1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Be2 O-O 6. Bg5 h6 7. Be3 Nbd7 8. Qd2 c5 9. Rd1 cxd4 10. Bxd4 a6?
   Nigel Short and I drew our games rather quickly and went to a pizza place nearby. There we discussed Bologan's difficult endgame. Nigel said that in his opinion 10...a6? was a poor move, which caused Black problems. After the game Victor confirmed that too. Much better was 10... b6 trying to tie White's pieces to the e4-pawn as soon as possible.
11. Nf3 b6 12. Be3 Bb7?!
   Here better was 12... Kh7 as 13. e5 dxe5 14. Nxe5 Qc7 is ok for black.
13. Bxh6 Nc5 14. Bxg7 Kxg7 15. Ng5 Ncxe4 16. Ncxe4 Bxe4 17. Nxe4 Nxe4 18. Qd4+ Nf6 19. g4! e5
   After 19... Kg8 20. g5 Nd7 21. h4 Black would soon be checkmated.
20. Qxd6 Qxd6 21. Rxd6 b5 22. h3 Rfd8 23. Rxd8 Rxd8 24. cxb5 axb5 25. Bxb5 Nd5
   Black's piece activity cannot compensate for two pawns.
26. a4 Rb8 27. Kd2 Nc7 28. Kc3 Nxb5+ 29. axb5 Rxb5
   White has an extra pawn, which is also a passed one. His king is well placed, while the enemy rook is not well posted (in front of the passed pawn). However, Black has counter-chances too: he might be able to create a passer on the e-file, supported by his king. I have a feeling that the ending should be winning for White, but in such positions it is always hard to support such feelings with variations - there are simply too many possibilities!
30. b4
   It is also possible to play 30. g5, trying to prevent ...f7-f5. However, advancing the b-pawn looks like a very sensible idea.
30... f5 31. Rg1?!
   I don't really understand this move - Black was going to play ...Kf6 anyway. I think that 31. Kc4 Rb8 32. b5 Kf6 33. Rb1 would be a much better way of playing.
31... Kf6 32. Kc4 Rb8 33. b5 e4 34. gxf5 gxf5 35. f4?!
   This is another move which is hard to understand... White wants to prevent ...f5-f4, but to me the remedy looks worse than the illness. Again, the simple move 35. Rb1 looks better to me. I do not see a draw for Black then.
35... Rc8+?
   It is tempting to keep the passed e-pawn, but this move has serious drawbacks too. First, most exchanges favour the defending side, so taking en passant would be better from the point of view of formal logic. Secondly, now Black's king is not active because of the f4-pawn. 35... exf3 36. Rf1 Ke5 37. Rxf3 Ke4 38. Rf1 Rc8+ 39. Kb4 Black has a decent choice.My preference is 39... Rh8 ... reducing the remaining material! (A similar idea 39... f4 40. b6 f3 41. Kb5 Rc2 42. b7 Rb2+ 43. Kc6 Rc2+ 44. Kb6 Rb2+ 45. Kc7 Rc2+ 46. Kb8 f2 should also be enough for a draw.) 40. b6 Rxh3 41. b7 Rh8 and the f-pawn saves Black.
36. Kd5 Rc2 37. b6 Rd2+ 38. Kc6 Rc2+ 39. Kd7 Rd2+ 40. Kc7 Rc2+ 41. Kb8 e3 42. h4!
   Suddenly, the h-pawn becomes very dangerous as it often queens with a check.
42... Rc4
   Here Jon Speelman, Nigel Short and I considered 42... e2 However, we soon saw 43. h5 Rd2 44. h6 Rd1 45. h7 which is obviously hopeless for Black. After the game Bologan said that he saw the same line, so he went for Rc4 knowing that is should not save the game either. (Immediately after the game in a brief post-mortem Korchnoi showed an amazing idea. 44... Rd8+ 45. Kc7 (45. Ka7 Rd1 46. h7 Rxg1 47. h8=Q+ Rg7+ 48. b7 e1=Q 49. Qh6+ Kf7 50. b8=Q Qa5+ 51. Qa6) 45... Rd1 46. h7 Rxg1 47. h8=Q+ Rg7+ and Black is fine! Great stuff - psychologically it is hard to find Rd8+ Black forces the enemy king from b8 where he wanted him to be just a few moves ago! However if you click on the green variation in italics above, beginning 45. Ka7, you will see that Black is in fact already lost after Rc8+. I remember that Bologan suggested it, but both players stopped at 48. b7, not spotting that after 49. Qh6+ Black's king is forced to block the 7th rank, so White gets another queen and wins.)
43. b7 Rxf4 44. h5 e2 45. Kc7 Rc4+ 46. Kd6 Rb4 47. Kc6
   This is it - the h-pawn is unstoppable:
47... Rb3 47... f4 48. h6 f3 49. h7 f2 50. h8=Q+ 48. h6 Rc3+ 49. Kb6 Rb3+ 50. Kc7 Rc3+ 51. Kd8 Rd3+ 52. Ke8 Rb3 53. h7 e1=Q+ 54. Rxe1 Kg7 55. Re7+ Kh8 56. Kd8 f4 57. Kc8 f3 58. b8=Q

1-0