The d4-bishop is the fulcrum of White's position. Placed solidly in the centre, it presents a potential challenge to the Dragon bishop on its favourite diagonal. White may later seek a trade via g4-g5, to advance his attack, while the current support offered to the c3-knight dissuades an exchange sacrifice. After 17...e5!, though, the bishop loses all of its privileges. Strictly speaking (a), (b) and (d) all apply but for tactical reasons, in the form of 18 iLe3 &xg4! 19 fxg4 Sxc3! 20 bxc3
Here Black has the bishop-pair but White's centralised bishop is trying to have a say in that matter. With the bishop-trapping 13 £5 a serious threat, White can hope that his queen and knight partnership will offer some serious attacking chances, particularly when twinned with his space advantage. By playing 12...e5!, Black chooses the perfect time to strike out at the centre. After 13 fxe5 dxe5, Black will find a way to activate his dark-squared bishop whereas the e4-pawn will always remain isolated. Best, therefore, is 13 Ae3 but, after 13...exf4, the b2-g7 diagonal is re-opened while the light-squared bishop need no longer be nervous.
After White plays d4 and Black ...cxd4, thereby defining the opening as an 'Open' Sicilian, Black's rooks have a means of entry into the game that should not be neglected. Of course, rooks love open files and seventh ranks, but half-open files are not to be sniffed at either and Black has one of those at his immediate disposal.
From the moment a black rook appears on the c-file, White must guard against the next-up thematic ...Sxc3 exchange sacrifice, as well as concern himself with a more obvious problem. Whether White castles kingside or queenside, there is always the danger that his c-pawn will come under attack. Depending upon the urgency of the position, Black can double or even treble his major pieces on the c-file, and remove the flimsy cover of the white c3 knight with ...b5-b4.
The following position is reached from a variation in chapter 5. Here Black has sacrificed two pieces to enable his rook to advance to its desired location.
Yes, the black rook is beautifully placed on c2 and if the white queen moves then next-up is the capturing of the juicy b2-pawn. I'm sure you haven't forgotten about that bishop on g7 and will notice how all of Black's remaining pieces combine perfectly, while a pawn armada waits in the wings in case of an endgame.
Occasionally Black's light-squared bishop may get the opportunity to pressurise c2 and even if White manages to move his knight and set up a barrier by c2-c3, there is always the old 'minority attack'. Indeed a plan of ...b5-b4 is likely to be successful either as an offensive weapon or as a means of reaching a favourable endgame.
To facilitate ...Sc8, Black must first move his light-squared bishop. I have known ...Axg4 (taking a pawn) to be its first move, but more likely is ...JLe6, with no white knight on d4, or ...jLd7 otherwise. In some of the 'quieter' lines (i.e. those that don't involve do-or-die attacks) sometimes the c8-bishop can also perform well on b7.
With the black queen out of the way (usually on a5), the rooks are connected and ready, if required, to double up. The most common way of achieving this 'doubling up' involves the ...i£>c4-e5(a5)-c4 manoeuvre, after which White is likely to exchange his light-squared bishop for it. This is because not only will he probably want to preserve his more important dark-squared bishop (assuming that it is on e3), but also the b2 pawn will be a problem for him. Indeed, even if it is protected, the chances are that there will be a combination involving ...4bxb2, for example undermining the defence of the white c3 knight.
After ±xc4 Sxc4, the black rook will be difficult to budge, with b2-b3 generally being unattractive for White. Instead, although ...Sc8-c7, to prepare doubling, is hardly ever used as it tends to walk into the likes of or £>d5, the alternative ...Sc8-c5 often fulfils a useful purpose. Though the rook could conceivably be vulnerable to White's dark-squared bishop or suffer in. the event of becoming available to White, on the c5-square it actually provides some extra bite along the 4 rank.
While the c-file is obviously a very useful asset for Black, the d-file is of less value to White as he usually finds negotiating the d-pawn too tough a task.
Let's face it, rooks are good pieces. They go left, right, up and down and they are capable of covering much ground, cutting off kings and capturing enemy pawns on light and dark squares alike. If your opponent gets an outside passed pawn then the best man for the job of halting it is a rook. Forking and pinning are nice, but little compares to that feeling of doubling your rooks on the 7"' rank and hoovering the opponent's position to your heart's content.
Yes, believe me, I, as much as the next man, needed to be convinced of the virtues of the thematic exchange sacrifice, particularly as I was quite materialistic as a junior.
Obviously, continuations such as the following are easy to fathom:
The justification is purely tactical. Often such sacrifices lead to mate but when (as above) the cushion of a clearly superior endgame exists, the pill is easy to swallow. But what of situations where there is no big hit and the harsh reality is that you are conceding a straight-line expert for a measly knight?
Basically then, I am merely informing the reader that I am well aware of the difficulty that less experienced players may have in coming to terms with this key concept. First off though, I need to point out that a piece is a piece and, with other bits around, it's not all just about a simple comparison of a rook versus a knight. The crux of the matter is that ...2xc3 (particularly characteristic of the Dragon but also frequently played in other Sicilian variations) is invariably played with Black already a pawn to the good. The pawn now on c3 (after bxc3) is occasionally won immediately and, even if it is not, White's queenside pawns are shattered beyond repair, while Black's pawn structure remains impeccable.
A. key feature, as illustrated in the above diagram, is that the positions reached are far from desirable for rooks. Black has only bagged one pawn for the exchange but the white rooks lack any serious action. Achieving activity on the half-open b-file is unlikely, especially when you consider that White's light-squared bishop will be a target as it is. He can double on the h-file but all that really does is prevent Black from making much of his extra outside pawn. In contrast, the black knights ensure that the enemy pawn structure remains fixed while Black's own rook sits pretty.
This sort of endgame is covered in more detail later in the book, as is the middlegame shown in the next diagram.
White has launched his kingside pawns with aggressive intent but things have backfired on him with the loss of his c3-knight. Black gave up his rook in order to grab the e4-pawn and now, with so much open space around the white king, one can see that this is now more of a position for minor pieces anyway.
The structural consideration is a major point and is something that even experienced tournament players clearly continue to neglect. Aiming for some sort of attack, White's last move to reach the following position was 14 W(el)-g3?!.
R.Willmoth White R.Felgaer Black Malaga Open 2001
Frankly this last move was like waving a red rag to a bull. White severely underestimates the positional danger he is in.
Text book (this one and others!) stuff. The exchange is a cheap price to pay for a pawn and the damage inflicted to White's queenside pawn structure.
16 &xe4 jfe,xe4
Soon came the likes of and
...Kc8 and, with all of Black's pieces quickly flooding into the game, it's no great surprise that it was Black who was ultimately victorious.
Once confident enough, rather than shying away from ...Sxc3, Dragon players may end up looking for any opportunity to bash this move out with a glint in their eye and maximum attention from spectators.
I must say, however, that although I will continue to preach the merits of this thematic exchange sacrifice, even I have been surprised by the apparently casual instances in which very strong players have been known to employ it. Three memorable (well for me at least!) occasions on which it has occurred almost out of the blue are:
E.Mortensen White V.Mikhalevski Black Copenhagen 2000
True, the h7-pawn might have suffered if White had been allowed to play ®d5, but with the queen on b8 there will be no ...iLxg4 follow up.
S.Beshukov White M.Tumer Black Hastings Premier 2001
B.Lalic White J.Hodgson Black East Kilbride 1998
White has not yet 'weakened' his kingside by advancing his pawns and Black is certainly not a pawn up. Nevertheless the queens now came off with the reigning British Champion apparently giving up the exchange just like that.
These examples have stuck in my mind and I'm not about to suggest to the reader that each one was a good decision. Nevertheless you can't argue with the practicality. Amazingly, Black scored 2V4/3!
On his way to another GM norm in this tournament, the talented young English IM confessed later to not knowing his theory. His logic: if in doubt, take on c3!
I guess a big part of this whole section is really to do with opening up the Dragon bishop to devastating effect. Occasionally, when White has advanced his kingside pawns in a threatening manner, Black will have the sacrifice ...iLxg4 available. This, in conjunction with a ...Hxc3, will leave the e-pawn en prise to a black knight after the f3-pawn captures the bishop. An alternative way to negotiate White's pawns and clear the long diagonal of its congestion (here being four knights!) is demonstrated below.
So, okay, the possibility to give up both knights in this manner won't come along every day but a single knight sac as below isn't that unusual.
13...£>xg4! 14 fxg4 iLxc3 and with White unable to recapture on c3 because of ...Wxa3-a2-al mate, things are looking pretty grim.
As you will discover, though, even in the Yugoslav Attack it's not all about checkmate and in fact the endgame should always remain a serious consideration. The next position is a long way from an ending but the principle still applies.
It always pays to watch out for e4-e5, as played on White's 17th move above. The idea is that 17...dxe5 runs into 18 g4-g5 winning either the f6-knight or the d7-bishop. Now, given that White has already offloaded his h-pawn in the cause of an h-flle assault, Black will of course secure a third pawn for the piece by taking on e5. However, although this has been known to occur, particularly when the queen isn't being pinned on d8 and the d7-bishop can usefully slide to f5, the drawback is that the Dragon bishop finds itself too obscured.
The not necessarily desirable retreat ...£)e8 is not available as Black placed his rook there last go to enable him to meet ilh6 with ...Ah8. Though the knight looks short of squares, the solution of course is 17...£«g4! 18 fxg4 ±xg4
Black ultimately ends up with three connected passed pawns for the piece and, as well as providing a usefiil barrier for the king, these will certainly prove very menacing when rolling down the board in the latter stages of the game.
Particularly in Yugoslav Attack variations, the opportunity to obtain a kingside pawn majority, as just described, is a serious reality. Sometimes it's three connected pawns vs a bishop or knight and sometimes it's four or even five vs a rook. Yes, that's a lot of potential queens and my advice to Black is: make the most of your king and have one piece put aside to stop any unlikely white passed pawn. Below, Black got everything spot on and no doubt the experience was immensely enjoyable.
Clearly White found it just too painful to play on, but Black could hardly go wrong.
Although it is dangerous to make too many generalisations, I would suggest that, more often than not, Dragon endgames are at least equal for Black. With or without queens, the c-file tends to be of more use to Black than the d-file is to White and the first player is also worse off if he has advanced his kingside pawns in a failed attack. Usually Black is encouraged to attack with 'pieces rather than pawns' and it therefore follows that his queenside pawns will probably not be similarly weakened. Actually even if they have been advanced, the a- and b-pawns are generally not so accessible to the enemy king. Regarding White's pawn structure, though, the black monarch invariably has squares such as e5 and f4 to invade, when safe, (i.e. when White has insufficient mating material floating around) in order to inflict some serious damage.
Should White prevent a simple king walk in the manner illustrated below then Black must find alternative ways to activate the required pieces.
F.Burgalat White P.Trifunovic Black Mar del Plata 1953
Typically, it is the black king that is more on top of the action and, with ...'4,f6-e5-f4 a serious threat, White's next move is as good as forced. Though certainly not 100% necessary it would be rubbing salt into the wound if Black should also fix White's pawns on the same colour as his bishop with moves such as ...h6 and ...g5. The game continued with:
Black must not simply accept that his king has no way in and here he rightly sets about negotiating White's pawn barrier.
19 M.h3 ilxh3 20 2xh3 hxg5
Keeping it simple although
20...Sh8 also looks promising.
Instead 23 4>cl would help towards centralising the king. It would also prevent the immediate back rank mate threat but of course is no use on the (soon to be) passed pawn pushing front.
Black has gained control of the only open file and now goes hunting for pawns.
Activating the rook, but, with his king so far away, Black will always win a pawn race. 24 f4 would have been met by 24...Sh4.
24...Sxg5 25 Ic7 Sg3 26 Hxb7 fixf3 27 Sxa7 g5
And Black was first to the finishing post.
Finally, some other bits and bobs to bear in mind both when reading through this book and playing your own Dragon games:
(a) To reiterate, if White eventually gets around to playing <5kl5, Black should try to avoid the reflex response of ...^xd5 even though the enemy knight controls several useful squares. It could well be that this exchange is the best reply, but at least consider the awkward pressure that might arise down the e-file after exd5. Life will be more uncomfortable while it remains there but there may be alternatives. Faced with this dilemma, you could possibly capture the white knight at some stage with your light-squared bishop, chase it away with ...e6, or play around it until you are ready to take it, and then round up the d5-pawn.
Indeed £ld5 is a White possibility that should always be on Black's mind. So long as it is, then you will never fall for the following trap. White played the popular quiet move 12 "¿>bl and Black responded with the inaccurate 12...Sac8? thus reaching the position below:
It was clear that Black wanted a rook on the c-file but he chose the wrong one. As the black king is unable to go to f8 at the end of the following sequence, he is punished by: 13 <SM5! Wxd2 14 £ixe7+. White regains the queen next and has succeeded in netting an important pawn.
(b) Something that I know novices often worry about when playing the Dragon is that White might play
£Xd4)xc6, and hence they unnecessarily prepare ...^c6 with ...ild7. When pushed for an explanation for their fear, they struggle, eventually pin-pointing the now isolated a-pawn as a weakness.
That is, of course, a rather flimsy excuse and this 'weakness' is significantly outweighed by the benefits. Besides, although this pawn is undeniably isolated, Black in fact has no more pawn islands than be fore the knight trade. What he does have, though, is a terrific half-open b-file for a rook to exert (in conjunction with the latent Dragon bishop power) some devilish pressure on b2. In addition, Black now has a firm grip on the d5-square, which is useful both for keeping out a white knight and perhaps helping to prepare a later ...d5.
If White can follow up with a quick e4-e5 then there could be problems, but the above position, which is a result of a Black move-order mistake, is a rare example. The point is that here 8...dxe5? would fall foul of 9 itxf7+!.
Although one must accept that this would win the house, I must confess that I also used to have an unjustified fear of the same position a move on:
I always felt more comfortable castling before playing ...£)c6 in order to avoid the likes of:
Note 10 ¿Lxf7+ doesn't work now as this time (after 10...<&xf7) the queen is protected by the rook.
However what I had failed to take into consideration is that now, although 11 J.xf7 leaves Black with a dodgy pawn structure, of more relevance is the bishop being trapped behind enemy lines via ll...e6!.
I would now go so far as to say that £Vxc6 is a move that Black should usually hope for, rather than fear and, even when having the choice, one cannot normally go wrong by recapturing on c6 with the b7-pawn rather than with a piece.
(c) Although Black plays ...d6, usually as early as move 2, it is generally accepted that if he can later achieve the pawn break ...d6-d5 without repercussion, then he has at least equalized. A brief explanation for this is that the inevitable opening up of the centre, following a timely ...d5, results in White's centrally posted pieces becoming targets for the black rooks. At present, this may sound somewhat vague but all will become clear as you read on.
(d) Always keep an eye out for the possibility of playing ...£lg4, as it is a definite bonus if you can manage to exchange off White's dark-squared bishop (although preferred, the light-squared bishop is no mean catch either). White should usually guard against this mainly post-development threat, as the e3-bishop has difficulty moving away due to its important role of defending the centre.
Note, though, please do NOT fall for jLb5+, one of the oldest tricks in the book (well this book anyway!). For example in the following position after 6...£>g4?? 7 &b5+.
Yes, seeing that 1..A&1 allows 8 Wxg4, the fact is that you would suffer heavy material loss and it certainly wouldn't be good for sales!
(e) Finally, in the same vein as ...®g4, another move to look out for is ...#b6. This may surprise the reader as, although here the queen may attack the (often poisoned) pawn on b2, on b6 it might be in the firing line of a potential ®(d4)-f5, unleashing the bishop on e3. Indeed, I am far from stating that ...#b6 is always a good move, rather I am pointing out, and I agree vaguely, that the move ...Wb6 can occasionally pose awkward problems for White — particularly when the opening has drifted off the well known tracks.
You'll know what I mean when the opportunity arises in your future encounters, but it's alright, you don't have to thank me: I'm just doing my job!
Was this article helpful?