 with a n equal
game (though perhaps we should call
such a position unclear), or 10 ... i.. e 6,
which transposes to A.Mista-M.5zelag,
1 e4 CtJc6 2 CtJc3
Koszalin 1999 (see G ame 85).
c) 4 i..g 5 h 6 ! (Mestrovic's move,
which he has had a chance to use four
times, with two wins and two draws
against very strong opposition) 5 i..h 4
(this doesn't seem consistent, but it is
the most common; also, if 5 i..xf6 exf6 !
6 d5 CtJe7, White has little to combat
the incredibly slow but effective plan of
7 ... a6, 8 ... g6, 9 ... i.. g 7, 10 ... 0-0 an d 11 .. .f5,
with approximate equality) 5 ... g5 6
i..g 3 i..g 7 (6 ... e 6 ! ?) is also equal. White
tried 7 h4 g4 8 h5 in l .Jakic-Z.Mestrovic,
Zadar 2001 (see Game 86), where 8 ... e5
9 d5 CtJd4 was quite sufficient; while
R.Zelcic-Z.Mestrovic, Nova Gorica 2003,
sa w 7 f3 0-0 8 i..f 2, but Black used the
same idea, taking advantage of White's
slow play: 8 ... e5 9 d5 CtJd4! 10 tbge2 cS!
11 dxc6 bxc6 12 tbxd4 exd4 13 i..xd4
and although Mestrovic was successful
with 13 ... C5?! (see Game 87), 13 ... :b8 ! is
stronger, with excellent compensation.
A : 4 f4 e5
5 dxe5
Other moves don't offer much:
115
Th e Dark Kn ight System
a) 5 lDf3 exd4 6 lDxd4 is some sort
of Philidor where White's 4 f4 is prema­
ture.
Black must be alert in order to prove
this - and it is worth noting that, until
now, he has not been up to the task:
6 ... .i.e7 7 .i.e2 0-0 8 .i.e3 lDxd4 9 i..xd4
and here the new move 9 ... dS ! 10 .i.xf6
.i.xf6 11 .xdS .i.xc3+ 12 bxc3 'ii'f6 13
'iVd2 :td8 14 .i.d3 (14 'We3 is met by
14 ...'it'h4+ 15 Wf2 'ii'f6 16 eS "c6, or 1 5
g 3 'it'h 3 and White h a s some trouble
getting his king safe and activating his
h l-rook; e.g. 16 i..f3 b6 17 eS l::t b 8 18
'ii e 2 .i.b7 19 .i.xb7 lixb7 20 :tdl l::tbb8
with compensation) 14 ... 'Wb6 15 'ii'f2
'ifb2 16 0-0 'ii'xc3 17 eS 'iVd4 18 'ii'xd4
:txd4 19 :tadl �f8 20 c3 :d7 2 1 i..e4
:tb8 and Black unravels safely.
b) 5 dS lDe7 6 lDf3 (6 fxeS dxes 7 .i.e3
lDg6 8 a3 .i.d6 is fine for Black; or 6 fS
c6 ! 7 dxc6 bxc6 8 lLlf3 'ii'C7 with usually
... d6-dS coming soon, and sometimes
.. �b8 and/or ... g 7-g6) 6 ... exf4 7 bf4
lDg6 8 .i.g3 .i.e7 9 Wd2 0-0 10 0-0-0 lDd7
intending ... i.f6, ..Jte8, ... a7-a6, ... b7-bS,
...lDcS, ....i.b7. White's h2-h4 must be
116
met by ...h7-h6 so that the enemy pawn
does not reach the h6-square. The posi­
tion is equal.
c) 5 fxeS dxeS 6 dS lDd4! (6 ...lDe7
transposes to 6 fxes in note 'b') 7 lLlf3
.i.cS 8 lDa4!? (or 8 .i.g s 0-0 9 'ii'd 3 Wd6
and since 10 O-O-O?! lDg4! is strong,
White is clearly worse) 8 ...lDxe4 9 lDxcs
lDxcs 10 lDxd4 'i'h4+ 11 g3 'ii'xd4 12
'ii'xd4 exd4 and although White's bish­
ops are nice, it is not easy for him to
recover the pawn with a good position
- Black is a little better.
d) 5 i..b S ? ! exd4 6 'ii'xd4 .i.e7 7 lDf3
0-0 8 i.xc6 bxc6 9 0-0 :e8 10 l:dl i..b 7
with two bishops and a big bull's-eye
on the e4-pawn - more than enough to
make up for our space disadvantage.
5 ... dxe5
Black often throws in S ... i.. g 4 6 lDf3,
but this does not help the situation.
6 'ii'x d8 �xd8 7 lDf3 .i.b4 8 .i.d3 ::te8
This position has been reached only
once: in Ma.Tseitlin-C.Barlocco, World
Seniors Championship, Port Erin 2004,
which continued 9 0-0 exf4 10 lDdS
lDxdS 11 exds lDe7, and White went on

to win . However, Black can equalize
with 9 ... h6! 10 fxe5 ttJg4 11 h3 (or 11
:dl �C5+ 12 �fl �d7) 11 ... ttJgxe5 12
�f4 (or 12 �b5 �d7 13 ttJd5 ttJxf3+ 14
gxf3 �d6) 12 ...ttJxd3 13 cxd3 .i.e6.
B: 4 d5 ttJb8
Don't get any funny ideas. 4...ttJe5 ? !
5 f4 ttJed7 6 ttJf3 c 6 7 dxc6 bxc6 8 e 5 i s
no good for Black.
After 4 ... ttJb8, White will usually
play ttJf3 soon (and Black ... g7-g6),
transposing to line B in Chapter Five.
Here we investigate some independent
set-ups by White.
5 f4
The most often played, but it is not
a strong move. The problem is that,
after 5 ... c6, White will be unable to
maintain the e4/f4 pawn duo, unless
he plays the otherwise undesirable 6
dxc6. Instead:
a) 5 .i.g 5 g6 6 �xf6 exf6 7 �d3 (to
hinder .. .f6-f5) 7 ... �g 7 8 'ilVd2 0-0 9
0-0-0 ttJd7 10 h4 ttJcS 11 h S fS with fun
counterplay.
b) S �e3 (this logical move has had
1 e4 ttJc6 2 ttJc3
only on e obscure trial) 5 ... c6 6 'ii'd2
ttJbd7 7 0-0-0 cxd5 8 exds a6 should
make for a lively game, with ... b7-bS
and ... �b7 coming, and probably ... g7-
g6 and ...�g7, with maybe .. ':c8, ... 'iic 7
and/or ... ttJb6. White will be unable to
organize the rapid pressure on the e­
file that would make this uncomfort­
able for Black.
5 c6
•..
6 dxc6
Other moves:
a) 6 ttJf3 cxdS 7 exdS g6 8 �e3 �g7
is equal . Black will continue with ... 0-0,
... ttJbd7, .. :ilC7, ... a7-a6, ...b7-bS, ... ttJcS
(or ...ttJb6), and ... .i.b7 (or ... .i.d7). Since
White's central pawns are split, the f­
pawn is mainly a weakness. Then 9
�e2 0-0 10 0-0 ttJbd7 transposes to
M.Drasko-Z.Mestrovic, Bosnian Team
Championship 2003 (see Game 24).
b) 6 �e2 g6 will probably transpose
to 6 ttJf3, after 7 ttJf3 cxd5 8 exds �g7 9
i.. e3 for instance. Instead, A.5tefanova­
M.Zielinska, Dresden 2004, continued 7
i.. e3 "as 8 i..f3 ! ? �g 7 9 ttJe2 cxd5 10
exds �g4, which was about equal, but
117
The Dark Kn ig h t System
Black did go on to lose. He can improve
slightly by 7 ...iLg 7 and 8 ... 0-0, rather
than committing the queen so early.
6 ... ltJxc6 7 ltJf3 g6
8 i.. C4
Altern atively:
a) 8 e s ! ? is a good try, though it has
never been played: 8 ...ltJh S (or 8 ..ltJd7 ! ?
. eS!, after which White was much better
9 exd6 iLg7 ! 1 0 dxe7 .-xe7+ 1 1 'iWe2
ltJb4 12 .-xe7+ �xe7 13 �dl :d8 with
compensation) 9 exd6 iLg 7 10 i..C4 0-0
11 0-0 exd6 and White's holes balance
Black's holes.
b) 8 iLe3 iLg 7 9 'iWd2 0 -0 10 0-0-0
ltJg4 11 i.. g l 'Was 12 h3 iLxc3 13 'it'xC3
118
'ifxc3 14 bXC3 ltJf6 with slightly the
more comfortable game for Black.
8 ... i.. g7 9 0-0 0-0 10 �h1 iLg4 11 i.. e3
:c8 12 iLb3
The position is equal, but Black
needs to find something more active
than 12 ... a6? ! 13 h3 i..d7 14 'ifel bS?! lS
in N.Rag havi-K.Szczepkowska, World
Ju nior Champions hips, Istanbul 200S.
Black can choose between 12 ...ltJd7,
12 ......d7, 12 ... ltJh s, and 12 ... ltJas (since
the complications after 13 iLxa7 ? ! b6
14 es ltJd7 ! lS h3 i..xf3 16 'ii'xf3 ltJc6 17
ltJb s dxe S favour Black).
 Section Th ree

Ot h e rs

1 c4 Chapter Seven
- he is experienced an d comfortable. In a
1 lLlf3 - Chapter Eight few cases, I am forced to admit that
Others Chapter Nine
- 1 ... lLlc6 is not suitable, in which case I
Miscellaneous Topics Chapter Ten
- like 1 . g6. I choose to believe that this is
..

not too much of a departure, since the

This section deals with almost eve­
rything except the opening moves 1 e4
and 1 d4, including some quite un­
usual openings. Except for 1 C4, which
n eeds to be taken seriously, the chal­
lenge for Black is to give himself
chan ces to wrest the initiative early,
and/or make sure that White does not
reach the type of position with which
fianchetto is so common in the Dark
Knig ht System.
I will warn the reader that the cov­
erage in section 3 is less detailed than
in the other sections. This is because
these openings are far less common,
and also because you may already have
your own systems of defen ce, which
there is n o need to abandon .

119
 C h a pter Seve n

1 c4 CLJc6

tLlC3 and now Black can certainly play

You could play the Dark Knight
against 1 e4 and 1 d4 and then play
S M.fS, transposing to line Bl; but S ... dS!
is even better, as in Ge. Lambert­
A. Labarth e, Vichy 2000. Not without
reason, White worried about the centre
opening up and played 6 CS?, when
6 ... h6 7 tLlh 3 .i.xh 3 or 6 ...b6! would have
been strong; but even after 6 a3 .i.xc3 7
bxc3 tLlf6 8 cxdS 'ii'x dS 9 f3 0-0 10 fxe4
tLlxe4 1 1 tLlxe4 'it'xe4 12 'iVd3 .i.fS Black
maintains an advantage along with his
grip on the e4-square.
2 e5
•••

any favourite system against 1 c4, but

it makes some sense to use 1 .. tLlc6 here
.

too, since there is a fair chance of

reaching positions you already need to
know to meet 1 d4.
2 tLlc3
Others:
a) 2 d4 eS returns to Chapter Two.
b) 2 g3 eS 3 .i.g 2 fS will transpose to
line A below once White (in evitably)
plays tLlc3.
c) 2 tLlf3 eS 3 d4 e4 4 tLlg s (4 tLlfd2 fS White will now choose from the fol­
S tLlc3 tLlf6 6 e3 is line B2) 4 ....i.b4+ S lowing:

120

A: 3 g3 121
B: 3 ttJf3 123
A: 3 g3 f5 11
A reversed Grand Prix Attack, which
is "not a good idea", according to Hi­
karu Nakamura. Nonetheless, it was
most likely the great disparity in play­
ing strength that was the difference in
our encounter - upon examination, th e
positions are just fine so long as Black
does not use the typical plan of trying
t o deliver checkmate right out of the
opening. Come to think of it, he may
have been saying that the opening was
a bad choice against him in particular,
which is no doubt true given that I had
nothing prepared other than the usual
recipe.
4 .i.g2 ltJf6 5 d3
After 5 e3, the gambit 5 ... d5 ! ? has
been very successful. The idea is to
open up the game to reach the weak
light squares on White's queenside,
particularly the gaping hole on d3. If
White tries to remain a pawn ahead,
1 c4 ltJc6
Black obtains dangerous compensa­
tion; e.g. 6 cxd5 ltJb4 7 'ii'h 3?! ltJd3+ 8
<Ml e4 9 ltJh 3 .i.d6 and White h as great
difficulties with both king safety and
development. Instead, 7 d3 ltJfxd5 8
ltJxd5 ltJxd5 9 ltJf3 is the normal con­
tinuation, when Black's position is a bit
loose but White neerls to create some
sort of off-board diversion to sneak his
e-pawn back to e2. L.Gofshtein ­
N.Mitkov, Lisbon 1999, continued
9 ... .i.d6 10 0-0 ltJf6 11 �3 'ike7 12 e4
fxe4 13 dxe4 and the game was equal
after 13 .. :iff7 14 'iVc3 'iWh 5 15 ltJd2 0-0,
though White went on to win (see
Game 88).
5....i. b4 6 .i.d2
Otherwise White's c-pawns will be
doubled, providing long-term compen­
sation for surrendering the bishop pair.
A brutal example is J. Ramirez­
J.5chuyler, Las Vegas 2007, which saw 6
a3? .i.xc3+ 7 bXC3 d6 8 1:lbl 0-0 9 e 3 ? e4
10 d4 b6, followed by ... .i.a6 (see Game
89).
6 0-0
•••
7 ltJf3
121
Th e Dark Kn ight System
Instead:
a) 7 e3 worked great for White in
H.Nakamura-J.5chuyler, Las Vegas 2008:
7 ... .i.xc3 8 .i.xc3 d6 9 lLle2 'iVe8 10 h3,
and since 10...'iVhS ? ! would be met by 11
f4! I was already out of ideas and lost
quickly. Meanwhile, Black is actuany
fine after 10 ...b6 ! ? 11 0-0 .i.b7 to oppose
White's strong bishop and work to­
wards relieving him of the bishop pair.
This is nearly equal according to Hou­
d;n;, and can be used just as wen
against 10 'i¥d2 (though 10 0-0 is a seri­
ous test).
Going back further, Black has 8 ... dS !
which is three for three in my database,
most significantly in O.Foisor-J.M.De­
graeve, Le Touquet 1996 (see G ame 90).
And if that's not enough, in J.lruzubieta­
B.Gulko, San Sebastian 1996, Gulko
casuany played 7 ... lLle7 ! ?, and after 8
lLlge2, 8 ... c6! to guard the ds-square and
preserve the dark-squared bishop. This
too was nearly equal (see Game 91), al­
though 8 a3 is more testing.
Last but not least, Black can try
7 .. .f4 ! ? which scored a full point in
N.5piridonov-K.5praggett, Cannes 1992
(see Game 92). In case it's n ot obvious,
my main recommendation is 7 ... .i.xc3 8
.i.xc3 dS!.
b ) 7 a3?! is very often played, and
with very bad results. It is hard to jus­
tify spending a tempo to force an ex­
change which Black is interested in
making anyway: 7 ....i.xc3 8 .i.xc3 d6
(again 8 ... dS ! ? 9 cxdS lLlxdS 10 .i.d2 h 6 !
11 lLlf3 e4! or 11 'iVh3?! .i.e6 12 'i¥xb7
122
lLld4 with excellent compensation) 9
lLlf3 'i¥e8, and with White having
squandered a tempo on a2-a3, Black's
"typical plan" (described below) has
done well, while if White delays cas­
tling, he runs other risks. F.Brun o­
B. Kurajica, Lugano 1985, was a short
lesson for White (see Game 93).
7 e4!
...
This rare line is best, taking advan­
tage of White's previous move to estab­
lish a knight on the e4-square.
For the record, the " typical plan" is
7 ... d6 8 0-0 .i.xc3 9 .i.xc3 �e8, intend­
ing 10 .. :i'h S, 11 .. .f4, 12 ... .i.h 3, 13 ... lLlg4,
14 .. .fxg 3 lS fxg3 .i.xg 2 16 �xg 2 :xf3
17 �xf3 l:tf8+ and wins! Some of the
most astute readers may have noticed
that five of White's moves were
skipped - indeed that is a little snag . To
see how Black's attack can work
ag ainst an opponent with a pulse,
check out M.5her-K.5praggett, Andorra
1993 (Game 94), though Spraggett did
require a large assist from his GM op­
ponent, and 7 ... d6 is not my recom­
mendation.

8 dxe4 .i.xc3!
8 ...ltJxe4 will normally transpose,
but this new move order cuts down on
White's options.
9 .i.xc3 ltJxe4 10 :C1
I expect White to try to preserve his
pawn structure as he has in the analo­
gous positions from the alternative
move order. However, if 10 O-O ! ?
ltJxc3 ? ! 11 bXC3, White's open lines
make it easy for him to pressure the
queenside. So Black sh ould retain the
strong knight: 10 ... d6 11 'iYC2 as 12
ltJd4 'ile7 and 13 ....i.d 7 with just a
slight edge for White.
10 ... d6 11 0-0
The only time this position has been
reached, Black started losing the thread
with 11 . .. .i.e6 ? ! 12 ltJd4 ltJxd4?! 13
'i'xd4 ltJxc3? ! 14 :xc3 cS 15 'iVd2, when
White had a pleasant long-term advan­
tage and went on to win in the game
Har.Becker-Joa.Franz, German League
1997.
Instead, after 11 ... 'iVe7 12 ltJd4 .i.d7
13 'Wb3 ltJcs 14 'ii'c 2 as, White's edge is
tiny.
1 c4 ltJc6
B: 3 ltJf3 f5
This move has a solid reputation
and is a popular alternative to 3 ...ltJf6.
4 d4 e4
And now main ly:
B1: S lb g s 124
B2: S tlJd2 125
But also:
a) 5 ltJes - only Larry Christiansen
(and Houdinn seems interested in play­
ing this move, though he has an im­
pressive 3-0 with it. After S ...ltJf6 6 .i.f4,
as in L.Christiansen-S.Conquest, Oviedo
(rapid) 1992, the new move 6 ... ltJhS ! ? 7
.i.d2 ltJf6 is fine for Black if he is con­
tent with a draw, because 8 e3 d6 9
ltJxc6 bxc6 10 .i.e2 .i.e7 11 0-0 0-0 is
nothing for White. Otherwise Black can
try 6 ... d6 7 ltJxc6 bxc6 8 e3 .i.e7 9 .i.e2
0-0 10 0-0 .i.e6, with a slight edge for
White.
b) 5 .i.gs .i.e7 6 .i.xe7 ltJgxe7 7 ltJd2
ltJxd4 8 ltJdxe4 ltJe6 9 ltJd2 b6! 10 e3
.i.b7 11 ltJf3 0-0 with easy equality and
chances for more; e.g. 12 .i.e2 f4! is
123
good for Black, as is 12 g3?! f4! 13 exf4
�4 14 gxf4 :Xf4 1 5 i..g 2 (or 1 5 i.. e 2)
15 ... 'iff8! .
81: 5 ttJgs i.. b4
6 ttJh3
This retreat is necessary.
a) G.McKenna-J.5chuyler, Richmond
2008, instead continued 6 f3 ? ! h6! 7
ttJh 3 exf3, threatening 8 ...'ifh4+ and
9 .. :.-xd4. My opponent found 8 e3!
which keeps White in the game,
though 8 .. :.-h4+ 9 g3 i.. x c3+ 10 bXC3
�h5 11 i.. g 2 'ii'f7 12 i.. xf3 'iVxC4 13
i..h 5+ �d8 would have allowed Black
an edge. (A. Beliavsky-V.Bagirov, Minsk
1983, also saw White play 6 f3 ?!, but
Black did not punish him and went on
to lose.)
b) 6 g 3 ? ! , as in lTimman-I.Sokolov,
Dortmund 1999, is n ot a good idea ei­
ther - the bishop should not be caught
dead on the g2-square. After 6 ...ttJf6 7
d5 ttJe5 8 'iWb3 'iVe7 9 i.. g 2 ? ! h6 10 ttJh 3
'Wc5 11 i..f4? ! ttJxC4 White had little for
the pawn.
After the text move, the position
124
again resembles a reversed Grand Prix
Attack. White's king's knight will soon
reach the excellent f4-square, but the
four moves it takes to get there is a
high price, even in a blocked position.
6 ... ttJf6 7 e3 i.xc3+ 8 bxc3 d6 9 ttJf4 0-0
White's c4-pawn is weak and his
bishops are not yet working, but he
should be able to force through th e c4-
c5 break. If Black is not careful, this will
bring White a significant advantag e.
However, White needs to be careful as
well, because if the position opens up
at the wrong time or in the wrong way,
he will be punished for his slow devel­
opment.
10 h4 b6 11 i.a3 :f71
This is a new move, and it makes us
more comfortable by stepping out of
the influence of White's a3-bishop and
avoiding ttJe6 forks. It is also useful to
guard the seventh rank and shield our
king. The similar 11 ... .:te8 was tried
successfully in R.Koch-M.Wiedenkeller,
Reggio Emilia 1982/83, but it accom­
plishes a bit less and leaves Black vul­
nerable to the i..b 5 pin . It is worth not-

ing that Black's results have been excel­
lent in this variation even without the
text move, though a few high-level
players have successfully advocated
White.
12 cS
If White does not play this now,
12 ... �a6 is coming, and the c4-c5 break
will be forever ineffective.
12 ... bxcS
12 ... d5 ! ?
13 �C4
Naturally, if 13 dxc5 then 13 ... d 5 ! 14
c4? d4 and White's structure is awful.
13 ... dS 14 �bS lLlb8 15 �xcS c6 16 �e2
i.. a 6
White has relieved himself of his
pawn weakness, but the coming bishop
exch ange brings equality. Black's plans
are to manoeuvre a knight to the lovely
c4-square, trade off White's f4-knig ht,
and fight for the b-file; e.g. 17 Abl
i..x e2 18 'iVxe2 lLlbd7 19 lLle6 .-c8 20
i.. d6 lLlb6 21 lLlc5 lLlfd7 22 lLlxd7 lixd7
(or 22 ... �xd7 23 i..f4 lIc8 and 24 ... c5)
23 i..f4 lLlC4 24 0-0 :b7, which is still
equal. Black has achieved his objec-
1 c4 lLlc6
tives, but White's "bad" bishop is very
strong . Black will not be better unless
he can arrange to exchange the other
pair of rooks.
82: S lLld2
This move is not common here, but
the position is important since it can
be reached from a few different move
orders.
s ...lLlf6 6 e3
6 lLldb l ! ? is a cute move, intending
7 i..g 5 . Houd;n; likes 6 ... b6 ! ? opening
paths for the bishop and sketching out
some territory on the queen side. Then
7 i..g 5 h6 8 i..xf6 'ifxf6 9 e3 (or 9 lLld5
i..b 4+ 10 lLllC3 i..x c3+ 11 bxc3 'iVd6 12
e3 lLle7 13 lLlf4 g5 14 'ii'h 5+?! �f8 15
lLle2 i..a 6 and Black is better) 9 ...1ff7 10
lLlb5 Wd8 11 i.. e 2 g 5 leads to a murky
position which Mr. H calls equal. Unfor­
tunately, there are no games to draw
on, and the computer convincingly
shot down all my sensible ideas, so we
are stuck with this nonsense. He con­
tinues 12 a3 h 5 13 lLld2 94 14 '1Wc2 h 4
15 0-0-0 h 3 16 g3, still equal.
1 25
The Dark Kn ig h t System

n ovelty that prevents Black from get­

ting overrun on the queen side, reach­
ing a satisfactory reversed Classical
French.
9 cxdS
Or 9 a3 i.. e 6 10 b4 a6 !, strongly dis­
couraging the further advance of
White's queenside pawns.
9 lbb4 10 'i'b3 lbbxdS 11 lbC4 c6 12
.•.

i.. d 2 'ii' C 7

6 ...i.. e 7!
6 ... g6 is just about always played,
but the fianchetto is slow and some­
what accommodating, releasing the
b4-square. Sokolov and Sigurjonsson
both got squashed by White's huge
queenside after l:lbl and b2-b4. The
text move keeps Black focused on
where the play wi1l actually take place.
7 i.. e2 0-0 8 0-0 dS!
Previously S ...'i'eS and S ...d 6 have Intending 13 ... i.. e 6, 14...:fdS, an d
been tried. The text is an unplayed Black is fine.

126
 C h a pte r E ight

1 tDf3 tDc6

4 i.g2 g61
An excellent recipe. White's extra
tempo amounts to little.
S 0-0 i.. g 7 6 e4
White challenges the centre with 6
c4 less often. In Bu Xiangzhi­
V.lvanchuk, Khanty-Mansiysk 2011,
Black played 6 ... dxC4 7 dXC4 'ii'xdl 8
:xdl e4 9 l2Jfd2 fS, when White needed
10 f3 to equalize. Instead, after 10 l2JC3
J.. e 6 11 l2JdS ? ! 0-0-0 Ivanchuk went on
If you play 1 ... l2Jc6 against 1 e4 and to win (see Game 9S).
1 d4, you may as wen play it against 1 6...l2Jge7 7 l2J bd2 0-0 8 c3 as 9 a4 h6
l2Jf3, since White will usually respond
with 2 d4, 2 e4, or 2 c4, transposing to
Ch apters One, Five and Seven respec­
tively. The only continuation with in­
dependent significance is:
2 g3 eS 3 d3 dS!
A reversed Pirc is quite safe for
Black, as long as he is careful not to
choose a sharp variation. I've played
3 .. .fS here in the past because it is ex­
tremely effective against the King's
Indian Attack, but its value is question- Black h as outscored White here at
able against other set-ups. all levels of play. K.Arakhamia Grant-

127
The Dark Knight Sys tem

A.Raetsky. Bern 1995 . show ed Black A.Baburin. San Francisco 1997. saw
expanding and using his space advan­ Bl ack's central control turning into a
tage after 10 lI el �e6 11 exds �xdS ! powerful tactical strike against White's
(see Game 96); while V.Frias Pablaza- king after 10 ex ds ltJxds (see Game 97).

128
 C h a pte r N i n e

Ot h e rs

I used to "pre-move" 1 .. .ltJc6 in online din; an d I have cooked up something

blitz games - after all, it's playable new and fun against Larsen's Attack: 1
against everything, right? Eventually I b3 g6 2 .i.b2 ttJf6 3 e4 i.g7 ! ? (3 ... d6 is
was embarrassed (several times!) by a much more common) 4 eS ttJdS 5 c4
player who sometimes opened 1 b4. ttJf4!
After 2 bS, I did not feel like I had found
the refutation to the Orang-utan !
2 ... ttJes 3 .i.b2 ? ! ttJC4! ? is not bad, but 3
e4 or 3 d4 leaves Black with no excuse
for his knight placement.
In all seriousness, 1 ..ttJc6 is just not
.

the best way to start ch allenging White

after 1 b3, 1 f4, 1 e3, 1 a3, or especially
1 b4. Play whatever you like, but in my
opinion, 1 ... g6 - the "Dark Bishop"? - is
simple and strong against all of these.
Against 1 b4 and 1 b3, 1 ... g6 may
come as a surprise to the opponent,
which is always nice. It also makes sense
to oppose White's bishop on the long
diagonal, especially since his will be
loose, while ours (after we castle short)
will be guarded - a tactical advantage.
La rsen's Attack
Based on the counter-fianchetto, Hou-
This move is provocative and strong .
White can keep advancing pawns, with
tempo in some cases, but he will not be
able to control the territory he imag­
ines he is conquering, and he will be­
come weak in an area of the board I like
to call the "deep centre" - d3, e3, and
the surroun ding squares. After 6 g 3
ttJe6 7 d 4 d 6 8 d S ? ! ttJcS 9 b 4 (otherwise
the knight is very powerful in conjunc-

129
The Dark Knight System
tion with ....i.fS) 9 ..lDcd7
. 10 f4 0-0 11
lDf3 c6 ! 12 'iWb3 cxdS 13 cxdS dxes 14
fxeS as 15 bxas lDcs, White's position
is approaching the later stages of a
long and painful decline. It is easy to
criticize White's play, but he was al­
ready worse very early; for instance, 8
lDf3 0-0 9 lDC3 dxes 10 dxes lDd7 11
lDe4 b6 12 'iVd3 i.b7 13 'ii'e 3 fS! 14 exf6
exf6 15 0-0-0 still leaves White strug­
gling after ls ...l1e8 or lS .. .fS.
Sokolsky's Opening (aka
The Orang-utan, aka The Polish)
1 b4 g6 2 i.b2 lDf6 3 c4 .i.g7 4lDf3 0 -0
5 e3 d6 6 d4lDbd7 7 .i.e2 eS 8 0-0 e4 9
lDfd2l:te8
I know, there are other ways the
game can go, but this is typical. Black
has reached a King's Indian Attack in
reverse, and in practice Black's attack
has proven stronger (+12 -2 =4 is
77.7%). White's extra tempo, .i.b2, is
just about useless because the bishop
belongs on the a3-square.
In spite of his successes, Black has
yet to produce a model for correct play.
130
In the games I can find, Black keeps
allowing the enemy pawn to a6, which
accelerates White's queenside initia­
tive. The reader is better off studying
the classic King's Indian Attacks by
Fischer, Petrosian, Bronstein, Larsen,
etc. They go (in reverse - I have trans­
lated here) ...lDf8, ...h7-hS, ....i.fS, ... a7-
a6 (when White's pawn reaches as),
...lD8h7, ... hS-h4 (and ...h4-h3 if possi­
ble), ...lDg 5 with scary invasions coming
on the light squares or (if White has
played h2-h3) scary sacrifices on the
h3-pawn.
Bird's Opening
Against 1 f4, I would warn against play­
ing ...d7-dS, which gives White what he
wants: the es-square. Instead, after
1 ... g6 2lDf3 .i.g7 3 e3 d6 4 d4lDd7,
Black is +17 -1 =3 according to my
database. Wow! Obviously White's po­
sition can't actually be that bad. What
the statistics mean is that, for those
interested in winning chess games, it is
extremely important to combat the
opponent's automatic plans. Then 5

�d3 es 6 c3 'i!ie7 was no fun at an for
White in A.Capaliku-J.Gombac, Nova
Gorica 2010 (see Game 98); while 5
.i.c4 e6 has scored 100% for Black, most
notably in A.5pichkin-D.Reinderman,
Rijeka 2010 (see Game 99).
Of course, White does not have to
play a Stonewan - just as popular is a
reversed Leningrad Dutch with 3 g3,
when one system has performed ex­
tremely well for Black: 3 ... b6 ! ? 4 �g2
�b7 5 0-0 e6 6 d3 li:Je7 7 e4 d6
with 8 ...li:Jd7, 9 ... C5, 10 .....c7, and
11 0-0 to follow. Dzindzi pithily states
...
that "any opening named after an
Oth ers
animal is bad", but here the Hippo­
potamus is wen-suited to combat both
White's kingside expansion and his
fianchettoed bishop. Black certainly
made it look easy in P.Auchenberg­
To.Christensen, Helsingor 1997 (see
Game 100). An important idea is to
play .. .f7-f5 just when White is ready
for his f4-f5 advance (though Christen­
sen did not find this necessary).
Other ot hers
As for 1 li:JC3 and 1 g3, 1 ... li:Jc6 should
transpose to our repertoire at some
point.
As I mentioned already, 1 ... g6 is a
good answer to 1 e3, and also to 1 a3,
when White is challenged to find any
use for his opening move. If 1 h3, 1 h4,
1 li:Jh 3, 1 li:Ja3, 1 a4, or 1 f3, you're on
your own. 1 f3 is particularly troubling
because it is not clear whether White is
intending 2 li:Jh 3 and 3 li:Jf2, or 2 g4!, or
2 �2 ! and 3 �g 3 ! !. May you be con­
fronted with these problems fre­
quently.
131
 C h a pter Te n

Misce l l a n eo u s Topics

Reducing the Workload - but anyone who prefers t o fianchetto

The Dark Knight for Dum m ies can study less an d play his own favour­
Perhaps some readers have g otten the ite positions more.
impression that the study material Notice that when playing the Sys­
(which in my opinion is very small) can tem this way, White's f- and c-pawns
be substantially simplified. Indeed, it are both free, so White could play a
can . There is no sen sible way to avoid " Four Pawns Attack" if he so chooses.
Chapter One, but Chapter Two is, in Without going into detail, I will point
some sense, unnecessary. After 1 d4 out t ha t this is not a good way for
ll)c6 2 c4, Black can simply play 2 ... d6, White to fight for an advantage: 1 e4
which will almost certainly transpose ll)c6 2 d4 d6 3 ds ll)b8 4 c4 g6 5 f4 .i.g7
to Chapter One. Likewise, Black need 6 ll)f3 tDf6 7 tDc3 0-0 8 .i.e2 e6!
not learn Chapter Three. After 1 d4 ll)c6
2 dS, Black can just play 2 ... ll)b8, with a
likely transposition to Ch apters One or
Five. Chapter Four? Toss it. Just play 1
e4 ll)c6 2 d4 d6, probably transposing
to Chapters One, Five, or Six.
In other words, don't play for an
early ... e7-eS and plan on fianchettoing
the kin g's bishop whether it is neces­
sary or not. The thing is, the positions
from Chapters Two, Three and Four, are
among my favourites in the Dark 9 0-0 (9 e s ! ?) 9 ... exds 10 cxdS c6! 11
Knight System, so I would rather know dxc6 ll)xc6 12 Wh l lIe8 13 .i.d3 �6, or
them and play them than avoid them, 13 ... a6 and 14 ... bS, with equality.

132

The Light Knight Attack!?
If the Dark Knight System is so great,
why not play it with an extra tempo?
Well, the good news is that nobody can
stop you; e.g. 1 lLlC3 ! ? eS 2 d3 dS 3 g 3
d4 4 lLlbl lLlc6 S i..g 2 with a reversed
Dark Knight Pirc. Furthermore, there
are many other possible move orders
White can use. The bad news is that it
will often be Black fighting for the ad­
vantage, not White. How can a system
be good for Black and not for White?
The answer is that Black and White
have different opening objectives. In
th e DKS Black often loses some time
with his knight, but this is mitigated by
th e fact that he has reached a position
where tempi are relatively unimpor­
tant. Therefore, playing such a position
a tempo up represents only a small
gain, sometimes not enough even to
equalize.
As for l lLlc3 dS 2 e4,
I believe White can fight for an ad­
vantage after 2 ... d4 (a practical one, if
not a theoretical one), but in my ex-
M iscella n eous Topics
perience Black usually plays 2 ... e6,
2 ... c6, or 2 ... lLlf6 instead, transposing to
his favourite defence (French, Caro­
Kann, and Alekhine's, respectively), or
just 2 ... dxe4 3 lLlxe4 and 3 ... .i.fS,
3 ... lLlf6, or 3 ... lLld7, with a good version
of Black's favourite Caro-Kann varia­
tion. While these positions are an play­
able for White, there is little chance of
getting Black out of his comfort zone.
Of course, real chess is not the same
as theory, and for the right player
and/or the right opponent, the "Light
Knight Attack" could be the perfect
weapon. The grandmaster Jorg Hickl
has played 1 g3 es 2 .i.g 2 dS 3 d3 lLlf6 4
lLlf3 i..d6 S lLlC3 c6 6 e4 (or 6 0-0 0-0 7
e4) many times with good results.
Readers might choose to view the
position as a Dark Knight Pirc reversed,
and Hickl's move order is an excellent
way to reach it since Black never had a
good opportunity to play ... dS-d4. There
is not much else to learn because
Black's set-up is an extremely common
one against 1 g3.
133
 I l l u st rative Ga mes
The games section serves several pur­
poses:
t It shows typical middlegame and
endgame ideas in the Dark
Knight System - there is a whole
game to be played, after all.
t It serves as a place to demon­
strate some alternatives to the
main recommendations.
t It shows by example why certain
popular variations were rejected.
Consider these games a warning.
t It shows how successful the Dark
Knight System can be in practice.
t It shows how much fun it can be
to play the Dark Knight System
in your games!
A "good" King's Indian
Came 1
J.Paasika ngas Tella­
T. Li ndqvist
F i n n i s h Tea m
Cham pionsh i p 1996
1 d4 d 6 2 c 4 e 5 3 e 3 lLlc6 4 d 5 lLlce7 5
lLlc3 f5 6 i.. d 3 lLlf6 7 i..C2 g6
134
We would reach this position via
the move order 1 d4 lLlc6 2 e3 eS 3 dS ? !
lLlce7 4 c 4 d 6 etc. Black now plays an
excellent version of a King's Indian -
not an uncommon occurrence against
an opponent inexperienced with the
Dark Knight.
B lLlf3 i.h611
Why not? Now if White plays e3-e4,
Black can relieve himself of a problem
piece. If White does not play e3-e4,
Black has the powerful eS/fs pawn
duo.
9 b4 0-0 10 i..b3 as 11 bxa5 l:txa5 12 a4
�hB 13 lLlb5 lLle4 14 0-011 c61 15 dxc6
bxc6 16 lLla3 i..e6
White has no compen sation for
Black's huge centre.

17 lLld2 lLlC5 18 .i.C2 :a6 19 i.b2 i.g7
20 e4 f411 21 �h1 'iVd7 22 f3?!
White weakens his dark squares
unnecessarily. Black should have to
work for this.
22 ...l:lfa8 23 lLlab1 lLlc8 24 lLlc3 lLlb6 25
i.b3 fif7 26 i.C2 i.f8 27 i.a3 lLlcxa4 28
lLlxa4 lLlxa4 29 i..x d6 i.xd6 30 l:lxa4
:xa4 31 i.. xa4 'iNc7
Black has somehow managed to
avoid winning any pawns, but his
bishop pair provides a nearly decisive
advantage.
32 h3?
3 2 "C2 .
Illustrative Games
32 ... i.b4! 3 3 lLlb1 i..xC4 34 Itg1 i..c5 3 5
:e1 i.f2
Black soon won.
An a nti-Stonewal l ­
rare light-squared play
Came 2
J.Vialatte-F.Giroux
P a r i s 2006
1 d4 lLlc6 2 f4 d5
Black has not signed a contract to
play on the dark squares, and White
has already given himself a big hole on
e4. Furthermore, with Black not having
played ...lLlf6, White is not guaranteed
an outpost on eS.
3 lLlf3 i..g4 4 e3 f6!?
Black may or may not get ...e7-eS in,
but he has certainly stopped lLle S !
5 i.. b 5 'ili'd6 6 0-0 a 6 7 i.xc6+11
This is not going to help the situa­
tion on the light squares.
7 .. :iVxc6 8 C311 lLlh6 9 lLlbd2
9 ... 0-0-011
Black's knight needs to get to the
13 5
Th e Dark Kn igh t System
d6-square as soon as possible - 9 ...lLJfS !
10 l:tel lLJd6 ! with a pleasant advan­
tag e. Black eventually gets the position
he should have, but only with some co­
operation from his opponent.
10 'iVel .i.f5 11 'ife2 e6 12 :el .i.e4 13
e4 .i.b4!?
Black's plan, which he will soon exe­
cute, is to give up both bishops and
play "g ood knight versus bad bishop".
This fails to take into account White's
development and queenside counter­
play.
14 exd5 exd5 15 a3 .i.xf3 16 'ifxf3 .i.xd2
17 .i.xd2 f5 18 :ael 'iVe6 19 'iVe2 l:td7
20 b4 �b8 21 b5 axb5 22 'ifxb5 lLJf7 23
.i.b4 lLJd6
24 'ifa4?
An incomprehensible positional
mistake. The knight that lands on c4 is
worth far more than the "tall pawn " on
b4. However, in this roundabout way
we see the proper fruition of Black's
early advantage.
24...lLJe4 25 .i.e5 b6 26 'ifa6 'ife6 27 l:tbl
'iVa8 28 'iVb5? We8! 29 a4
29 .i.b4 c6 traps the queen .
136
2 9... bxe5 3 0 'iVxe5 'ifa5 31 'ife6 J:td6 3 2
'ifb7+ Wd7 33 neel na8 3 4 'iVb5+ 'iVxb5
35 l:txb5 :xa4 36 l:tb8 lLJxe3 37 l:tel
l:te6 38 :g8 l:te7 0-1
Can Wh ite be made to
pay for delaying d4-d 5 ? -
Game 3
M. Trata r-M.Srebrnic
S l ove n i a n C h a m pi o n s h i p ,
Lj u b lj a n a 2010
1 d4 d6 2 lLJf3 g6 3 e4 .i.g7 4 lLJe3 lLJe6
We would reach this by 1 d4 lLJc6 2
lLJf3 d6 3 c4 g6 4 lLJC3 .i.g7 - in this par­
ticular game White did not even have
the option of playing 4 dS.
5 d5 lLJe5 6 lLJxe5 .i.xe5 7 e4 lLJf6
It seems odd to cut off the retreat
for the dark-squared bishop, but the
piece is very active where it is, an d the
danger is far less than it appears.
White's lame attempt to trap the
bishop on move nine goes nowhere.
Such lovely dark square control !
8 .i.d3 0-0 9 lLJe2
In addition to the threat of f2-f4,
White makes sure Black cannot dam­
age his structure with ... .i.xc3 - how­
ever, this retreat is still not the most
accurate.
9 ... lLJd7 10 h4 h5 11 lLJf4 lLJe5 12 .i.e2
e6!
White has forgotten about his de­
velopment and now stands worse.
White is a GM, by the way. The rest of
the game is kind of brutal .

13 11bl as 14 f311
Be careful on the dark squares,
White !
14 ... exds 15 cxd S c6 16 dxc6 bxc6 17
ttJe2 �a6 18 �gs 'i'c8?! [18 ... 1i'b6] 19
ttf2 :b8 20 b3 :te8 21 :tel ltJe6 22 �e3
dS! 23 exds cxd S 24 :Cl 'i'd8 25 �d3
'ifxh4+ 26 �fl �b7 27 'iVd2 d4 28 �gl
ltJf4 29 ltJxf4 �xf4 30 'ii'xas �xf3 31
1:c2 �g3 0-1
Making White pay
for delaying d4-d 5
Game 4
A. lpatov- R.Antoniewski
G e rm a n Lea gue 2011
1 d 4 d 6 2 ltJf3 g 6 3 c 4 �g7 4 ltJC3 ltJc6 5
ds ltJes 6 ltJxes .i.xes 7 e4 ltJf6 8 .td3
0-0 9 ltJe2 ltJd7 10 0-0 ltJcs 11 f4 �g7
12 fS?!
No sooner has White conquered the
es-square than he immediately relin­
quishes it. White (another grandmas­
ter) receives th e proper punishment -
eventually.
Illustra tive Games
12 . e6 13 �C2 exfs 14 exfs ltJd7?! 15
..
�f4? .i.xb2 16 .i.h6 'iVh4 17 �xf8 ltJxf8
18 ltJf4 .txa1 19 'ii'x al �d7
White has sacrificed a pawn to
trade off Black's powerful bishop, only
to find th at the exchange has not
helped him. This is a typical result.
20 'i'c3 :te8 21 cS �bS 22 g3 'ii'e 7 23
:tbl �a6 24 .ta4 'ife4 25 �C2 '6'e3+ 26
'i'xe3 :xe3 27 cxd6 cxd6 28 �a4 :a 3
29 i.. b 3 ltJd7 30 :el �f8
31 ltJe6+??
White's position was very unpleas­
ant anyway, but he gets no compensa­
tion for the knig ht.
31...fxe6 32 dxe6 ltJes 33 :Cl gxfS 34
137
Th e Dark Knight System

:c8+ We7 3S :C7+ wf6 36 :xh7 .i.C4 37 tbxe4 19 -.e3 .i.xd4 20 Vi'xd4 tbcs 2 1
e7 .i.bS 38 h4 dS 39 hS tbf3+ 40 Wg2 : e l .f8 YI-YI
tbgs 41 :h6+ Wxe7 42 :g6 tbf7 43 h6
:a6 0-1 White avoids liJxe5

White plays an ea rly e4-e5, Game 6

accomplishing nothing
Game 5
Y.Balashov-G.Kuzmin
USSR Cha m p i o nsh i p
Vi l n i u s 1980
1 tbf3 tbf6 2 c4 g6 3 tbc3 .i.g7 4 e4 d6 S
d4 0-0 6 .i.e2 tbc6 7 dS tbb8
We would reach this by position
with the move order 1 d4 tbc6 2 tbf3 d6
3 c4 g6 4 dS tbbS 5 tbC3 .i.g7 6 e4 tbf6 7
.i.e2 0-0.
8 h3 tba6 9 .i.e3 tbcS 10 eS tbfd7 11
exd6 exd6 12 .i.d4
R.Fischer-J.Sch uyler
Rich m o n d 2008
1 tbf3 tbc6 2 e4 d6 3 d4 tbf6 4 dS tbes S
tbc3 tbxf3+ 6 -.xf3 g6 7 .i.gs .i.g7 8
.i.bS+
White's bishop has no squares so he
trades it, but I appreciate the extra
breathing room.
8 ... .i.d7 9 .i.xd7+ .xd7 [9 ... tbxd7 !?] 10
Vi'e2 0-0 11 0-0 e6

12 l:tadl
On 12 .f3, I intend 12 ... tbeS fol ­
lowed by 13 ... c6.
12 ... exdS 13 exdSl!
12 tbf6
••• Instead 13 .i.xf6 ! .i.xf6 13 tbxds is
A valid choice, though Black could equal .
just as comfortably have allowed the 13 ... l:tae8 14 .f3 tbg4 lS h3 tbes 16
exchange with 12 ... aS or 12 ... l:r.eS. "g31!
13 0-0 as 14 :Iel %:te8 lS .i.fl .i.d7 16 16 Vi'e2 is the lesser evil.
:xe8+ -.xe8 17 -.d2 tbfe4 18 tbxe4 16 ... 'iVfS !

138

Black is better here, but there are
many mistakes in the coming compli­
cations - White's next move really stirs
things up.
17 �b51? h6 18 ii.e3 'iixc2 19 ii.d4?1 [19
.i.xa7! ?] 19... a6? [19 .....c4! 20 �xa7
'ifxdS] 20 :C1? 'ii e 4 21 :fe1 'ii'xd5 22
f?JxC7 'ii'xd4 23 �xe8 :xe8 24 �h1 :d8
25 :C7 'ii'b6 26 :e7? rj;f8 0-1
White's a lternate plan
with g2-g3
Game 7
L.Altounian-J .Schuyler
La s Vega s 2008
1 �f3 �c6 2 d4 d6 3 d5 �e5 4 �xe5
dxe5 5 c4 e6 6 �c3 �f6 7 g3 exd5 8
cxd5 ii.c5 9 ii.g2 0-0 10 0-0 ii.g4
White has managed to keep his
space plus without committing his e­
pawn, which leaves me in some doubt
as to the best squares for my pieces. My
last is intended to provoke White to
weaken his king position, a plan which
I continue throughout the game.
Illustrative Games
11 'ii'b3 ii.b6 12 h3 ii.f5 13 �h2 h6 14
Vi'C4?1
This only helps me find the right
plan. After this game, the knight trans­
fer became the standard way for me to
combat this whole variation with 3 dS.
The alternative try 14 a4 would at least
have given me a little problem to deal
with.
14 ... �e81 15 �a4 �d6 16 1Vb3 ii.e4
White is intending to grab the
bishop pair, so I already start working
to relieve him of it. It is awkward for
him to avoid the trade.
17 f3 ii.f5 18 ii.d2 ii.d7 19 �xb6 axb6
20 ii.b4 :e8 21 e4
139
Th e Dark Kn ight System

Levon mentioned to me after the

game that he played this to stop
21 ... lbfs, oddly overlooking the alter­
nate route.
21 ... lbb5 22 .i.C3 cS 23 dxc6 bxc6 24
:fd1

Whatever you do, don't let this hap­

By n ow I am actuany better if I just
play the simple moves 24 ... cs and
2 s ... lbd4.
24... 'i'C711 25 .i.f1 .i.e6 26 .i.C4 .i.xC4 27
-.xc4 lbxC3 28 "xc3 :ed8 29 a3 ::td6 30
:ac1 l:lad8 31 :xd6 'li'xd6 3 2 :C2 'iVd1
33 :f2 :d3 34 'ii'x c6 'iVe1 35 "c2 :d1
Yt-Yz
After 36 "c8+ �h7 37 'iVfs+ g6 3 8
'iVxf7+ the result i s a draw by perpetual
check.
pen to you!
7 ... .i.xb5 8 'ii' h 5! .i.b4+!?
Following the disaster on move 5,
this is the best chance in a terrible posi­
tion.
9 lbC311 "d41
Much better is 9 ... .i.xc3+ 10 bxc3
-.f6 when the continuation 11 exf7+
�8 12 fxg 8+ �xg8 is not so terrible for
Black.
10 'iVxf7+ �d8 11 .i.d2 .i.xC3 12 O-O-O!

Don't play 5 e6??..•

Carne 8
J.Bonin-J.Schuyler
New Yo rk 1988
Owl
1 d4 lbc6 2 lbf3 d6 3 d5 lbe5 4 lbxe5 12 ... .i.xb2+ 13 �b1 lbe7 14 .i.e3 .i.C3 15
dxe5 5 e4 e6?? [5 ... lbf6 !J 6 .i.b5+ .i.tb 7 'ii' x g7 :g8 16 'i'xe5 1-0
dxe611 The carnage is unspeakable.

140

Equal plus symmetrical eq uals
win - fight the good fight
Game 9
H.Keska r-J.Sch uyler
H a m pton 2011
1 d4 the6 2 thf3 d6 3 d5 the5 4 thxe5
dxe5 5 e4 thf6 6 .td3 e6 7 dxe611
Obviously White can forget about
an advantag e after this - the question
is, how do I beat my lower-rated oppo­
nent from this nearly symmetrical po­
sition ? The fact is, if I play at all dy­
namically and keep my eyes open,
there will be opportunities to create
imbalances. One way to start is to delay
castling.
7....txe6 8 0-0 .te5 9 .tg5 h6 10 .th4
'iVe7 11 a3 :d81! [11 ...0-0-0] 12 thd2 g5
13 .tg3 .tg4 14 'iVel thh5 15 .te2 thxg3
16 hxg3 .te6 17 thf3 f6 18 :dl h5
Black is nowhere near winning, but
White certainly has a lot to worry
about with ... h 5 -h4 and (after g3xh4)
... g5 -g4-g 3 coming.
19 :xd8+ 'ii'x d8?! [19 ...Wxd8 !] 20 'ii'e 3
Illustrative Games
.. e7 21 .te4 .te8 22 b4 .tb6 23 a4 a6 24
as .ta7 25 b51! axb5 26 .txb5+ e611
[26 ... \t>f8 !] 27 a61
I had overlooked this.
27 ...'iVe5 28 'iVxe5?! [28 axb7 ! .txb7 2 9
.tC4] 28....txe5 29 axb7 .txb7 30 .ta4
<it>d7 31 :dl+ �e7 32 .tb3 .te8
In spite of all the trades, this renews
Black's king side threats.
33 .tf7?! h4 34 �fl hxg3 35 fxg3 .tg4
36 .tb3 wb6 37 l:td2 .tb411
It is best for Black to cash out with
37 - .:th 1 + 38 'ite2 :g 1 39 �d3 .txf3 40
gxf3 :txg 3. Instead, I kept trying to
squeeze out a clear win with no risk,
but I h ave underestimated White's
ability to generate counterplay.
38 :dl .te5 39 :d2 i.. e 31 40 :d61
Somehow I overlooked this. Sud­
denly Black has no advantage at all.
40 .. .f5 1 41 exf5 e4 42 lbe6+! Wxe6 43
the5+ We5 44 thxg4 i..d4 45 f6
My opponent reasonably offered a
draw here. I turned over my scoresheet
and wrote the numbers 51 through 100
on the back.
45 ... i..xf6 46 thxf6 IU8 47 �e2 lhf6 48
141
Th e Dark Kn ight System
�e3 1i>d6 49 1i>xe4 �f2 50 �e3 :f1
If 50 ... :xg 2 51 1i>f3 :g l 52 1i>g4 and
53 �xg 5 draws.
51 �e2 :f8 52 1i>e3 1i>e5 53 .i.e4 :d8
Somehow I will need to win one of
the g-pawns and then generate
zugzwang and/or mating threats. I
wasn't sure how, but I kept on playing.
To make the draw easier on himself,
White needs to be alert for the chance
to push th e c-pawn, though he is un­
derstandably reluctant to loosen his
bishop.
54 .i.d3 :d4 55 �3 :a4 56 <ite3 :d4 57
�f3 1i>d5 58 �e3 �e5 59 1i>f3 1i>b4 60
1i>e3 1i>e3 61 1i>e2 :a4 62 1i>e3 �b2 63
�3 1i>e1 64 <ite2 :b4 65 .i.f5 l:ta4 66
.i.d3 :d4 67 1i>e3 :d5 68 1i>f3 :d4 69
<ite3 :b4 70 �e2 l:ta4 71 .i.f5 �a5 72
.i.d3 g41?
It was important to try everything
before playing this committal move,
which presents slightly different chal­
lenges to White.
73 1i>e3 :ta4 74 .i.e4 [74 C4 !] 74 ... 1i>d1
75 �4?1 <ite2 76 �f5 �2 77 �4 :e4
78 �e5 <ite3 79 .i.d3 l%a4 80 1i>f5 �2 81
142
il.e4 1i>xg3 82 1i>e5 <itf2 83 .i.d5 1i>e3 84
e4 :a5 85 1i>f5 g3 86 1i>g4 1i>f2 87 1i>f4
:e5 88 1i>e5 �e3 89 �5 lile8 90 1i>e5
Unfortunately, the time situation
demanded that I stop notating here. In
addition to zugzwang and mating
threats, there is now a third winning
possibility, which is somehow to drive
the white king to d8 (or thereabouts),
sacrifice the rook on d5, and win the
resulting king and pawn endgame. This
is, in fact, how I won, somewhere
around move 120-130. Naturally,
White could have held the position
with correct play, but this is hardly
relevant from a practical standpoint.
Winning the draw
Game 1 0
D. Haessel-J.Sch uyler
Pawt ucket 2 008
1 d4 ttJe6 2 ttJf3 d6 3 d5 ttJe5 4 e 4 ttJf6
I am planning 5 ... ttJxf3+, but first I
want White to figure out how to de­
fend his e-pawn .
 Illustrative Ga mes

5 lLlxe5 dxe5 6 i.. b 5+ i.. d7 7 "d3 a6 8 22 �C2 hxg4 23 hxg4 :th2+ 24 :d2 i.. c 5
i..x d7+ 'i'xd7 9 tt.Jc3 e6 10 �g5 i.b4 11 25 ttJC3 i..d4 26 :fdl i.e3
0-0-0 0-0-0 12 f3 'iVe7 13 "C4 h6 14
i..xf6 gxf6 15 g4

As the game develops, I am starting

to understand better why the bishop
Th e position has been equal since and rooks work well together - one
move 9, and now I am faced with the reason is that the bishop disrupts the
problem of generating some winning co-ordination of the enemy rooks.
chances against my somewhat lower­ 27 :te2 %:l8h3 28 ':xh2 ':xh2+ 29 �b3
rated opponent. Somewhere in my :f2 30 :d3 i.. d4 31 a4 �d7 32 as �d6
chess education I learned that Q+N 33 ttJa4 <it>e7 34 �a3 ':C2 35 �b3 :f2 36
work better together than Q+B, while �a3 f5!?
R+B co-ordinate better than R+N. It is In spite of my hard work, Black is
time to arrange the ideal exchange - still nowhere close to winning. This is
which, fortunately, is not too difficult. as good a try as any.
15 ... "cS! 16 "xC5 i.xc5 17 l:td3 h5 37 gxf5 exf5 38 ':b3 fxe4 39 fxe4 f5 40
My half-open files will not suffice :xb7 fxe4 41 :txc7+ <it>e8
for this position .
18 h3 l:th7 19 :fl :dh8 20 ttJdl i.. d 6
My idea is to block the d-file an d
cross my king over to the e7-square - a
small improvement to my position. My
opponent is sufficiently worried about
this to weaken his dark squares in or­
der to stop it.
21 C411 i..e 7
I can't explain this move, but the
tempo is not important.

143
The Dark Kn ight System

42 CS?? White has kindly released Black's fi­

Oddly, White can hold with the anchettoed bishop from its prison.
"grovelling" manoeuvre 42 l:th 7 ! e3 43 13 0-0-0 0-0 14 es e6 15 d6 cxd6?!
l:th l l:td2 (43 ... e2 44 :el) 44 ttJC3. For­ Too co-operative, allowing White to
tunately for me, my opponent could land his knight on the juicy d6-square.
not bring himself to consider this de­ Instead, lS ...•a4 16 <iitb l bS 17 ttJa3
fence seriously. cxd6 and Black is better.
42 ... e3 43 d6 e2 44 ttJb6 e1'if 45 d7+ 16 ttJxd6 'lWa4 17 <iitb1 f3 18 'ii'xf3 i..x e5
We7 0-1 19 l:the1 i..x d6 20 l:txd6 l::t a c8 21 b3
.-a5 22 :ed1 'lWc5 23 li1d2 b5 24 h4 as
Mestrovic tries 7 ••. g6!? and wins! 25 l:t6d3 'ife5 26 l:te2?! 'ifh2! 27 h5
"g1+ 28 �b2 a4 29 a3 b4?1 30 axb4
Game 1 1 :ta8?! 31 l:ted2 axb3 32 l:td1
D.Rasic-Z.Mestrovic
Croatia n Tea m
C h a m pi o ns h i p 2001

1 e4 ttJc6 2 d4 d6 3 ttJf3 ttJf6 4 d 5 ttJe5 5

ttJxe5 dxe5 6 i.. b 5+ i.. d7 7 'ii'e2 g6 8
i..x d7+ 'ifxd7 9 ttJd2?1
9 c4 and 10 ttJc3 is simpler and
stronger.
9... i.. g7 10 ttJC4

By now White has regained the ad­

10 ... ttJh5
10... 0-0 ! ? 11 ttJxes 'ifa4! and Black
recovers the pawn on e way or another.
11 g4?1 ttJf4 12 i..xf4 exf4
vantage, but it is still not simple to play
against heavy pieces with poor kin g
protection .
32 ..:.h2 3 3 l:txb3 'iWe5+ 34 '-c3 "g5 3 5
f 3 :tfe8 36 'ifd41 [ 3 6 b S ! ) 3 6 ... :ed8 3 7
l:t a 3 l:tab8 38 'iWe3 l:txb4+ 3 9 c;t>a 2 'iff6
40 l:txd8+ 'iVxd8 41 h6 'i'd5+ 42 :b3
l:ta4+ 43 c;t>b2 :la8 44 'iff4 e5 45 'iVb4
'ifd8 46 'ii'C 5 ifd1 47 'iVa7 :tc8 48 :c3
:f8 49 l:td3 'iVe1 50 'ife3 'ifa5 51 �a3
:b8+ 52 WC1 'ii' C 7 53 'iWC3 'iWe7 54 'ii'e3
'iWh4 55 l:tb3 l:tc8 56 c;t>b2 'iVf6 57 �b6
'ife7 58 'ifb3 e4 59 f41

144

After a long defence, White cracks.
S9 ...'iYc7 60 :Ib7 'i'xf4 61 'iYdS? e3 62
:ta7? 'iYf6+ 0-1
Mestrovic wins again with 7 g6!? •••
Game 12
J.Barle-Z.Mestrovic
S l ove n i a n C h a m pion s h i p,
Krsko 1997
1 ltJf3 ltJc6 2 d4 d6 3 e4 ltJf6 4 dS ltJes S
ltJxes dxes 6 i.. b s+ i.. d7 7 'ii'e2 g6 8 0-0
i.. g 7 9 i..x d7+ 'i'xd7 10 c4 0-0 11 ltJc3
ltJhs 12 g3 cS 13 dxc6 'ii'xc6 14 i.. e 3 e6
lS b4 fS
Illustrative Games
White is still better. Black's kingside
pawn majority is not as dangerous as it
looks. Mestrovic soon drifts into a bad
position without making any obvious
mistakes, which is why I'm not crazy
about this variation.
16 b5 "e8 17 i.. cs :f7 18 f3 b6 19 i.. d 6
l:ld7 20 :adl :c8 21 cS bxcS 22 '6'C4
ltJf6 23 i..x es :cd8 24 l:txd7 ltJxd7 2S
i..xg7 �xg7 26 f4?1
26 ltJe2 ! is stronger, intending 2 7
ltJf4.
26 ... 'iWe7 27 exfs exfs 28 ltJd S 'iYe6 29
'ii'c 3+ �h6 30 g4 fxg4 31 'ifd2 .-e4 32
fS+ gs 33 Jitel 'ii'xfs 34 ltJe7 'ii'f6 3S
'6'd s �hS 36 ltJfS ltJf8 3 7 'ii'x cs '6'b6 38
'ifxb6 axb6 39 ltJg7+ �h6 40 ltJfS+ �g6
41 ltJe7+ �f6 42 ltJc6 l:ld2 43 :fl+ �g7
44 ltJes ltJg6 4S :f7+? �g8 46 :fS11
:dl+ 0-1
Not 46 .. J:tdS ?? 47 ltJxg6 l:lxfS 48
ltJe7+ and White wins, but after
46 .. J:�dl+ 47 �2 (or 47 �g 2 ltJh4+)
47 .. J:�dS 48 ltJxg 6 :xfS is check.
Beating the London System
Game 13
P.B.Pedersen-D. Bekker Jensen
Da n i sh Tea m
C h a m pion s h i p 2 008
1 d4 ltJf6 2 ltJf3 g6 3 i..f4 i.. g 7 4 e3 0-0 S
i.e2 d6 6 h3 ltJfd7 7 0-0 ltJc6 8 c3 eS
We would reach this by 1 d4 ltJc6 2
ltJf3 d6 3 i..f4 ltJf6 4 e3 ifJ 5 i.. e 2 i..g 7 6
h3 0-0 7 c3 ltJd7 8 0-0 eS.
9 i.. h 2 fS 10 dxeS dxes 11 ltJa3 gs 12
145
Th e Dark Kn ight System

lLJc4 .f6 13 b4 f4 Beating the London System again

Game 14
R.Va lenti-V.Tkach iev
Corsica (ra p i d ) 1997

1 ltJf3 ltJf6 2 d4 g6 3 i..f4 i.. g7 4 e3 d6 5

h3 0-0 6 i..e2 lLJc6 7 0-0 lLJd7?! [7 ... e s !] 8
i.. h 2 eS 9 c3 'ilie7 10 a4 fs 11 lLJa3 �h8
12 b4 e4

Houdini completely hates Black's

position, but an aggressive stance on
the kingside often reaps benefits. It is
very difficult for White to solve the
problem of liberating his entombed h2-
bishop while at the same time keeping
his king safe.
14 lLJfd2 lLJb6 1S lLJa s 'ii'g6 16 e4 :d8
By now even Mr. H understands
that White is not better. Black decides he will be able to get
17 'ii'c2 lLJxa S 18 bxa 5 lLJd7 19 :ad1 .. .fS-f4 in later - this is slightly optimis­
:e8 20 'ii' a 4 �h8 21 f3 hS 22 ltJb3 lLJf6 tic, but it pays off.
23 a6 g4 24 hxg4? hxg4 25 axb7? g3!! 13 lLJd2 g5 14 'ii'C2 lLJf6 1S bS lLJd8 16 c4
lLJe6 [16 .. .f4 ! ?] 17 :fC1 f4 18 llab1 'ii'e8
19 lLJxe4?? fxe3 20 lLJxf6 exf2+ 21 <iit h1
i..xf6 22 -'e4 lLJxd4 0-1

Losing to the London System

Game 1 5
V.Golod-E.Sutovsky
N ata nya (ra p i d) 2 009

1 d4 lLJf6 2 ltJf3 g6 3 i..f4 i.. g7 4 e3 d6 S

26 bxc8'ii' :axc8 27 i..x g3 fxg3 28 'ii'xa7 i..e2 0-0 6 h3 lLJc6 7 0-0 lLJd7?1 8 c3 eS 9
'ii'h 6 29 :fe1 l:ta8 30 'ii'c s i..f8 0-1 i.h2 fS 10 a4 'ife7 11 b4 e4?!

146

Opening up th e game for White's
dark-squared bishop. This is clearly bad
since Black has little chance of achiev­
ing .. .f5 -f4 to close the diagonal again.
12 ttJfd2 ttJdS 13 as a6 14 c41 cS?
This stops White's planned 15 CS!,
but it weakens too many squares, most
especially d5 and d6.
15 tbc3 cxd4 16 ttJd5 .f7 17 exd4 ttJe6
is i.xd6 tbxd4 19 bfS ttJxfS 20 tb b31?
Giving back the exchange is not
necessary, but it does help to clarify the
position.
20...ttJxb3 21 -.xb3 i.xa1 22 :xa1 i.e6
23 :d1 ladS 24 'iVc3 i..xd5 25 :xd5
25 ... :txd5?
Illustrative Games
This gives White's queen easy ac­
cess, though Black's position was lousy
anyway.
26 cxd5 -g7 27 'iVC5 'ii'e 5 2S i..c 4 'ii'a 1+
29 �h2 'ii'e 5+ 30 g3 �7 31 'ii'b 6 �h6
32 'iVdS? 'lfi>g7??
My computer kindly tells me that
Black can save the game with
32 ... tbd7 ! ! 33 'iVxd7 'iVd4. The text move
is hopeless.
33 d6 "b2 34 'ife7+ �h6 35 'ii'xfS+ �h5
36 i..f1 e3 37 d7 e2 3S i.. g 2 1-0
How to fight in the critica I Ii nes
of the 1 d4 Dark Knight System
Game 1 6
H.Kmoch-F.Yates
H a stings 1927/28
1 d4 tbf6 2 tbf3 g6 3 c 4 i.. g7 4 g 3 0-0 5
i..g2 d6 6 0-0 ttJc6 7 d 5 ttJbS
Here 7 .. /has ! ? is recommended by
theory, and is objectively best, though
Yates makes a good case for the text
move - and note that we would nor­
mally not have this option, reaching
the game position by 1 d4 tbc6 2 tbf3
d6 3 g 3 g6 4 d5 tbbB 5 i..g 2 i..g 7 6 0-0
ttJf6 7 c4 0-0 or one of many similar
move orders.
S tbc3 a51 9 tbd4 tba6 10 h3
White prevents 10 ...tb g 4 but, as I
am so fond of saying, a tempo is a
tempo. This also necessitates a later
�h 2.
10...tbc5 11 i.e3 i.d7 12 'ii'd 2 'ifcs 13
�h2 e5
147
Th e Dark Kn ight System
14 lhdbS
White should have preferred 1 4
dxe6, though Black is fi n e after
14 ...lhxe6, or even 14 ... i.xe6 15 lhxe6
'iVxe6 16 .lxcs dxcs 17 .lxb7 l:tad8 18
'iVf4 'iVb6 19 .lf3 'ii'xb2, when Black's
strong bishop and activity make up for
the crippled queen side.
14 ... b6 15 l:tacl lhhS 16 b3 fS
Yates's play has a very modern feel
to it, but his next move is too optimis­
tic.
17 f4 gS?! 18 .lxcS?!
After 18 fxg s f4 19 .lf2 fxg 3+ 20
.lxg 3 lhxg 3 21 �xg 3 White a t least has
a pawn for his trouble.
18 ... gxf4?
Yates bluffs again with this fake
(and unnecessary) zwischenzug . How­
ever, the rest of the game he conducts
masterfully.
19 gxf4? bxcS 20 e3 �h8 21 .lf3 lhf6 22
:gl '6'd8 23 �g3 .lh6 24 :fl .le8 2S
lhe2 .lg6 26 :ggl 'ile7 27 lhbc3 l:tg8
28 lhg3 �af8 29 .ldl :g7 30 l:tg2 lUg8
31 :ff2 lhd7 32 lhce2 _h4 33 lhgl?!
exf4 34 exf4 lhf6 3S �hl l:te7 36 :fl
148
l:tge8 37 lhle2?
The losing mistake, though White
was about waist-deep in it anyway.
37 ... lhg4 38 l:tf3 'ilxh3+ 39 �gl .lg7 40
l:td3 l:te3 41 l:txe3 l1xe3 42 .lC2 l:txg3!
43 lhxg3 i.d4+ 44 �l lhe3+ 0-1
s ltJas l ? a strong alternative
••• -
to S ltJb8
•••
Game 1 7
A.Gal liamova-M.Krasenkow
Kosza l i n 1997
1 d4 d6 2 lhf3 g6 3 g3 .lg7 4 i.g2 lhc6 S
ds lha5
It is possible to consider this square
any time White h as fi an chettoed,
though it is more usual with a white
pawn on c4 to harass. As it turns out,
White does not enjoy th e omission of
C2 -C4, as the ds -pawn becomes a tar­
get. In any case, Black must be active
quickly on the queenside to justify the
knight's position . The typical method is
...c7-cs to gain space and make sure the
knight doesn't get trapped - in this

game Bl ack has other ideas.
6 0-0 c6 7 e4 ttJf6 8 'ii'e 2 cxdS 9 exdS
i.d7 10 ttJfd2 bS 11 b4 ttJC4 12 ttJxC4
bxc4 13 'iVxC4 :c8
14 'ife2?!
It was already dangerous to win the
pawn, and now White chooses the
wrong retreat: 14 'Wb3 ttJxdS ? ! 15 i.xdS
i.xal 16 i.xf7+ Wf8 17 c3 is good for
White.
14 ...ttJxdS 15 i.xds i.xal 16 c3 0-0 17
i.gs i.c6?! [17 ... i.fS !] 18 i.xc6 l:txc6 19
i.xe7 l:te8 20 'iVf3 'iVd7 21 i.f6 l:ta6 22
ttJd2 �xa2 23 'iVf4 'iVfs 24 'iVxd6
24... i.b2?!
Bl ack should make space for his
Illustrative Ga mes
king with 24 ...h6 or 24 ... hS. Against the
text White is right back in th e game
after 25 i.d4.
25 ttJb3? .e6 26 'iVxe6 �xe6 27 i.d4 �a3
28 ttJas?
28 ttJd2 was a better try, intending
28 ... i.xc3? 29 ttJbl.
28...i.xc3 29 ttJc4 l:tal! 30 i.xc3 :xfl+
31 Wxfl Itc6 32 i.d4 �xC4 33 i.xa7
:Xb4 34 h4 wg7 35 Wg2 Wf6 36 i.e3
Wes 37 Wf3 hS 38 Wg2 We4 39 i.cs :bS
40 i.e3 Wd3 41 i.f4 We2 42 i.e3 l:tb3 43
i.cs l:tf3 44 i.d4 fS 45 wgl l:td3 0-1
Black, given plenty of rope,
tries to hang himself but fails
Game 18
R.Aghasarya n-A.Chibukhchian
Kaj a ra n 2 0 1 1
1 d 4 ttJc6 2 c 4 eS 3 dxes ttJxes 4 e3
White's unambitious play has al­
ready left him with no trace of an ad­
vantage. Frankly, the rest of the game
sees Black try too hard to win, for
which he was unjustly rewarded.
149
Th e Dark Kn ight System
4.. .tt�f6 5 lLle3 i..b4 6 i.. d2 0-0 7 i.. e2 e6
8 lLlf3 d6 9 0-0 J:te8 10 'ii' b3 i.. a s 11
l:tadl i.. e 7 12 lLld4 'fIe7 13 h3 lLlg6 14
�fel lLlh4 15 g3 lLlg6 16 i..fl hS 17 f4?!
17 ... h4?! [17 ... i..b6 18 lLla4 lLle4!] 18 g4
i..b6 19 lLla4 i..e 7?! 20 e4 lLlxe4?! 21
i.. g 2?!
2 1 lLlc3 ! was stronger, with th e idea
21 .. :iff6 22 lLlxe4! 'i¥xd4 23 i.. e 3! iVxe4
24 i..f2.
21 ... dS? [21 ...iVf6!] 22 exds 'fIf6 23 gs
�xd4+ 24 i.. e3 'ifxdl 25 'fIxdl exds 26
fS? [26 iVxdS] 26 ... .txfS 27 �xds l'ild6
28 l:tel? l:txe3 29 J:txe7 :d3 30 'ifes?!
lLlf4! 31 i..fl? J:tdl 32 lLle3 lLlxh3+ 33
�g2 :d2+ 34 �hl lLle4 0-1
150
Crushed on the d-file -
don't let this ha ppen to you!
Game 19
A. Hoffman-A.Ferna ndez
Ma r del Plata 1996
1 d4 lLle6 2 e 4 eS 3 dS i..b4+ 4 i.. d 2
i..x d2+ 5 �xd2 lLlee7 6 d6 lLle6?
Circumstances may change, but
right now the best square for this
knight is e7, from which it controls dS
and h as the option to attack d6 with
... lLlfS . Black shoul d only consider ... lLlc6
after White has played e2-e4, clamping
on dS but opening a hole on the d4-
square. Even after the correct 6 ... cxd6,
Bl ack can still be crushed in this fash­
ion if he is not alert to all th e methods
of counterplay, so study Position Two.
In this game, Bl ack never got on the
board.
7 lLle3 exd6 8 lLlbS lLlf6 9 lLlxd6+ wf8 10
lLlf3 h6 11 e3 e4 12 lLld4 'fIas 13 �xas
lLlxas 14 l'il4bS b6 15 i.. e 2 �e7 16 0-0-0
lLlb7 17 l:ld4 lLles 18 lLle3 lLle6 19 l:td2
as 20 l:hdl lLles 21 lLldS+ lLlxdS 22

:XdS g6 23 lbxe4 lbxe4 24 lIes+ �d8
2S :Xe4 l:te8 26 :ed4
26 ...:e6 27 i.. f 3 l:ta7 28 h4 hS 29 :f4
tite7 30 -'.dS :tf6 31 l1xf6 <it>xf6 32 :d4
d6 33 �d2 �e7 34 �c3 :c7 3S a4 i.. e6
36 lId2 :cS 37 b3 c;t>d7 38 f3 lIC7 39 e4
:c8 40 eS -'.xd S 41 :xd S l:te8 42 :xd6+
r3;C7 43 �d4 :tg8 44 f4 :te8 4S <it>d s l:te7
46 :c6+ �b7 47 e6 1-0
Midd lega me and endga me
ideas in the 3 .i.b4+ structure
•••
Game 20
H. Titz-C.Ba rlocco
Dresde n 2 004
1 c4 lbc6 2 d4 eS 3 ds i.. b4+ 4 i.. d2
i..xd 2+ S .xd2 lbce7 6 lbc3 d6 7 e4 fS 8
exfs i..xfs 9 i.. d 3 lbf6 10 lbge2 0-0 11
0-0 i..x d3 12 'ii'xd3 lbhS 13 g3 Vi'd7 14
f3 a6 lS lIadl l:tae8 16 lbe4 h6 17 cS
lbf6 18 lb2C3 lbfs 19 b4 lbxe4 20 lbxe4
gs 21 <it>g2 lbd4
This is th e position where we left off
wh en analysing in the th eoretical sec-
tion.
Illustra tive Games
22 cxd6 cxd6 23 :Cl :f7 24 lbd2 <it>g7
2S lbb3 lbxb3 26 axb3 .fS 2 7 'ii'xfS
lIxfs 28 lIC7+ :f7 29 :xf7+ <it>xf7 30
1:cl lte7 31 Wf2 e4 32 f4 gxf4 33 gxf4
e3+ 34 <it>e2 �f6 3 S :c3 lIe4 36 :xe3
<itfS 3 7 �d3 :xb4 38 �c3 l:txf4 39 :e6
:h4 40 l:txd 6 c;t>es 41 lId7 bS 42 d6 b4+
43 Wd3 Wd s 44 l:ta7 l:th3+ 4S �c2
l:txh2+ 46 'iitd 3 llh3+ 47 <it>d2 �xd6 48
l:txa6+ �dS
49 :b6? �cs
49 ... l:txb3 (or on the next move) 50
lhh6 :'c3 reaches the winning end-
game rath er sooner.
SO :f6 l:th4? Sl <ite3 �dS S2 l:US+ �e6
S3 l:tbs <ili>d7 S4 l:tb6 hS SS <it>d3 :g4 S6
151
Th e Dark Kn ight System
:h6 h4 S 7 �e3 �e7 S8 �f3 :g3+ S9
�f4 :xb3 60 :xh4 �d6 61 l:thS �c6 62
�e4 :c3 63 �d4 <it>b6 64 :g s �a6 6S
:hS :Cl 66 l:lh8 �bS 67 �d3 �a4 68
�d2 :c7 69 :h3 b3 70 :h6 �a3 71 l:tb6
�b2 72 �d3 :d7+ 73 �c4 �c2 74 �b4
b2 7S :c6+ �d2 0-1
A wa rning: beware the static
king - a nice attack by White
Game 2 1
M.Gu revich- D.Zoler
Antwe rp 1998
1 d4 tLJc6 2 c4 eS 3 dS i.. b4+ 4 tLJd2
tLJce7 S a3 i..xd2+ 6 i..x d2 d6 7 e4 fS 8
'ii' h s+!?
8 ...�f8?!
8 g6 9 'ifh4 fxe4 is not the most
...
fun ever, but it does leave Black with a
pawn for his troubl e. The text move is
less accurate an d demands more preci­
sion from Black in the coming moves in
order to avoid a disaster like the one in
the game.
9 f3 [9 exfs !] 9...tLJf6 10 'ii'h 4 tLJg6 11
1 52
-.f2 �f7?!
11 .. .fxe4 12 fxe4 �f7 will allow
Black to finish castling artificially with
approximate equality.
12 exfS i..xfs 13 g4 i..d7 14 h4 :e8 lS
hS tLJf4 16 0-0-0 i.. a4 17 liel bS 18 tLJh3
tLJxh3 19 lixh3 -.d7 20 lth4 bXC4 21 gs
tLJxds 22 �xc4 -.c6 23 �bl i.. b s 24
i.. a 2 �f8 2S f 4 e4?
Black's position has been bad for a
long time, but this ends th e game.
26 'ii'd4 i.. d 3+ 27 �al e3 28 h6 l:tes 29
hxg7+ �e7 30 fxes exd2 31 exd6+
�xd6 32 :h6+ i..g6 33 'ii'x d2 1-0
No problems for Black
after 3 .i.b4+ 4 ..td2
•••
Game 22
E.Arlandi-M.Lanzani
Sa n Ma rino 1998
1 d4 tLJc6 2 c 4 eS 3 d S i..b4+ 4 tLJd2
tLJce7 S a3 i..x d2+ 6 i..x d2 d6 7 e4 fs 8
exfS �xfs 9 tLJe2 tLJf6 10 tLJg3 i..g6 11
i.. e2 0-0 12 0-0

12 ... llJe4 [12 ... ttJfS] 13 ttJxe4 i.Xe4 14 f3
i..fS 1S i..e3 b6 16 '6'd2 '6'd7 17 1:Iac1
�h8 18 J:tfd1 as 19 b3 h6 20 'Wb2 i.h7
21 �h1 ttJg6 22 i..f1 ttJf4 23 g3 ttJh3 24
i.g2 ttJgs 2S 1:1 f1 e4 26 i.xgs exf3 Vz-Vz
Fight for the e4-square
or suffer the conseq uences!
Game 23
G G rigore P Brochet
.
-
.
Creon 1999
1 d4 ttJc6 2 c 4 eS 3 dS i..b4+ 4 ttJd2
ttJce7 S a3 i..x d2+ 6 i..x d2 d6 7 e4 fS 8
exfs i..xfs 9 llJe2 ttJf6 10 ttJg3 O-O!?
Illustrative Ga mes
Black offers his other bishop, an in­
vitation White is wise not to accept: 1 1
ttJxfS ? ! ttJxfs 1 2 i. d 3 e 4 13 i. c 2 "e8 1 4
0-0 .g 6 and ls . . .:tae8 with a harmoni­
ous and menadng position. Who's
afraid of the bishop pair?
11 i.. e2 "e8 [11 . ..'i¥d7] 12 0-0 ttJe4
[12 ... i.g6] 13 i.e3 ttJxg3 [13 ...•g6] 14
hxg3 b6
Although this was also pl ayed in
Game 2 2, I don't think it helps Bl ack If
he wants to, White can break through
with b2-b4 and c4-cS anyway, and now
Black h as to worry about penetration
on both the c7- and c6-squares. Fur­
thermore, it rules out the possibility of
Black breaking with ... C7-c6. 14 ...i.. d 7 is
more flexible.
1S g4 i.. d 7 16 i.. d 3
Taking advantage of tiny inaccura­
des, White has built an advantage, tak­
ing over the b1-h7 diagonal, along with
the critical e4-square. It's not so easy
playing a Grandmaster. Remember to
fight for e4! As it h appens, Bl ack still
had plenty of chances.
16 ...ttJg6 17 i.e4 ttJf4 18 g3 'ii' e 7 19 f311
153
Th e Dark Kn ight System

[19 .C2! ] 19... �h3+ 20 �g2 �g5 21 39 Af7 _eS??

.ltxg5 "xg5 22 'ifd3 g6 23 11h1 as 24 39 ... i.e8! holds.
11h4 �g7?! [24.. J�tf7] 25 :ah1 :hS 26 40 "'f6+?
'l'c3 h6 27 b31 [27 f4!] 27 ......f6 2S f4 40 :xd7+! wins immediately.
g51! 40 ... cit>cS 41 .lth7 'ii'd S 42 'ifg7 11eS 43
28 ... exf4 29 '6'xf6+ �xf6 30 l::lxh 6 g4 .ltxg4 44 'ifxg4+ �bS 45 11g7 lUS 46
11xh6 3 1 11xh 6 %lg8 holds the balance. 'ii'd 7 'it'xd7 47 11xd7 11f4 4S :dS+ �a7
29 fxg5 _xg5 30 11h5! 'l'xg4 31 'ii'e3 49 a4 e4 50 :gS :h4 51 .ltf5 :f4 52
�7 32 115h4 'ii'g 7 33 :xh6 :xh6 34 .ltd7 e3 53 l:teS 1-0
11xh6 :hS 35 11g6 '6'fS 36 'l'g5 �eS 37
:f6 '6'e7 3 d6!? an interesting
... -

alternative to 3 e6 .•.

Game 24
M.Drasko-Z.Mestrovic
Bosn i a n Tea m
C h a m pio n s h i p 2003

1 d4 �c6 2 d5 �e5 3 e4 d6
Previously Mestrovic h ad success­
fully played 3 ... e6, but maybe he no
longer believed in it? The text move
3S 'ii'g 6+?! winds up leading to a position consid­
38 .It fS maintains more pressure. ered in Chapter Six (see 6 �f3 in line B).
Presumably the mistakes here are due 4 f4 �d7 5 �c3 c6 6 �f3 cxd5 7 exd5
to a time scramble. �gf6 S .lte3 g6 9 .lte2 .ltg7 10 0-0 0-0
3S ... �dS 11 '6'd2 a6 12 a4 "c7 13 .ltd4

1 54

13 ... ttJb6?!
The position was equal, until this
time-waster hands White an advan­
tage. 13 ... ttJcS is better.
14 as ttJbd7 15 ttJa4 ttJe4 16 �e3 .i.xd4
17 ttJxd4 ttJef6 18 ttJc3 �c5 19 �fd1 b5
20 axb6?! [20 b4] 20 ... ttJxb6 21 b4 YI-YI
Perhaps nobody was in the mood
fOT a fight - th e position is equal again
anyway.
GM uses Diebl's novelty 4 exd S !
•.•
Game 2 5
V. Erdos-R. Rapport
H u nga r i a n Tea m
C h a m pi on s h i p 2012
1 d4 ttJc6 2 d5 ttJe5 3 e4 e6 4 f4 exd5 5
fxe5 Vi'h4+ 6 We2 .h5+ 7 Wd2 �h6+ 8
WC3 �c6+ 9 Wd2 Vi'h6+ 10 �d3 'iWa6+
11 �d2 Vi'h6+ 12 �e1 Vi'h4+ 13 �d2
I can't help but get the feeling these
guys are messing around. Black's next
constitutes a risky attempt to play for
th e win.
13 ..."'f4+ 14 �C3 'ifxe5+ 15 'iVd4
Illustrative Games
15 ... .i.b4+ [ls ...'iVxe4!?] 16 �d3 dxe4+
17 �xe4 �xe4+ 18 Wxe4 ttJf6+ 19 We3
.i.C5+ 20 �f3 b5 21 .i.d3?!
According to Houdini, White is bet­
ter after 21 .i.e3 ! .i.b7+ 22 �e2 .i.xe3
23 Wxe3, though it hardly looks like a
safe advantage. Then again, the text
move doesn't look safe either, and
White clearly has no advantage there.
23 ... .i.b7+ 22 �g3 .i.d6+ 23 �f2 ttJg4+
24 �1 ttJxh2+ 25 �f2 ttJg4+ 26 �f1
ttJh2+ 27 �f2 ttJg4+ YI-YI
An a lternative to 4 fxe6!? -
•••
the "endgame" with 4 dxe6 •••
Game 2 6
S.Gordon-N.Short
B riti sh C h a m pi on s h i p,
S h effi e l d 2011
1 d4 ttJc6
I must say, this warms my heart.
2 d5 ttJe5 3 e4 e6 4 dxe6 dxe6 5 Vi'xd8+
Wxd8 6 f4 ttJc6 7 c3 .i.c5 8 ttJf3 as 9 a4
ttJh6
In this position, 9 ... ttJf6 is also good.
155
Th e Dark Kn ight System
However, it is nice for this knight to
h ave a stable home, controlling impor­
tant squares. Short's lo ...f6 ensures
th at he will not be squeezed on the
king side, as does 14 ... hS.
10 i.. d 3 f6 11 "'e2 C;;e7 12 ltla3 i.. d 7 13
ltlC4 ltlf7 14 h4 h5! 15 i.e3 i.xe3 16
ltlxe3 b6 17 :hg1 g5 1S g3 g4 19 ltld4
ltld6 20 l:tac1 1::thcS 21 b3 ltleS 22 :gf1
ltldS 23 es f5 24 :fd1 1t1b7 25 l:td2 ltlc5
26 :cd1 ltlg7 27 ltlC4 ltlxd3 2S l:txd3
l:tdS 29 ltla3 ltleS
30 "'e3 l:tacS 31 lI3d2 l:tbS 32 l:td3 :tbcS
33 l:1d2 l:bS 34 <iW2 :bcS 35 �e2 :as
36 "'e3 :tabS 37 �e2 lIaS 3S <iW2 :acS
39 :d1 :bS 40 :3d2 l:tbcS 41 �e1 l:taS
156
42 l::td 3 l:IabS 43 l:t1d2 :as 44 c;i;>d1 ltlg7
45 �C2 ltleS 46 �b2 ltlg7 47 ltlC4 i..eS 4S
�a3 l:ldcS 49 ltlb5 i..c 6 50 :d4 ltleS 51
ltlbd6 :dS 52 ltlxeS l:txd4 53 :xd4 Yz-Yz
Wow, that was bOring ! Are you sold
on 4 .fxe6 yet?
..
A better advertisement
for 4 dxe6•••
Came 2 7
M.G urevich-M.Rohde
Ph i l a delph i a (bl itz) 1989
1 d4 1t1c6 2 d5 ltle5 3 e4 e6 4 dxe6 dxe6
5 'iYxdS+ �xdS 6 i.f4 ltlg6
Although ... ltlc6 is normally the best
retreat for this piece, the difference is
not so great th at Black would decline
th e tempo gain. Here 6 ... i.. d6! is also
good, threatening 7 .. 1t1d3 + or 7 ... 1t1f3+.
.
7 i..e 3 ltlf6 S ltlC3?!
This careless move allows Black to
h arass White's poor dark-squared
bishop further, disrupting White's de­
velopment. 8 f3 was called for.
S ... ltlg4 9 i.. d 2 i.C5 10 ltlh3 i.d7 11
 Illustrative Games

l.e2 ltJf6 12 0-0-0 �e7 13 f4?! dxe6 fxe6 6 e4 i.c511 [6 ... ds ! ] 7 i.d3?!
Again 13 f3. ltJh6 8 'it'e2 0-0 9 g3 a6?! [9 ...dS!] 10
13 i.c6
.•• ltJC3 b5 11 e5 i.b7 12 ltJe4 i.b6 13 i.d2
ltJf5 14 0-0-0 h61! 15 :hf1 CS?!

By now Black is a little better -

White's position is too loose. Black gets tired of sitting aroun d
14 i.d3 ltJh4 15 e5? waiting for the brutality, but this only
This is a terrible idea, releasing the makes things worse due to the new
strong c6-bishop. White needed to let hole on d6.
go of the g-pawn and seek counterplay 16 ltJd6 i.d5
against Black's king with 15 fs ! .
1 5...ltJd5 1 6 ltJe4 i.e3 1 7 c4 i.xd2+ 1 8
lixd2 ltJb4 19 i.b1 ltJxa2+ 2 0 �d1 lthd8
21 g3 ltJf3 22 lixd8 :xd8+ 23 �e2 ltJd4+
24 �2 ltJb4 25 :c1?! ltJb3 26 :c3 ltJd2
27 ltJxd2 ltxd2+ 28 �e3 :xb2 29 i.xh7
:xh2 30 ltJgs :g2 31 ltJe4 a6 32 'it;>d4 b6
33 ltJg5 ltJC2+ 34 �d3 ltJe1+ 35 �d4 as
36 i.g8 a4 37 i.xf7 i.d7 38 cS 0-1

Don't play 6 i.cS ?!?!

•••

Came 28
A.Onisch uk-I.Sh kuro
U kra i n i a n Tea m
17 C41
White should play 17 i.xfs !, main­
taining the wonderful d6-knight.
17 bxc4 18 ltJxC4 ltJfe7?! [18 ...:b8] 19
•••

C h a m p i on sh i p 2 009 h4 ltJf5 20 :g1 ltJge7 21 g4 ltJd4 22

liJxd4 cxd4 23 ligf1 lib8 24 <it>b1 i.C5 25
1 d4 ltJc6 2 d5 ltJe5 3 f4 ltJg6 4 ltJf3 e6 5 f5 ltJc6 26 g5 exf5 27 gxh6 .xh4 2 8

157
Th e Dark Kn ight System
hxg7 lif7 29 ttJd6 ..txd6 30 exd6 ..te4 31
':f4 ..txd3+ 32 'ii'x d3 'ii'f6? 33 l::t h l lixg7
34 ':xfS 'ii'e6 35 'ii'h 3 1-0
In spite of th e inaccuracies, there is
feeling of inevitability about the result
of this game due to White's uncon­
tested superiority in th e centre and on
th e king side.
4 iLe3 fS ! ? - ta ke one
Game 29
A. Beliavsky-A.Miles
E u ropea n C h a m pi o n s h i p
S a i nt Vi nce nt 2000
1 d4 liJe6 2 e4 eS 3 ds ttJee7 4 ..te3 fs
5 f3
White resigns himself to allowing
Bl ack a good Kin g's Indian with an
early and easy - .f7-fS. Instead, if 5 ttJC3
liJf6 6 liJf3 d6 7 exf5 c6 (7 ... a6 ! ?),
White's centre disintegrates. (In fact,
White can still play for an advantage,
but it is easy to see why this was not
appealing.)
S ...liJf6 6 liJh3 d6 7 ttJf2
158
The great Beliavsky has not suc­
ceeded in casting any doubt on Black's
idea. Now Miles should just get on with
his comfortable King's Indian: 7. g6. ..
The c-pawn can wait.
7 ... e6 8 e4 cS 9 g3 g6 10 ttJe3 ..tg7 11 g4
f4 12 ..td2 gs 13 b4 b6 14 bxeS bxeS 15
..td3?!
Doubtful . 15 h4 looks more to th e
point.
lS ... hS! 16 h3 rtt f 7 17 rtt e2 ttJg6 18 1i'a4
..tf8 19 :tabl ..te7 20 ..te2
20... hxg4?!
I suppose it's a draw either way, but
there is no disadvantage to maintain­
ing the favourable tension. After this,
I'm not sure who's playing for the win,
or why.
21 hxg4 l:txhl 22 l:txhl ..td7 23 'iYa3
ilb8 24 liJbl 'ii' b6 25 ..te3 J:th8 26 l:!gl
..te8 27 ttJd2 'iYa6 28 'iYxa6 ..txa6 29
l:tbl ..td8 30 ..ta4 ..te7 31 ..tbS ..te8 32
ttJfl rtt e7 33 ..te6 ttJh4 34 ttJd2 ttJg2 35
..tb7 ..td7 36 ..ta6 ..tb6 37 ttJfl ttJh4 3 8
..tel liJg2 39 ..td2 ttJh4 4 0 ..tbS ..te8 41
..tel ttJg2 42 ..td2 ttJh4 43 l:tb3 <it>d8 44
..te6 Yz-Yz

4 ..te3 f5 !? - take two
Game 30
D.Cam pora-A.Miles
Sevi l le 1993
1 e4 ttJc6 2 d4 e5 3 d 5 ttJce7 4 i.e3 f5 5
f3 ttJf6 6 ttJC3 d6 7 �d2 g6 8 0-0-0 i.g7
9 h3?! f4 10 i.f2
The pawn structure on the king side
makes it impossible for White to open
lines there without creating a complex
of weak squares and pawns. Bl ack does
have ways to create play on the queen­
side, as we will see.
10...0-0 11 <iit b1 a6 12 ttJge2 i.d7 13
ttJC1 b5 14 a3 'ii'b8 15 ttJb3 J:td8 16 ttJa5
cS 17 dxc6 ttJxc6 18 ttJxc6 i.xc6 19 i.h4
:1d7 20 i.xf6 i.xf6 21 ttJd5 i.. h 4 22 ttJb4
i.b7 23 c4
White tries to free his terrible fl­
bishop, but in th e process opens lines
against his own king.
23 ...�g7 24 i.d3?! as 25 ttJa2?! [25
ttJd5] 25 b4! 26 axb4 axb4 27 'ii'xb4?!
••
i.f2?!
Bl ack's "bad" bishop turns into a
Illustra tive Games
monster. Nevertheless, Black should
insert 27 .. .illa7 2 8 ttJc3 i.c6 29 'i'b3 and
then 29 ...i.f2 perhaps, though Black
has other good ways to continue the
attack here.
28 ttJc3 i..d4 29 �C2 �c8 30 ttJb5 i.c5
31 'ii' b3 i.a6 32 :a1 i.e3 33 l:thd1
White has actually done an excel­
lent job repairing his position and
Black shoul d no longer be able to break
down the fortress.
33 ...:b7 34 ttJxd6??
Oops!
34...:xb3 0-1
White realizes too l ate th at 35 ttJxc8
will be met by 35 ...l:tab8! intending 36
l::txa6 :xb2+ 3 7 �c3 :8b3 mate. Of
course White can defend the mate, but
then he will remain a piece down.
Fighting the eternal CDg5
Game 31
M.Kravtsiv-K.Ta rlev
Evpatoria 2 007
1 e4 ttJc6 2 d4 e5 3 d 5 ttJce7 4 ttJf3 ttJf6?
159
Th e Dark Kn ight System
Perhaps this is a database error? In
any case, th e reader shoul d know that
this is not an acceptable move order: 5
ttJxe5 ttJxe4 6 'ife2 ttJd6 (not 6 ... ttJf6? 7
d6! cxd6 8 ttJC4 'iIIC7 9 �f4 and Black is
busted) 7 ttJC3 already leaves Black with
no good way to complete his develop­
ment.
S ttJc3? ttJg6 6 h4 hS 7 ttJgS!? i..b4 S
i..e2 d6 9 'ii'd 3 i.. d7 10 a3
10....i.xc3+
As I mentioned in the theoretical
section, it is better to retain this strong
bishop with 10 ... i..C 5. This is particu­
l arly true since Black will not inflict any
structural dam age with th e exchange.
White's queenside stays mobile and his
bishops should come into their own
sooner or l ater. Frankly, from this point
on, I do not care for Black's position
until the game is nearly over.
11 'ii'x c3 c6 12 'ii'b 3?!
It makes more sense for White to
open the g ame with 12 dxc6.
12 ...cxdS 13 exdS O-O?! [13 ..... c7] 14
t'Lle6! fxe6 1S dxe6 i..xe6 16 'ii'xe6+
�hS 17 i..gs -.b6 1S 0-0-0 'ii'xf2 19 i..f3
160
e4 20 iJef6 :aeS
21 Vi'd7?! [ 2 1 �e7!] 21 gxf6 22 i..xh s
•••
:e7 23 'ifxd6 'ii'f4+ 24 'ii'd2 'ifxd2+ 2S
:Xd2 ttJf4 26 i.. d 1 e3 27 l:td4 t'Llxg2 2S
.i.e2 fS 29 �d1 :eeS 30 .i.f3 :dS 31
:xdS ':xdS+ 32 Iiit C 1 :d2 33 1:g1 ttJxh4
34 i..xb7 l:td7? [34 . .J::th 2] 3S :h1?! [35
�a6] 3S ...l:txb7 36 l:lxh4+ �g7 3 7 �d1
Iiitf6 3S liite 2 �gs 39 :d4 l:te7 40 C4?? f4
41 :d1 lih7 42 Iiitf3 l:th3+ 43 liite4 e2 44
l%e1 l:te3+ 4S �d4 l:teS 0-1
Game 32
H.Meissner-A.Miles
E u ro pe a n C u p, Slough 1997
1 e4 t'Llc6 2 d4 eS 3 ds t'Llce7 4 ttJf3 t'Llg6
S h4 hS 6 g3 .i.cs 7 .i.gs f6 S i..d2 d6 9
ttJc3 i.. d7
Or 9 ... a6 - Bl ack must play to pre­
serve the dark-squared bishop.
10 .i.e2 ttJ6e7
Miles prepares to defend th e h­
pawn, which was about to drop off.
11 ttJh2 g6 12 'ifC1 c6 13 i.. e3 i..xe3 14
'ifxe3 'ii'b6 1S 'ii'xb6 axb6 16 dxc6 bxc6

17 0-0 b5 18 a3 lDc8 19 lUdl lDh6 20
lDfl We7 21 lDe3
21 a4.
21 lDb6 22 l%acl lDg4 23 i.. x g4 i..xg4
•..
24 �d3
Very slowly White is being out­
played. Miles makes th e most out of a
small thing - his single, uncompro­
mised pawn group. White needs to or­
ganize a pawn break, but he is showing
no inclination to do so.
24 lDc4 25 lDxc4?!
•.•
Now Bl ack has b2 as a target. The
immediate 25 :b1 was better.
25 bxC4 26 %:te3 :hb8 27 :bl :tb7 28
•..
f3 i..e6 29 !Ddl d5
Illustrative Games
Yikes! Black sure h as a lot of centre
pawns!
30 %:tal wd6 3 1 %:te2 d4 32 �f2 cS 33 c3
i.. d7 34 %:td2 We6 35 We2 :tab8 36 �el
i.. a 4
37 %:ta2?
White grows tired of th e thankless
defensive task
37 i..b3 38 �al i..x dl 39 Wxdl :xb2
.•.
40 :xb2 :Xb2 41 a4 :g2 42 %:ta3 wd6
43 as Wc7 44 a6 wb8 45 a7+ Was 46
:a6 :xg3 47 :xf6 Wxa7 0-1
Game 33
L.Ch ristiansen-J.Benjamin
US C h a m pion s h i p,
Seattle 2000
1 e4 !Dc6 2 d4 e5 3 d5 lDce7 4 lDf3 lDg6
5 h4 h5 6 i..g 5 lDf6 7 lDc3 i.. b4 8 a3
This seems odd. Doesn't White have
anything better to do? Indeed 8 lDd2 is
better, but one way or another White
does need to force our dark bishop off
the board or we will be h appy to pre­
serve it with 8 . . a6.
.
161
Th e Dark Kn ight System
8... i..x c3+ 9 bxc3 c6!
I am always reluctant to change th e
pawn structure on th e queenside after
saddling White with th e doubled
pawns - after all, the structure is al­
ready favourable, right? Th ere is some
logic there, but better players th an me
have demonstrated time and time
again that Black should be willing to
play on both sides of the board in this
type of position, particularly since
White has already made inroads on the
king side. Therefore 9 ... c6 !, the only
convenient pawn lever for either col­
our. Bl ack targets the centre and pre­
pares to break the annoying pin on f6.
10 c4 d6 11 ttJd2 "as 12 i.. d3 ttJg4!
Benjamin reminds his opponent
that 5 h4 cuts both ways - the g4-
square is lovely this time of year. If 13
f3 ? f6 14 fxg4 iLxg4 15 iLe2 i.. xe2 16
'ifxe2 fxg 5 17 hxg 5 ttJf4 18 'ii'f3 cxd5 19
exd5 0-0 with an obvious advantage
for Bl ack.
13 -.e2 f6 14 i.. e3 ttJf4
Houd;n; doesn't like this, but Ben­
jamin acquires the gorgeous e5-square
1 62
for his other knight.
15 i..xf4 exf4 16 0-0 cS 17 ttJf3 ttJes 18
l:tfbl 'i'c3 19 ttJxes -.xes 20 a4 gs
White is in trouble. Even the coming
endgame gives littl e relief.
21 hxgs fxgs 22 :a3 g4 23 'Wd2 l::t h 7 24
-'c3 -'xc3 25 %1xc3 :e7 26 g3 f3 27 as
rj;f7 28 Wfl l:Ib8 29 �el?!
(rossing to the queenside with th e
king loses the e-pawn, though staying
put is no bargain either. Black will
bring his king to e5 and prepare to
break with ... h5-h4.
29 ... i..fS 30 rj;d2 i..xe4 31 i..xe4 l:xe4
32 :e3 l:xe3 33 Wxe3 wf6 34 Wf4 wg6
35 a6 b6 36 c3 :f8+ 0-1
Came 34
W. Weisser-L. Trumpp
Ge r m a n League 2 004
1 d4 ttJc6 2 dS ttJes 3 e4 ttJg6
I don't believe in this move order
because of 4 h4!.
4 ttJf3 eS 5 c 4 i.. c S
5 ... ttJf6 is more accurate here, and if
 Illustra tive Games

6 ltJc3 then 6 ... i..b4. In addition to the

structural damage we can inflict, it is
not convenient for White to defend the
e-pawn. On the other hand, White could
force the game variation using the
move order 4 c4 eS S ltJC3 1.cs 6 ltJf3.
6 i.e2 ltJf6 7 ltJC3 0-0 8 0-0 as! 9 ltJe1
d6 10 ltJd3
This is an odd way to pursue the
bishop - and quite unsuccessful, as
White h as released th e d4-square.
10 ... .i.d4 11 .i.f3 ltJh4?! 18 ltJfS 'ii'f6 19 ltJe3 'ii'g6 20 b3 fS 21
i.a3 ltJe4 22 ltJd1 i.C2 23 i.C1 f4 24
'iVe6+ 'ifxe6 2S dxe6 ltJcS?! [2s .. Jife8]
26 e7 llfe8 27 ltJc3 ltJd3?
Bl ack h as let slip most of his advan­
tage.
28 l::te2 ltJxc1 29 I1xc1 .i.fs

Setting a trap, which White fall s

right into, but it is best prepared by
11 ...h6 to prevent 12 i.gS .
12 ltJbS? ltJxf3+! 1 3 'ii'xf3?! 1.g4 14 'ii'g3
i.e2!
Suddenly Black's light-squared
bishop is all in White's business. 30 CS?! [30 ltJdS] 30 11xe7 31 ltJdS 11f7
.•.

1S lle1?! 32 cxd6 cxd6 33 lld2 ltd8 34 f3 i.e6 3S

15 ltJxd4 i..xfl 16 ltJfs ltJh s 17 'Wf3
i.xd3 18 "xd3 is a better chance.
1s ltJxe4 16 'ifh3 .i.xd3 17 ltJxd4 ltJcS
..•
The smoke h as cleared and Bl ack
has an extra centre pawn for no com­
pensation. 17 .....f6 ! would h ave been
ltJC7? l::t c 8 36 l::t d c2 i.fS 37 l::tc4 :H8 3 8
�f2 �f7 39 ltJbs �e7 40 llxc8?
Equivalent to resignation, even if
the game somehow lasted 27 more
moves.
40 :Xc8 41 llxc8 i.xc8 42 ltJc3 �e6 43
.••

even stronger. ltJe4 h6 44 �e2 dS 4S ltJc3 i.d7 46 �d2

1 63
Th e Dark Kn ;ght System
c;t>d6 47 c;t>d3 �c5 48 c;t>d2 i.. c 6 49 �d3
i.. b 5+ 50 Wd2 i..f1 51 g3 fxg3 52 hxg3
h5 53 tbd1 g5 54 tbe3 i.. b 5 55 tbf5 i..d 7
56 tbe3 b5 57 Wd3 h4 58 gxh4 gxh4 59
a3 h3 60 tbf1 i..f5+ 61 Wc3 b4+ 62 axb4+
axb4+ 63 'iitd 2 'iitd4 64 tbg3 i..d 3 65 tbh1
h2 66 tbg3 i..f1 67 tbf5+ Wc5 0-1
Game 35
B.Perrusset-I.Moullier
Pa r i s 200 5
1 d4 tbf6 2 c4 tbc6 3 tbC3 es 4 d5 tbe7 5
e4 tbg6 6 i..e2 i..c5 [6 ... i..b4!] 7 tbf3 0-0
8 0-0 d6 9 a3 as 10 i.. d2 'iVe7
This generally useful move often
makes ... tbh s possible; e.g. 11 'ifC2
tbh S ! ? 12 tbxes?? 'fixes. In this game
Bl ack h ad other ideas.
11 :tb1 i.. d 7 12 b4?! axb4 13 axb4 i.. d 4!
Once again the slippery dark bishop
finds a home on the lovely d4-square.
14 i..d 3 J:[a3 15 :b3 nxb3 16 'iWxb3 :ta8
17 -.c2 i.. b6 18 'ifb2 h6 19 lla1 nxa1+
20 'ii'x a1
White has finally dealt with the can
1 64
of worms he opened up on the a-fil e,
but his remaining pieces are not very
well co-ordinated. Suddenly Black
starts attacking. All the attacking
moves are typical of th e variation, and
th e rooks are not needed.
tbg4 21 i..e1 tbf4 22 i.. C2?
Th e only satisfactory defence is 22
i..f1, though Bl ack will still have some
initiative.
22 .. :.f6!
Black is threatening the crushing
2 3 ... tbh 3+! 24 Wf1 i..xf2 ! 2S i..xf2
tbxh 2+ 26 �e2 tbf4+ 27 �d1 tbxg 2.
2 3 h3? 'ii'g6
23 ... tbxh3+! an d 23 ... tbxf2 ! also win.
24 hxg4 'ii' xg4 25 tbh4 �xh4 26 i.. d 1
tbd3 27 i.. f 3 i..xf2+ 28 i..xf2 'it'xf2+ 29
�h1 tbf4 30 tbd1 'ii' h4+ 31 �g1 i..g4 32
'iia 3 i..xf3 33 �xf3 'ii'e1+ 0-1
Game 3 6
D.Bara midze-E.Griezne
B a u n ata l 1999
1 e4 tbc6 2 d4 e5 3 d5 tbce7 4 c4 tbg6 5

8C3 i..C5 6 ltJf3 ltJf6 7 i..e2 0-0 8 0-0 d6
9 'i'C2 a6
In th e theoretical section I recom­
mend 9 ... i..d7 and usually 10 ... aS, but
th e game move is quite reasonable. In
either case, Black uses th e same attack­
i ng ideas on the king side.
10 a3 'ike7
In addition to preparing ... ltJh s, this
holds up White's c4-cS advance.
11 b4 i.. a 7 12 ltJdl ltJh5 [12 ... ltJf4 ! ?]
13 ltJe3?!
Here 13 g3 is best, which certainly
makes it clear why White needs to omit
h2-h3.
13 ...ltJgf4 14 ltJf5?! 'iWf6?!
At some cost, White has arran ged
not to be checkm ated on the kingside.
Bl ack shoul d slow down and play
14 ... i..xfs lS exfs e4 or l S ....:tfe8 with
positional advantages based on space,
activity, and White's weak fs-pawn.
15 g3! i..xfS 16 exf5 e4 17 i..xf4 ltJxf4
18 gxf4 exf3 19 i..xf3
You don't see this every day. Black is
still slightly better, and his position is
simpler to play, as we will see.
Illustra tive Games
19 ...'i¥h6
20 l:tael?
White decides that with three f­
pawns, he won't miss one. However,
even tripled pawns control squares,
and the f4-pawn is particularly impor­
tant since it covers weak dark squares.
Far better to give up the fs-pawn which
is one more obstruction to White's sad
bishop (though in fact White should
not be eager to give up any of his
pawns). After this his game slowly de­
teriorates.
20 .. :.xf4 21 'if e4 'iWg5+ 22 'if g4 'iff6 23
l:e4 :ae8 24 :tfel l:[xe4 25 l:txe4
Since the f4-pawn's disappearance
White h as had to be careful not to al­
low ... i..d4 and ...i.. e s, but he can't de­
fend everything.
25 .. :ii' b 2 26 i.. e 2?! 'ifxa3 27 f6?
Time trouble?
27 .. :.al+ 28 1i>g2 'ifxf6 29 f3 .d8 30
.f4 g6 3 1 i..d 3 �g7 32 h4 h5 3 3 'i'g5
lte8 34 .xd8 Ilxd8 35 l:te7 i.. b 8 36 b5
axb5 37 cxb5 �f6 38 l:te4 i.. a 7 39 :f4+
�g7 40 l:te4 i.. b6 41 �g3 :a8 42 �f4
:a3 43 i.. e2 i..f2 44 Ile7 l:Ia4+ 45 :e4
1 65
Th e Dark Knig h t System
lixe4+ 46 Wxe4 i.. x h4 47 f4 Wh6 48
i.d1 i.e1 49 i.f3 i.. c 3 50 i.g2 i.b2 51
i.f1 f5+ 52 Wf3 g5 0-1
Game 3 7
G.Kaidanov-A.Miles
Pa l m a d e M a l l o rca 1989
1 d 4 ttJc6 2 e4 e5 3 d5 ttJce7 4 c4 ttJg6 5
i.e3 i..b4+ [s ... ttJf6] 6 ttJd2 ttJf6 7 f3
'fIe7 8 g3 0-0 9 i.. h 3 c6!
The strongest players are much
more likely to do this than look to
blockade th e queen side.
10 a3 i.cS 11 ttJf1?! b5!
White's slow manoeuvring does not
take into account Black's option to
open th e position up. Black was better
already, but now White is in serious
troubl e.
12 b4 i..d4!
That square again ! When will White
learn not to play c2-c4 - ?
13 i.xd4 exd4 14 'iWxd4 bxc4 15 d6 .-es
Un derstandably, White tries to close
things again. Miles has correctly seen
166
th at h e will be winning the endgame if
it com es to th at.
16 ttJe2 as 17 ttJe3?! [17 bxas] 17 ... axb4
18 ttJxc4 'i'hS 19 i.g2 cS 20 'fIe3 i.a6
21 ttJb6 :ab8 22 axb4 :xb6 23 bxcS
:c6 24 :as ttJes 25 ttJf4 'iVgs 26 'iWd4
:b8 27 :a1 h6 28 lIxa6 :xa6 29 0-0
ttJc6 30 'fIC4 l:tas 0-1
Brutality in the .te3 variation
Game 38
S.Brudno-J. Benjamin
Boston 2001
1 d4 ttJf6 2 c4 ttJc6 3 ttJc3 eS 4 d5 liJ e 7 5
e4 liJg6 6 i.. e3 i.. b4 7 f3 i.xc3+ 8 bxc3
d6 9 cS 0-0 10 i.d3 liJd7! 11 cxd6 cxd6
12 liJe2 'iWas 13 0-0 liJcs 14 i.c4 i.d7 15
i.b3 .:tac8 16 g3 fS
After constructing an ideal position
on the queenside, Benjamin opens up
th e second front. The more I look at the
Dark Knight System, th e more I'm
struck by how often it is Bl ack who h as
th e convenient pawn breaks, and how
usefu1 th at is.
 Illustrative Games

As I mentioned in the theoretical

section, this dubious move h appens
like clockwork. Incidentally, I worked
out Black's plan independently during
this game.
9 ...0-0 10 i..d 3 ttJd7 11 cxd6 cxd6 12
ttJe2 'ii'a 5 13 0-0 tDC5 14 'ii'd 2 ttJxd3?1
This is a big commitment. 14...b6 is
more accurate - White's bishop isn't
going anywhere.
15 'ii'x d3 b6 16 c4 i.. a 6?1
17 .i.c2 fxe4 18 fxe4 :xf1+ 19 �xfl Having taken the bishop, 16 ... fS (or
�h3+ 20 �gl :f8 21 'ii'd 2 -.xa21 lS .. .fS ) is the logical follow-up. Black is
Owl better on the queenside, but it is h ard
22 :el to play against a single weakness.
If 22 l:txa2 :f1 mate. 17 "b3 :ac8 18 :fCl :C7 19 :[c2?1
22 .. :iVC4 23 i.. f 2 as 24 ttJCl :f3 25 ttJe2 White intends a perpetual on my
a4 26 'ii' g5 ttJxe4 0-1 queen, but there is a tactical snag. The
right way to implement this plan is 19
.i.f2 l:Uc8 20 i.. e 1 "cS+ 2 1 .i.f2 etc.
Game 39 19 ...�fc8 20 .i.d2 .c5+ 21 i.. e 3 i..xc4 22
R.Mitchell-J.Schuyler i..x c5 i.. x b3 23 axb3
B l oom i n gton 1991

1 d4 ttJc6 2 c 4 e 5 3 d5 ttJce7 [3 ....i.b4+1]

4 e4 ttJf6?1 [4...tDg6] 5 tDc3 ttJg6 6 i.. e3
.ib4 7 f3 .i.xc3+ 8 bxc3 d6 9 cS?!

23 ... bxc5?
This game was a long time ago, but
I remember that I never even consid­
ered th e right move, which is 23 ... as l,
even if Black is actually not much bet-

167
Th e Dark Knig h t System
ter after 24 ttJg 3 ! ttJh4 25 ttJf1 h6 26
ttJe3. Still, White must find these
moves. After the text, Black gets all
tang led up and is lucky to extricate
himself. The rest of th e game is full of
inaccuracies, most of which I will not
bother to mention.
24 .:ta6 :d8 25 :ca2 [25 b4! c4 26 1:c6]
25 .. Jtb7 26 l:t2a3 l:tdd7 27 ttJC3 [27
11c6 !] 27 ...'iitf8 28 ttJa4 lbf4 29 lbb2 rj;e7
30 �f2 f5 31 g3 lbh5 32 lbc4 fxe4 3 3
fxe4 ttJf6 34 lba5 :b6 3 5 lbc6+ �f7 36
'lite3 lbg4+ 3 7 �d3 lbxh2 38 :xa7 l:1xa7
39 l:txa7+ c;t>f6 40 �C4 lbfl 41 l::td 7
lbXg3 42 1:txd6+ �gs 43 l:td7 g6 44
:xh7 lbxe4 45 l:te7 �f4 46 b4 cxb4 47
lbxb4 lbf2 48 11f7+ �g3 49 rj;c5 l:tb8 50
lbc6??
Things have been approximately
even the whole time despite my fre­
quent pawn advantag e. Here White
falls into a trap. The coming knight and
pawn endgame is winning for Black in
all variations because White's king and
knight are both so far from the passed
pawns.
50 lbe4+ 51 WC4 lbd6+ 52 �d3 :b3+
•••
1 68
53 WC2 lbxf7 54 �xb3 e4 5 5 �C3 Wf2
56 ttJe7 g5 57 lbf5 e3 0-1
Game 4 0
l.Jelen-E.Dizdarevic
Lj u b lj a n a 1992
1 d 4 lbf6 2 c 4 lbc6 3 lbc3 e 5 4 d5 lbe7 5
e4 lbg6 6 .i.e3 .i.b4 7 f3 .i.xc3+ 8 bxc3
d6 9 cS 0-0 10 h4?!
White must attack or Black will
tighten the screws one by one. How­
ever, this is too soon and too crude.
10 ... lbh5 ll .i.f2 c61
Over and over we see this move,
typically at th e most surprising and
inconvenient tim e for White.
12 g4 lbhf4 13 h5 ttJe7 14 cxd6?!
The complications are dizzying,
even for my computer. It's hard for
things to work out well for White with
his king in the middle, but he needs to
try 14 h6 or 14 ttJe2.
14 .. :ifxd6 15 c4 cxd51 16 cxd5 .i.xg4! 17
h6
17 fxg4 runs into 17 .. .'i'b4+ and

18 .. :iVxe4+, winning the rook
17 gS 18 'ifb3 .td7 19 tiJe2
•••
19 tiJeg61
..•
19 ... J.. a4! 20 'iVe3 lUc8! and White
cannot l ast long.
20 :gl f6 21 tiJe3 :fe8 22 a4 wh8 23
Wd2 b6 24 i.. a 6 tiJh3 25 :gfl tiJxf2 26
J..xe8 :lxe8 27 l:txf2 'iVes 28 :ffl tiJf4 29
ttJe2 'iVas+ 30 tiJe3 tiJe2 31 :fel tiJxel
32 l:txel <ii;g8 33 f41
White isn't h aving any fun at all,
but this only makes things worse.
33 exf4 34 eS? fxes 35 d6+ <ii; f8 36
..•
'ii'b l �g8 37 'iNb3+ <ii;f8 38 'iVbl 'ifes 0-1
Black loses because
he's playing Karpov
Game 41
A.Karpov-D.Cheva l l ier
Fra nce 1993
1 d4 tiJf6 2 e4 tiJe6 3 tiJe3 eS 4 dS tiJe7 5
e4 tiJg6 6 i..e 3 i.. b4 7 f3 .txe3+ 8 bxe3
d6 9 'iVd2 tiJd7 10 h4 h6 11 g3 b6 12
�h3 tiJes 13 tiJf2 .d7
Th ere is nothing wrong with this
Illustra tive Games
move, but 13 .. .fS can be played, and so I
believe it shoul d be played.
14 :bl as 15 f41!
In his annotations, Palliser gives this
an ' !' and indicates th at it is the begin­
ning of the end for Chevallier.
ls exf4 16 gxf4
.••
16 'ife71!
•••
After this Bl ack is a little worse.
Meanwhile, should Bl ack find 16 ...h S,
he is a little better! In one stroke, Black
stabilizes his knight, clamps down on
White's weak g4-square, and immobi­
lizes White's isol ated h-pawn. Of
course, with a small edg e (or even a
small disadvantage) Karpov will proba­
bly win - he does outrate Chevallier by
430 points.
17 i..x eS! bxes 18 hS tiJf8 19 :gl f6 20
i..d 3 tiJd7 21 'iWe2 i.. a 61
Black's situation will not improve
with his king in the centre. For better or
for worse, he must castle and try for
.. .f6-fS.
22 tiJg4 tiJb6 23 tiJe3 i..e8 24 'ii'g2 l:[g8
25 .g6+ �d81! [2S ... �f8] 26 �d2 i..d7
27 'iVh7 'ift8 28 :g2 <ii;e 8 29 l:tbgl �b7
1 69
Th e Dark Kn ight System
30 :txg7 J:th8 31 �g6 i.. e 8 32 'iVfs l:td8
3 3 �e6 �xhS 34 ltJfS ltJc8 35 eS fxeS 36
fxes 'iVe8 37 J:tbl+ Was 38 :xC7 'irxe6
39 dxe6 dxes 40 llxcS e4 41 ltJd4 l:txd4
42 cxd4 exd3 43 l:txhS l-0
Overall, an impressive game by Kar­
pov, but not one that casts any doubt
on Black's opening.
Game 4 2
W.M.Buehl-J. Benja m i n
R e n o 1999
1 d4 ltJf6 2 C4 ltJc6 3 ltJc3 eS 4 ds ltJe7 5
e4 ltJg6 6 ltJf3 i..b4 7 �d3 �xc3+ 8 bxc3
d6 9 0-0 0-0 10 h3? ltJhs 11 Wh2!?
11...ltJhf4
Because of 11 Wh 2, Black's norm al
attacking plan needs to be modified -
11 ... ltJgf4?! 12 i..a 'irf6? ? 13 g4 traps a
knight.
12 ltJgl?! ltJxd3!
They say bad bishops defend good
pawns - but they don't if they're dead.
13 �xd3 fS 14 exfs?! [14 f3] 14 ... i..xfS
15 -'g3?1 'ii'd 7! 16 f4?! -.a4
1 70
This is why I like doubling White's
pawns so much - it provides a ready­
made target on the other front if the
king side play is running out of steam .
17 fxes -.xc4 18 lldl ltJxes 19 l:td4 'ifcs
20 ltJe2 llae8 21 ltJf4 ltJg6 22 ltJxg6
�xg6 23 �gs :e2 24 a4?! laff2 25 :gl
i..e 4 0-1
Game 43
A.Hahn-J.Bonin
New York (ra p i d ) 2003
1 d4 ltJc6 2 c4 eS 3 dS ltJce7 4 e4 ltJf6 5
ltJc3 ltJg6 6 ltJf3 i.. c s
I prefer 6 .. ib4 here, but we must
know this position anyway via 4 ... ltJg6
S ltJC3 �cS 6 ltJf3 ltJf6.
7 i.. d 3 0-0 8 0-0 as 9 a3 d6 10 :bl ltJhs
11 b4 axb4 12 axb4 �b6 13 ltJa4 �a7
14 i.. c 2 ltJhf4
Here again, the h-knight goes first.
This is to allow ... i.. g 4 and ... ltJh4 - an
effective plan with White's bishop al­
ready on c2. Black h as a significant ad­
vantage and a powerful initiative.

15 c5 i..g4 16 g3 .f61
Black is almost winning at once
with the alternative 16 ...ttJh 3+ 17 �g2
-.d7 18 i.. e 3 f5 !, bringing a rook into
the attack.
17 l:tb3 i..x C5 18 ttJxc5 dxC5 19 bXC5
l:ta1 20 gxf4 ttJxf4 21 Wh1 ttJh3 22 'it'e2
l:tfa81 [22 ... l:ta2!]
23 :c311
Instead, with 2 3 'ife3 ! White h as ex­
cellent chances to realize her extra ma­
terial. After the text, she is completely
busted.
23 ... :8a2 24 �g2 :XC1 25 :XC1 ttJf4+
26 litf1 ttJxe2 27 Wxe2 'iff4 28 :e3
i..xf3+ 0-1
Illustrative Games
The catastrophe that is
White's plan of g2-g3 and f2-f4
Game 44
E.Schiendorfer­
D.Recuero Guerra
E u ropea n J u n i or C h ' s h i ps,
H e rceg N ovi 2006
1 d4 ttJf6 2 c4 ttJc6 3 ttJc3 e5 4 d5 ttJe7 5
e4 ttJg6 6 g3 i..b4 7 i..g2 i..x c3+ 8 bxc3
d6 9 ttJe2 0-0 10 0-0 'iVe7 11 �d3 i..d 7
12 i.. a 3 :fe8 13 :ae1 b6 14 ttJC1 .d8
Not the most inspiring play so far.
Now comes typical h ara-kiri by White.
15 f4? exf4 16 gxf4 ttJh5 17 ttJe2 ""'4 18
i..C1 f5 19 Wh1 :e7 20 exf5?? iJd5 21 �f3
2 1 'ifxf5 ':xe2 2 2 l:[xe2 ttJg3+ 2 3
�g l ttJxe2+ 2 4 �h l ttJg3+ 25 �g l ttJxf5
is hilariously brutal.
21 .i.g4 22 'it'f2 :xe2 0-1
..•
Game 45
C.Baluta-A.Cioa ra
Buch a re st 1996
1 d4 ttJc6 2 c 4 e5 3 d5 ttJce7 4 e4 ttJg6 5
1 71
Th e Dark Kn ;ght System
a3 tbf6 6 lbC3 .i.c5 7 h3 d6 8 tbf3 as 9
g3
This is amazing . I'm wondering
when White will get around to moving
his b- and f-pawns.
9 .0-0 10 .i.g2 c6 11 0-0 cxd5 12 cxd5
••
.i.d7 13 l:[e1 'iVc8 14 h4? tbg4 15 :e2 b5
16 .i.d2 b4 17 tba4
17 ... .i.b5?!
Black has built his position admira­
bly so far, but now he starts to lose the
thread. 17 ... .i.a7 18 axb4 .i.bS was bet­
ter.
18 tbxc5 'iVxc5 19 h5 tbe7?! [19 ... .i.xe2]
20 .i.h3?! [20 .i.e3 !] 20... .i.d7?
After this lemon, Black is actually
worse for a moment.
21 .i.g2?
But only for a moment. 21 h 6 ! g6 22
'iVb3 was correct.
21 bxa3 22 :xa3 f5?! [22 ... .i.bS!] 23
•..
.i.e3?! tbxe3 24 1:texe3 fxe4 25 :xe4
:ab8 26 ..-d2 a4 27 tbh4 :b7 28 :ee3
l:lb4 29 1:tac3?! ..-b6 30 l:le2 h6 31 elth2?
l:ld4? [3 1 ... .i.bS] 32 :d3?? :xd3 33
..-xd3 .i.b5!
Finally!
1 72
34 ..-e3 ..-xe3 35 :xe3 l:lxf2 36 eltg1
':xb2 37 .i.h3 :b1+ 38 eltf2 l:tb3 39
.i.e6+ eltf8 40 :1e1 a3 41 :C1 a2 42
tbg6+ tbxg6 0-1
A loss for Black -
ca n this be right?
Game 4 6
B u Xia ngzh i-L.Ch ristia nsen
Dei zisa u 2 000
1 d4 tbf6 2 c4 tbc6 3 tbc3 e5 4 d5 tbe7 5
e4 tbg6 6 a3 .i.c5 7 .i.d3 as 8 :b1 d6 9
h3 0-0 10 b4 axb4 11 axb4 .i.a7 12 g3?!
12 lbe8?!
•••
This awkward move prepares the
•.f7-fS break, but Black never gets to
execute this plan. Besides, 12 ... c6! is
just sitting there waiting to be pl ayed.
Yet again. One benefit to ... c7-c6 is th at
after ... c6xdS c4xdS White can no
longer shut out the monster a7-bishop.
Black would then be significantly bet­
ter.
13 tbf3 .i.d7
Black realizes th at 13 .. .fS ? ! hurts his

position, trading off White's lame light­
squared bishop and allowing White to
castle without dropping the h 3-pawn.
After th e text move, White h as equal­
ized - though not for long.
14 i.. g 5?!
14 �fl! 1t'c8 15 �g 2 was correct.
14 ... "eS 15 g4 e6 16 ttJe2 h6 17 i.. d 2
exd 5 1S exd 5 1t'dS 19 ttJg3 i.. b 6 20 0-0
ttJf4! 21 i..xf4 exf4 22 ttJh5 ttJf6 23 ttJxf4
lieS
Black's pawn sacrifice h as removed
White's best minor piece from play and
opened the gates for Black's pieces.
This is a thematic idea in the Dark
Knight System, and indeed in any posi-
Illustra tive Games
tion where the opponent has put most
of his pawns on one colour (especially
in the vicinity of his king).
24 �e2 lieS
Houd;n; points out 24 ... g 5 ! 25 ttJg 2
(the knight's idiot square; but if 25 ttJe2
h 5 !) 25 ... :c8 and 26 ...ttJxe4. In the
game White's knight becomes a real
pain in the keister.
25 �b2 ttJxe4 26 ttJh5 ttJe3 27 :be1 g6
2S "ii'd 2
Had White played 27 l::tfcl! he could
have gone for 28 lIxC3 :xC3 29 'iWxc3
gxh 5 30 gxh 5 with no fear of 30 ... .txh 3.
2S...ttJe2+ 29 �g2 lIxe1 30 .-xh6 ttJf4+
31 'ii'xf4 lbf1 32 �xf1 gxh5 3 3 'ii'h6 f5
34 "g6+ �hS?!
Christiansen, who has been defend­
ing perfectly until now, slips. It is not
good to allow 'ifxh 5 with check, and by
winning the pawn White defends his
own peon on h 3 .
35 gxf5 lIfS 36 'iWxh5+ �gS 3 7 "-g6+
�hS 3S 'ii' h 6+ �gS 39 f6
39 ...:f7??
39 ... .txh 3+ 40 'ii'xh 3 lIf7 or 40 ...lIxf6
offers legitimate chances of survival.
1 73
Th e Dark Kn igh t System
40 ttJg51 i..x h3+ 41 ttJxh3??
Suddenly Black is back in it. 41
�e1 !! is completely winning; e.g.
41 .. :ifd7 42 'iVg 6+ �8 43 J..b S ! .
4 1 ..'iWxf6 4 2 ..th7+ �hS 4 3 ..tg6+ �gS
44 i..xf7+ c.t>xf7 45 'ii' h 7+ �eS 46 'iVxb7
'it'd4? [46 ... 'ii'fS !] 47 "eS+ ..tdS 4S 1t'e6+
i..e7 49 b5 'iVd3+ 50 �g2 'iVxb5
Black has recovered one pawn, but
his queen is out of pl ay an d White
takes the opportunity to enter a win­
ning pawn endgame.
51 ttJf4 'iVb2 52 ttJg6 'ii'f6 53 'ii'xe7+
'iVxe7 54 ttJxe7 �xe7 55 �f3 �f6 56
�4 �g6 57 �g4 �f6 5S f4 1-0
Hort plays the Kevitz,
and lets Hu bner off the hook
Game 4 7
R.H ubner-V.Hort
G e r m a n League 19 84
1 e4 ttJe6 2 d4 e5 3 dxe5 ttJxe5 4 f4 ttJe6
5 i.. e3 ..tb4+ [S ... dS!] 6 e3 ..ta5 7 ttJf3
ttJf6 S e5 [8 'ii'a4! ?] S ... ttJg4 9 ..te5? d6!
10 exd6 0-0
1 74
White is much worse here.
11 i.. e2 :leS 12 h3 exd6 13 hxg4 dxe5
14 'ii'x dS l:txdS?1 [14 ... ..txd8] 15 ttJfd2
ttJe7 16 ttJa3 ttJd5 17 g3 ..te7 1S ..tf3
ttJe3 19 g5
15 ..tf5?!
•.•
19 ...ttJfS ! is better, forcing 20 ttJf1
(or 20 ttJe4, which is not very good ei­
ther), because 20 l::tg 1? loses to ttJxg 3 !
2 1 l:txg 3 l:txd2 ! 22 <ittxd2 J..xf4+ and
2 3 ... J..xg 3.
20 ttJde4 ttJxe4 21 ttJxe4 ..td3
There is still some play, but Black
has no real advantage.
22 b3 l:tabS 23 �f2 b5 24 ttJe3 e4 25
ttJd5 ..tb6+ 26 �g2 �S 27 l:the1 �eS 2S
b4 ..tf5 29 llxeS+ !txeS 30 a4 bxa4 31
l:txa4 h6 32 gxh6 gxh6 3 3 ttJxb6 axb6
34 �f2 :dS
By now it is White who is trying for
a win, though it is unlikely.
35 llaS l::tx aS 36 ..txaS �e7 37 �e3 �d6
3S �d4 b5 39 i.. d 5 f6 40 ..tf7 ..td7 41
�e4 �e7 42 i..g6 ..te6+ 43 �d4 �d6 44
..th5 f5 45 i..g 6 i.. d7 46 i.. f 7 i.. eS 47
..teS i.. a 6 4S �e3 �e7 49 i.. g 6 ..teS 50
�d4 �d6 51 ..teS i.. a 6 Vz-Vz

Game 48
M.Orso-G.Bordas
B u d a pest 2 000
1 e4 �c6 2 d4 eS 3 dxes �xes 4 f4 �c6
5 i..c 4 �f6 6 �c3 i..b 4 7 e5 d5 8 exf6
dXC4 9 'ii'e 2+ �f8!? [9 ...ii.e6! ] 10 ii.e3
Had White been tempted by 10
fxg7+? �xg7, Black's last move would
have been amply rewarded. The rook's
rapid arrival on the e-file is a serious
problem for White.
10 ...'iYxf6 11 0-0-0 i..e 6
The inconvenience suffered by
Black's king is not enough to make up
for White's missing pawn and shortag e
of light-squared bishops.
12 �e4 'ii'e 7?! [12 .....f5] 13 �f3 c3 14
b3?! [14 bxc3] 14... aS lS �bl a4 16
'ii' bS??
White is now lost. 16 .i.f2 is essen­
tial, though Bl ack has a clear edge.
16 ...axb3 17 axb3
17 ...i.. x b3 18 cxb3 'iVxe4+ 19 :d3 l:[a5
20 'ilfC4 'ii'xC4 21 bXC4 We7 22 �d4
�xd4 23 i..x d4 l:lha8 24 :e3+ <iitd 7? 25
Illustra tive Ga mes
lthel?
25 .txc3! ii.xc3 26 l:tdl+ and Black
still has to show good technique.
25 ... l:tal+ 26 <iit c2 l::t la2+ 27 �d3 J::t d 2+
28 �e4 fS+ 29 �xfS lIxd4 30 cS i.. x cs
31 :xc3 :f8+ 32 �g5 i.. e7+ 0-1
Game 49
S.Fedorchuk-A.Miles
E u ropea n C h a m p i o n s h i p,
Ohrid 2001
1 e4 �c6 2 d4 e5 3 dxe5 �xe5 4 �c3
i.. cs S f4 �c6 6 ttJf3 d6 7 �a4 i.. b 6 8
ii.d3 d5?!
An interesting attempt to take ad­
vantage of the decentralized knight,
but Black is spending a tempo to open
the position when he is behind in de­
velopment and White will soon have
the bishop pair.
9 �xb6 axb6 10 e5?!
Not terrible, but it makes no sense
to close the position. The fs-square will
be useful for all of Black's pieces.
10 ...�ge7 11 0-0 i.. f 5 12 �h4 i..x d3 13
1 75
The Dark Kn ight System

-.xd3 -.d7 i.. c 5 5 ttJb3 i.. b 6 6 ttJC3 d6

Black h as equalized.
14 i.. d 2 -'g4 15 i.. e1 0-0 16 h3 'ifd7 17
i.. d2 :a4 18 i..c 3 ttJg6 19 ttJxg6 fxg6 20
g3 ttJe7 21 �h2 h5 22 b3 l:[e4 23 :ae1
'ifc6 24 i.. d 2 ttJf5
For our purposes, it is important to
use the precise move order 6 ...ttJge7 7
-.e2 d6.
7 'iie2 ttJge7 8 i..e 3 0-0 9 0-0-0 f5!

There's no tim e like the present.

25 'ilVc3?!
The black knight becomes strong af­
ter this. 2 5 94 was best.
25 ... h4 26 gxh4 ttJxh4?!
26 ...'iWxc3 27 i.. xc3 �7 ! was better.
27 'ilVxc6 bxc6 28 �g3 ttJf5+ 29 �g4 �f7
30 :xe4 dxe4 31 I:te1 :d8 32 i..e 3?!
:d1! 33 l1e2 �e6 34 i..f2 :f1 35 a4 c5
36 a5?! bxa5 3 7 �g5? l:[h1 38 i..x c5
:xh3 39 �xg6 l:1g3+ 40 �h7 e3 41 �g8
g5 42 fxg5 :xg5+ 43 �h7 :g7+ 44 �h8
:g3 45 l:Ih2 �xe5 46 i.. a 3 :g4 47 i.. b 2+
�e4 48 i.. c 3 :h4+ 49 l:[xh4+ ttJxh4 50
i..xa 5 e2 51 �g7 ttJf3 52 �f6 �e3 0-1
10 g3 fxe4 11 ttJxe4 ttJf5 12 i..f4 -.e8
Mr. H says 12 ... �h 8 or 12 ... h6 with
advantage to Black.
13 i.. g2 a5 14 l:the1 'iig6 15 c3?! h6 16
a4 i..e 6?! 17 'ilVC2? [17 ttJec5!] 17 ... ttJh4
18 i.. h 1 i..g4 19 l:td3 :ae8 20 f3? ttJxf3?
[20 ... i.. e 6] 21 i..xf3 i..xf3 22 :xf3 d5 23
ttJbd2 dxe4 24 :ff1 ttJe5 25 l:[xe4 ttJg4
26 l:tfe1 :d8 27 :e7?1 'iiXC2+ 28 �XC2
g5 29 :1e6 :f7

Game 50
Goh Wei Ming-F.Bellini
T u r i n Olym p i a d 2006

1 e4 e5 2 ttJf3 ttJc6 3 d4 exd4 4 ttJxd4

1 76

30 h3??
Instead of this blunder the simple
30 :g6+! �h 7 31 l:tee6 keeps White in
the game.
30...:xe7 31 l:txe7 gxf4 32 hxg4 fxg3
0-1
Game 51
P.Bonte mpi-O.Jova nic
N ova G orica 2008
1 e4 eS 2 ttJf3 ttJc6 3 d4 exd4 4 tiJxd4
..tcs S .i.e3 .f6 6 ttJbs .i.xe3 7 fxe3
�h4+ 8 g3 "d8 9 'it'g4 gs!
This Basmaniac-esque move is a bit
funny-looking at first, but if we think
about it, it is the dark squares that
need coverage since we have a light­
squared bishop and we are about to
have a knight posted on the lovely eS­
outpost. To be a little more concrete, it
is beneficial to take the f4-square from
White's queen.
10 tiJ1C3 ttJes 11 'ii'e 2 d6 12 h3 c6 13
ttJd4 ttJf6 14 0-0-0 'iVe7 lS 'Wf2 .i.e6 16
.i.e2 0-0-0 17 ttJfS .i.xfS 18 exfS
Illustra tive Games
White's pawn structure has im­
proved somewhat, but his bishop has
not, nor h as the gaping hole on eS. Fur­
thermore, White no longer has the fS­
hole to use.
18 ... ttJed7 19 .i.f3 ttJcS [19 ... dS ! ?] 20
l:ihel hS 21 :d4 'ii'e s 22 h4 l:thg8 23
:edl gxh4 24 gxh4 �C7 2S :f41! l:de8
26 �d21
26 ...tiJcd71!
Black is reluctant to put any pawns
on light squares for long-term posi­
tion al reasons, but White won't reach
the long term after 26 ... dS! 27 :el
ttJce4+ 28 i.. x e4 ttJxe4+ 29 ttJxe4 dxe4,
which exposes the ludicrous placement
of White's king; e.g. 30 �Cl "as 3 1
�bl ':g2 32 'iYfl 'ii'd2 33 :Cl 'Wxe3 34
'i¥xg 2 'ifxf4 with a winning position.
Instead, soon comes an avalanche of
blunders that is presumably time­
induced.
You know what? You don't want to
see th e rest of this. It's ridiculous and
embarrassing.
Black eventually won the game on
move 81.
1 77
Th e Dark Kn ight System
Game 52
S.Vajda-S.Skembris
N a ujac 1999
1 e4 eS 2 ltJf3 ltJc6 3 d4 exd4 4 ltJxd4
-.f6!1
Apparently this is a pl ayable move
order, though it gives White th e option
of 5 ltJbS ! ?, which is good enough for
some edg e.
5 i.. e3 i.. c S 6 c3 'i¥ g6 7 f3 ltJge7
In the theoretical section, I recom­
mend th e immediate 7 ... a6.
8 -.d2 a6 9 ltJc2 i..xe3 10 ltJxe3 d6 11
i.. e2 fS
12 exfs i..x fS
12 ...ltJxfS is equally good.
13 0-0 0-0
13 ... 0-0-0 is at least as good too.
14 ltJa3 �ae8 15 �ae1 �h8
It's time to take away White's op­
tion to snap off the bishop; i.e.
ls ...i.. e 6.
16 ltJac2 -.h6 17 i..d 3 i..x d3 18 -.xd3
ltJg6 19 ltJdS1! ltJces 20 -.d4 c6 21
ltJde31! [21 f4!] 21 ... ltJf4 22 �h1 'ii'g 61!
1 78
[22 ... ltJed3 !] 23 g31 [23 'iWd2] 23 ... ltJfd3
24 l:ld1 ltJxb2 25 �b1 ltJed3 26 'ii' b6
ltJa4 27 'ii'x b7 ltJxc3 28 �b61! [28 l:1b3]
28...ltJes 29 �xa6 lixf3 30 ::txf3 'i¥e4 31
ltJe1 ltJxf3 32 ltJ1g2 ltJbS! 0-1
Game 53
A.Motylev-S.Gligoric
Yugoslav Tea m
Cha m pion s h i p 2 000
1 e4 eS 2 ltJf3 ltJc6 3 d4 exd4 4 ltJxd4
i..cS 5 i.. e3 'iYf6 6 c3 -'g6 7 ltJd2 ltJf6 8
f4 ltJxd4 9 cxd4 i.. b4 10 fS 'iYg4 11
-'xg4 ltJxg4 12 i..f4 dS 13 h3 ltJf6 14 eS
ltJe4 15 g4 g6 16 fxg6 fxg6 17 a3
i..xd2+ 18 i..x d2 hS 19 l:tg1 hxg4 20
hxg4 i..e 6 21 0-0-0 �d7 22 i..g 2
22 ... ltJf21!
Black keeps an edge with 22 ...ltJxd2
23 1txd2 lhf8 or 23 Wxd2 i.. x g4.
23 lidf1 �af8 24 i.. gs ltJxg4 25 l:lxf8
l:txf8 26 i..x dS i..x dS 27 l:txg4 :f1+ 28
�d2 l:tf2+ 29 �c3 as 30 i.. h 4 l:tf3+ 31
�d2 i..f7 32 dS i..x dS 3 3 l:txg6 l:lb3
Yz-Yz

Game 54
P .Hromada-L.Ostrowski
Moravi a n Tea m
C h a m pion s h i p 2 003
1 e4 eS 2 tbf3 tbc6 3 d4 exd4 4 tbxd4
�f6 5 i.e3 i.cs 6 c3 'ii'g6 7 tbbs i.xe3 8
tbxC7+ �d8 9 tbxa8 i.f4 10 .f3 i.b8
I recommend lo ...i.h 6 ! in th e theory
section.
11 tbd2 tbf6 12 0-0-0 l:te8
13 l:te1?!
White shoul d be willing to sacrifice
th e e-pawn, for which he will gain easy
development; e.g. 13 g3 tbxe4 14 tbxe4
'ii'xe4 15 'ii'xe4 :xe4 16 .i.g 2 l:te7 17 f4
and, for one thing, it is no longer clear
how Bl ack will ever corral the wayward
knight. The text move is far too passive,
a recurring problem for White as the
game goes on.
13 ... b6 14 i.. b s i.. b7 15 tbxb6 axb6 16
'iVe3 i..c 7 17 f3 tbds [17 ...Wc8] 18 'iVf2
tbf4 19 g3 tZJe6 20 i.xc6?!
A bad idea, which weakens the light
squares around White's king and sim-
Illustrative Games
plifies Black's task of using his bishop
pair.
20 ... i..x c6 21 tbb3 tbcS [ 2 1 ... We7!] 22
tbxcs 'iVgS+ 23 'iVe3 'iVxcs 24 'iVxcs bxcS
25 b3 fS 26 exfs 1%18 27 l1e3 l:txfS 28
l:the1 i.. d6
29 :tf1? [29 f4] 29 ... gS 30 �b2 hS 31
:d1 rtt c 7 32 l:tf1 rtt b 7 33 a3 c4 34 b4
i.. C 7? [34 ... h4!] 35 �C2?! [35 l:[e7 !]
3S...i.b6 36 l:te7 �c8 3 7 f4 i.g2 3 8 lld1
gxf4 39 l:tdxd7??
Here 39 gxf4 d5 is no bargain for
White, but the text ends the game im­
mediately.
39 .. .f3 40 l:tb7 i..d8 41 l:tf7 f2 0-1
Game 55
D.Ca mpora-V.Tkachiev
B i e l 199 5
1 e4 es 2 tbf3 tbc6 3 d4 exd4 4 tbxd4
i..c S 5 i..e3 'iYf6 6 c3 'it'g6 7 tbbS i..xe3 8
tbxC7+ wd8 9 tbxa8 i..f4 10 'iVf3 i.. h 6 11
'iVfS?! 'ii'xfs 12 exfs b6 13 tba3 i.. b7 14
tbxb6 axb6 15 tbC4 �c7 16 l:td1 dS!
A clever thrust.
1 79
Th e Dark Kn igh t System
17 lLla3
If 17 llxdS?! lLlf6 18 lld1 l:te8+ 19
.i.e2 .i.a6, White regrets his pawn grab.
17 ... lLlf6 18 lLlb5+ �d7 19 .i.e2 :te8 20
�fl :e5 21 h4 lLle4 22 :h3 lLld2+ 23
�el .i.a6 24 a4 lLlC4 25 b3 lLld6 26
l:lhd3 d4! 27 cxd41!
White's rooks will not enjoy lan­
guishing behind the isolated d-pawn.
However, the position is not rich
enough to offer Black good winning
chances (in GM play).
27 . .:e4 28 g4 .i.f4 29 �1 lLlxb5 30
. this system is known.
axb5 .i.xb5 31 :3d2!
Unlikely as it seems, this move saves
the day, since 31... .i.xd2 3 2 .i.xbs
1 80
(threatening both 33 l:txd2 and 33 dS)
32 ...�xd4 33 �e2 regains the piece.
31 ... :xe2 32 :xe2 �d6 33 �el i..xe2
34 �xe2 �d5 35 �d3 h6
Not 3 s ... lLlxd4?? 36 �c3 i..e5 37 f4
i..f6 38 g5.
36 :el i.. d6 3 7 :e8 .i.e7 38 h5 .i.f6 39
:c8 lLlxd4 4 0 :C7 lLlc61!
4o ...lLlxb3 is a better try.
41 f4 lLle7 42 :a7 �c6 43 �e4 Yz-Yz
Game 5 6
I.Grynfeld-A. Bisguier
H el s i n ki Olym p i a d 1 9 52
l e4 e5
Once, when GM Bi sguier was ana­
lysing one of my games, he said he'd
known all about ...lLlc6 an d ...e7-eS
since long before I'd been born. This
makes sense because he was coached
by Alexander Kevitz, by whose name
2 lLlf3 lLlc6 3 d4 exd4 4 lLlxd4 .i.C5 5
lLlxc6 �f6 6 1Wf3 iWxf3 7 gxf3 bxc6 8
i..e 3 .i.xe3 9 fxe3

As I mentioned in the theoretical
section, White has done pretty well
here. Bisguier's plan takes care of any
potential problems, though I think it's
simpler to start with 9 ... d6.
9 ...ttJf6 10 ttJc3
After 10 e s ! ? ttJg 8 11 :g l g6 12 ttJc3
f6 13 exf6 ttJxf6 14 0-0-0 0-0, White h as
littl e to nothing either, but I still see no
reason to allow him this option.
10 ... 0-0 11 0-0-0 [11 eS ttJdS] 11 ..Jie8
12 .th3 d6
Also possible is 12 ... g6 (a useful
semi-waiting move), intending ...:e S­
hS. After the text, Black slips into a pas­
sive position.
13 i..xc8 :axc8 14 :d4 as?! [14 ... ttJd7]
15 :C4 cS 16 l:ta4 :ta8 17 ttJbS l:tec8 18
ttJa 3?! [18 e s ! ?] 18 ... ttJd7 19 ttJC4 ttJb6
20 ttJxb6 cxb6
Black h as equalized (again). Nothing
else h appens.
21 c4 �8 22 :a3 riit e 7 23 :b3 :tab8 24
a4 �e6 25 :d1 :c6 26 :g1 g6 27 riit d 2
�f8 28 J:tg3 fS 29 exfs+ :XfS 30 e4 l:thS
31 h3 gs 32 riit e 3 h6 33 l:ld3 :h4 34 l:lb3
�es 35 �d3 :c7 36 �b3 :b7 37 :d3
Illustra tive Ga m es
�e6 38 :b3 �d7 39 l:ld3 Wc6 40 �f2
:tf7 41 �g2 Yz-Yz
Game 5 7
J.Smeets-A. Beliavsky
N eth e rl a n d s-Slove n i a
ra p i d m atch, M a ri bor 2 004
1 e4 eS 2 ttJf3 ttJc6 3 d4 exd4 4 ttJxd4
i.. c S 5 ttJxc6 'ir'f6 6 '6'd2 dxc6 7 ttJC3
.td4 8 .td3 ttJe7 9 0-0 ttJg6 10 ttJe2
i..x b2!
Beliavsky courageously grabs th e
pawn - th e "principled" move (what­
ever th at means).
11 i.. x b2 .xb2 12 f4 'ii'a 3 13 fS ttJes 14
-gs i.. d 7
I a m recommending 14... 'iif8!, when
White h as worries about proving full
compensation, though Black had his
ch ances in this game as well.
15 .xg7 "cS+ 16 �h1 0-0-0 17 "f6 hS
18 h3 �b8
1 8.....d6 ! ? with a structural advan­
tage in the endgame.
19 :f4?1 i..c 8 20 l:lh4?
1 81
Th e Dark Kn ight System

if Bl ack wants to win ? Pokazanjev's

move is very risky - in fact, it's objec­
tively poor, but sometimes it's worth
taking such risks. Perhaps Bl ack should
h ave deviated earlier: 10 ... 0-0 is not a
bad move, and avoids the present di­
lemma; 10 ... 'Wh4 is also reasonable ac­
cording to Houd;n;.

20... l:thg8?
Black is nearly wmnmg with
20 ... llJxd3 2 1 cxd3 'iWf2 22 llJf4 :hg8,
because White's queen, knight, and
king's rook are stuck in a very un­
healthy co-dependent relationship.
21 :f1 J:td6 22 .e7 "a s?! 23 :xhS :d7
24 .f6 :td6 2S 'ife7 :d7 26 'ii'f6 l:Id6 27
'iII e7 :d7 Vz-Vz 12 ... .td7?! 13 llJa4?!
Instead of repeating moves, Black Out of book, White makes a poor
was much better after 27 ...llJxd3 28 choice. As we will see, Black does not
cxd3 (or 2 8 'ilxf7 llJf2+!) 2 8 ... 'lixa2 29 even have to allow his bishop to be
llJf4 :Lxd3 ! (and not 30 llJxd3 ? ? 'ilxg 2 forced off the strong diagonal .
mate).

Game 58
B.Sultimov-N.Pokazanjev
R u s s i a 2007

1 e4 eS 2 llJf3 llJc6 3 d4 exd4 4 llJxd4

.i.cs S llJxc6 'iff6 6 'ifd2 dxc6 7 llJC3
.td4 8 .td3 llJe7 9 0-0 llJg6 10 �h1 llJes
11 f4 llJg4 12 'ii'e1
White h as certainly lost the theo- 13 ... a6 14 c3 .ta7 1S b4 'iVe7 16 :b1
retical battle since 12 ... llJxh 2 13 'ittxh 2 0-0-0
1i'h6+ is a draw by perpetual. But what Bl ack is not fooled by White's

1 82

queenside demonstration, which
mainly bark and has little bite.
17 'iVg3 h5 18 h3? ..tf5!!
19 ttJC5
Th e bishop shoul d not be taken: 19
exfs ? ttJf2+! 20 :Xf2 'iWel+ 21 ..tfl (or
21 �h 2 ..txf2 an d 22 ... 'iVg l mate)
21 . .. ..txf2 22 ..td2 'iixbl 23 "xf2 l:txd2
and wins - though White is just about
lost anyway.
19 ... ..txc5 20 bXC5 ..txe4 21 ..txe4 'iVxe4
22 :b4 'it'e2 23 .f3
Th e en dgame will bring White little
relief - he cannot possibly hol d onto his
six isol ated pawns.
23 'iVxf3 24 gxf3 ttJf6 25 f5 l:td5 26 ..te3
•..
Illustrative Games
is lhf5 27 ..td4 ttJd5 28 :b2 :116 29 c4 ttJf4
30 :fb1 ttJe6 31 ..tgl :xt3 32 l:txb7
lhh3+ 33 ..th2 l:d3 34 l:ta7 �f6 35 I:txa6
:f2 36 a4 :th3 37 :a8+ �d7 38 �gl
:hxh2 39 :el l:tfg2+ 40 �fl ttJf4 0-1
Dea ling roughly with 7 i..f4?!
Game 59
D.Von Wa ntoch
Rekowski-J.Peric
Yugoslav J u n ior C h ' s h i ps,
Tivat 2001
1 e4 e5 2 ttJf3 ttJc6 3 d4 exd4 4 ..tC4 ttJf6
5 e5 ttJg4 6 0-0 ..te7 7 ..tf4 g5! 8 ..tg3
h5! 9 h3 h4!
10 ..th2?!
10 hxg4 hxg 3 1 1 fxg 3 d6 is rel atively
best, though still with advantage to
Black
10 ... ttJxh2 11 �xh2 d6 12 ..tb5 dxe5 13
ttJxe5 'lWd6 14 f4 ..te6 15 C4? gxf4 16
.f3 �d8?! [16 ... �f8!] 17 ttJxc6+ bxc6 18
..txc6 :b8 19 b3 :g8 20 ttJd2 l:tg3 21 cS
'lWxC5 22 Itfcl :Xf3 23 ..txC5 :xh3+ 24
gxh3 ..txC5 25 ttJe4 :b6 26 ..ta4 ..td6 27
1 83
Th e Dark Knight System
Adl .i.e5 28 lbg5 �e7 29 lbf3 f6 30
lbxd4 :d6 31 lbc6+ �f8 32 :xd6 .i.xd6
33 lbd4 f3+ 34 Whl .i.d5 0-1
Bl ack did not have a high rating, but
sh e produced the model game.
Game 60
J.Becerra Rivero-A.Miles
Andorra 199 5
1 e4 lbc6 2 lbf3
Like most high-rated players, White
tries to narrow the knowledge gap with
thi s second move. Instead, 2 d4 eS (or
in some cases 2 . . dS) fights in Black's
. ness dark-square torture yet again.
home territory. Of course, there is noth­
ing stopping Black from also knowing 2
lbf3 d6 positions well.
2 ... d6 3 d4 lbf6 4 c3 g6 5 .i.d3 .i.g7 6
lba3
Becerra Rivero is the only one to
h ave tried this move h ere. Although
decentralizing, it has the merit of not
interfering with White's other pieces.
6 ...0-0 7 0-0 e5 8 dxe5?! lbxe5 9 lbxe5
dxe5
1 84
White's decision to exchange in the
centre is not consistent with his previ­
ous pl ay. Th e c3-pawn is intended to
shorten the diagonal for Black's ft­
anchettoed bishop and/or hol d the d4-
point, neither of which are meaningful
anymore. White's d3-bishop is also
misplaced, obstructing the open d-file.
He no longer has any trace of an ad­
vantage. Perh aps White is trying to
make a draw, in which case he has cho­
sen the wrong opponent.
10 'iVe2 'ii e7 ll .i.g5 h6 12 .i.xf6?!
This small inaccuracy leads to
White's future problems. We will wit­
12 ...'ilVxf6 13 .i.c4 .i.d7 14 :adl :ad8
15 :d3 h5 16 'iVe3 a6 17 :fdl .i.g4!1 18
f3 .i.c8
Miles's bishop manoeuvre softens
up the dark squares for his unopposed
dark bishop. Th ere is certainly nothing
concrete yet, but it is getting to where
White needs to be careful .
19 .i.b3 �d3 20 :xd3 h4 21 h3 'iIIe 7 22
lbC4 Wh7! 2 3 f4?!
This game reminds me of th e old

story of the ol d l ady who swallowed a
fly, and a spider to catch the fly, and a
bird to catch the spider, etc. White pre­
vents ... .i.h6 for now, but the new prob­
lems are bigger than the ol d ones.
23 ... exf4 24 "xf4 b5 25 ttJe3 i.e5
26 .f3 [26 ttJdS !] 26 .....g5 27 lId1 �g7
28 lIdS 'iVg3 29 ttJg4 'Wxf3 30 gxf3 i..f4
31 cj;g2 :e8 32 :d1 f5 3 3 ttJf2 i.g3 34
ttJh1 i.d6 35 ttJf2 �f6 36 i.. d 5 :h8 37
ttJd3 :h5 38 f4 fxe4 39 i.xe4 l:th8 40
i..f3 :e8 41 i.g4?
In principle, this is a favourable
trade for White, who relieves Black of
th e bishop pair, but th e devil is in the
details.
41... i.xg4 42 hxg4 :e3 43 �f1 i.xf4 44
ttJb4 wg5 45 ttJd5 :f3+ 46 �e2 :g3 47
J:td4 i.d6 0-1
Game 61
R.Zelcic-A.Miles
P u l a 1994
1 e4 ttJc6 2 ttJf3 d6 3 d4 ttJf6 4 c3 g6 5
i.. b 5 i.d7
Illustra tive Games
S ...a6 is also fine, or even S ....tg7 6
dS a6 7 i.a4 bS 8 dxc6 bxa4 9 'iWxa4 0-0
with compensation.
6 -.e2 i.. g7 7 0-0 0-0 8 lld1 'iWe8 9 c4 e5
10 dxe5 ttJxe5 11 i..x d7?1 ttJxf3+ 12
'6'xf3 ttJxd7
Black is already better.
13 'iWe2? f5 14 ttJd2 ttJc5 15 l:e1 "f7
[ls ...'iWa4!] 16 exf5 :ae8 17 fxg6??
hxg6?
Overlooking 17 ... '6'xg 6 18 '6'fl ':xel
19 .xel i.d4, which wins immediately.
18 'Wf1 :xe1 19 'Wxe1 ttJd3 20 'iff1
"xf2+ 21 .xf2 :xf2 22 ttJf3 :C2 23
i.e3 ttJxb2 24 llb1 ttJxC4 25 i.xa7 b6 26
i.. b8 ttJe3 27 ttJe1
1 85
Th e Dark Knight System
27 ...:e2?
Miles puts the game in jeopardy
with this mistake. The c-pawn is Black's
most valuable.
28 iLxC7 iLc3 29 lLlf3 l:txg2+ 30 �hl
:f2 31 lLlg5 iLe5 32 iLxb6 l:txh 2+ 33
�gl l1e2 34 lLlf3 iLf4 35 l:.el l:txel+ 36
lLlxel lLlC4 37 iLd4 Wf7 38 Wf2 We6 39
lLlg2 iLd2 40 �f3 g5 41 We4 d5+ 42
�d3 iLa5 43 lLle3 lLlxe3 44 iLxe3 g4 45
iLf4 iLb6 46 iLg3 �d7 47 iLf4 Wc6 48
iLg3 iLc5 49 a4 iLd6 50 iLel? [50 iLh4!]
50 ... g3 51 �e2 g2 0-1
Game 62
L.Rozman-J.Schuyler
Wa s h i ngton 2012
1 e 4 lLlc6 2 lLlf3 d 6 3 d 4 lLlf6 4 c3 g6 5
iLd3 iLg7 6 0-0 0-0 7 h3 e5 8 lLla3 d5!
White has the slow c2-c3 and h 3-h 3
under his belt, and the decentralizing
lLla3, so I felt it was time to hit back in
the centre.
9 iLg5?! dxe4 10 iLxe4 exd4 11 iLxc6?!
White was hoping to damage my
186
structure instead of accepting th e iso­
l ated d-pawn. This would have been
fine for Black, but I saw something I
liked better.
11 ... dxc3! 12 'ifa4?!
The critical tries involved saving th e
bishop, but since Black was getting
three pawns for it, including the dan­
gerous b2-pawn, I was willing to take
my chances (an d White was not).
12 ...bxc6 13 'ifxc6?
Either 13 bXC3 or 13 ':ad1 was cor­
rect.
13 ... cxb2 14 :adl iLd7 15 'ifC2 'iVc8?!
I was eager to break the pins, but
analysis shows I coul d h ave safely
made it harder for White to recover th e
b-pawn with 15 ...l1b8 ! .
1 6 'iVxb2 lLle4 17 .-b4 :te8
Incidentally threatening 18 ... iLxh 3
19 gxh 3 'ifxh 3 , which will win the piece
back on the kingside.
18 �h2?
Not a working defence, though nei­
th er one of us noticed the obvious refu­
tation : 18 ... iLxh 3 ! 19 gxh3 'iif5, again
recovering the piece.

18 ... .i.f81 19 .-a5 i.. d 6+ 20 Wgl .i.xh3
21 gxh3 'ir'xh3 22 l:td3 ttJg3 1 23 fxg3
'ii'x g3+ 24 Whl l:te2 0-1
Do not try this at home,
or even when away from home
Game 63
A.Za pata-A.Miles
Mata nzas 199 5
1 e4 ttJc6 2 ttJf3 d6 3 d4 ttJf6 4 ttJbd2 g6
5 c3 i.g7 6 i.. b 5 0-0 7 0-0 .i.d7
7 ... a6 is better, trying to pick up the
bishop pair. Black's bishop may prefer
to deploy on b7 or a6.
8 :el 'ii' e8 9 i.. f l e5 10 h3 .-d8 11 a4
:e8 12 as a6 13 d5 ttJe7 14 c4 l:tf8
While each of Black's moves had its
purpose, the overall impression is very
poor. This is a bad King's Indian.
15 cS ttJe8 16 ttJc4 ttJc8 17 'ii' b 3 .i.b5 18
i.. e3 ltb8 19 :acl ttJf6 20 ttJfd2 ttJd7 21
cxd6 cxd6 22 .-a3 ttJf6 23 ttJb6 i..xfl 24
'1ttxfl ttJa7 25 .-b4 ttJb5 26 ttJbl ttJe8 27
ttJc3 ttJxc3 28 :xc3 ttJC7 29 ttJC4
Without making any obvious mis-
Illustrative Games
takes, most of White's pressure has
dissipated.
29 ttJb5 30 :ccl f5 31 ttJa3 ttJxa3 32
••.
'ir'xa3 .-d7 3 3 .i.b6 :bc8 34 'ii' b 3 i.. h 6
35 Iial i.. g 5 36 :a4?1
By now, White is actually worse.
36 ..i.d8 37 l:tb4 fxe4 38 i..x d8?1 .-xd8
•.
39 :xb7 �xa5 40 'ii'd l? [40 'iVb4]
40. . :.c5 41 'ifd2 :f4?
41 ... l:txf2+! 42 �xf2 1U8 wins.
42 Wg l 'ii'd4 43 :dl 'ii'xd2 44 :xd2 as
45 :b6 :cl+ 4 6 '1tt h 2 :f6 47 l:tb5 :f1 48
�g3 h5 49 l:txa5 :f4 50 f3?? exf3?
Here 50 ... e3! 51 :C2 :h 1 52 h4 g 5.
51 gxf3 :lxf3+ 52 �g2 h4 53 l:ta6 :f6
54 :a4 :3f4 55 b4 �g7 56 l:tb2 'lith6 57
%la7 :f3 58 b5? [58 '1tth 2] 58... :g 3+ 59
�h2 :H3 60 :a4 :xh3+ 61 �gl l:thg3+
62 11g2 :xg2+ 63 �xg2 :b3 0-1
Try this at home
Game 64
F.Nijboe r-A.Miles
Li n a re s 1 9 9 5
1 e4 ttJc6 2 ttJf3 d6 3 d4 ttJf6 4 ttJbd2 g6
187
Th e Dark Kn ig h t System
5 .i.b5 a6 6 .i.xc6+ bxc6
As we will see, this position is much
easier to play than ... .i.d7 of the previ­
ous g ame. Miles's methods of play are
worth remembering and repeating.
7 0-0 .i.g7 8 e5 tbd5 9 c 4 tbb6 10 'li'e2
0-0 11 :ld1 as 12 tbf1 .i.a6 [12 .. .f6 !?] 13
b3 'ti'd7 14 tbg5 .i.c8 15 f4 a4 16 .i.e3
f6! 17 tbf3 'ile8 18 h3 fxe5 19 fxe5 h6
20 :lac1 axb3 21 axb3 g5
Black has been preparin g ....i.fS,
though after White's next the plan
changes.
22 d5 cxd5 23 cxd5 ..ta6 24 'ild2 .i.xf1
25 :xf1 'it'b5 26 :fd1?! 'ilxb3 27 l:1xC7?
tbC4 28 'ii'd 3 'ii'x d3 29 :xd3 :ta1+ 30
�f2 tbxe5 31 l:[d2 :1f7 32 :ta7? llb1?
[32 ... l::th 1] 33 lla4 �h7 34 �g3 :b3 3 5
':e4 :ta3 36 h4?! tbxf3 37 gxf3 gxh4+
38 <iitg4 .i.f6 39 f4 llg7+ 40 <iitf5 h3 41
:th2 ':g3 42 ..tC5? dXC5 0-1
Game 65
E.Sveshnikov-T.Gelashvi li
C a p pe l le l a G ra nd e 2009
1 88
1 e4 tbc6 2 tbf3 d6 3 d4 tbf6 4 tbc3 g6 5
..tb5 a6 6 ..txc6+ bxc6 7 0-0 .i.g7 8 'ti'e2
0-0 9 Iid1 ..tg4
Th is makes it easier to fight for the
central dark squares and advances a
clear plan. Nonetheless, I prefer 9 ... aS
and 10 ... ..ta6.
10 h3 ..txf3 11 'ti'xf3 tbd7 12 .i.e3 e5 13
'ti'e2 lle8 14 'ti'C4?!
White could have bottled up Black's
bishop and rook with 14 dS!.
14. exd4 15 i..xd4 ..txd4 16 'ii'x d4 'ti'g5
. •
17 'iVd2 'ii'h4 18 lle1 tbc5 19 f3 :te5 20
'ii'f2 'ile7 21 tbe2 tbe6 22 tbd4 tbxd4 23
'iixd4 as 24 'ii'c 3 1:[c5 25 �d2 :te8 26 a4
'ti'f6 27 c3 llb8 28 :te2 lle5 29 l:[ae1 g5
30 'ti'e3 cS 31 'ii'd 3 'ti'e6 32 'ti'a6 c4 33
'i¥a7 :c8 34 'i¥d4 l:e8 35 �f1 f6 36 �d2
�f7 37 <iitg1 We7 38 �h2 Wd7 39 'i¥d4
'WIf7 40 'ii'f2 h5 41 <iitg1 cS 42 :td2 f5 43
lled1 ll8e6 44 'i¥e2 fxe4 45 'ii'xC4
Bl ack has been pressing h ard for the
win, but without anything to work
with. Now he is in danger.
45 ...'ti'f4?! 46 'ii'a 6?!
46 'i!fbS+ �e7 47 'ifb7+ wf8 48 fxe4
is stronger.
 Illustrative Ga mes

46 ... exf3 47 ':xd6+ �e7 48 -.a7+ �f6 i..c4 thc6 6 d5 thb8 7 h3 0-0 8 0-0 c6
49 :xe6+ l:txe6 SO 'ii'x C5 �g6 51 'ii'd4 [8 ... es!] 9 a4! as 10 :e1 thfd71
"iff5 52 g4 hxg4 53 hxg4 'iYf7? 54 l:lf1?!
And here S4 b4!.
54...:f6 5 5 Wf2?! 'ikb3 56 'iVd3+ �g7 57
:d1 'iYxb2+ 58 l:ld2 'iYb8 59 'i'd7+ �h6
60 �1 'ii'b1+ 61 :d1 -.e4 62 "d2??
After thi s move White is losing. 62
:d2 held th e balance.
62 .. .f2 63 'iYd5 'ii'xg4 64 'ii'h 1+ wg6
[64 ... �g7! ] 65 :d4

BlacK clamps down on the darK

squares, incidentally stopping White
from considering the e4-eS break
11 i..e 3 tha6 12 i..d4 [12 i.. xa6 !]
12...thb4 13 i..xg7 wxg7 14 the2 'ikC7 15
c3 cxd5 16 i..xd5 thc6 17 thed4 thf6?1
WeaKening the darK square control.
White now h as 18 i..xc6 bxc6 19 eS
with a pleasant advantage.
65 .. :iVc8? 18 'iYb3?! ttJe5 ! 19 thxe5 dxe5 20 thf3
6s ...'it'e6 (or 6s ... 'iVfS) 66 'iYe4+ :a6 21 :ac1 thh5 22 i..c 4?1 lif6 23
(forced) 66 ... 'iWxe4 67 :xe4 Wh S wins, thd2?! 'lWc5 24 rLe3 :d8 25 thf1?! thf4
since if 68 l:Le2 then 68 ...�g4 69 :1xf2 26 _b5 "c7 27 "b3 h5 28 -.a2 :fd6 29
:xf2 + 70 Wxf2 Wf4 etc. 'ir'b3 i.. d 7 30 i.. b 5 i.. e 6 31 -.C2 h4 32
66 'ii'e4+ �h6 67 'ii'e 3 'ii'f 5 68 'iWd3? :ee1 "c8 33 l:te3 g5 34 f3?1 thh5
'ir'xd3+ 69 :xd3 :g6?? [34.. :ifcS!] 35 ':ce1 [3S :eel] 3 5 .. :iic 5
69 ...�h S still wins. 36 �h2 wf8 37 :d3?1 :xd3 38 i..x d3
70 c4 Yz-Yz i..x h3! 0-1

Game 66 Game 6 7
N. Pra z ni k -A. Be liavsky K.Nemcova-F .Olafsson
Bled 199 9 M a ri a n s ke La z n e 2008

1 e 4 d 6 2 d 4 thf6 3 thC3 g 6 4 thf3 i.. g7 5 1 e4 d6 2 d4 g6 3 thc3 i.. g 7 4 i..e 3 thc6 5

1 89
Th e Dark Knight System
d5 It)b8 6 'ii'd 2 c6 7 lt)f3 It)f6 8 h3 b5?!
This is a great idea if it works out
tactically, but it doesn't. 9 dxc6 b4 10
i..b s is a problem for Bl ack. So let us
im agine instead that the game went
8 ... 0-0 9 i.. d3 bS 10 a3 a6 11 dxc6 etc.
9 a3 a6 10 dxc6 It)xc6 11 �d3 0-0 12
0-0 i.. b 7
Black's opening h as been extremely
successful - White's game has a
cramped feel in spite of her central
space advantage.
13 :tfel d5?!
This "freeing" move is quite unnec­
essary. Black has many improvements
he can make (such as ... ':c8, ... It)d7,
... 'iWas, or even ... e7-e6) before he needs
to consider taking action. It is White
who will be left without anything con­
structive to do.
14 �h6?
I've never heard anyone say that it's
a good idea to react to central action
with play on the wings.
14 dxe4 15 �xg7 �xg7 16 lt)xe4 lt)d4
.•.
17 It)xd4 'iYxd4 18 "c3 'ii'xc3 19 It)xc3
e6
190
White i s slightly worse due t o her
very restricted bishop, though she
solves this problem quickly.
20 It)e4 l::tfd8 21 lt)c5 �c6 22 a4 bxa4 23
It)xa4 .i.b5 24 lt) c 5 :ac8 25 b4 :d4 26
.i.xb5 axb5 27 c3 :C4 28 :a7 �g8 29
:b7?! [ 2 9 :e3 It)ds 30 :f3] 29...:xc3
30 l::tx b5 It)d5 31 lIb7 It)xb4 32 :xb4
:8xc5
A theoretical draw, but Black has
some practical chances.
33 :e2 :5c4 34 l:tbb2 g51
This stops White from constructing
the ideal defensive pawn formation for
this endgame: f2/g3/h4. The point of
th e form ation is th at Bl ack cannot cre­
ate a passed pawn without trading off
all th e rest of the pawns.
35 �h2 �g7 36 Itb5 �g6 37 l:tbb2 h5
38 l:[a2 h4 39 i:tel :c5 40 �gl <MS 41
<ittfl e5 42 f3 :3C4 43 :ee2 <it>f4 44 �f2
f5 45 :a3 l::t b 5 46 l::tae3?
The endgame h as been getting
harder to defend. Now, White is lost.
46 ... l::tCl 47 :el l:tb2+ 48 l:t3e2 :cc2 49
:XC2 :tXC2+ 50 �1 l:.b2 51 <it>gl �g3
0-1

Game 68
J.Hjartarson-F .Olafsson
Reykjavik 19 9 5
1 e4 g6 2 d4 d6 3 liJc3 i..g7 4 i.. e3 liJc6 5
d5 liJb8 6 'ifd2 c6 7 liJf3 liJf6 8 h3 0-0 9
..te2 b5! 10 a3
White has nothing better than this
l ame try. Bl ack h as equalized already.
10 a6 11 0-0 i.. b 7 12 l:.fel liJbd7 13
•••
dxc6 i..x c6 14 i..d3 i.. b7 15 i..h6 :c8 16
l:.adl 'ii'b6 17 ..txg7 Wxg7 18 l:.e3 liJe5
19 liJxe5 dxe5 20 a:e2 :fd8
By now Black is a little better, but
the players soon decided to call it a day.
21 1fe3 'ii'd4 22 :del liJh5 Yz-Yz
Game 69
Y.Gruenfeld-I.Smiri n
Isra e l i Tea m
Cha m pi o n s h i p 1997
1 e4 d6 2 d4 liJf6 3 liJc3 g6 4 liJf3 i..g7 5
i..e2 0-0 6 0-0 liJc6 7 d5 liJb8 8 h3 e5 9
dxe6 i..xe6 10 ..tg5 h6 11 i.. e 3 liJc6 12
Illustrative Ga mes
'ii'd 2 Wh7 13 :adl :e8 14 :fel i.. d7 15
i..c 4 i.. e 6 16 i..fl a6
17 1fCl
Very soon Black has all the play he
needs. White can make him work
harder with 17 a4.
17 1i'c8 18 a3 b5 19 liJd4 liJxd4 20
...
i..xd4 'fWb7 21 'ii'f4 liJd7 22 i.. x g7 wxg7
23 'ii'd 2 'ii' b6 24 :e3 l:.ad8 25 g3 liJf6 26
i.. g 2 as 27 b3 :!b8 28 a4 bxa4?! [28...b4]
29 liJxa4 'ii'b4 30 liJC3 liJd7 31 Wh2 liJc5
32 f4 f6?! 33 1fe2?! [33 fS !] 33 ... i..f7 34
liJd5 i..xd5 35 exd5 l:!xe3 36 'it'xe3 :f8
37 f5 gs 38 'ii'e2 liJd7 39 :al liJe5 40
J:[a4 'ii'c 3 41 J:[e4 :b8 42 'ii'h 5 :b4 43
:e2 a4 44 'ii'e8 axb3 Yz-Yz
Game 70
B.Chata lbashev-M.Popchev
Caca k 1991
1 d4 d6 2 e4 liJf6 3 liJc3 g6 4 liJf3 i..g 7 5
i..e2 0-0 6 0-0 ..tg4 7 ..te3 liJc6 8 d 5
liJb8 9 liJd4 i..xe2 10 'ii'xe2 cS 1 1 liJf3
'ii'b6 12 l:tabl 'iVa6
Chernin points out th at this queen
191
Th e Dark Kn ight System
manoeuvre is both typical and strong.
13 "d2 llJg4 14 �g5 ::te8 15 "f4 llJe5
16 llJxe5 �xe5 17 'ifh4 "iVC4
18 :be1
White gains nothing by grabbing
th e pawn : 18 �xe7 llJd7 19 f4 i..d4+ 20
'it>h l �xc3 21 bXC3 'iVxe4 is equal.
18 ... �xc3 19 bxc3 f6 20 �h6 llJd7 21
'ifg4 e6 22 f4 �ad8 23 dxe6 'i¥xe6 24
'ii' g3 'iWxa2 25 f5 'ii'f7 26 fxg6 hxg6 27
'ii'x d6 �h7 28 'i¥f4?!
It is tim e to extricate the bishop.
28 l:[e5 29 �g5 lU8 30 'iYh4+ wg8 31
.••
�f4 :e6 32 l:[f3 g5 33 'ii'g4 :fe8 34
�g3 llJe5 35 �xe5 :xe5 36 h4?! 'iWe6
37 'ii'h 5 'it>g7 38 l:Ief1??
Oops! Bl ack had an advantage, but
this throws the game.
38 J�h8 39 'ii'x h8+ �xh8 40 �xf6 'ii'C4
••
0-1
Game 71
R.Ziatdi nov-A.Chernin
N ew York Open 1998
1 e4 g6 2 d 4 �g7 3 llJf3 d6 4 �e2 llJf6 5
192
llJc3 0-0 6 0-0 �g4 7 �e3 llJc6 8 d5 llJb8
9 llJg5 �xe2 10 'i¥xe2 c6 11 �ad1 'ii'a 5
12 f4 'ii'a6 13 .f3 llJbd7 14 'iVh3 h6! 15
llJf3 h5! 16 �d4?! cS 17 .i.xf6 liJxf6 18
'ii'h 4?1 b5
White h asn't gotten anything done
on th e kingside, but he has serious
problem s on the queen's wing.
19 l1Je2?! 'iVa4 20 llJg3 'ii'x c2 21 e5 dxe5
22 fxe5 llJg4 23 ::td2 .C4 24 'iVxe7 :ae8
25 d6?!
A desperate try for complications
which fail s after Chernin's accurate
treatm ent.
25 ... �h6! 26 b3 'i¥b4 27 e6 ':xe7 28
dxe7 �xd2 29 exf7+ 'itxf7 30 exf8'ii'+
�xf8 31 llJxd2+ 'itg7 32 llJge4 'ii'd4+ 33
'ith1 c 4 34 bxc4 bXC4 35 h3 llJe3 36 :f4
c3 3 7 llJf3 'ii'd 1+ 38 Wh2 llJf1+ 39 �g1
llJd2+ 40 'itf2 llJxe4+ 41 l:[xe4 c2 0-1
Game 72
So. Polga r-J.Fries Nielsen
R i m a vs ka Sobota 1991
1 e4 g6 2 d4 �g7 3 llJc3 llJc6 4 llJf3

If 4 i..e 3, Fries Niel sen intended
4... dS, an interesting and surprising
move that I don't believe in.
4... d6 5 h3 tDf6 6 i..gs 0-0 7 .d2 dS
In th e next game, Bl ack tries the en­
terprising 7 ... a6.
8 exd S tDxds 9 i.h6 tDxc3 10 i..xg7
�xg7 11 Vi'xC3 'iVds 12 O-O-O!
12 i.C4? ! 'ifas is in effective.
12 ... i.e6 13 b3 i.. f S 14 tDh4 i.d7 15
:d2?1 Wg8 16 'ife3 'ifd6 17 c3 as 18
tDf3 'iVa3+
19 �d1?!
Asking for trouble. 19 l:tb2 was cor­
rect.
19 ...a4 20 -.h6 f6 21 i.c4+ e6 22 dS
tDe7?
Since White h as counterplay in ei­
ther case, it does not help to give up
the pawn. Hence 22 ... ttJd8 !'
23 dxe6 i.c6 2 4 ttJd4 axb3 25 ttJxc6
bxc6 26 i.xb3 l:tab8 27 :te1?!
The way to extricate th e king is 27
�e2 ! l:txb3 28 axb3 'iixb3 29 l:th d1.
27 .. ':xb3 28 axb3 'it'xb3+ 29 We2 -'C4+
30 l:td3?
Walking into a dangerous pin. 3 0
Illustrative Ga mes
�3 ! is still approxim ately equal.
30 ... tDds 31 g 3?
My computer tells me th at White is
okay after 31 e7!! lte8 32 '1tt d 2 and th at
an other moves lose brutally. I won't try
to explain, especially since I don't un­
derstand it myself.
31 l:ta8 32 'ii'd 2 :a2 3 3 :b1 tDxc3+ 34
•.•
'1ite1 'ii'e4+ 0-1
3 S ... 'iVh 1 mate foll ows.
Game 73
M. Yilmazyerli-D.Arutinian
I sta n b u l 2007
1 e 4 ttJc6 2 tDf3 d6 3 d 4 tDf6 4 tD C 3 g6 5
h3 i..g7 6 i.gs 0-0 7 "d2 a6 8 0-0-0 bS
9 a3 :b8 10 i.h6 b4 11 axb4 ttJxb4 12
i.xg7 wxg7 13 eS ttJfdS 14 ttJxdS ttJxdS
15 h4 h S 16 tDgS?!
The knight can stay here as long as
it wants, but it does not h ave impor­
tant targets in Black's position.
16 ... c6 17 i..c 4 'iVb6 18 b3 i.fS 19 exd6
exd6 20 g4 i.xg4 21 l:tdg1?1 i..fS 22
tDh3 'ifb4 23 'it'xb4 lixb4 24 :g3 as 25
193
Th e Dark Kn ight System
.i.xdS cxdS 26 �d1 :c8 27 l:.d2 a4 28
�b2 axb3 29 cxb3 �cc4 30 ttJf4 �xd4
31 l:.xd4 l:.xd4 32 ttJxh 5+ �f8 3 3 ttJf6
:d1 34 hS gxhS 3 5 ttJxhS d4 36 ligS?!
l:.d2+ 3 7 �C1 l:.xf2 3 8 ttJg3 .i.g6 39 :dS
:g2 40 ttJhS �f2 41 ttJg 3 :f3 42 ttJe2 d3
43 t'LlC3 �e7 44 b4 �e6 45 lid4 :f2 46
ttJbS �c2+ 47 Wd1 .i.hS+ 48 We1 :e2+
49 Wf1 d2 0-1
Gam e 74
N.Ryba-J.Schuyler
Wa s h i n gton 2012
1 e4 ttJc6 2 ttJf3 d6 3 d4 ttJf6 4 t'Llc3 g6 5
.i.e2 .i.g7 6 0-0 0-0 7 h3 a6!?
Against a lower-rated opponent
there is some concern th at the n atural
move, 7 ... es, wm lead to a position with
too few winning ch ances, in which case
7 ... a6 is often a good alternative.
8 a4 eS
It is nice for me to have the bS­
square covered and his b4-square
weak. Therefore, if 9 dxes, Bl ack will
play 9 ... dxes, keeping th e extra pair of
knig hts on th e board.
9 .i.e3?!
It is almost always a bad idea for
White to try to preserve th e central
tension in the Dark Knight Pirc. This is
especially true when he has played h2-
h3. Black is already at least equal.
9 ... exd4 10 ttJxd4 %le8
One Pirc author recommends
lo ... .i.d7 first in this type of position,
but I prefer to h ave my c-pawns dou-
1 94
bled! The h alf-open b-file co-ordinates
well with the dark-squared bishop, and
it is nice to have the ds-square securely
guarded.
11 ttJxc6 bxc6 12 .i.f3 ttJd7 13 'iYd2 :b8
14 :ab1 'ti'f6 15 .i.e2 'iYes 16 l:.fd1 'ifas
It is often defl ating to computer­
check your own games. This queen
manoeuvre, which I was proud of dur­
ing the game, accomplishes less th an
nothing after 17 b4! :xb4?! 18 ttJdS !
:xa4 19 'iYxas :xas 2 0 ttJxC7 l:txe4 2 1
':xd6 with an edge for White.
17 .i.d4 .i.xd4 18 'ti'xd4 J:tb4 19 'ti'd2
ttJcS
This will not get anything done ei­
ther. Bl ack is not even threatening
20 ...ttJxa4 because of 21 :tal an d 22 b3.
20 b3 ttJxe4 21 ttJxe4 l:.exe4 22 .i.f3 l:[es
23 .i.xc6 .i.fS 24 .i.f3 'ifcs 25 c3 l:.b8 26
b4 'iYC4 27 J:tb2 l:.be8 28 'iWd4 'iWe6 29
Wh2 .i.e4!
It isn't much, but I will h ave a little
something to play for after stripping
White's king of its best defender. Who
knows? The e-file might even be worth
something.

30 i.xe4 l:tXe4 31 'ii'd 3 as 32 .dS 1Vf6
33 'iVxas 'iVxc3 34 l::tbb1 cS?!
This is another move that seemed
very strong to me when I played it, but
White can equalize starting with 3 5
'ifb 6 ! (the text move i s not bad, eith er).
34_ .:e2 is a better try. But by now we
were running low on time to reach
move 40.
35 'iYbS cxb4 36 :txd6 :t8es 37 .b8+
�g7 38 :d8 :fS 39 :g8+ �h6 40 'ii'f8+
<it>g s 41 'iVd8+ f6 42 l:te8 l:td4 43 'iVe7?
Houd;n; finds 43 f4+! l:tfxf4 44 lIe5+
with equality. White is lost after any
other move.
43 ...l1xf2 44 :te1 l:td3??
Illustrative Games
Threatening 45 ...:xh 3+ 46 �g l
llxg 2+ and wins! Somehow I over­
looked both of White's defences.
45 'ii'x h7
Or 45 h4+! �xh4 46 'iYxh 7+ �g 5 47
::t1e4, when Bl ack h as to find
47 ... l:txg2+! 48 �xg 2 l1d2+ 49 l::[ e 2 l:td4!
and draws (apparently).
4S ...:tdd2
This is the move th at would h ave
won last turn.
46 :g1??
Snatching defeat from th e jaws of a
draw. My opponent saw the correct
continuation but misevaluated it: 46
h4+ �f5 47 :8e5+ 'iix e5 (47 .. .fxe5 48
1i'f7+ �g4 49 1:.e4+ l:lf4 50 'ii'xg6+ �xh4
51 'iVh6+ draws) 48 ':'xe5 fxe5 49 'ii'f7+
'iite4 50 'iVb7+ is no worse for White.
46 .. :iVc6 47 h4+ �fS 0-1
Not the most beautiful game, but
the opening was a success.
Game 75
N .J h u nj h n uwala-S.Gligoric
Luce r n e Olym p i a d 1982
1 ttJf3 g6 2 e4 d6 3 d4 i.g7 4 ttJC3 ttJf6 5
Jie2 0-0 6 0-0 ttJc6 7 h3 eS 8 dS ttJe7 9
i.e3 c6!
White is unable to keep his centre
intact, nor is he fast enough on the d­
file to inconvenience Black.
10 dxc6 bxc6 11 'ii'd 2 'iliC7 12 :ad1 d S
13 exdS ttJexdS 14 ttJxdS?!
Helpfully completing the opposing
centre. Black starts building his edge.
195
Th e Dark Kn ight System
14... cxd5 15 c3 .i.b7 16 .i.h6 ltJe4 17
'if cl :fe8 18 .i.x g7 .tt xg 7 19 ltJd2 ltJc5
20 ltJb3 ltJe6 21 l:lfel as
22 .i.b5 %:te7?!
Black does not h ave a good reason
to disconnect his rooks. However,
White's next makes it easy to fix the
problem.
23 .i.fl? [23 a4] 23 ...a4 24 ltJal :d8 25
ltJc2 f6 26 ltJb4 'ifb6 27 'ifC2 ttJ C 5 28 a3
l:ted7 29 l:te3 ltJb3 30 %:teel :d6 31 'ife2
d4 3 2 Wb5 'iVxb5 33 .i.xb5 dxc3 34
:xd6 l:lxd6 35 bxc3 ttJC5 36 ite2 l:tdl+
37 �h2 e4 38 1i>g3 f5 39 h4 <tW6 40
.i.e8? h6? [40 ...:h l!] 41 .i.b5? [41 Wh 2 ! ]
41. .Jlhl 42 f4 g5 43 hxg5+ hxg5 44
fxg5+?! �xg5 45 �f2 f4 46 .i.c6? .i.xc6
0-1
Since after 47 ltJxc6 the reply
47 ... ttJd3 is mate.
Game 76
G.Bastri kov-E.Gel le r
Tash kent 1 9 5 8
1 e4 e5 2 ltJf3 ltJc6 3 ttJC3 d 6 4 h3 g6 5
196
d4 exd4 6 ttJxd4 .i.g7 7 .i.e3 ltJf6 8 'ifd2
0-0 9 0-0-0 l:te8 10 f3 ttJxd4 11 .i.xd4
.i.e6 12 .i.f2 a6 13 �bl b5 14 h4 cS 15
g4?
White can play th e aggressive 15
h s ! ? or the defensive 15 a3 - th e text
move is too slow.
15 ... b4 16 ltJe2?!
16 ......a5?
Black gets a big advantag e with
16 ...ttJxe4! 17 fxe4 'iff6, threatening
mate and White's bishop.
17 ltJCl .i.xg4?! 18 .i.g2?
Black's sacrifice is thematic, but not
quite sound. 18 fxg4 ttJxe4 19 'it'f4
ttJc3+? ! 20 bXC3 bXC3 2 1 i.c4 is defence
and counterattack.
18 ... .i.e6 19 'ifxd6 :tac8 20 ltJb3 'ir'a4 21
.i.h3? .i.f8?
Unnecessarily removing the bishop
from th e main diagonal and misevalu­
ating the most direct continuation:
2 1 ... i.xh 3 2 2 lixh 3 c4 23 ttJcs 'iVbs
threatens 24 ...:c6.
22 'ifh2?! .i.xh3 23 '6'xh3 c4 24 ltJcl c3
25 .i.d4 cxb2 26 ltJb3 .i. g7 27 .i.xb2 'ifc6
28 %:th2 lIed8 29 :hd�? l:txd2 30 l:txd2

30 ... �xe4! 31 fXe4 .i.xb2 32 h5 [32
<itxb2 ? ! 'ifC3+] 32 Ji'c3 3 3 'ii'x c3 .i.xc3
•.
34 :dl c;t>g7 35 :d6 :c4 0-1
Game 77
D.Saduakassova­
Art.Minasian
D u b a i 2011
1 e4 �c6 2 �f3 d6 3 �C3 �f6 4 d4 g6 5 h3
.i.g7 6 .i.e3 0-0 7 -.d2 e5 8 d5 �e7 9 .i.h6
White trades her best minor piece
for Bl ack's obstructed bishop in the
hope of weakening his king position.
9 .i.d71
•.•
Illustrative Ga mes
Preparing ...c7-c6 an d/or ...b7-bS.
10 .i.d3 c6
lo ...bS 11 a3 as and 12 ...b4 is also
good.
11 dxc6 .i.xc6 12 �xg 7 litx g7 13 0-0
'ilC7 [13 ... b S ! ?] 14 'ile2 :fe8 15 .i.b5 a6
[lS ... dS!] 16 .i.xc6 1Wxc6 17 :adl :ac8
18 �d2?! d5 19 exd 5 �xd5 20 �xd5
�xd5 21 c4 �f4 22 'ii'e4 f5 23 'ii'x c6
lhc6 24 b3 �e2+ 25 �h2 �c3 26 l:tal
1id6 2 7 �bl �xbl 28 l:taxbl l:td2
Rook endg ames are notoriously
drawish, but according to Jesus de l a
Villa i n 100 Endgames You Must Know,
this reputation is unearned. With his
better rooks and king, Black h as his
winning chances.
29 &14 It>f6 30 litgl e4 31 lifel l:ted8 32
c;t>fl c;t>e5 33 cS :C2 34 :edl lld3! 35 b4
:ta3 36 :dcl l:[xcl+ 37 lixcl l:[xa4 38 c6
bxc6 39 :xc6 Wd5 40 :C7 :xb4 41
:xh7 :b6 42 h4 <itt C 4 43 g 3 as 44 l:tc7+
litb5 45 h5 gxh5 46 :f7 c;t>c4 47 :txf5
l:la6 48 :txh5 a4 49 l:th8 a3 50 :c8+
Wd4 51 :d8+ �e5 52 :dl a2 53 :al
litf5 54 lite2 :a3 55 litfl lites 56 �e2
Wd4 57 g4 We5 58 Wd2 <iW4 59 WCl
197
Th e Dark Knight System
c;i(xg4 60 c;i(b2 l:ta8 0-1
Came 78
A.Sa kha rov-A.Adorj a n
Soch i 197 6
1 e4 d6 2 d4 li.Jf6 3 li.JC3 g6 4 li.Jf3 i..g7 5
i..e2 0-0 6 0-0 li.Jc6 7 i..g5 h6 8 i..f4 li.Jg4
9 h3 e5 10 dxe5 li.Jgxe5 11 li.Jxe5 dxe5
12 i..e 3 li.Jd4 13 i.. c 4 'iVh4 14 li.Jd5?!
Not appreciating the danger. By re­
linquishing control over e4, White sub­
jects him self to a powerful attack 14 f3
or 14 .i.d3 is still approximately equal,
but nobody likes to play such m oves.
14 ... c6! 15 li.Jc7 i..x h3!
Now White must defend accurately
to survive, and even then he will be
worse - there is no good way for him to
retain a material advantage.
16 li.Jxa8 'iYxe4
The queen's arrival on this square is
a nightmare for White, who h as far too
many bishops lying around.
17 gXh3 li.Jf3+ 18 c;i(hl li.Jh4+ 19 f3 -.xe3
20 li.JC7 'iWf4 21 'iVd3 'iVg 3 22 "'e2 e4?!
198
[22 ... li.JfS !] 23 �h2 'iVxh2+ 24 c;i(xh2
li.Jxf3+ 25 �hl li.Jd2 26 li.Je8?? [26 li.Je6 ! ]
26 ...li.Jxc4 27 li.Jxg 7 c;i(x g7 28 :fel f S 29
l:adl 1:[f7 30 b3 li.Jes 3 1 :d6 li.Jf3 32
1:tedl l%e7 33 :d7 �f7 34 �g2 f4 35
l:tld6 li.Jh4+ 36 �fl e3 3 7 c 4 e2+ 0-1
Came 79
Z.Bratanov-B.Chatal bashev
B u l ga ri a n C h a m pion sh i p,
Sofia 2 004
1 e4 g6 2 d4 i..g 7 3 li.Jf3 d6 4 li.Jc3 li.Jf6 5
i..e2 0-0 6 0-0 li.Jc6 7 i.. e3 eS 8 dxes
dxeS 9 'ii'x d8 :xd8 10 i..C4
One might think th at White plays
for a win with little risk However,
White (a GM) also lost in M.Hebden­
E.5utovsky, Isle of Man 1999. Black's
next is to prevent th e annoying 11
li.Jg S.
10... h6 [10... li.Je8 ! ?] 11 h3 b6 12 li.Jds
li.Jas! 13 li.Jxf6+ i..xf6 14 i..e2 i.. b7 15 b4
li.Jc6 16 c3 as 17 a3 li.Je7 18 li.Jd2 i..g S!
50 far it's not much for Bl ack, but
trading the worst minor for White's
 Illustrative Games

best is the first step on the road to a where 7 ... es! works a little better than
win. usual and 7... a6 works a little worse
19 i.xgs hxgs 20 :ad1 axb4 21 axb4 than usual.
�g7 22 f3 l:la2 23 lLlC4 l:txd1 24 i.xd1 8 d5! lLlb8 9 a4!
i.a6 25 i.b3 i.xC4 26 i.xa2 i.xa2 27
l:ta1 i.C4 28 l:la7 bS 29 l:1xc7 �f6
Perhaps White should hold this, but
he has a long and thankless defensive
task ah ead of him. Black's task is to
penetrate with his king.
30 l:ia7 �e6 3 1 l:la6+ �d7 32 :a7+ �d6
33 l:Ia6+ lLlc6 34 �f2 �d7 35 �g3 lLld8
36 h4 gxh4+ 37 �xh4 �e7 38 �g3 lLle6
39 �f2 lLlf4 40 g 3 lLle2 41 lia3 �f6 42
'itg 2 �gs
And now White is lost. How did th at Since a2-a4 is norm ally met by ... a7-
happen? as, it stands to reason th at Black's a­
43 �2 fS 44 exfS gxfs 45 'litg2 f4 46 pawn is mispl aced.
gxf4+ �xf4 47 �f2 lLlC1 48 %:ta8 lLld3+ 9 c6
.•.

49 �g2 �e3 50 '1tg3 �d2 51 :a3 �C2 I now believe th at 9 ...b6 is the best
52 '1t g4 '1tb2 53 :a 7 �xc3 54 �fS 'iiixb4 move in the position, intending to fol­
0-1 low with ... C7-c6, ... i.b7, ... lLlbd7. At the
time I was reluctant to try this since
White had not spent a tempo on h2-h 3
Game 80 to prevent ...i.g4.
K. Wang-J.Sch uyler
Wa s h i n gton 2 0 1 2

1 e4 lLlc6 2 lLlf3 d 6 3 d4 lLlf6 4 lLlc3 g6 5

i.e3 i.g7 6 'iVd2 0-0 7 i.e2
White tries to play without h2-h 3 -
an uncommon pl an. I was aware that
7 ... eS was the main move, but as in
Game 74, I wanted to spice things up. I
couldn't remember for sure, but 7 ... a6
is usually a reasonable option.
7 ... a6?! 10 as! lLlbd7 11 0-0 'ikc7 12 l:Ue1 lLlcs
As it turns out, this is one position This picks up the bishop pair, but I

1 99
Th e Dark Knight System

will not be enjoying my pawn struc- �xe2 %:txg7 12 i.. h 6 �g8

ture.
13 i..xcS dxcS 14 dxc6 bxc6
Ugly as this is, it is worse to allow
White access to the ds-square.
15 i..C 4 :b8?!
Instead, 1S ... thg4! frees the bishop
and control s White's e-pawn.
16 es thg4 17 'ife2 :xb2
losing a bit of material, but it is no
worse th an anything el se.

" Everybody knows" three pieces are

18 h3 thh6 19 i.. b 3 C41! [19 ...thfS] 20
tha4 cxb3 21 thxb2 bXC2 22 '6'xc2 i.. e6
23 thd3 i.. d S 24 thd4 l:Id81 [24 ... e6] 25
iVcS thfs 26 thxfs pfS 27 'ii' b6 '6'xb6 28
axb6 :b8 29 :xa6 i.. C 4 30 :a31! e61 31
thcS i..f8 32 thd7 i..xa3 33 thxb8 i..d S
34 b7 c S 3 5 thc6 1-0
much better than a queen, but is this
true regardless of how many pawns
the queen h as? Regardless of structural
problems ? Regardless of king position?
A s much fun a s these positions can be
to play, White is asking too much from
his minors.
13 h4
This is played mainly to keep the
bishop from getting cut off by an even­
tual ... g6-g S.

Game 81
R.Zelcic-M.Dj ur kovic
P u l a 2001

1 e4 g6 2 d4 i.. g7 3 thc3 d6 4 i..c 4 thc6 5 13 ... c61

thf3 thf6 6 '6'e2 i..g4 7 eS i..xf3 8 gxf3 An important move to control
thxd4 9 exf6 thxe2 10 fxg7 :g8 11 White's minors, and uses those extra

200

pawns well.
14 J:lhe1 �d7! 15 :ad1 �f5
Activating the queen quickly is cer­
tainly the right idea, but 15 .. :i'h 3 ! (16
l:h 1 "f5) is more accurate.
16 i.. g 5 0-0-0
17 �f1?
The queen is supposed to be h arder
to play, but not when th e pieces h ave
an exposed king. It is the GM who
makes the first big mistake.
17 ... d5 18 i.. d 3?! 'ith3+?1
This check would h ave been more
useful to Bl ack after grabbin g the f­
pawn.
19 �g 1 f6 20 i..f1 "xf3 21 Ild3 "g4+
22 1:[g3 'iWb4!
Black's queen is not done making
trouble.
23 a3?1 'iWa5?
White's bluff works. Every white
pawn that disappears de-stabilizes
White's pieces more. They are running
short on support points, and they can't
be left lying around loose. Hence
23 ...'iWxb2 ! .
24 i.. d 2 e5?! 25 ttJe4 'iWb6? [ 2 5. . .'iWc7] 2 6
Illustrative Ga mes
ttJxf6 :1gf8 27 :b3?
Just as White is getting back into
the game, he plays this awful interme­
diate move, forcing Bl ack to improve
his queen position. White is not so
badly off after 27 :f3, threatening 2 8
i..h 3+ and 29 lbd7, while i f 27 . . .:f7
then 28 C4! �b8 ( 2 8 .. ltdf8 29 i..h 3+
�c7 30 CS!) 2 9 lbxd5 :xf3 30 lbxb6
l:txd2 3 1 lba4 l1dxf2 and Black is "only"
better.
27 ... 'iWC7 28 l:tf3?
White is in a bad way after 28 lbg4
e4, but at least he doesn't lose more
material.
28 .. J�f7??
28 ... "g7 29 i..h 3 + �C7 30 i.. g 5 :d6
wins.
29 i.. h 3+ �b8 30 1:[xe5 'iWxe5 31 i..f4
'ii'xf4 3 2 1hf4 �df8 3 3 ttJd7+ �C7??
Yikes! When it rains, it pours.
33 ... �a8 ! is still approximately even.
The text walks into 34 lbxf8 lhf4 3 5
ttJe6+.
34 ttJxf8 1-0
Game 82
H.H ughes-K.Richardson
British League 2 004
1 e4 g6 2 d4 i..g7 3 ttJc3 d6 4 i..C 4 ttJc6 5
ttJf3 ttJf6 6 0-0 0-0 7 i..e3 ttJg4 8 i..f4?1
ttJxd4! 9 ttJxd4 e5 10 i.. e 3 ttJxe3 11 fxe3
exd4 12 exd4 i.. e6 [12 ... c6!] 13 "d3
i..xC4 [13 ... c5] 14 'iWxC4 c6 15 �h1?!
'iWb6 16 :ad1 l:tad8?! [16 ...'ifxb2] 17 b3
"a5 18 �f3 d5
201
Th e Dark Kn ight System

tbd4 cS 11 i.. b 5+ It>f8!

The king is fine here. Bl ack keeps the
sensitive fs-square under control.
12 tbf5 i..x f5 13 exf5 ..wa5 14 i..e2 tbxg3

15 hxg3 tbd7 16 0-0 e4 17 :fel :e8

19 exd5 exd 5 20 'iVe5 'it'xe5 21 dxe5 d4

22 tbe2 lUe8 23 tbel :e8 24 b4 b6 25
exb6?!
It is no good to open the c-file for
Bl ack and allow him to press on the
weak c2-pawn. 25 tbb3 bxcS 26 bxcS Mo hr h as no shortage of ways to
offers White better chances of a suc­ improve his position and h as gained
cessful defence. the advantage.
25 ... axb6 26 :f2 :e4 27 l2Jd3 l:lee8 28 18 ltabl :e7 19 'ife3 i..x e3 20 bxe3 tbf6
It>gl?1 [28 a4) 28 ...:xe2 29 ':dd2? [29
a4) 29 ...:xd2 30 :xd2 i.. h 6 31 :b2 :e3
32 tbe5 d3 33 tbxd3 l:txd3 34 It>fl :d2
0-1

Game 83
A.G rilc-G.Mohr
Slove n ia n Tea m
Ch a m pi on s h i p 2008

1 e4 tbe6 2 tbf3 d6 3 d4 tbf6 4 tbe3 g6 5 21 i..xe4?

i..g5 i..g7 6 'if d2
Thanks to thi s move, an d the fact
th at White is giving Black ...h 7-h 6 for
free, the bishop no longer h as enough
squares to escape.
6 ... h6 7 i..f4 g5 8 i.. g 3 tbh5 9 d5 tbb8 10
White is understandably unh appy
with his position, but things are not yet
as desperate as this. There will be no
compensation for the piece.
21 ...:xe4 22 -.xe7+ �g7 23 'it'e3 �f8
24 'ife7+ �g7 25 :xb7 :f8 26 'ii'x d6

2 02

'ii'xd5 27 '6'a6 lIxc3 28 g4 l:xc2 29 l:[xa7
'ii'd4 0-1
Game 84
O.Janowski-F. Yates
M ar i en ba d 1 9 2 5
1 d4 l2Jf6 2 l2Jf3 g6 3 l2Jc3 i..g7 4 e4 d6 5
i..g 5 l2Jc6 6 'ii'd 2 h6 7 i.. e3 l2Jg4 8 0-0-0
If the white bishop tries to slip away
with 8 i..f4, Bl ack equalizes immedi­
ately with 8 ... l2Jxd4!.
8...l2Jxe3 9 'ii'xe3 0-0 10 h3 a6
With the two players castled on op­
posite wings, the race is on.
11 g4 b5 12 1:[g1 tiJa5 13 i.. d 3 c6 14
tiJe21!
White is "racing" a little slowly. 14
e5 or 14 h4 is better.
14 ... tiJC4 15 i..xC4 bXC4 16 'ii'c 3 as 17 e5
'iWb6 18 exd6 exd6 19 l2Jg3
White is much worse, with no bish­
ops and a weaker attack
19 ... c511 [19 .. J�b8] 20 l2Jf51 cxd4 21
l2J3xd4 i..xf5 22 gxf5 d5 23 lIg2?1 lIae8
24 f3 :e7?!
Illustrative Games
It is important to prevent 2 5 'ifd2,
which allows White to escape the un­
pleasant pin and reorganize his de­
fence. Hence 24 ... a4! 2 5 a3 l1b8, threat­
ening 2 6 ...l:fe8 and 2 7 ... l:e3.
25 'iWd2 a4 26 a3 l:b8 27 c3 :eb7 28
fxg6 fxg6 29 'iWc2 g 5 30 "xa4 "cS 31
I1dd2 'ii'f8 32 '6'a5 �h8 33 �xd5?
'iVxa3! 34 'ii'e4 'ii' a1+ 3 5 'ii'b1
What's the best move in this posi­
tion?
35 ...1ia4?
Yates overlooks th e problem-like
35 ...'6'a8 ! ! , which sets White the unen­
viable task of defending against ...l:ta7-
al. (In fact Yates didn't play the second­
best move either - 36 ... '6'a5.)
36 lIge2 'iWa5 37 :e3 i..f8 38 ct>d1 'ifc7
39 'ii'f5 :ta7 40 tiJb5 :a1+ 41 �C2 'ii'h 7
42 'ii'xh7+ ct>xh7 43 l2Jd6 i..x d6 44 :xd6
:a2 45 :e7+ �g8 46 �d2 :axb2+ 47
�e3 :c2 48 :dd7 :xc3+ 49 �e4 eM8 50
%:[f7+ �g8 51 :lg7+ �f8 52 f4 gxf4 53
lIdf7+ �e8 54 :c7 :g3 55 :h7 �f8 56
:h8+ :g8 57 :xh6 :te8+ 58 Wf5 :g7
59 l:[h8+ lIg8 60 :hh7 :g1 61 :h8+
Vz-Vz
203
Th e Dark Kn ight System
Game 85
A.Mista -M.Sze l a g
Kosz a l i n 19 99
1 e 4 d6 2 d4 lLlf6 3 lLlc3 e S 4 lLlge2 Ji.e7
S f3 0-0 6 Ji.e3 exd4 7 lLlxd4 lLlc6 8 �d2
lLlxd4 9 Ji.xd4 Ji.e6 10 0-0-0 c6 11 g4 bS
12 b3 lLld7 13 :g1 b4 14 lLle2 cS is
Ji.b2
1s ... lLles?!
Initiating complication s that shoul d
not work out well for Bl ack. lS ...lic8 ! ?
was better, with ideas of ... cS-c4.
16 'ilfe3 'ilfas 17 f4 c4 18 fxes c3 19
lLlxc3 bxc3 20 Ji.xc3 'ilfxa2 21 exd6?!
2 1 �d2 ! 'ii' a3 (or 21 ...Ji.xb3 22 Ji.d3
Ji.xC2 23 .:tal) 22 :tal 'ii'c s 23 'ilfxcs
dxcS 23 �e3 is good for White.
21 ... Ji.xd6 22 Ji.xg7 Ji.a3+
22 .. JUc8! was stronger.
23 �d2 Ji.b4+ 24 �c1??
Better is 24 Ji.c3 l:tfd8+ 25 �el
:xdl+ 26 �xdl :d8+ 27 �el when
White can still defend.
24 ...f6 2S Ji.xf6 :xf6 26 'ilfd4 :tf2 27
�b2 Ji.a3 0-1
204
Game 8 6
I J a kic-Z.Mestrovic
Zad a r 2001
1 e4 lLlc6 2 d4 d6 3 lLlc3 lLlf6 4 Ji.gs h6 S
i.. h 4 gs 6 i.. g3 Ji.g7 7 h4 g4 8 hS eS 9 dS
lLld4 10 f4?! exf4 11 Ji.h4?! cS 12 'ilfd2?!
White's gambit was a poor choice,
as in this position he h as no real com­
pensation for th e pawn he has sacri­
ficed.
12 ... 'ii b6 13 0-0-0 0-0 14 'ilfxf4 lLlxhS is
'1Wd2 Ji.d7?
lS .. .fS ! is just good for Bl ack,
whereas now White gains time for his
attack.
16 Ji.f2 lLlf6 17 Ji.d3 :tfe8 18 Ji.e3 hS 19
i..h6 �es 20 :f1 l:iae8 21 'ii'f4 "d8 22
lLlge2 lLlxe2+ 23 lLlxe2 c4 24 lLlg3 cxd3
2S Ji.xg7 �xg7 26 lLlxhS+??
Missing th e much stronger 26
l:txh S ! l:ixh S 2 7 '1Wxf6+! 'ilfxf6 28 lLlxh S +
�g6 29 lLlxf6, wh en White comes out
on top.
26 .. JlxhS 27 :Xxhs lLlxhS 28 �xf7+ �h6
29 cxd3 'ilfe7 30 'ilff2 Ji.bS 0-1

Game 8 7
R.Zelcic-Z.Mestrovic
N ova G o rica 2003
1 e4 ltJc6 2 d4 d6 3 ltJc3 ltJf6 4 .i.g 5 h6 5
.i.h4 g5 6 .i.g3 .i.g7 7 f3 0-0 8 .i.f2 es 9
dS ltJd4
White has been getting very little
done, so it is easy to justify Black's ag­
gression.
10 ltJ ge2 cS 11 dxc6 bxc6 12 ltJxd4 exd4
13 i..xd4 CS?! [13 ... l:tb8!] 14 .i.xf6?!
'iVxf6
When will they ever learn ? From
now on White h as a severe weak colour
complex on the dark squares, and
Mestrovic's play from thi s point on is
impeccable.
15 :b1 .i.e6 16 i..e 2 :ab8 17 0-0 .f4
18 �h1 i..x c3 19 bxc3 l::t x b1 20 'ii'x b1
'iVe3 21 .i.d3 c4 22 l:e1 'iVd2 23 i..f1
'ifxc3 24 h3 l:Id8 25 'ifc1 l:Ib8 26 l:Ie3
'iVe5 27 :a3 :b2 28 .-d1? as 29 .-C1 c3
30 l::t a4 'ii'f4 31 'ii'xf4 gxf4 32 i.. c 4 :xc2
33 i..d 5 .i.xd5 34 exd5 llb2 3 5 :C4 c2
36 �h2 :xa2 0-1
Illustrative Games
Game 88
L.Gofs hte i n - N .Mitkov
Li sbon 1999
1 c4 e5 2 ltJc3 ltJc6 3 g3 f5 4 i.. g2 ltJf6 5
e3 d 5 6 ltJxd5 ltJxd5 7 cxd5 ltJb4 8 d3
ltJxd5 9 ltJf3 i.. d 6 10 0-0 ltJf6 11 .-b3
.e7 12 e4 fxe4 13 dxe4 'fIf7 14 '-c3
'iVh5 15 ltJd2 0-0 16 f3 .i.h3 17 ltJb3 l::tf7
18 .i.e3 ltaf8 19 '-C4
19 ... h6?!
After this, White has an edge. It is
simpler and better to keep the knight
out with 19 ... b6.
20 ltJa5 i..x g2 21 �xg2 g5?1 [21 ...b6] 22
h3 '-g6? 23 ltJxb7 h5 24 l:Iad1 g4 25
ltJxd6 cxd6 26 l:txd6 1-0
Black h as no real counterpl ay.
Punishing the dou bled c-pawns
Game 89
J . Ra m i rez-J.Sch uyler
La s Vega s 2007
1 c4 ltJc6 2 ltJc3 e5 3 g 3 f5 4 i..g2 ltJf6 5
2 05
Th e Dark Kn ight System
d3 i.. b4 6 a3?
White greatly overestimates the
value of the bishop pair and h alf-open
b-file, spending a tempo to reach a po­
sition Bl ack would h appily play with a
tempo less.
6 ... i..xc3+ 7 bxc3 d6 8 :lb1 0-0 9 e3?
White's previous sins are minor
rel ative to this positional catastrophe
(which is nonetheless an extremely
common mistake among amateurs ) .
White's pawn structure will self­
destruct in 5 .. .4 . .. 3 ... 2 . .. 1 ...
9 ... e41 10 d4 b61 11 tbe2 i.. a 6 12 tbf4
i..xC4
13 dS?
2 06
A hopeless try at complications.
13 ...tbes 14 tbe6 'ife7 1S tbxf8 tbd3+ 16
",d2 "xf8 17 h3 tbxf2 18 'iWa4 bS 19
'ifa s tbxh1 20 i.. x h1 'fIf7 21 l:Ib2 'ifhs
0-1
White's play was poor, but not un­
common for a club player. That was
how to win.
Game 90
O.Foisor-J.M. Degraeve
Le To u q u et 1996
1 c 4 eS 2 g3 tbc6 3 i..g2 f S 4 ttJc3 tbf6 S
d3 i..b4 6 i.. d 2 0-0 7 e3 i..xc3 8 i.. x c3
dSI 9 tbe2 i..e6 10 b3 'iVe7 11 a3 :ad8
Simple chess has brought Black a
small advantage, but this will dissipate
unless he takes action now.
12 'ikC2 i..f7?!
12...dxC4 13 dXC4 tbe4! w as better.
13 0-0 i.. h S?!
This bishop transfer, typical of th e
Stonewall Dutch, makes far less sense
without a closed centre.
14 :fe1 'iVf7 1S i.. b2 "'h8 16 :ac1 lid6
 Illustra tive Games

17 b4 dxe4 is dxe4 :fdS 19 i.. x e6?

Bl ack's pieces are already poised to
jump into White's holes after thi s ill­
conceived trade.
19 ... J:xe6?
After 19 ...bxc6 ! White doesn't even
get th e e-pawn for his troubl e (20
.txe s ? .:td2).
20 .txe5 lbg4 21 :ed1 :leS 22 .tf4
lbxh 2!? 23 �xh2 .tf3 24 g4 fxg4 25
lb g 3 .f6
Black is already better.
15 dxe6 i..xe6 16 e4 :eS 17 i..e3 i..xe4
is i..x e4?!

26 'iid 3??
26 �g l ! 1fh4 27 'litfl hol ds on. The
text just loses.
26 ... g5 27 'iid4 gxf4 0-1 1S ...lbxe4?
After lS .. .fxe4, Black keeps his
pawn, and with it a larg e advantage.
Game 91 19 lbxe4 fxe4 20 lbe3 i..bS 21 lbxe4 "c6
J . l ruzu bieta 22 lbd2 lbf5 23 'ii'e2 i..e7 24 bs 'iVd7 25
Vi l l a l uenga-B. G u l ko .txa7 b6? 26 as bxa5 27 b6 lbd4 2S Vi'e4
Sa n Sebasti a n 1996 lbe6 29 bxe7?! lbxa7 30 ':xa5 lbe6 Yz-Yz

1 e 4 e 5 2 lbe3 lbe6 3 g3 f5 4 .i.g2 lbf6 5 Game 92

d3 i..b4 6 .i.d2 0-0 7 e3 lbe7 S lbge2?! N . Spiridonov-K.Spraggett
S a3 or s lbf3 is preferable. Ca n ne s 1992
S ... e6 9 0-0 d6 10 d4 .t>hS 11 a3 i.. a 5 12
b4 i.. e 7 13 d5 i.. d 7 14 a4 "eS 1 e4 e5 2 g 3 lbe6 3 i..g2 f 5 4 d3 lbf6 5

207
Th e Dark Knight System

ltJc3 ..tb4 6 ..td2 0-0 7 e3 f4 dxe5 1 4 'ii'g 5, and there is no attack.

13 fxes dxes 14 d4 ..th3

This is not the best, but it is aggres­

sive and sound. If 8 exf4 exf4 9 i.xf4,
Bl ack gets good pl ay with 9... d5!.
8 ltJf3 .e8 9 0-0 ..txc3 1 0 ..txc3 d6 11
exf4 exf4 12 d4 'ii'h S 13 ds ltJe7 14 ltJd4
..tg4 15 f3 ..td7 16 %le1 ltJg6 17 ltJe6
..txe6 18 :xe6 fx g3 19 hx g3 "g s 20
'ii' e1 hS 21 �h2 h4 22 gxh4 ltJxh4 23
Vi'g3 'ii'h 6 24 �g1 ltJfS 25 'ii'h 3 'ii'g s 26
l:tae1 11f7 27 111e2 %laf8 28 ..td2 _g6 29
�h2?1 [29 :2e4] 29...ltJd4 30 ..tc3? [30
:2e4] 30... ltJxe2 31 %lxe2 'iid 3 32 :d2
'ii' h 7 33 ..td4 b6 34 b4 l:te7 35 cS? ltJhS
3 6 cxd6 cxd6 3 7 �g1 �b1+ 3 8 ..tf1 ltJf4
39 'ii'h 2 :fS 0-1
15 dxes?!
Black's aggression usually pays
dividends, but it is not without risk.
Even at this late stage, White can es­
cape to a good position with 15 i.xh 3
"xh 3 16 ltJxe5 ltJxe5 17 dxe5 ltJg4 (or
17 ...%lad8 18 �f4 and 19 '6'g3) 18 '6'd5+
l:.f7 19 'ifg2 .
1s ... lIad8 1 6 '6'f4 ..tx g2 1 7 �xg2?! [17
exf6 !] 17...ltJdS

Game 93
F Bruno B K u rajica
.
-
.

Luga n o 1985

1 c 4 e S 2 ltJC3 ltJc6 3 g3 fS 4 d3 ltJf6 5

i. g2 ..tb4 6 ..td2 0-0 7 ltJf3 d6 [7 ...e4!] 8 18 'ii'c 1??
a3 ..txc3 9 ..txc3 'ii'e8 10 0-0 �hS 11 White is worse now, but he survives
'ifd2 f4 12 gxf4 h6 after 18 1i'h4.
Otherwise White will play 13 fxe5 18 ... ltJf4+ 0-1

208

Game 94
M.Sher- K.Spraggett
Ando rra 199 3
1 c 4 e S 2 g3 ltJc6 3 �g2 f S 4 ltJc3 ltJf6 5
ca �b4 6 �d2 0-0 7 ltJf3 d6 8 0-0 �xc3
9 �xc3 'iVe8 10 e3 .i.d7
Bl ack, who is down a tempo on the
previous game, is not ready for 10 .f4?!.
•
Instead, 10...e 4 1 1 dxe4 ltJxe4 is fine for
Bl ack, but this is not why a player like
Spraggett plays the reversed Grand Prix
Attack.
11 :Cl �h8 12 b4 ltJd8 13 bS l:tb8 14 a4
ltJe6 lS ltJd2 f41 16 exf4 exf4 17 �xf6
Removing one of Black's most dan­
gerous attacking pieces, but pulling his
rook into the action. Besides, th at was a
very good bishop!
17 ... l:txf6 18 ltJe4 :h6 19 d4 'ii'g6 20
:el :f8 21 l:tc3 b6 22 l:ld3 �c8 23 as
:hS 24 �f3 :lh6 25 axb6 axb6
According to Houdini, White has
been at least a little better the whole
time, but that doesn't take into consid­
eration the difficulty and fatigue of
Illustrative Games
defence. White makes a move he cer­
tainly wouldn't h ave made if he were
fresh - but he isn't fresh, and that is no
accident.
26 'ii'e 2?? fxg3 27 fxg3 ltJf4 28 'li'd2
ltJxd3 29 'li'xd3 .i.g4 30 �g2 'WhS 31 h4
:e6 32 1Ifl :'xfl+ 33 'ii'xfl 'iffS 34 -.al
:e8 3 5 'iVa7 .i.f3 0-1
Game 95
Bu Xia ngz h i -V.lva n c h u k
FIDE World C u p,
K h a nty·Ma n s iysk 2011
l ltJf3 d5 2 g3 g6 3 .i.g2 �g7 4 0-0 e S S
ca ltJc6 6 c4 dXC4 7 dxc4 'ii'x dl 8 lixdl
e4 9 ltJfd2?1 [9 ltJ g S] 9 ... f5 10 ltJc3 .i.e6
11 ltJdS? 0-0-0
This is just awful for White, who
can't activate any of his pieces. He soon
pitch es a pawn to free himself, but
there is no compensation for this sacri­
fice.
12 ltJb3 ltJf6 13 .i. gs �xdS 14 cxdS
:xdS 15 f3 exf3 16 :xdS ltJxdS 17 .i.xf3
ltJdb4 18 �fl l:te8 19 ltJcS �d4 20 a3
209
Th e Dark Kn ight System
i..xc5 21 axb4 ttJxb4 22 g4 ttJC2 23 :a5
i.. e3 24 gxf5 i..xg5 25 fxg6 ttJe3+ 26
�gl h6 27 l:xa7 c 6 28 g7 �C7 29 l:[a4
l:tg8 0-1
Game 96
K.Ara k h a m i a G ra nt­
A. Raetsky
B ern 1 9 9 5
1 ttJf3 d 5 2 g3 g6 3 i..g 2 i..g7 4 0-0 e5 5
d3 ttJc6 6 ttJbd2 ttJge7 7 e4 0-0 8 c3 as 9
a4 h6 10 :el i.. e6 11 exd5 i..x d5!
This prevents the active 12 ttJC4 be­
cause of 12 ... e4, exploiting the vulner­
able situation of th e white knight at c4.
12 'ii'C2 f5 13 b3 'ii'd 7 14 i..a 3 lUe8 15
l:tadl g5 16 ttJC4 ttJg6 17 ttJe3 i..f7 18 d4
e4 19 ttJd2 h5 20 f3 f4 21 ttJec4 fxg3 22
hxg3 exf3 23 ttJxf3 'ii'g4 24 ttJe3 J:xe3
25 l:txe3 'ii'xg3 26 'ii' f2 'ii'xf2+ 27 �xf2
g4 28 ttJg5 i.. h 6 29 ttJxf7 i..xe3+ 30
�xe3 �xf7 31 :fl+ <i!tg7 32 i.. d 5 J:e8+
33 <i!td3 ttJd8 34 :f5 c6 35 i.. e4 ttJe6 36
C4?? ttJef4+ 37 �e3 ttJg2+ 38 �d3 ttJel+
39 <i!te3 ttJc2+ 40 �d3 ttJxa3 0-1
210
Game 9 7
V . Frias Pa blaza-A. Bab uri n
Sa n Fra ncisco 1 9 9 7
1 ttJf3 d 5 2 d3 g6 3 g3 i..g7 4 i..g2 e 5 5
0-0 ttJe7 6 e4 0-0 7 ttJbd2 ttJbc6 8 c3 as
9 a4 h6 10 exd5 ttJxd5 11 ttJC4 �f5 12
l:tel :te8 13 ttJh4 �e6 14 �d2 'ii'd 7 15
'ii' b3 ttJde7 16 i..fl l:tad8 17 l:tadl b6 18
i..C l g5 19 ttJg2 i..g4 20 i..e 2 i..xe2 21
l:txe2 ttJf5 22 l:teel??
Bl ack is already better, but this al­
lows a winning shot as Black quickly
exploits the weakened White kingside.
22 ...ttJfd4! 23 cxd4 ttJxd4 24 'ii'a 2 ttJf3+
25 �fl [25 �h 1? 'i'h3J 25 ... ttJxh2+ 26
�gl ttJf3+ 27 �fl 'ii'h 3 28 ttJce3 ttJd4 29
�gl l:te6 30 b3 l:ted6 31 i.. b 2 ttJf3+ 32
�fl e4 33 dxe4 l:d2 34 :txd2 :txd2 0-1
Game 98
A.Ca p a l i k u-J.Gom bac
N ova Gorica 2010
1 f4 g6 2 ttJf3 i..g7 3 e3 d6 4 d4 ttJd7 5
 Illustra tive Games

.i.d3 es 6 c3 'iie 7 7 e4 ttJgf6 8 fxes dxeS .i.c4 e6

9 0-0 0-0 10 .i.gs h6 11 .i.h4 cS 12 It is most important to blunt the
ttJbd2? ! bishop.
1 2 d S c4 13 .i. c 2 'iVd6 is best, though 6 0-0 ttJe7 7 ttJc3 0-0 8 .i.b3 cS 9 'i'e2 ds
this is not a problem for Black
12 ... cxd4 13 cxd4 exd4 14 ttJxd411 ttJes
15 .i.C2 lId8 16 ttJ4b3?! [16 ttJ2b3)
16 ... .i.e6?!
White is in trouble, but 16 ... as! is
stron ger; e.g. 17 a4 b6 and 18 ....i.a6.
17 '6'e2 :ac8 18 l:lacl .i.g4 19 "f2 gs
20 .i.g3 ttJhS

White's light-squared bishop is now

both passive and in danger.
10 a4 b6 11 :dl .i.b7 12 .i.d2 a6 13
.i.el ttJfs 14 .i.f2 'iYC7

I sense a dark-square catastrophe

on the horizon for White.
21 �hl ttJxg3+ 22 'ii'Xg3 aSI 23 ttJf3 .i.xf3
24 gxf3 ttJg6 25 ttJxas :d2 26 .i.b3 llXCl
27 llxcl .i.es 28 'ii'el l:xh2+ 29 �gl ttJh4
30 'ife3 .i.d4 31 :c8+ �h7 32 llh8+ �g6
33 llg8+ �hS 34 'ifxd4 ttJxf3+ 0-1
15 g311 ttJd6 16 h4 c4 17 .i.a2 bS 18 hS
b4 19 ttJbl ttJb611 [ 1 9...as!] 20 hxg6
Game 99 fxg6 21 as ttJa4 22 c3 bxc3?1 23 ttJxC3
A.Spich ki n - D. Rei nderma n ttJxc3 24 bxc3 :ab8 25 .i.bl
E u rope a n C h a m pi o n s h i p, White, who was practically lost, i s
R ij eka 2010 now back i n the game.
2S ic8 26 ttJgS1! .i.f6 27 ttJf3 llbS 28
••

1 f4 g6 2 ttJf3 .i.g7 3 e3 d6 4 d4 ttJd7 5 'iYa2 .i.d8 29 .i.c2 l:[xas 30 'iWbl lIxal 31

211
Th e Dark Knight System

'ii'x al ttJb5 32 ..ta4 ..te7 33 ..tel ..td7 34
ttJe5 ..te8 35 :Cl ..td6 36 ttJf3 h6 37
..tc2 ttJa3 0-1
Did White's flag fall? Bl ack h asn't
made any progress on the board since
winning the a-pawn.
Bl ack needs t o start using the c-file
as soon as possible. His next few moves
do not work towards th at, and he starts
drifting.
14...i.a6?! 15 i.e3 ttJc5 16 %:[Cl 'iWd7 17
ttJel ttJc6 18 'iWd2 ..td4?! 19 i.xd4 ttJxd4

Game 1 00
P.Auc henberg­
To.Ch ristensen
H e l s i ngor 1997

1 f4 g6 2 ttJf3 ..tg7 3 g3 b6 4 ..tg2 i.. b 7 5

0-0 e6 6 d3 d6 7 e4 tiJe7 8 'iVe2 ttJd7 9
ttJbd2 0-0 10 g4 cS
Black has actually gained the ad­
vantage with his simple development
scheme. White's pl an to attack on the
kingside is slow.
11 :f2 'ilc7 12 ttJfl c4 13 ttJg3 cxd3 14
cxd3
20 "b4? [20 fS ! ] 20 ... e5 21 ..tf1? "xg4
22 fxe5 dxe5 23 :g2 'iWf4 24 'ii'd 2 'irxd2
25 :xd2 liac8 26 lIcdl ttJce6 27 ..th3 f5
28 exf5 gxf5 29 Wf2 ttJf4 30 i..fl h5 31
h4 �h8 32 ttJg2? ttJde6? [32 .. ttJh 3+!] 33
.

ttJe3 ttJg7 34 i.e2 :cd8 3 5 i.f1 Ad4 36

ttJC2 :d6 3 7 ttJb4 i.. b 7 38 ttJC2 l:fd8 39
ttJb4 :d4 40 ttJC2 :4d6 41 ttJb4 as 42
ttJc2 ttJg6 0-1

Adieu
With this, I bid my readers adieu. I hope
this book was more fun for you to read
than it was for me to write. Har ! I wish
you many successes with 1 ... ttJc6 and
the Dark Knight System .

212
 I nd ex of Va riatio n s

Chapter One
1 d4 �c6 2 �f3 (others - 15) 2 ... d6

A: 3 c4 g6 4 d5 �b8 5 �c3 i.. g 7 6 e4 �f6 7 i.. e 2 0-0 - 19

Ai: 8 h 3 - 20
A2: 8 0-0 - 2 1
8 : 3 d 5 �e5 4 �xe5 dxe5 5 e 4 �f6 - 2 2
8 1 : 6 �C3 - 2 3
8 2 : 6 i..b 5+ - 2 4
c : 3 i..f4 �f6 4 e 3 g6 5 i.. e 2 i.. g 7 - 2 6

213
Th e Dark Kn ight System

Cl: 6 0-0 - 27
C2: 6 h3 - 2 8
0 : 3 g3 g 6

01: 4 d5 - 30
02: 4 �g 2 - 3 1
E : 3 1.g 5 - 33

Chapter Two
1 d4 llJc6 2 c4 e5 3 d5 1.b4+

214
 Index of Va riations

A: 4 i..d2 i.. x d2+ 5 'Wxd2 liJce7 - 3 6

A l : 6 d 6 - 36
A2: 6 liJc3 - 38
B: 4 liJd2 - 40

Chapter Three
1 d4 liJc6 2 dS liJes

A: 3 e4 - 43
B: 3 f4 - 46

215
Th e Dark Kn ight System

Chapter Four
1 e4 lt)c6 2 d4 (others - 5 5 ) 2 ... e5

A: 3 d5 It)ce7 - 57
Al: 4 lt)f3 lt)g6 - 5 8

All: 5 h 4 f - 5 9
A12: 5 i.. e 3 - 6 1
A 2 : 4 C4?I lt)g6 - 6 2

A21: 5 It)c3 - 6 2

216
 Index of Variations

A22: 5 .i.e3 - 64-

A23: 5 lLlf3 - 6 7
A24: 5 .i. d 3 - 6 8
A25: 5 g 3 - 70
A26: 5 a3 - 71
8: 3 dxe 5 lbxe5 - 73

81: 4 llJf3 lLlxf3+ 5 'ii'xf3 lLlf6 - 73

811: 6 lLlc3 - 74
8 1 2: 6 .i.c4 - 74
813: 6 e 5 - 75
814: 6 .i.d3 - 7 5
815: 6 .i.e2 - 7 5
816: 6 .i. g 5 - 7 6
8 2: 4 f4 lLlc6 - 7 6
821: 5 .i.c4 - 77
822: 5 lLlf3 - 78
The Dark Knight System

(: 3 lDf3 exd4 - 79

(1: 4 lDxd4 1.C5 - 80

(11: 5 1.e3 'i¥f6 6 c3 "96 - 81

(111: 7 lDd2 - 82
(112: 7 lDb 5 - 84
(12: 5 lDxc6 - 8 5
( 2 : 4 1.c4 lDf6 - 8 8

(21: 5 0-0 - 8 8

218
 Index of Variations

(22: 5 eS ttJg4 - 89
(221: 6 'We2 - 89
(222: 6 0-0 - 90

Chapter Five
1 e4 ttJc6 2 ttJf3 d6 3 d4 ttJf6 4 ttJC3 g 6

A : 5 i..b S - 9 4
B : 5 d S ttJb8 6 i.. e 2 i..g 7 7 0-0 0-0 - 96

Bl: 8 h3 eS 9 dxe6 .i.xe6 - 9 7

B l l : 10 .i.g s - 9 8
B 1 2 : 10 ttJd4 - 99
B2: Others - 99
(: White doesn't play d4-dS - 102
(1: 5 h3 - 102
(2: 5 i.. e 2 - 106
(3: 5 i.. e3 - 108
(4: 5 .i.C4 .i.g7 - 109
(41: 6 0-0 - 110

219
The Dark Kn igh t System

(42: 6 'i'e2 - 1 10
(43 : 6 -'.e3 - 1 1 2
(5: 5 -'.g 5 - 112

Chapter Six
1 e4 t'Llc6 2 t'Llc3 t'Llf6 3 d4 d6

A: 4 f4 - 115
B: 4 d5 - 117

Chapter Seven
1 c4 t'Llc6 2 t'Llc3 e5

A: 3 g3 - 121
B : 3 t'Llf3 f5 4 d4 e 4 - 1 2 3
B l : 5 t'Llg 5 - 1 2 4
B 2 : 5 t'Lld2 - 1 2 5

Chapter Eight
1 t'Llf3 t'Llc6 - 1 2 7

220
 In dex of Variations

Chapter Nine
Others - 1 3 1
1 b3 - 129
1 b4 - 130
1 f4 - 1 30

221
 I n dex of G a mes

Aghasaryan.R-Chibu khchian.A, Kajaran 2011 . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

Altounian.L-Schuyler.J, las Vega s 2008 . ................................................. .......... .... ...... ....... 139
Arakhamia Grant.K-Raetsky.A, Bern 1995 . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Arlandi.E-Lanzani.M, San Marino 1998 . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Auchenberg.P-Christensen.To, Helsingor 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . 212
Balashov.Y-Kuzmin.G, USSR C h a m pion ship, Viln i us 1980 . . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Baluta.C-Cioa ra.A, Buch a rest 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 71
Baramidze.D-Griezne.E, B a u n atal 1999 ............................................................................. 1 64
Barle.J-Mestrovic.Z, Slovenian C h a m pionship, Krsko 1997 . . . . . . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . 145
Bastrikov.G-Gel ler.E, Tashkent 1 958 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . 196
Becerra Rivero.J-Miles.A, Andorra 1995 . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Beliavsky.A-Miles.A, E u ropea n C h a mpionship, Sa int Vincent 2000 .......................... 158
Bonin.J-Schuyler.J, New York 1988 . . . .. . . . . .. . . . . .. . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Bontempi.P-Jovanic.O, Nova Gorica 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . 1 77
Bratanov.Z-Chatalbashev.B, B u lga ria n C h a mpionsh ip, Sofia 2004 . . . . . . . . .. . . . . . . . . . . . . . .. . . 198
Brudno.S-Benjamin.J, Boston 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Bruno.F-Kurajica.B, lugano 1985 .............. .... .... . . . . . ........ ...................................................... 208
Bu Xiangzhi-Christiansen.L, Deizisa u 2000 . . . . . .. . . .. . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 72
Bu Xiangzhi-lvanchuk.V, FIDE World C up, Kha nty-Ma nsiys k 2011 . . ............ ...... ........ 209
Buehl.W.M-Benjamin.J, Reno 1999 . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . 1 70
Campora.D-Miles.A, Sevi lle 1993 ......... ...... . . . . . . .................................................................... 159
Campora.D-Tkachiev.V, Biel 1995 . .... . . .... . ................ . . ..... . . .... . . .. . . . . .... . .. . . . . .......... ......... . . . . .... 1 79
Capaliku.A-Gombac.J, Nova Gorica 2010 .......................................................................... 210
Chatal bashev.B-Popchev.M, Cacak 1991 . .............................................. ...... . .. . . . . .... ........... 191
Christiansen.L-Benjamin.J, US C h am pions hip, Seattle 2000 ............................... ......... 1 61
Drasko.M-Mestrovic.Z, Bosnian Tea m Cham pionship 2003 . ................... . .... ............... 154
Erdos.V-Rapport.R, H u nga ria n Tea m Cham pionship 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . 155
Fedorc huk.S-Miles.A, E u ropean Championship, Ohrid 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . 1 75
Fischer.R-Schuyler.J, R ichmond 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . 138

222
 Index of Games

Foisor.O-Degraeve.J.M, Le Touq uet 1996 . ...... ... ...... ......... ... . . . . . . . . . . . . . . . . . . . . . . . . ...................... 206
Frias Pablaza.V-Baburin.A, San Francisco 1997 ............................................................... 210
Galliamova.A-Krasenkow.M, Koszal i n 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 148
Gofshtein.L-Mitkov. N, Lisbon 1999 ..... ... ... . . . . ........ . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Goh Wei Ming-Bellini.F, T u rin Olymp iad 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 76
Golod.V-Sutovsky.E, Natanya (rapid) 2009 . . . . . . . . .. . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Gordon.S-Short.N, British C h a mpion s h i p, Sheffield 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . ... . 155
Grigore.G-Brochet.P, Creon 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Grilc.A-Mohr.G, Slove n i a n Tea m Championship 2008 . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Gruenfeld.Y-Sm irin.l, I srae l i Team Champions h i p 1997 . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . .. . . . . . . . . 191
Grynfeld.I-Bisguier.A, He l s i n k i Oly mpiad 1952 . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... 180
Gurevich.M-Rohde.M, P hi ladel phia (bl itz) 1989 . .. . . . ........ . . . . . . . . . . . . . . . . . . . . . ......... ... ... ... ... ...... 156
Gurevich.M-Zoler.D, Antwerp 1 9 9 8 .................................................................................... 152
Haessel.D-SchuylerJ, Pawt ucket 2008 ......... ... ....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Hahn.A-BoninJ, New York (rapid) 2003 . . . . ..... ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 70
Hjartarson.J-Olafsson.F, Re ykjav ik 1995 ......................................... ................................... 191
Hoffman.A-Fernandez.A, Mar del Plata 1996 . ...... ......... ....... .............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 50
Hromada.P-Ostrowski.L, Moravian Team C hampionship 2003 . .. . . . . .. . . .. . . . ......... ... ...... 1 79
Hubner.R-Hort.V, Germ a n Leag ue 1984 . .. . . . . . . . . . . . . . . . . . . . . . . . . .. . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 74
Hughes.H-Richardson.K, British Lea g ue 2004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... ... ... ... ...... 201
Ipatov.A-Antoniewski.R, Ge rman Leag ue 2011 . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Iruzubieta Villaluen ga.J-Gulko.B, S a n Sebast i an 1 9 9 6 . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Jakic.I-Mestrovic.Z, Zadar 2001 ............................................................................................ 204
Janowski.D-Yates.F, Marienbad 1925 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Jelen.I-Dizdarevic.E, Lj u bljana 1992 . . . . . . . . . . . . . . .... .... . . . . . . . . . .. ...... ......... ... ... ............. . . . . . . . . . . . . . . . . . 168
Jhunjhnuwala.N-Gligoric.S, Lucerne Olym piad 1982 ........... ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Kaidanov.G-Miles.A, Pal ma de Ma l lorca 1989 . ...... ... ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... ......... 166
Karpov.A-Chevallier.D, Fra nce 1993 . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 69.
Keskar.H-Schuyler.J, H ampton 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... ... ... . . . . . . 141
Kmoch.H-Vates.F, Hastings 1927/28 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . 147
Kravtsiv.M-Tarlev.K, Evpatoria 2007 . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Meissner.H-Miles.A, E urop'n C u p, Slough 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 60
Mista.A-Szelag.M, Kosza l i n 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... ... ... 204
MitcheII.R-Schuyler.J, B l oo m ington 1991 ..... .. ........ . . . . . . . . . . . . . . . . . ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
Motylev.A-Gligoric.S, Y u goslav Team Cha m pion s hi p 2000 ......................................... 1 78
Nemcova.K-Olafsson.F, M a rianske Laz ne 2008 ....... ........................................................ 189
Nijboer.F-Miles.A, Li n a res 1 995
. . . . . . . . . . . . . . ... ......... ... ..... ... ... ......... . . . . . . . . . . . . . . . . . ... .... . .. . . . . . . . . . . . . . . . . 187
Onischuk.A-Shkuro.l, U kra i n i a n Tea m Champions h i p 2009 ... ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

223
The Dark Knig ht System

Orso.M-Bordas.G, Buda pest 2000 ........................................................................................ 1 75

Paasikangas Tella.J-Lindqvist.T, Fi nnish Team C hampionship 1996 . . . . . . . . . . . . . . . . . . . . . . ... 134
Pedersen.P.B-Bekker Jensen.D, Danish Tea m Championship 2008 . ........ .................. 145
Perrusset.B-Moullier.l, Pa ris 2005 ........................................................................................ 164
Pol gar.So-Fries NielsenJ, Rimavska Sobota 1991 . .................................... ...... ................ 192
Praznik.N-Beliavsky.A, Bled 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . 1 89
Ramirez.J-Schuyler.J, Las Vegas 2007 . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Rasic.D-Mestrovic.Z, C roatian Tea m Cha mpionship 2001 . . . . . .. . . . . . . . . . . . . . . . . . . .. . . . .. . . . . . . . . . . .. 144
Rozman.L-Schuyler.J, Washington 2012 . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . .. . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Ryba.N-Schuyler.J, Washington 2012 . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . . . 194
Saduakassova.D-Minasian.Art, D u bai 2011 . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 197
Sakharov.A-Adorjan.A, Soc hi 1976 . . . . . . . . . . . . . . . . . . . ......... .... . . . . . . . . . . . . . . . ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Schiendorfer .E-Recuero Guerra.D, E u ro J u n ior C h ' sh ips, H erceg Novi 2006 ..... ....... 1 71
Sher.M-Spraggett.K, Andorra 1993 ................................. ...... .................................. ...... ...... 209
SmeetsJ-Beliavsky.A, Netherl a nd s-S love nia rapid match, M aribor 2004 ......... ...... 181
Spichkin.A-Reinderman.D, E u ropea n C ham pionsh ip, Rijeka 2010 . .. ................. ......... 211
Spiridonov.N-Spraggett.K, C a n nes 1992 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 207
Sultimov.B-Pokazanjev.N, R u ss ia 2007 .............. ................................................................ 182
Sveshnikov.E-Gelashvili.T, Cappe lle la Grande 2009 . .. . . . . .. . . . . .. . . . . .. . . . . .. . . . . . . . . . . . . .. . . . . .. . . . . .. 188
Titz.H-Barlocco.C, D re sden 2 004 . .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Tratar.M-Srebrnic.M, Slove n ia n Championship, Lj u bljana 2010 . . . . . . . . . .. . . . . . . . . .. .. . . . . . .. . . . 136
Vajda.S-Skembris.S, N a ujac 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . 1 78
Valenti.R-Tkachiev.V, Cor sica (ra pid) 1997 . .. . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Vialatte.J-Giroux.F, Paris 2006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . . . . . . 135
Von Wantoch Rekowski.D-Peric.P,Yugoslav J u n ior C h' sh i ps, Tivat 2001 . . . . . . . . . . . . . . ... 183
Wang.K-Schuyler.J, Washi ngton 2012 ....... ...... ...... ............. .................... ...... ..................... 199
Weisser.W-Trumpp.L, German League 2004 ......................... ........................................... 1 62
Yilmazyerli.M-Arutinian.D, I sta n b u l 2007 . ....................... ................................................. 193
Zapata.A-Miles.A, Mata nzas 1995 . . . . . . . . . . . . ..... ... ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Zelcic.R-Djurkovic.M, P ula 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 200
Zelcic.R-Mestrovic.Z, Nova Gorica 2003 . . . . . . . . . .. . . . . . . . . . . . . . . . . . . .. . . . . .. . . . . .. . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . 205
Zelcic.R-Miles.A, P u l a 1994. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
Ziatdinov.R-Chernin.A, New York Open 1998 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . ... 192

224