1.d4 Nf6 2.c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5.e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 Nc6 9.Be3 0–0 10.0–0 Qc7 11.Rc1 Rd8 12.Bf4 Be5< A not often played sideline. Svidler played the 12...Qd7 Main line when he beat Topalov at M-TEL 2006> 13.Bg3 Bxg3 14.hxg3< Novelty I think. [14.fxg3 Ne5 15.Bd5 e6 Has been played before.]> 14...e5 15.Bd5 Be6 16.dxe5 Bxd5 17.exd5 Nxe5 18.c4 a6 19.Re1 b5 20.cxb5 axb5 21.Nc3 c4!? <This is a very interesting move. If Svidler played this against Kramnik, I would probably give it a question mark. Playing it against Shirov it should probably be an exclamation mark! Svidler gives up his b5 pawn for piece activity which I'm not convinced is enough compensation. Svidler may have assessed that Shirov would play 22 Qd4 because of his attacking style, and Svidler had calculated deeply that it was unsound, Only Svidler can answer this, but 21...c4 definately gives chances for white to go wrong.> 22.Qd4?< Shirov as one would expect sees an attack. It seems he also saw an opportunity to force a draw if the attack fizzled out, but this time his calculations had bg flaws. [22.Nxb5 Qb8 (22...Qb6 23.Rxe5 Qxb5 24.Qc2 Rac8 25.Qe4 with equal chances) 23.Rb1 Nd3 24.Re2 Qc8 25.d6 Qf5 26.a4 Is the way to go for white.]> 22...Nd3 23.Ne4?? <A duel winning threat of 24 Nf6+ Kf8 25 Nxh7+ Kg8 26 Nf6+ Kf8 27 Ne8! winning. Also white is threatening the simple 24 Qxd3! It all seems ready for a Shirov brilliancy prize, BUT there is a move that Shirov missed which defends everything - 23...Qa7! I just wonder if Svidler saw all of this when he played 21...c4 [23.Nxb5 Qa5 24.Nc3 Qa7 25.Qf6 Rd6 26.Qf3 Nxc1 27.Rxc1 and black should eventually cause problems with his 2 rooks on an open board.] >23...Qa7! <Solves every threat problem >24.Nf6+ Kh8 25.Qh4 Kg7 <[25...h5?? 26.Qg5 Kg7 27.Nxh5+ Kh7 28.Nf6+ draw]> 26.Re3 <Shirov still has idea's of bailing out with a draw 27 Nh5+ gxh5 (27...Kf8 28 Qf6 mates) 28 Qg5+ draw.> 26...Rd6< But there are defences to all Shirov's hopes for a win or a draw.> 27.Rf3 <Now threatening 28 Qxh7+> 27...h6 28.Ne4 <[28.Rf1 Ne5 29.Re3 Rxf6 30.Rxe5 Qc5 is a more stubborn defence, but black will have 2 connected passed pawns and should win.]> 28...Rxd5 29.Qf6+ Kg8 30.Rc3?< Threatening 31 Ra3 and the queen will be removed from guarding his f7 pawn. But 30 Rc3 is not a good move because black has a good counterthreat.> 30...Ne5!< With a counterthreat of 31...Rd1+ 32 Kh2 Ng4+ forking king and queen.> 31.Rf4 Qxa2 <With 32..Qb1+ and 33...Ra1 threats >32.Kh2 Qe2 33.Qh4 Ra6< Threatening 34...g5> 34.g4 g5 35.Nxg5 hxg5 36.Qxg5+ Rg6 0–1

Shirov's won many great attacking games, but when you miscalculate it can all go horribly wrong. This is one of the games where it went horribly wrong.

My one question to Svidler would be - 'Did you expect Shirov to play 22 Qd4 after your 21...c4. Or did you sacrifice the b5 pawn for better piece activity. Anyway, it was a good game by Svidler who's vision and calculating was on form today.

If that was really Svidler's assessment, then 21...c4 was indeed a devilish trap, even though it might not be the best move objectively. To illustrate how well Black has to calculate when playing that move - in order to make sure that he's ok after the sequence reaching to 25.Qh4 Kg7:

click for larger view

When playing 21...c4, Svidler had to see that in the diagram position, there's no real danger to his king at the middle of the board after 26.Qxh7+ Kxf6, e.g. 27.Re3 Nxc1 28.Rf3+ Ke5 29.Rxf7 Qd4 30.Qg7+ Kxd5.

That White cannot force a perpetual by 26.Nh5+ because of 26...Kf8, when 27.Qf6 is refuted by 27...Qxf2+! 28.Qxf2 Nxf2 29.Kxf2 gxh5 (that's why 26.Re3 by Shirov, which blocks the queen's diagonal as well as bringing the rook into the attack, allows White to draw by the Nh5+ idea in case of 26...Nxc1?).

It was this game that made me realise how deep these players think. It was one of my favourite draws ever. Hope you like my exciteable comments. I just love this game :-)

I'll be happy to kick some Brady ass for you, if he shows up.