[Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.26"] [Round "13.1"] [White "Wang Hao"] [Black "Caruana, Fabiano"] [Result "0-1"] [ECO "B40"] [WhiteElo "2762"] [BlackElo "2842"] [Annotator "Monokroussos,Dennis"] [PlyCount "84"] [EventDate "2020.03.17"] [EventRounds "14"] [EventCountry "RUS"] [EventCategory "21"] [SourceTitle "The Week in Chess 1381"] [Source "Mark Crowther"] [SourceDate "2021.04.26"] [SourceVersion "1"] [SourceVersionDate "2021.04.26"] [SourceQuality "2"] {[%evp 0,84,36,25,39,63,74,6,14,14,33,15,9,42,18,18,42,21,24,14,3,6,17,25,16,6, 12,3,0,-2,4,-17,8,-4,-3,7,7,-2,0,-25,-13,0,0,0,24,9,17,-17,5,-17,0,-2,-2,-41,3, 0,-8,-53,-26,-25,-42,-51,-37,-37,30,35,35,-15,-12,-42,-36,-75,-46,-75,-76,-65, -65,-122,-144,-152,-145,-145,-162,-160,-156,-212,-199]} 1. e4 c5 2. Nf3 e6 3. c3 {The product of a year's preparation?} Nf6 4. e5 Nd5 5. d4 cxd4 6. cxd4 d6 7. Bc4 Be7 {A normal move, but hitherto Caruana had preferred other options.} ( 7... Nb6 8. Bd3 (8. Bb3 dxe5 9. Nxe5 Nc6 10. Qf3 Nxe5 11. dxe5 Bb4+ 12. Nc3 Qd4 13. Qg3 Qe4+ 14. Be3 Qg6 15. Qh4 Be7 16. Qe4 Qxe4 17. Nxe4 Bd7 18. O-O-O Bc6 19. Nd6+ Bxd6 20. Rxd6 Nd5 21. Bd2 Rd8 22. Rxd8+ Kxd8 23. f3 Kd7 24. Kb1 Bb5 25. Rc1 Bd3+ 26. Ka1 Rc8 27. Ba4+ b5 28. Rxc8 Kxc8 29. Bd1 Kc7 30. b3 Kc6 31. Kb2 a6 32. b4 Nb6 33. Kc3 Bf1 34. g3 Kd5 35. f4 Bg2 36. Be3 Nc4 37. Bc5 g6 38. a4 h5 39. Bc2 Be4 40. Bb3 Bg2 41. Bg1 Bf3 42. axb5 axb5 43. Bxc4+ bxc4 44. b5 Be2 45. b6 Kc6 46. Kb4 Bf1 47. Bd4 Be2 48. Ka5 c3 49. Bxc3 Bf1 50. Bd4 Be2 51. Kb4 Bf1 52. Kc3 Be2 53. Kd2 Ba6 54. Ke3 Kd5 55. Bc3 Kc5 56. Ba5 Kb5 57. Be1 { 1/2-1/2 (57) Vachier Lagrave,M (2789)-Caruana,F (2804) Leuven 2016}) 8... Nc6 9. O-O Nb4 10. Bg5 Be7 11. Bxe7 Qxe7 12. Nc3 dxe5 13. dxe5 Nxd3 14. Qxd3 Bd7 15. Ne4 O-O 16. Nd4 Rad8 17. Rad1 Bc6 18. Nxc6 bxc6 19. Nd6 Nd5 20. Qc2 f6 21. Qxc6 fxe5 22. Ne4 Rc8 23. Qa4 Rb8 24. Nc3 Rxb2 25. Nxd5 exd5 26. Rxd5 Qf7 27. Rd7 Rxa2 28. Qb5 a6 29. Qb1 Qxd7 30. Qxa2+ Qf7 31. Qxa6 h6 32. Qe2 Qf5 33. Qe3 Qf4 34. Qxf4 exf4 {1/2-1/2 (34) Liang,A (2590)-Caruana,F (2828) Saint Louis 2019}) (7... Nc6 8. O-O Be7 9. Qe2 O-O 10. a3 Bd7 11. exd6 Bxd6 12. Nc3 Nce7 13. Ne4 Bc6 14. Ne5 Rc8 15. Rd1 Bb8 16. Ng5 Be8 17. Bd3 Nf6 18. Ngf3 Ned5 19. Bd2 a6 20. Rac1 Rxc1 21. Rxc1 Qb6 22. Nc4 Qd8 23. Nce5 Qb6 24. Nc4 Qd8 { 1/2-1/2 (24) Vachier Lagrave,M (2789)-Caruana,F (2816) Leuven 2018}) 8. O-O O-O 9. a3 (9. Qe2 {is (by far) the most commonly chosen move.}) 9... Bd7 {Rare, but the computer's preference. Caruana is even prepared for obscure lines in the unpopular (for the super-GM level) c3 Sicilian.} 10. Bxd5 $146 exd5 11. Nc3 Be6 12. Bf4 Nc6 13. exd6 Bxd6 14. Bxd6 Qxd6 $11 {Sometimes, positions like this with both sides having isolated d-pawns can favor the side with the extra knight, when the bishop is restricted by its own center pawn. Here it seems to be equal, in part because Black can solve the problem of that bishop pretty easily, so it's at least a decent piece.} 15. Re1 Rac8 (15... Bg4 {would be the most straightforward solution, but Caruana doesn't feel any need to do this.}) 16. Qd3 f6 17. h3 Bf7 18. Rac1 Qd7 19. Nh2 Rfe8 20. Rxe8+ Rxe8 21. Nf1 g6 $6 {This keeps White's knight out of f5, but it restricts his bishop. White is a little more comfortable now.} (21... Na5) (21... Ne7) 22. Ne3 Kg7 23. Na4 Qd6 24. Nc3 {Implicitly offering a draw. Perhaps Wang Hao thought that Caruana would just want to finish out the tournament after his all-but-official elimination in round 12. If so, he was mistaken.} (24. Nc5) 24... h5 $1 (24... f5 $1 {is also good, marching the pawn on to f4.}) 25. Rd1 a6 26. Rc1 (26. Qd2 {is still a touch better for White, who hasn't done anything really wrong so far. But it looks as if he has no ambition and no ideas, while Caruana is still playing, open to whatever the course of the game will bring about.}) 26... h4 27. Rd1 {Is this psychology - a bit of "rope-a-dope" by Wang Hao? Or does he think that whatever he does is irrelevant, and therefore such an overt display of time-wasting is going to induce Caruana to realize the pointlessness of playing real moves? Whatever the story, White's woodshifting gives Caruana the chance to keep making microimprovements in his position.} Na5 28. b4 $5 {A sudden overcommitment?} (28. Qd2 $142 $11) 28... Nc4 29. Nxc4 dxc4 30. Qd2 (30. Qf3 $142) 30... Bd5 (30... g5 $142 $15) 31. b5 Bf7 (31... g5 $142) 32. bxa6 bxa6 33. a4 g5 34. d5 (34. Rc1 $142 $11) 34... Bg6 $15 35. Re1 Qf4 $1 36. Qd1 (36. Re3 $142) 36... Bc2 $1 37. Qa1 $8 Re5 {Black has made some progress, but it's still plenty drawish.} 38. Rxe5 $6 (38. a5 $142) 38... Qxe5 $17 39. a5 Kg6 $6 (39... Bg6 $142 40. Qc1 $8 Qd4 $1 41. Qa3 Bf7 42. d6 Be8 43. Qb4 Bd7 $17 {White is tied up, but Black must still win the game.}) ({Note that Black cannot yet round up the d-pawn, e.g. with} 39... Be4 {, because - and only because - of} 40. Qe1 $1) 40. Kh1 $4 {On the last move of the time control, as it often happens, White blunders his d-pawn.} Be4 41. d6 {The best option.} ({Now} 41. Qe1 $4 {is impossible, thanks to White's 40th move.} Bxg2+ $1 42. Kxg2 Qxe1 $19) 41... Bc6 42. Qb2 (42. Qa3 $4 {saves the d-pawn without immediately blundering the knight, but} Qe1+ 43. Kh2 Qxf2 {wins the knight anyway.} 44. Qa2 Qg3+ 45. Kh1 Qxc3 $19) 42... Qxd6 {Although Black's passer is securely blockaded, White's position really is bad enough to resign. White must worry about his a-pawn, the f-pawn, the possibility of tricks like ...g4 hxg4 h3, and of course the c-pawn, even if it is blockaded for the meantime. There are too many problems, and just because he's not getting mated or losing a piece doesn't mean that the position isn't bad enough to resign.} 0-1 [Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.26"] [Round "13.2"] [White "Nepomniachtchi, Ian"] [Black "Vachier Lagrave, Maxime"] [Result "1/2-1/2"] [ECO "A15"] [WhiteElo "2774"] [BlackElo "2767"] [Annotator "Monokroussos,Dennis"] [PlyCount "83"] [EventDate "2020.03.17"] [EventRounds "14"] [EventCountry "RUS"] [EventCategory "21"] [SourceTitle "The Week in Chess 1381"] [Source "Mark Crowther"] [SourceDate "2021.04.26"] [SourceVersion "1"] [SourceVersionDate "2021.04.26"] [SourceQuality "2"] {[%evp 0,83,39,18,34,-24,27,6,24,12,35,33,33,17,13,27,9,2,9,12,10,1,5,10,13,2, 15,18,13,1,3,-1,-16,-24,-14,4,58,58,58,63,66,25,38,55,23,47,78,43,97,77,102,66, 92,80,70,54,49,66,70,65,65,53,56,47,50,44,69,52,58,56,53,47,52,52,53,53,53,24, 47,24,36,11,24,31,38,24]} 1. Nf3 Nf6 2. c4 b6 3. g3 Bb7 4. Bg2 g6 {Objectively dubious, but Vachier-Lagrave is in a must-win situation, and has to play something that will at least force White to figure things out on his own.} 5. d4 Bg7 6. d5 {Rare here, though after first castling it's not uncommon to follow up with d5, if permitted.} (6. O-O) (6. Nc3) 6... Na6 7. Nc3 Nc5 8. O-O O-O {Back to a bigger pool of games.} 9. Qc2 (9. Nd4 {is more usual, and it also prevents Black from putting a knight on e4.}) 9... a5 10. Rd1 $146 {+/=} Ne8 11. Bf4 (11. Be3 $16 {followed by Bd4 looks strong.}) 11... Nd6 12. b3 $16 Re8 13. Be5 (13. Nd4) ({and} 13. Bxd6 {may be even better. White is spoiled for choice.}) 13... e6 14. Bxg7 (14. Bxd6 $1 cxd6 15. Nd4 $16 {/+- followed by Ndb5 leaves Black with a pretty rancid position.}) 14... Kxg7 15. Rab1 { Intending a3, b4, c5 and so on.} e5 16. Nd2 (16. a3) 16... f5 17. a3 Qf6 18. b4 axb4 19. axb4 Na6 20. e4 (20. c5 bxc5 21. bxc5 Nxc5 22. Nb5 Nxb5 23. Qxc5 Nd6 24. Qxc7 Ba6 25. Rb6 Nb5 26. Qxd7+ Re7 27. Qc6 Qxc6 28. dxc6 $16) (20. Nb5 Nf7 21. c5 $16) 20... f4 21. Ne2 (21. gxf4 $142 $1 exf4 22. Re1 Nf7 (22... Re7 $142 $16) (22... Kg8 $142 $16) 23. e5 $1 Nxe5 24. Nce4 Qe7 25. Qc3 d6 26. Nf3 $18) 21... fxg3 22. fxg3 $6 (22. hxg3 $142 $16) 22... c5 $1 $14 23. bxc5 (23. b5 Nb4 24. Qc3 $14) 23... Nxc5 $6 (23... bxc5 $142) 24. Rxb6 Qd8 $6 (24... Rf8 $142 25. Rf1 Qd8 26. Rbb1 Rxf1+ 27. Rxf1 Qa5 $14) 25. Rdb1 Qc7 26. Nc3 Rf8 27. R6b2 Ba6 28. Nd1 Rab8 $6 {Both objectively and practically an error. With the rooks off, White's king is instantly safe, and White can live a life of ease. Maybe he can make something of his extra pawn, or maybe not, but there's no risk for him whatsoever. And that likewise means that MVL's winning chances sink to approximately zero, so he is in effect resigning the Candidates with this move. } 29. Rxb8 Rxb8 30. Rxb8 Qxb8 31. Qb2 Qxb2 32. Nxb2 $16 Nc8 33. Kf2 Nb6 34. Bf1 d6 35. Ke3 Kf6 36. Be2 Ke7 37. Nb1 Bc8 38. Nc3 Bd7 39. Bd1 Be8 40. Bc2 Bd7 41. Bd1 Be8 42. Bc2 $14 {Giri's game was still going by this point, but he was clearly lost. There was therefore no need for Nepo to bother continuing, as the draw (assuming Grischuk didn't go insane) meant the Candidates was over, and he had won it. (It should be emphasized that we're only talking about *trying* to win. By now, it's far from clear that Nepo would succeed in winning even if he felt fully motivated to do so and cranked away for several more hours.} 1/2-1/2 [Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.26"] [Round "13.3"] [White "Alekseenko, Kirill"] [Black "Ding Liren"] [Result "0-1"] [ECO "C50"] [WhiteElo "2698"] [BlackElo "2805"] [Annotator "Monokroussos,Dennis"] [PlyCount "122"] [EventDate "2020.03.17"] [EventRounds "14"] [EventCountry "RUS"] [EventCategory "21"] [SourceTitle "The Week in Chess 1381"] [Source "Mark Crowther"] [SourceDate "2021.04.26"] [SourceVersion "1"] [SourceVersionDate "2021.04.26"] [SourceQuality "2"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. d3 Bc5 5. c3 d6 6. O-O O-O 7. h3 h6 8. Re1 a5 9. b3 $146 {Something new, in a position that's very popular in recent years.} (9. Nbd2 {is usual.}) 9... Bb6 10. Nbd2 {Transposing to a couple of games.} Be6 $146 {A novelty, but one which transposes to (a few) other games.} 11. Nf1 d5 { This is likely to result in the disappearance of Black's e-pawn, when the game will take on the character of a Marshall Gambit.} 12. exd5 Nxd5 13. Qc2 $146 { The final "novelty" of the game.} (13. Bd2 Qf6 14. N1h2 Rad8 15. Ng4 Bxg4 16. hxg4 e4 17. g5 hxg5 18. Bxg5 Nxc3 (18... Qg6 19. dxe4 Nxc3 20. Qc2 Nd4 21. Qxc3 Nxf3+ 22. Qxf3 Qxg5 23. Rad1 Qf6 24. Qxf6 gxf6 25. Kf1 c6 26. Ke2 Rxd1 27. Rxd1 Rd8 28. Rxd8+ {1/2-1/2 (28) Tobor,M (2350)-Kudr,R (2343) ICCF email 2018}) 19. Bxf6 Nxd1 20. Raxd1 gxf6 21. Rxe4 Rfe8 22. Rg4+ Kf8 {1/2-1/2 (22) Wunderlich,H (2596)-Zidu,J (2594) GER email 2017}) 13... Qf6 14. Ng3 Qg6 15. Kh2 Rad8 16. Ba3 Rfe8 17. Nxe5 Nxe5 18. Rxe5 c6 19. Bxd5 Rxd5 20. Rxd5 Bxd5 (20... cxd5 $11) 21. c4 Be6 22. Re1 Rd8 23. Ne4 Bc7+ 24. Kh1 Qh5 (24... b6 $11) 25. Re3 Qe5 26. Ng3 Qa1+ 27. Kh2 Qf6 28. Bb2 Qg6 29. Kh1 (29. Be5 Bxe5 30. Rxe5 Rxd3 31. f3 $14 {/+/- is one very decent option, returning the extra pawn to eliminate Black's bishop pair and break his pressure.}) 29... Bb6 30. Rf3 Bc7 31. Bc3 (31. Qe2 $142) 31... b5 (31... h5) 32. Qe2 bxc4 33. bxc4 a4 34. Qe3 Qg5 35. Qe1 Qg6 36. Qe3 Qg5 37. Qe1 Qg6 38. Qe2 {Playing on, and rightly so. White's advantage isn't large, but it is real, and justifies continuing.} a3 39. Bb4 $6 Rb8 ( 39... Bxc4 $1 40. dxc4 Qb1+ 41. Be1 Rd1 42. Re3 Ba5 43. Re8+ Kh7 44. Qf3 Rxe1+ 45. Rxe1 Bxe1 46. Kh2 $11 {If Black takes on a2, White has an immediate perpetual with 47.Qf5+; if he doesn't, then there's nothing for him to do to make any progress.}) 40. Qe1 Rd8 41. Qc3 h5 (41... Rb8 $142) 42. Bxa3 h4 43. Ne2 Bf5 44. Nd4 Be4 $2 {A miscalculation, or a calculated risk? Ding goes on to win, but could easily have lost.} 45. dxe4 Qxe4 46. Rd3 Be5 {The key moment. } 47. Bc5 $4 {Played immediately, and only after Black's next move did Alekseenko sink into thought. Lots of thought - but it was too late.} (47. Bb2 $1 {wins. It's possible to miss this in a quick game or without much time on the clock, but for a player of Alekseenko's level it would have been easy to work it out, had he just taken the time to think.} c5 48. Nf3 $1 {The key. It's nice, but again, not all that difficult for a 2700.} Rxd3 (48... Bxc3 $4 49. Rxd8+ Kh7 50. Ng5+ $18 {is the tactical justification, but White is doing so well that even} (50. Bxc3 $18 {would win.})) (48... Qxd3 {is an inferior version of the 48...Rxd3 line.} 49. Qxe5 f6 50. Qe6+ Kf8 51. Bxf6 gxf6 52. Qxf6+ Ke8 53. Qe6+ Kf8 54. Qh6+ Kg8 55. Qg5+ $18 {and unless Black wants to throw away his queen (55...Kf7?? 56.Ne5+), he must surrender either the c-pawn or the h-pawn with check, and then a few moves later he'll lose the other pawn as well. Black might get one pawn back, but four pawns for the exchange is more than enough.}) 49. Qxe5 Qxe5 50. Bxe5 Rd1+ 51. Kh2 Rf1 52. Nxh4 Rxf2 53. Kg3 Rxa2 54. Nf5 f6 55. Bd6 $18) 47... Rb8 $1 $19 48. Bb4 Rxb4 $1 (48... c5 $2 {lets the win slip, thanks to} 49. Qe1 $1 Qxd3 50. Qxe5 Rxb4 51. Nf5 f6 52. Qe8+ Kh7 53. Qh5+ Kg8 54. Qe8+ $11) 49. Qxb4 Qxd3 50. Nf3 Qf1+ 51. Ng1 Bd4 52. Qb8+ Kh7 53. Qf4 Bxf2 54. Qf5+ {There's no perpetual, though it's close.} Kh6 55. Qf4+ Kg6 56. Qg4+ Kf6 57. Qf4+ (57. Qf3+ {forces a perpetual or a queen trade. Unfortunately for White, the queen trade wins for Black, as his king is so much more active than its counterpart.} Ke5 58. Qe2+ Qxe2 59. Nxe2 Ke4 $19) 57... Ke6 58. Qe4+ Kd6 59. Qf4+ Kc5 60. Qe5+ Kxc4 61. Qe4+ Kb5 {The checks will run out soon, so White will have to swap queens - with a lost minor piece ending - or get mated (or lose the queen by taking on f2 at some point).} 0-1 [Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.26"] [Round "13.4"] [White "Grischuk, Alexander"] [Black "Giri, Anish"] [Result "1-0"] [ECO "E16"] [WhiteElo "2777"] [BlackElo "2763"] [Annotator "Monokroussos,Dennis"] [PlyCount "101"] [EventDate "2020.03.17"] [EventRounds "14"] [EventCountry "RUS"] [EventCategory "21"] [SourceTitle "The Week in Chess 1381"] [Source "Mark Crowther"] [SourceDate "2021.04.26"] [SourceVersion "1"] [SourceVersionDate "2021.04.26"] [SourceQuality "2"] 1. d4 Nf6 2. c4 e6 3. Nf3 b6 4. g3 Bb7 5. Bg2 Bb4+ 6. Bd2 c5 7. Bxb4 cxb4 8. O-O O-O 9. Nbd2 d6 10. Qb3 a5 11. a3 Na6 12. Rfd1 Qe7 13. Ne1 Bxg2 14. Kxg2 $146 (14. Nxg2 {has been played almost a dozen times, including a couple of high-level wins by all-time greats.} Qb7 (14... Rfd8 15. Ne3 Qe8 16. Qd3 e5 17. Ne4 Nxe4 18. Qxe4 exd4 19. Qxe8+ Rxe8 20. Rxd4 Rad8 21. Rad1 Re6 22. Kf1 b3 23. Nd5 Nc5 24. Nxb6 h6 25. f3 g5 26. R1d2 Kg7 27. Nd5 Na4 28. Nc7 Rf6 29. Nb5 Kf8 30. Nxd6 Rfxd6 31. Rxd6 Rxd6 32. Rxd6 Nxb2 33. Rb6 Nxc4 34. Rb8+ Kg7 35. a4 Kf6 36. e4 Ke5 37. Ke2 {1-0 (37) Karpov,A (2710)-Kortschnoj,V (2625) Amsterdam 1987 }) 15. Qd3 Rfe8 16. e4 e5 17. f3 bxa3 18. bxa3 Nc7 19. Ne3 Ne6 20. Nf5 Qc7 21. d5 Nc5 22. Qe3 Qd8 23. Rab1 g6 24. Nh6+ Kg7 25. Nf1 Nfd7 26. h4 Rf8 27. Rb2 f5 28. exf5 gxf5 29. Qg5+ Qxg5 30. hxg5 Kg6 31. f4 Rae8 32. Rf2 exf4 33. gxf4 Re4 34. Nd2 Rd4 35. Re1 Ne4 36. Nxe4 Rxe4 37. Rxe4 fxe4 38. Ng4 Rc8 39. Ne3 b5 40. Rh2 h5 41. gxh6 Kh7 42. cxb5 Rc3 43. Nf5 Rc1+ 44. Kf2 Nb6 45. Nxd6 Rc2+ 46. Kg3 Rc3+ 47. Kg4 Nxd5 48. Nxe4 Rc4 49. Kf3 Ra4 50. Ng5+ Kh8 51. Rd2 Rxf4+ 52. Kg3 Rf5 53. Kg4 Rf4+ 54. Kh5 Rf5 55. b6 {1-0 (57) Kasparov,G (2790)-Polugaevsky,L (2640) Paris 1992}) 14... h5 $5 15. Nc2 bxa3 16. bxa3 $14 Rab8 17. e4 (17. Ne3 $142) (17. Rab1 $142) 17... e5 18. Qd3 (18. Rab1) 18... Nc7 (18... exd4 $142 $11) 19. Rab1 (19. Ne3 $14) (19. dxe5 dxe5 20. Ne3 $14) 19... Ne6 20. Rb5 Rfe8 (20... h4 $142) 21. h4 g6 (21... exd4 $1 22. f3 Rbc8 23. Nxd4 Nxd4 24. Qxd4 Rc5 $1 25. Rxb6 d5 $1 26. cxd5 Rxd5 $11 {Did Giri spot this? I don't know, but it's possible that he did but rejected it. It's clear that he's trying to start some trouble in this game, if possible, believing himself to be in a must-win situation. So an equalizing idea like this, that leads to a dry position, is unacceptable unless the alternative is overtly awful.}) 22. f3 Nd7 23. Nf1 $1 exd4 24. Nxd4 Ne5 25. Qe2 Nxd4 26. Rxd4 {White's knight is headed to d5, and unless Black finds something brilliant he'll have a miserable position in perpetuity.} Nc6 $2 (26... Rec8 $1 27. Ne3 Rc5 28. Nd5 Qe6 29. Rxb6 Rxb6 30. Nxb6 Kh7 $11 {This is the best Black can do. He'll never win, barring a blunder, but White seems to be stuck as well.}) 27. Rd1 $16 Qe6 28. Ne3 Ne7 $2 (28... f5 $6 29. Qd2 fxe4 30. f4 Ne7 31. Rg5 Rf8 32. Qxd6 Qxd6 33. Rxd6 $18) 29. Qd2 $18 f5 $4 {If this was Giri's intention from several moves ago, then he just miscalculated.} 30. Qxd6 {It's over. The rest is easy for White, and agony for Black.} Nc6 31. exf5 gxf5 32. Qxe6+ Rxe6 33. Nxf5 Ne5 34. Rd6 Ree8 35. Rd4 Nc6 36. Rd2 Rbd8 37. Rxd8 Rxd8 38. Rd5 Rxd5 39. cxd5 Ne5 40. Nd6 Kf8 41. Kf2 Ke7 42. Nb5 Kf6 43. Ke3 Kf5 44. Nd6+ Kf6 45. Ke4 Nd7 46. Kd4 Ke7 47. Nb5 Kf6 48. Nc3 Kf5 49. Ne4 Kg6 50. g4 b5 51. Nc5 1-0