Wedging with Alexei Shirov

October 20, 2017 Matthew Sadler No comments exist

While casually browsing through a compilation of positions, I came across this unusual position from the game Shirov-Topalov Bosnia 2000. The game was given as an example of a pawn wedge in the enemy camp which is quite understandable when you see the rest of the game! It’s worth playing it through without comments just to enjoy the sheer power of Shirov’s play. As always, a playable version of this game is available at



14.d6 Nc6 15.d7 Bb7 16.Qd6 e4 17.Nd5 Bg7 18.Ng5



18…Nd4 19.Ne7+ Kh8 20.Rh3 f4 21.Kb1 b4 22.Be2



22…f3 23.gxf3 Nxe2 24.Qxc5 Nf4 25.Qf5 Ng6 26.h5 Qxe7 27.hxg6







However, something troubled me. If you ask me what the best move in the position is after Black’s 13th move, then my intuition screams out 14.g4!



14.d6 feels wrong to me, despite Shirov’s impressive victory. It seemed worthwhile then to examine the position a little further and evaluate how far off my intuition is!


First of all, let’s take a good look at the position after 13…b5. The most notable features of this position are:

  1. Black’s central / kingside pawn structure e5-f5-f7.

A normal Black position would see Black’s f5 pawn back on g6 and the bishop on g7. Black’s position doesn’t look any worse for it, but there is a subtle difference: Black’s e-pawn is less mobile. If Black plays …f7–f5 to support …e5–e4, then White’s reply Ng5 eyes a panoply of weak central and kingside light squares. In particular of course, he has the e6 square in his sights which can be very unpleasant for Black.

In our pawn structure, Black’s damaged kingside pawn structure nurses 1 big advantage: he has achieved …f5 supporting …e5–e4 without weakening e6! As we shall see in many variations, in this structure a knight on g5 can only go backwards when it is chased away! The mobility of the e-pawn in this structure is the enabler for the next key feature of the position

  1. Black’s dark-squared bishop

Of course, Black does have the advantage of the bishop pair, but the bishop on c8 is not a particularly impressive piece. After …e5–e4 however, Black’s dark-squared bishop would make any King’s Indian player’s heart beat faster. The bishop is unchallenged along the long diagonal and its field of influence reaches as far as b2 where it affects the White king which leads to our next feature

  1. The kings have castled on opposite wings

As we saw in the previous point, Black has an easy way of starting up some worrying pressure against White’s king, however the fact that White’s king is on the queenside gives White some satisfyingly violent ways of storming Black’s kingside.

The more space an opponent has, the easier he is to attack with your pawns which is why the move g4 springs out at you. White exploits the advanced doubled pawn on f5 as a nearby pressure point. White has a substantial amount of light-square control in the position. The pawn on d5 covers 2 key development squares for Black’s queenside minor pieces: c6 (for the knight on b8) and e6 (for the bishop on c8).

g4 aims to loosen Black’s control of the e4 square by removing or deflecting the f5 pawn and by so doing to open up the b1–h7 diagonal for White’s pieces (the bishop on f1 for example) which is a nice incision route into Black’s kingside. If Black’s pawn leaves f5, then …e5–e4 is difficult to achieve; if Black ignores the attack and plays …e5–e4 anyway, then the stakes are raised because Black’s support of his e4 pawn will collapse very quickly after gxf5. Black will need to develop his counterplay very quickly and achieve something concrete before his centre melts away

  1. The disruptive …b5–b4

Something that took me a while to notice. Black’s last move …b7–b5 feels a little like a desperate attempt to open up the queenside, but in fact its main point is the irritating …b5–b4, forcing the knight to move either to the side (Na4) where it’s influence is dubious and where it can also be a target or to e2 where it (temporarily) gets in the way of White’s other pieces.


Weighing up all these factors, how do we look at the move 14.d6? First of all, I still don’t like giving up so much of White’s light-squared control. After 14.d6, Black gains extra development options for his pieces (although as we saw in the game, it’s not as simple as it seems to make use of that). Moreover, I still don’t like the fact that Black can play …e4 and set up his centre without yet facing any resistance from White. However, what 14.d6 does offer – which 14.g4 does not – is a square on d5 for the knight on c3 when it gets tickled by …b4. So this seems to be the fundamental positional point you have to weigh up. What is worth more: a good central position for the knight on c3 after …b4, or the time gained by starting violent kingside play immediately?


Let’s start by taking a look at the natural continuation that just popped into my head when looking at 14.d6 and fuelled my doubts:


14.d6 e4



The only serious move as far as I’m concerned. Sakaev also suggests 14…Nd7 which feels all wrong to me, but conceals a nice tactical point:


14…Be6 15.g4


Not Komodo’s choice, but he likes it a lot once it’s played on the board. The bishop on e6 is in the way of gxf5 with tempo and Ng5


14…Nd7 15.g4 was (of course) what I wanted but Komodo finds a lovely little counter 15…Bb7 16.Nd5 Bxd5 17.Qxd5 Nb6 The key point. I hadn’t appreciated how strong Black’s position has become 18.Qxc5 Rc8 19.Qe3 e4 gets Black’s position going very impressively.




Again, we’ll start with White’s most natural idea





This was it. After opening the h8–a1 diagonal with tempo, Black places his bishop in the middle of the board, cutting off the d-pawn from the rest of White’s army. Moving the bishop to d4 also has some sneaky points when it comes to dealing with White’s attacks against his f5–e4 pawn chain.


First of all 16.f3 can be met by 16…e3. Secondly, after 16.g4, we notice that the bishop has made way for the f-pawn with 16…f6. It’s quite a computer-like defence and I was – perhaps slightly sadly! – very pleased that it had occurred to me immediately (a sign of the times that a GM is proud when he thinks of a computer move!)




16.Nd5 Qxd6 17.g4 f6 Black’s key defensive idea is really strong here 18.Nh3 fxg4


16.Ne2 Nc6 17.Nxd4 cxd4 (17…Nxd4 18.b4 winning is Komodo’s astonishing line!) 18.c5



with mutual chances is Komodo’s final line, but I do like Black’s central pawns!


16.Qf4 f6


was more or less where I stopped my analysis: after 17.Nh3, White’s pieces don’t look very clever. However Komodo found something astonishing!


17.cxb5 fxg5 18.Bc4+ Kh8 19.hxg5 Ra7 20.Rxd4



is an amazing win. I stopped at …Ra7 thinking Black could defend himself!


20…cxd4 21.Qe5+ Rg7 22.Rxh7+



22…Kxh7 23.Qh2+ Kg6 24.Qh6#



Gorgeous! Still don’t quite believe it


16…Nc6 is Komodo’s improvement


17.Ncxe4 f6


Komodo likes Black!


18.cxb5 Ne5


White has too many things hanging!







Again, my feeling for danger seems to have been a little underdeveloped. I didn’t see much problem with Black’s position after this piece sacrifice but all the engines seem rather concerned!


17.Nh3 fxg4 18.Nf4 Nc6


18…Qxd6 19.Nxe4 gives White some time and play

18…f5 was my thought, but Komodo sees some danger 19.cxb5 Qxd6 20.Bc4+ Kh8 21.h5 with the threat of Ng6+)


19.Nxe4 Ne5 is Komodo’s unobvious favourite line, when …Nf3 really cements Black’s pieces in White’s camp




17…fxg5 18.Bc4+ Kh8 19.hxg5 Ra7 20.g6 followed by Qh6 is winning according to Komodo


18.Nxb5 Nc6 19.Bc4+ Kg7 20.Bd5 Qb6



is Komodo’s main line which is still very murky. Black is certainly not worse however.


During my analysis of 15.Ng5, I became discouraged at a number of points and tried to make 15.Qd5



work. White isn’t trying to use d5 for his knight, but for his queen! The main justification is that 15…Ra7 16.Qxc5 is strong for White as the queen cunningly steps aside and frees d5 for the White knight on c3 on the next move.




15…b4 attracted me a lot when I saw it




16.Nxe4 fxe4 17.Qxe4


I thought this would be quite pleasant for White, but I’d obviously missed something


17…Re8 18.Qxa8 Qa5 wins for Black! 19.Kb1 Bxb2


16…bxc3 17.Qxb8 cxb2+


17…Qa5 18.Qb3 cxb2+ 19.Kb1 Bd7 worried me for White, but it’s nothing 20.Nd2 Ba4 21.Qg3+ the use of having an open Black king!


18.Kb1 exf3 19.gxf3 looked quite good for White


15…Qa5 is similar to 15…Bxc3. I liked to have the option of playing …Qf6 but Komodo doesn’t see the need




16.Qxa8 Qb6


Komodo, and rather strong!


16…Qf6 17.bxc3 Qxc3+ 18.Kb1 Qb4+ (18…Be6 19.Rc1 ruined the perpetual I thought 19…Qb4+ 20.Ka1) 19.Kc2 Be6 was what I was looking at bit it won’t be any more than a draw


17.Qd5 Be6 The White queen can’t get back!







Komodo. Hadn’t occurred to me!


17.Kb2 b4 looked good for me

17.Qd2 Nd7 looked nice for Black (17…exf3 18.Qg5+ Kh8 19.Qf6+ Covering c3 19…Kg8 20.gxf3)




with …Nc6 to follow is pretty decent for Black:




18.Rh3 Nc6


18…Nc6 19.Qb7 Qxc3+ 20.Nc2 Nb4 21.Rd2 Nxc2 22.Rxc2 Qe1+


is a draw by repetition according to Komodo.


And now my move…








chasing the White knight away from e4


14…e4 15.Ng5 h6 16.gxf5 hxg5 17.hxg5 Bxg5 18.Rg1 f6 19.Nxe4 threatening f4




15.Na4 e4 16.Ng5 h6



Again a proud moment for me as I understood quite quickly that this was fine for Black. It feels very strange, but the idea is simply to stop the knight attacking e4 so that Black can take the pawn on g4




17.Nh3 fxg4 18.Qxh6


18.Nf4 Qd6 followed by …Be5. So hard for White to get at the Black king!


18…Bg7 19.Qh5 Qf6


19…gxh3 20.Bxh3 might offer chances I thought, but I wasn’t happy about 19…Qf6. Komodo’s evaluation shocked me though: completely winning for Black!


20.Ng5 Qf4+ 21.Kb1 Bf5


…Nd7–f6 trapping the queen is coming!


17…hxg5 18.hxg5 Bxg5 19.Rg1 f6 20.Nxc5


20.d6 Kh7




Just looked good for Black. Komodo shocked me with a perpetual!




Komodo (missed that)


21…Bxh3 22.Nxe4 Nd7 23.f4 Qc7 24.b3 Kf7 25.fxg5 Qe5 26.gxf6 Qa1+ 27.Kc2 Qxa2+ 28.Kc1 Qa1+ is Komodo’s miracle perpetual!


15…e4 16.Ng5 h6 17.gxf5


17.Nh3 fxg4 18.Nhf4 Bxh4 Komodo! Followed by …Bg5 cementing the Black kingside!


17…hxg5 18.hxg5 Bxg5 19.Rg1 f6 20.d6



Interested me. Komodo thinks White is doing quite OK as his knight can come into f4




seemed most natural to me. Now I was looking for all sorts of violent ideas with White but couldn’t make them work, but Komodo just wants a slow build up


20…Bxf5 21.Rxg5+ fxg5 22.Qd5+


20…Kh8 21.Nf4 Bxf5 22.Rh1+ Kg7 23.Bh3 Qd7 24.Rdg1 Rh8 25.Qd5 Rxh3 26.Rxh3 Bxh3 27.Rxg5+ fxg5 28.Qxg5+ Kf8 29.Qf6+ Ke8 30.Qh8+ Kf7 31.Qh5+ Kg7 32.Qg6+ Kf8 33.Qf6+


21.Kb1 Rg7 22.Qd5+ Rff7 23.Qxc5 Nd7 24.Qxb4 a5



was the kind of thing I got analysing with Komodo. Komodo really likes White, but it feels very fraught to me. A slight mistake and Black’s 2 bishops and extra piece are going to tell!


All in all, I’m still somewhat undecided as to which is the absolute best move but I have certainly got a better idea of the dynamic potential of Shirov’s move. 14.g4 tries to broaden out White’s existing light-square predominance, but 14.d6 introduces all sorts of new elements into the position: a knight jump to d5 (aiming for e7 with the support of the pawn on d6) and the opening of the a2–g8 diagonal for the bishop on f1 after cxb5. It’s interesting for me that I underestimated White’s attack from both sides, maybe over-influenced by the optical aspect of that enormous bishop on d4! The more ideas you can shove into a position, the more confusing the defensive task and from that point of view, 14.d6 is clearly a much better move than 14.g4!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.