[Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.19"] [Round "8.1"] [White "Caruana, Fabiano"] [Black "Vachier-Lagrave, Maxime"] [Result "1-0"] [ECO "B97"] [WhiteElo "2820"] [BlackElo "2758"] [Annotator "Monokroussos,Dennis"] [PlyCount "147"] [EventDate "2020.03.17"] [SourceVersionDate "2020.07.21"] {[%evp 0,147,28,26,60,60,88,74,74,37,37,29,64,63,90,29,22,16,12,12,4,-4,-35, -35,-77,-55,-32,-33,-130,-127,-73,-37,-84,-90,-105,-117,-116,-291,-248,-262, -136,-136,-168,6,-31,-31,-26,-67,-38,-27,-14,-14,-14,-5,-12,-21,0,0,65,49,36, 33,34,23,49,53,61,70,75,51,51,37,70,58,82,86,89,95,82,93,170,174,191,180,192, 192,185,142,129,132,140,153,239,235,239,153,239,175,254,263,254,250,254,245, 268,255,268,254,249,255,268,255,256,255,255,254,254,246,254,247,254,250,246, 246,254,254,254,254,254,254,254,254,254,254,254,254,254,254,256,254,254,254, 254,246,254,254,254,229,395,440]} 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bg5 e6 7. f4 Qb6 8. Qd2 Qxb2 9. Rb1 Qa3 10. e5 h6 11. Bh4 dxe5 12. fxe5 Nfd7 13. Ne4 Qxa2 14. Rd1 Qd5 15. Qe3 Qxe5 16. c3 (16. Be2 {is far more popular - it has been played at least 2162 times, compared to 12 previous trials for 16.c3.}) 16... Bc5 (16... Be7 {has also been tried with apparent success in correspondence play, and Stockfish (running on my not particularly amazing laptop) prefers it to 16... Bc5.} 17. Bxe7 Kxe7 18. Nf3 {has been automatic, and now Black has generally played} Qc7 19. Be2 Nf6 {without experiencing any particular difficulties.} 20. Nd6 Rd8 21. Nf5+ Kf8 22. Rxd8+ Qxd8 23. Qc5+ Kg8 24. Qxc8 Nbd7 25. Qxd8+ Rxd8 26. Ng3 Nc5 {and Black is not worse - the three pawns are fully a match for the piece here.} 27. Nd2 a5 28. Bf3 Nd3+ 29. Ke2 Nf4+ 30. Kd1 N6d5 31. Bxd5 Nxd5 32. c4 Ne3+ 33. Ke2 Nxg2 34. Rb1 Nf4+ 35. Kd1 Rd7 36. Nge4 Nd3 37. Rb5 a4 38. Nc5 Nxc5 39. Rxc5 a3 40. Kc2 f5 41. Ra5 Kf7 42. Rxa3 g5 43. Ra8 e5 44. Nf1 Rc7 45. Kc3 b5 46. Nd2 bxc4 47. Ra5 g4 48. Rxe5 Kf6 49. Rd5 Re7 50. Nxc4 Re2 51. Rd6+ Kg5 52. Rd2 Rxd2 53. Kxd2 f4 54. Kd3 {1/2-1/2 (54) Petters Merino,O (1924)-Cintins,I (2239) ICCF email 2012}) 17. Bg3 Qd5 18. Bc4 $146 {Something new and startling!} ({In previous games, White had preferred a different (and also temporary) piece sacrifice:} 18. Bd6 Bxd6 19. Nb5 Qxd1+ 20. Kxd1 Be5 (20... axb5 $1 $146 {looks like the reason MVL was willing to choose this path, and the reason Caruana avoided it.} 21. Nxd6+ Ke7 22. Nf5+ (22. Nxb5 Rd8 23. Kc2 Nc6 $11) 22... Kf8 23. Nd6 Ke7 $11 ) 21. Nbd6+ Ke7 22. Nc4 {White may be able to claim a very slight edge here.}) 18... Qxc4 {Forced.} 19. Bd6 {Black has a lot of options here, so if you want to play this way with White you'll have to do some R&M - Research and Memorization. Let's take a start at it.} Nf6 {Apparently best. MVL was ready for Caruana's novelty.} (19... f5 {This may not be Black's best option, but since it may lead to an immediate draw it needs to be taken into account.} 20. Nxc5 Kf7 21. Nxf5 (21. Nd3 $5 {is a playable way to keep the game going, but the engine very slightly prefers Black after} Re8 22. O-O Nc6 23. Rf4 $1 e5 24. Re1 $1 Nxd4 25. cxd4 Qc6 $1) 21... exf5 22. Rd4 Qa2 23. Rd2 Qc4 24. Rd4 $11) ( 19... Bxd4 20. Rxd4 Qb3 21. Qg3 Qb1+ (21... g6) (21... g5) ({and even the ridiculous} 21... Rh7 {are all acceptable alternatives.}) 22. Kd2 $8 Qb2+ ( 22... Qxh1 $4 23. Qxg7 Rf8 24. Bxf8 Nxf8 25. Rd8+ Kxd8 26. Qxf8+ Kd7 27. Qd6+ Ke8 28. Nf6#) 23. Ke3 $8 g5 $1 24. Bb4 Nc6 25. Nd6+ Kd8 26. Nxf7+ Ke8 {and it appears that sooner or later White will have to give perpetual by} 27. Nd6+ Kd8 28. Nf7+ Ke8 $11 {and so on.}) 20. Nxc5 Nd5 21. Qe5 Rg8 22. Ndxe6 $1 {Flashy, strong, and the only move in any case.} fxe6 23. Nxe6 Qxc3+ (23... Nf6 { was also possible.} 24. Nf4+ Be6 25. Qxe6+ ({If you enjoy a spectacle, here's a move for you:} 25. Rd5 $5 {/?} Qe4+ $8 26. Qxe4 Nxe4 27. Nxe6 Nxd6 28. Nc7+ Kd7 29. Nxa8 Kc6 $17 {/-+ Unfortunately, all the brilliance of the move leads to nought, and Black is probably winning here.}) 25... Qxe6+ 26. Nxe6 Kf7 27. Nc7 Ra7 $15 28. O-O (28. Ra1 Rd8 29. Nb5 axb5 30. Rxa7 Rxd6 31. Rxb7+ Nbd7 32. O-O {Better late than never.} Rd3 33. Rxb5 Rxc3 $15 {/-/+ It's almost certainly a draw, but Black can play forever in hopes of more.}) (28. Bg3 Rc8 29. O-O Kg8 $15 {/-/+}) 28... Rd8 29. Bc5 Rxd1 30. Rxd1 b6 31. Bxb6 Rb7 32. Na8 $8 Nc6 {gives Black the slightly sunnier side of equality.}) 24. Qxc3 Nxc3 25. Nc7+ Kf7 ({The brave and paradoxical} 25... Kd7 {is also playable.} 26. Rd3 Ne4 27. Bf4+ Kc6 28. Nxa8 Nd7 29. O-O Ndc5 30. Rc1 b5 31. Nc7 Rf8 32. Be3 Kxc7 33. Bxc5 Nxc5 34. Rxc5+ Kb6 35. Rc1 $11 {and chances are equal in this ending.}) 26. Rd3 Ne4 $6 {A minor inaccuracy. This was the correct follow-up in the 25... Kd7 line, but not here.} (26... Ra7 27. O-O+ Kg6 28. Rxc3 Kh7 $11 {Black's king is safe at last, and despite his unimpressive development he is two pawns ahead. White can reclaim that pawn with 29.Bc5 or 29.Nd5 Nc6 30.Bc5, in both cases inducing ...b6, and that will be enough to draw. But not more.}) 27. O-O+ Kg6 28. Nxa8 $14 Nc6 $6 (28... Rd8 $142) 29. Nb6 $16 {My guess, based on time usage, is that this is it's only here that Caruana was on his own.} Rd8 $6 ( 29... Bg4 $142) 30. Nxc8 Rxc8 31. Ba3 $2 {Caruana's play had been perfect up to here, but this is apparently an error. It was more important to gain time activating his rooks than to preserve the bishop.} (31. Re1 $142 $18) 31... Rc7 $16 {/+-} 32. Rf4 (32. Bb2 $142) 32... Nf6 33. Bb2 Ne7 34. Bxf6 gxf6 $16 35. h4 {This looked suspicious to me at the time, as it rigidifies White's structure. Caruana has a concrete idea in mind, but the engine agrees with my dislike for the move.} (35. Rd6 $142) 35... h5 36. Rg3+ Kf7 37. Rg5 {The point. If it won Black's h-pawn for free it would be a great idea, but alas, it does not.} Rc1+ $1 (37... Ng6 $2 38. Rxh5 $1 Rc1+ (38... Nxf4 $2 39. Rh7+ $18) 39. Rf1 $18) 38. Kh2 Ng6 39. Rf2 Nxh4 40. Rxh5 Ng6 41. Rh7+ Ke6 42. Rxb7 {White has won a pawn after all, and this was undoubtedly Caruana's idea with 35.h4. The problem is that the rook vs. knight ending with just the kingside pawns remaining is objectively drawn. From that point of view, White needed to keep his h-pawn. As we'll see, however, not all tablebase endings are easy for humans to hold.} Ne5 (42... Rc6) 43. Rb6+ Rc6 {A good move, as MVL recognizes that the ensuing R vs. N ending is a draw; the a-pawn will be lost, but it's immaterial. (Immaterial material? Apparently so.)} 44. Rxc6+ Nxc6 {"Officially" equal, but not in real life.} 45. Kg3 Kf7 46. Rc2 Nb4 47. Rd2 Nc6 48. Kf4 Kg6 $1 49. Rd6 Ne5 50. Rxa6 {If you have access to a tablebase, you can confirm everything for yourself. At this point, it's a draw with best play.} Nf7 51. Ke4 Nh6 52. Ra5 Nf7 $8 53. Ra3 Nd6+ 54. Kf4 Nf5 55. Rd3 Nh6 $2 $18 {Losing! Not all tablebase endings are easily held.} (55... Ne7) ({and} 55... Ng7 {kept the game officially drawn.}) 56. Rg3+ Kf7 57. Ke4 Ng8 58. Kf5 (58. Ra3 $142) 58... Ne7+ 59. Kf4 $6 (59. Ke4 $142) 59... Nd5+ 60. Kg4 $2 {Nor, it seems, are all tablebase wins easily converted.} ({White had to repeat, and then meet} 60. Kf5 Ne7+ {with} 61. Ke4 {, e.g.} Ng8 62. Rh3 Kg6 63. Rb3 Ne7 64. Rb5 $1 Nc6 65. Kd5 Ne5 66. Ke6 $18) 60... Kg6 61. Kf3+ Kf7 62. Ke4 Ne7 $2 (62... Nc7 $1 {was apparently the only move.}) 63. Kf4 (63. Rb3 $142) 63... Nd5+ 64. Kf5 $1 { Obviously White would play this without an engine at hand, but it's also the only move that wins.} Ne7+ 65. Ke4 Ng8 66. Rh3 Kg6 67. Ra3 Kf7 68. Kf4 Nh6 69. Rg3 $1 {Not the only move, but a nice one. Black is in zugzwang.} Ng8 70. Kg4 Ne7 (70... Kg6 71. Kh4+ Kf7 72. Kh5 {Progress!}) 71. Kh5 Nd5 72. Rf3 Ke6 (72... Kg7 73. g4 Ne7 74. Ra3 {guarantees further progress.}) 73. g4 Ke5 74. Kg6 { White threatens Rf5+ followed by taking on d5 and then f6. If the knight moves away, then obviously 75.Rxf6+ will win. That leaves three options, none of which appealed to MVL enough to bother continuing.} (74. Kg6 Nf4+ (74... Ne7+ 75. Kf7 f5 (75... Nd5 76. Rf5+ Ke4 77. Rxd5 Kxd5 78. Kxf6 $18) 76. g5 $1 $18 ( 76. Kxe7 $4 fxg4 $11)) (74... Ke6 75. Rf1 Ne7+ 76. Kg7 Nd5 (76... f5 77. Re1+ Kd6 78. Rxe7 Kxe7 79. gxf5 $18) 77. Re1+ Kd6 78. Rd1 Ke6 79. Rxd5 Kxd5 80. Kxf6 $18) 75. Kf7 Ke4 (75... Nd5 76. Rf5+ Ke4 77. Rxd5 Kxd5 78. Kxf6 $18) 76. Rf1 Ke5 (76... f5 77. g5 $18) 77. Re1+ $18) 1-0 [Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.19"] [Round "8.2"] [White "Wang Hao"] [Black "Ding Liren"] [Result "1/2-1/2"] [ECO "C45"] [WhiteElo "2763"] [BlackElo "2791"] [Annotator "Monokroussos,Dennis"] [PlyCount "55"] [EventDate "2020.03.17"] [SourceVersionDate "2020.07.21"] {[%evp 0,55,19,38,28,25,17,-31,27,-8,2,-8,-12,5,6,17,0,-10,-6,-45,16,-74,25,57, 26,15,-23,-3,10,10,0,0,-8,0,0,0,0,0,0,-13,-3,-13,-3,-3,0,0,0,0,0,0,0,0,0,0,0,0, 0,0]} 1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Nf6 {That Ding Liren plays this line is not a surprise.} 5. Nxc6 bxc6 6. e5 Qe7 7. Qe2 Nd5 8. c4 Ba6 {Nor is this a surprise from Ding.} 9. b3 g6 {This is the main move here, so this wouldn't be a surprise from anyone. But Ding has played this move as well.} 10. Ba3 {Unless Wang Hao has a new idea here, and it quickly transpires that he doesn't, this is a draw offer. But why?} (10. g3) (10. f4) (10. Bb2) 10... Nb4 $1 (10... Qxa3 11. Nxa3 Bb4+ 12. Qd2 Bxd2+ 13. Kxd2 $11 {/+/=}) (10... d6 11. exd6 Qxe2+ 12. Bxe2 Bg7 13. cxd5 Bxe2 14. Kxe2 Bxa1 15. Rc1 $11 {/+/=}) (10... Qh4 11. Bb2 (11. Bxf8 $4 Qd4 $17 {/-+}) 11... Bb4+ 12. Nd2 Nc3 13. Qe3 Ne4 14. g3 Qg4 15. f3 Bxd2+ 16. Qxd2 Qxf3 17. Qg2 Qe3+ 18. Qe2 Qxe2+ 19. Bxe2 $14) ( 10... Qg5 11. Bxf8 (11. h4)) 11. Bb2 Bg7 12. a3 Nd5 13. Nd2 O-O 14. O-O-O Rfe8 15. Qf3 Nb6 16. Ne4 (16. Re1 $6 d5 $1 $15 (16... d6 $6 17. c5 $1 Bxf1 18. cxb6 Ba6 19. exd6 Bxb2+ 20. Kxb2 Qxd6 21. Qc3 Qf4 $11) 17. cxd5 $2 Bxf1 18. Rhxf1 cxd5 $19) 16... Bxe5 (16... d5 $146 17. Nf6+ Bxf6 18. exf6 Qc5 19. Qf4 dxc4 20. Qh6 Qf8 21. Qxf8+ Kxf8 22. a4 $11 {/+/=}) 17. Bxe5 Qxe5 18. Nf6+ Kf8 19. Nxe8 Qa1+ {This position had already happened a dozen times, including seven GM-GM battles, and all (but one correspondence game where Black got careless) were drawn by perpetual over the next few moves.} 20. Kc2 (20. Kd2 $4 Qb2+ 21. Kd3 Rxe8 $19) 20... Qa2+ 21. Kc1 Qxa3+ (21... Qa1+ {was chosen in almost every case, and why not? Ding decides to have a little fun, which makes the game appear more dramatic to the spectators, but it changes nothing.} 22. Kc2 Qa2+ $11) 22. Kb1 (22. Kc2 Qa2+ $11) 22... Na4 $1 (22... Rb8 $4 23. Qc3 $1 Nxc4 24. Qh8+ Ke7 25. Re1+ Kd8 26. Nd6# {1-0 (28) Sos,F-Bender,Z ICCF email 2016}) 23. Qf6 $1 (23. Rd2 $2 Rb8 24. Kc2 Qb2+ 25. Kd1 Qb1+ 26. Ke2 Rxe8+ $19) 23... Qxb3+ (23... Rb8 $4 24. Qh8+ Ke7 25. Re1+ Kd8 26. Nd6#) 24. Kc1 Qa3+ 25. Kc2 (25. Kd2 Qa2+ 26. Kc1 $8 (26. Ke3 $2 Rxe8+ 27. Kf3 Re6 $19) 26... Qa3+ $11) 25... Qa2+ 26. Kc1 {All this had already occurred in a correspondence game, which was abandoned as a draw here.} Qa3+ (26... Nc5 27. Qh8+ Ke7 28. Re1+ Ne6 29. Rxe6+ (29. Qb2 $2 Qxb2+ 30. Kxb2 Kxe8 $17) (29. Qf6+ Kxe8 30. Qh8+ Ke7 31. Qxa8 Qa1+ 32. Kd2 Qa5+ 33. Kd1 Qa1+ $11) 29... dxe6 30. Qf6+ Kxe8 31. Qh8+ Ke7 32. Qxa8 Qa3+ 33. Kc2 Qa2+ $11 34. Kc3 Qa3+ 35. Kd2 Qb2+ 36. Kd1 (36. Ke3 $2 Qc1+ 37. Kf3 Qc3+ 38. Kf4 e5+ 39. Kg5 Qd2+ 40. Kh4 Qxf2+ 41. g3 Qf3 42. Kh3 Qxh1 $19) 36... Qb1+ $11) 27. Kc2 Qa2+ 28. Kc1 1/2-1/2 [Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.19"] [Round "8.3"] [White "Nepomniachtchi, Ian"] [Black "Giri, Anish"] [Result "1/2-1/2"] [ECO "B33"] [WhiteElo "2789"] [BlackElo "2776"] [Annotator "Monokroussos,Dennis"] [PlyCount "55"] [EventDate "2020.03.17"] [SourceVersionDate "2020.07.21"] 1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e5 6. Ndb5 d6 7. Bg5 a6 8. Na3 b5 9. Nd5 Be7 10. Bxf6 Bxf6 11. c3 Rb8 (11... Bg5) 12. Nc2 Bg5 13. g3 (13. a4) 13... O-O 14. h4 Bh6 15. a3 (15. Bh3) (15. Bg2) 15... a5 16. Qd3 $146 (16. Bg2 Ne7 17. Nce3 Bxe3 18. Nxe3 Be6 19. O-O {1/2-1/2 (19) Vitiugov,N (2722) -Sevian,S (2660) chess.com INT 2020}) 16... Ne7 (16... f5 {seems a sensible alternative.} 17. Bg2 f4 (17... Ne7 18. exf5 Bxf5 19. Be4 (19. Qe2 {is also fine, maintaining the tension and preparing Rd1.}) 19... Bxe4 {Not forced, and not bad.} 20. Qxe4 Nxd5 21. Qxd5+ Kh8 {If the queens were off the board White would be very happy. Here, it's equal.} 22. O-O Rf6 $11) 18. gxf4 Bxf4 { The position is very complicated, strategically, with play all over the board. It may be equal, but there's no tidy summary available to explain this. White could play a4, expanding the front. He could take on f4, castle queenside, and run the h-pawn. He could play Rd1 and go for the d-pawn. Note by the way that White can win a piece with Nxf4 followed by Qd5+ and Qxc6, and yet he shouldn't - at least right now.} 19. a4 (19. Nxf4 exf4 (19... Rxf4 {is also fine.} 20. Qd5+ $6 Kh8 21. Qxc6 $4 Bb7 22. Qxb5 Bxe4 $19) 20. O-O-O (20. Qd5+ Kh8 21. Qxc6 $6 Bb7 22. Qxb5 Bxe4 23. Qg5 Qxg5 24. hxg5 Bxg2 $15) 20... Qb6 $11 ) 19... bxa4 20. Nxf4 Rxf4 21. Qd5+ Kh8 22. Qxc6 Rxb2 (22... Bb7 $2 {is pointless here, thanks to 19.a4 bxa4.} 23. Qxa4 $18) 23. O-O-O Rb6 24. Qxa4 Be6 $44 {Black is at least equal in this crazy position.}) 17. Nce3 Bxe3 {All according to the standard Sveshnikov recipe, not allowing White a permanent knight-grip over d5.} 18. Nxe3 Be6 (18... f5 {is also playable - it's a matter of taste here.} 19. exf5 (19. Bg2 f4 20. gxf4 Rxf4 $15) 19... Bb7 20. Bg2 { Else 20...d5.} Bxg2 21. Nxg2 Nxf5 $11) 19. Rd1 Rb6 (19... b4 20. axb4 axb4 21. Qxd6 Qxd6 22. Rxd6 bxc3 23. bxc3 Rfc8 24. Nd1 {is probably equal, but there's no need for Black to sac a pawn and hope his compensation will endure.}) 20. Bh3 {Black can swap the bishops and get away with it, but why make it easier for White to use the d5 square?} Bb3 (20... Bxh3 21. Rxh3 Qd7 22. Rh1 (22. g4 Ng6) 22... f5 (22... b4)) 21. Rd2 Qc7 22. O-O Rfb8 23. Rc1 Rc6 24. Bg4 h6 25. Bd1 Be6 26. Bg4 Bb3 27. Bd1 Be6 28. Bg4 1/2-1/2 [Event "FIDE Candidates 2020"] [Site "Yekaterinburg RUS"] [Date "2021.04.19"] [Round "8.4"] [White "Alekseenko, Kirill"] [Black "Grischuk, Alexander"] [Result "1-0"] [ECO "C11"] [WhiteElo "2696"] [BlackElo "2777"] [Annotator "Monokroussos,Dennis"] [PlyCount "147"] [EventDate "2020.03.17"] [SourceVersionDate "2020.07.21"] 1. e4 e6 2. d4 d5 3. Nc3 Nf6 4. e5 Nfd7 5. f4 c5 6. Nf3 Nc6 7. Be3 cxd4 8. Nxd4 Qb6 9. Qd2 Qxb2 10. Rb1 Qa3 11. Bb5 Nxd4 12. Bxd4 a6 13. Bxd7+ Bxd7 14. Rb3 Qe7 15. Rxb7 Qh4+ 16. Qf2 (16. Bf2 {has been played many more times than 16.Qf2.} Qd8 17. Bb6 Qc8 18. Rc7 Qd8 19. Qd4 Ba3 20. Nb1 Be7 {and so on.}) 16... Qxf2+ 17. Kxf2 Be7 18. Nd1 $146 {The engine is not impressed by this novelty, which probably wasn't prepared.} (18. Rhb1 $14 Bc8 19. R7b3 O-O 20. Na4 a5 21. Nb6 Ra6 22. Rc3 Rxb6 23. Rxb6 f6 24. Rc7 fxe5 25. Bxe5 Bh4+ 26. g3 Bf6 27. Bxf6 gxf6 28. Rbc6 {1-0 (28) Wang,J (2380)-Zhang,X (2374) China 2016}) 18... Bd8 ({ Stockfish 13:} 18... Bd8 19. Nc3 Bc6 20. Rb3 O-O 21. Rhb1 f6 22. g3 Rc8 23. Rb8 Ba5 24. Rxc8 Rxc8 25. Rb3 Kf7 26. Ke3 Kg6 27. h3 Bd7 28. Kd3 h5 29. Rb7 Rc7 30. Rb8 Rc4 31. Rb7 Bxc3 32. Bxc3 Bb5 33. Kd2 Re4 34. Kd1 Ra4 35. exf6 gxf6 36. Re7 {[%eval -28,33] [%wdl 47,818,135]}) ({Stockfish 13:} 18... Bc6 19. Rc7 Bd7 20. Rb7 Bd8 21. Nc3 Bc6 22. Rb3 O-O 23. Rhb1 f6 24. Bc5 Re8 25. Bd4 Be7 26. Ke3 Rec8 27. Kd3 Kf7 28. g3 h6 29. h4 h5 30. Kd2 a5 31. Nb5 Rd8 32. a4 Rab8 33. c3 Rg8 34. Kd3 Rgc8 35. R3b2 Bd7 36. Nd6+ Bxd6 37. exd6 {[%eval -25,33] [%wdl 50, 823,127]}) ({Stockfish 13:} 18... h5 19. Ne3 Bb5 20. f5 Bc6 21. Rc7 Bb5 { [%eval 0,32] [%wdl 80,839,81]}) ({Stockfish 13:} 18... Bh4+ 19. g3 Bd8 20. Ne3 Bc6 21. Rb2 Rc8 22. Rc1 Bb5 23. c4 Ba5 24. Rbc2 Ba4 25. Rb2 {[%eval 0,32] [%wdl 80,839,81]}) 19. Ne3 Bc6 (19... Bb5 20. g3 (20. Ra1 O-O 21. c4 dxc4 22. a4 Bc6 23. Rb4 Be7 24. Rxc4 Rfc8 25. Rac1 Bd7 $11) 20... O-O 21. Ra1) 20. Rb2 O-O 21. Bc5 Re8 22. c4 $1 $11 f6 23. cxd5 exd5 24. Bd6 fxe5 25. fxe5 (25. Bxe5 $11) 25... d4 26. Nf5 d3 27. Ke3 Bb5 28. Ke4 (28. Nd4 Bb6 29. Ke4 Bxd4 { is the reason for 28.Ke4. Both lines are objectively equal, but this approach lets Black kill the game, while Alekseenko's keeps it going (for good or ill).} ) 28... Rc8 ({Trying to be extra-clever with} 28... Bb6 $2 {results in Black getting a lost game:} 29. a4 Bc6+ 30. Kxd3 $18) 29. Nd4 Bc4 (29... Rc4 30. Kxd3 {Not forced.} Rb4+ 31. Kc3 Rc4+ $11) 30. Rd1 a5 (30... g6) 31. a3 (31. Rb7 $142 ) 31... g6 (31... Bf6) 32. Rb7 Bf6 33. Nf3 Bg7 34. Ra7 (34. g4 $142) 34... Rcd8 (34... Bb3 $1 35. Rd2 Bc2 36. Rxa5 Bh6 37. Rxd3 Rc4+ 38. Nd4 Bf8 $1 $15) 35. Kd4 $2 {This isn't as bad as it looks, and in the end everything worked out swimmingly. Still, it's an error.} (35. Kf4 $11) (35. Re7 $11) 35... Bb3 { Winning material, thanks to the two threats of ...Bxd1 and ...Rxd6+.} 36. Rxg7+ $8 Kxg7 37. Rxd3 a4 (37... Bg8 {may have been better, resisting the temptation to anchor the bishop at the cost of giving White's pieces a permanent parking place on b4.}) 38. Nd2 Be6 $6 (38... Bc2 $142) 39. Ne4 $11 {White's extra space and grip on the dark squares is so strong that he enjoys full compensation for the exchange.} Bf5 40. Re3 Bxe4 41. Rxe4 {If Black could safely park his king on e6, White would be in serious trouble. But that doesn't seem to be possible.} Rf8 (41... Kf7 $2 42. Kd5 $18) 42. Ke3 Rfe8 43. Kf4 {White is playing for the win?! That's completely unjustified here, so a likelier explanation is that he believes the king is better placed here, so there's no reason to repeat moves and wait for Grischuk to find an improvement. } Ra8 (43... Kf7 {now the king gets where it wants, but at the cost of the a-pawn.} 44. Rxa4 h6 45. Ra7+ Ke6 46. Rg7 Rg8 47. Rh7 Rh8 48. Rg7 $11) 44. Rb4 Ra7 45. Ke4 Kf7 46. Rb6 Rc8 47. Bb4 Rac7 48. Kd5 Rc2 49. e6+ Kf6 50. Rb7 g5 ( 50... R8c7 51. Rxc7 Rxc7 52. Kd6 Rc4 53. e7 Kf7 54. Kd7 Re4 $11 {is a dead draw, with neither player able to improve his position.}) 51. Rxh7 Rxg2 52. Kd6 Kg6 53. Rh3 g4 (53... Re2 54. e7 Ra8 $11) 54. Re3 Rxh2 {Grischuk hasn't crossed the line yet in his attempt to win the game, but he's meandering ever closer to it.} 55. Kd7 Rhh8 56. e7 Rce8 $6 (56... Ra8 $1 57. e8=Q+ Raxe8 58. Rxe8 Rxe8 59. Kxe8 {is a draw, as White's king will not be able to grab the a-pawn without letting Black's king reach the a8 corner. These endings can be tricky (there was a game from the 1978 Karpov-Korchnoi match in which the side with the extra piece forced perfect defense from his opponent), but it is objectively drawn.}) 57. Bd6 Ra8 58. Bc7 (58. e8=Q+ Raxe8 59. Rxe8 Rxe8 60. Kxe8 {is a slightly better version of the same ending, but Dr. Tablebase continues to assure us that it's a draw.}) 58... Rag8 $2 (58... Rh7 $142) 59. Re4 Kf5 60. Rf4+ Kg5 61. Rxa4 {The reason to keep the rook on a8 was to prevent this. Now White will bring the rook back to the e-file and push the a-pawn, and that point the e8Q liquidation will win immediately.} Kh4 $2 (61... Rh7 $1 62. Be5 $1 Rf7 63. Bd6 Rh8 {and it's at least not clear to me if White is "officially" winning.}) 62. Re4 (62. Bd6) (62. Rf4) 62... Ra8 $1 63. a4 $1 Kg5 64. a5 {Now e8Q is a winning threat, so Black must try to stop it.} Rh7 65. Kc6 (65. Kd6 $1 Rh6+ 66. Re6 Rhh8 67. Re1 Kf5 68. Kd7 Rh7 69. Rf1+ Kg5 (69... Ke4 70. Rf4+ $18) 70. Bd8 g3 71. Kd6 Ra6+ 72. Kc5 Rxe7 73. Bxe7+ Kg4 74. Kb5 $18) 65... Rhh8 66. Kd7 Rh7 67. Kc6 $2 (67. Kd6 $1 $18) 67... Rhh8 68. Re3 Rhe8 $2 {The final error, and now Alekseenko wins.} (68... Kf5 $1 {and it seems that Black is still holding.}) 69. Kd7 Kf6 70. Re6+ Kf7 71. Re4 Kf6 72. Bd8 Kf5 73. Kxe8 Kxe4 74. Kf8 1-0