| Once: Hmm. We may be in danger of over-praising black's play simply because he won. It's a fairly common thing - history tends to be written by the victors and chess games tend to be written up as if every move by the winner helped to seal the win. The reality here is that white should have held this game, and even possibly won it. The first key position (after 23...d5):
click for larger viewWhite has a good position. He is safely castled, he has the bishop pair, although the board is temporarily better for knights because it is full of prawns. Black's king looks artificially placed and his pieces not well coordinated. That's why I'm not wholly convinced by 24. Nh4 from this position. Black's main plan is to push d5-d4 to make the most of his queenside space advantage and open up his Bg7. So I really want to keep my knight on f3 or sink it into the outpost on d4. Fritzie says that white is ahead by the equivalent of nearly two prawns (+1.85). He prefers 24. Nd4. He's not afraid of the maneouvre Nc5-Nxa4 for black because of this line: 24...Nc5 25. Qf2 Nxa4 26. f5! followed by a piece sac on f5. The point is that black's poorly coordinated pieces can't defend well enough and if his queen comes back to defend he loses his knight on a4. Funnily enough, Fritz doesn't mind 24. Nh4 although it comes a way down his top ten. He still rates it as winning for white. The next key position comes after 27...Bxe5
click for larger viewNow Fritz wants to play the bold 28. g6 where I would like to hold d5 with 28. Qf2 or 28. Qd1. Fritzie rates his move as +2.19 and my two as +1.something. He also quite likes 28. Kh1. But neither of us likes 28. Rf7. White's eval falls to around equality because of 28...Bxc3! 29. bxc3 Qxc3. After 28...Rg7, white is back in the lead. I would have snaffled the rook with 29. Rxg7 (+1.67) but Fritz is made of sterner stuff. He wants to play the cool 29. Rf1 (+2.06). But 29. Qxh5? No, no, no! White assumes he is winning by oodles and has the luxury to commit both his queen and king rook to an "attack". But in doing so, he leaves his queenside relatively undefended. Again, the best riposte by black is 29...Bxc3! Black doesn't see it. Instead he plays the thematic 29...d4 - the move that white should have been defending against ever since 23...d5. Now we can forget about a white initiative. Black is the one in the driving seat. Fritzie rates the position as -0.39. A slight edge to black but almost level. In practice, I'd say this was a stronger black initiative than that. OTB I'd expect black to win more often than not from this position. Let's fast forward to 31...Qe5
click for larger viewFrom here, 32. Bd2 (my choice) or 32. Bf4 (Fritzie's) keep the game level. But 32. Bf2? hands the initiative to black on a silver platter. White totally abandons his queenside. The inevitable happens... Incidentally, let's take a look at the position where white resigned: 
click for larger viewFritzie analyses this as a draw with best play - eg: 34. Bxc2 Qxb2 35. Bb6+ Rc7 36. Qe7+ Kc8 37. Be4 Bd4+ 38. Kh2 Be5+ 39. Kh1 Qxa1+ 40. Bg1 Kb8 41. Qb4+ Kc8 42. Qe7 A keenly fought game, with chances and mistakes on both sides. Of course, they don't have the luxury of silicon assistance as we do. But a positional masterpiece? I don't think so. Sorry. | 
Kibitzing
