[Event "FIDE World Championship 2023"] [Site "Astana"] [Date "2023.04.21"] [Round "9"] [White "Nepomniachtchi, Ian"] [Black "Ding, Liren"] [Result "1/2-1/2"] [ECO "C65"] [WhiteElo "2795"] [BlackElo "2788"] [Annotator "Europe-Echecs"] [PlyCount "163"] [EventDate "2023.04.09"] [EventType "match"] [EventCountry "KAZ"] [SourceTitle "europe-echecs.com"] [Source "ChessBase"] [SourceQuality "1"] {[%evp 0,163,33,29,29,25,25,13,26,14,26,-2,-1,-1,18,26,18,6,10,-7,-5,0,0,4,-13,-13,16,16,3,16,16,15,21,14,5,-10,-21,-9,35,8,68,-7,19,-31,-16,-16,-16,17,24,24,31,60,66,52,54,-6,-20,-25,-14,-14,-17,0,14,0,0,0,10,17,37,32,31,31,32,32,26,31,23,28,108,113,99,90,92,95,95,89,93,85,139,77,92,72,75,81,89,84,90,88,101,96,101,105,105,94,109,114,137,125,129,111,97,97,150,144,155,122,120,120,102,101,101,87,87,88,101,101,101,101,101,101,103,101,101,101,101,101,101,101,101,82,101,101,101,101,101,101,101,101,122,101,101,35,101,0,0,0,0,0,0,0,0,0,0,0,0,0] C'est la partie de tous les dangers pour Ding Liren. Après avoir laissé échapper un gain qui lui aurait permis d'égaliser le match dans la huitième partie, une défaite dans la neuvième lui porterait sans aucun doute un coup fatal.} 1. e4 e5 2. Nf3 Nc6 3. Bb5 {La troisième partie Espagnole du match.} Nf6 {La variante berlinoise. Une surprise de la part de Ding Liren. La Pléiade berlinoise est le nom donné à un groupe de sept champions d'échecs constitué à Berlin au milieu du XIXe siècle avec Ludwig Bledow (1795-1846), Karl Schorn (1803-1850), Bernhard Horwitz (1808-1885), Carl Mayet (1810-1868), Wilhelm Hanstein (1811-1850), Paul Rudolf von Bilguer (1815-1840) et Tassilo von der Lasa (1817-1899), qui ont élaboré une nouvelle ouverture, la célèbre défense berlinoise, basée sur la contre-attaque du pion-roi des blancs par le Cavalier en f6 dans la partie espagnole.} 4. d3 Bc5 5. c3 O-O 6. O-O d5 7. Nbd2 dxe4 8. dxe4 {Le choix de cette Berlinoise démontre que le Chinois joue seulement pour le demi-point.} a5 ({La principale alternative est} 8... Qe7 9. b4 Bd6 10. Bb2 a5 11. a3 Be6 12. Qc2 h6 13. Rac1 Rfd8 14. Rfd1 {½-½ (80) Nakamura,H (2736)-Ding Liren (2799) Speed 2021 INT blitz}) 9. a4 $5 (9. Qe2 {donne l'égalité aux Noirs. Par exemple :} Qe7 10. Nh4 Bg4 11. Ndf3 Nh5 12. Qc4 Na7 13. Bg5 Qd6 14. Nf5 Bxf5 15. exf5 h6 {½-½ (32) Vachier Lagrave,M (2757)-Giri,A (2768) Tashkent FIDE GP 2014}) (9. Bxc6 {est connu pour faire nulle.} bxc6 10. Nxe5 Re8 11. Nb3 Qxd1 12. Rxd1 Bb6 13. Nc4 Nxe4 14. Nxb6 cxb6 15. f3 Nc5 16. Nxc5 bxc5 17. Rd6 {½-½ (32) Anand,V (2779)-Grischuk,A (2771) Gashimov Memorial 2019}) 9... Qe7 10. Qc2 {Viswanathan Anand : « La différence entre 10.Dc2 et 10.De2 est que le Fou peut reculer en f1. »} Nb8 {Jusqu'ici, « à l'est rien de nouveau ».} 11. Re1 (11. Be2 Nbd7 12. Nc4 b6 13. h3 Bb7 14. Bd3 Rfd8 15. Re1 h6 16. b3 Qe8 17. Be3 Bxe3 18. Rxe3 Qe6 19. Rae1 Ba6 20. Bf1 Bb7 21. Ncd2 Ne8 22. Nc4 Nd6 23. Rd1 Nc5 24. Nxd6 Rxd6 {0-1 So,W (2773)-Nakamura,H (2760) RCC INT rapid 2022}) 11... Rd8 12. h3 (12. Nf1 c6 13. Bc4 h6 14. Ng3 Ng4 15. Rf1 Be6 16. Bxe6 Qxe6 17. h3 {½-½ (35) So,W (2771)-Nakamura,H (2768) CGC KO 2022 INT rapid}) 12... h6 13. Nf1 (13. Bf1 Nbd7 14. Nc4 b6 15. Be3 Bb7 16. Bxc5 Qxc5 17. Rab1 Qe7 18. b4 Bc6 {½-½ (51) Nakamura,H (2768)-So,W (2771) CGC KO 2022 INT rapid}) 13... c6 (13... Nh5 14. Ne3 c6 15. Bc4 Qf6 16. Nf5 Bxf5 17. exf5 Nd7 18. Be3 Bxe3 19. Rxe3 {1-0 (45) Gopal,G (2580)-Gupta,A (2625) Pardubice Czech op-A 28th 2017}) 14. Bc4 Na6 {Viswanathan Anand : « La préparation de Ding semble en effet très efficace. »} 15. Ng3 Qc7 {Il ne reste qu'une partie dans les bases, jouée par email. Comme Ding Liren a blitzé ce coup et que Ian Nepomniachtchi prend une première respiration, le Russe est donc hors préparation.} 16. Ba2 $5 {Après 20 minutes de réflexion ; une éternité pour Ian. 1h42-1h23. Ce type de positions, très positionnelles, avec toutes les pièces encore sur l'échiquier, pourrait plutôt correspondre au stratège Ding Liren.} (16. Qe2 Bf8 17. Be3 Nc5 18. b4 Ncxe4 19. Nxe4 Nxe4 20. Bb6 Qxb6 21. Qxe4 Re8 22. Nxe5 Be6 23. Bd3 f5 24. Qf3 axb4 25. Bxf5 Bxf5 26. Qxf5 Qc7 27. Nd7 Rxe1+ 28. Rxe1 bxc3 29. Qe6+ Kh7 30. Qe4+ Kh8 31. Ne5 Bd6 {½-½ De la Rosa Solorzano,J (2426)-Genga,S (2424) W-ch ICCF email 2021}) 16... b5 {Selon Vladislav Tkachiev : « L'équipe chinoise a dominé [dans la préparation] l'équipe russe. »} 17. Qe2 {Attaque le pion b5, mais pour l'entraîneur Tom Rendle : « ... les Noirs peuvent simplement sacrifier le pion avec ...Ff8!, car prendre b5 permettrait alors ...Cc5 avec un excellent jeu pour le pion. » De son côté, Viswanathan Anand : « 17.De2 et Ch4, suivi de Df3, est un plan sérieux maintenant pour Ian. »} Rb8 $6 {Après 16 minutes de réflexion, le Chinois choisit un coup inattendu.} ({La ligne critique était} 17... Bf8 $1 18. axb5 cxb5 19. Qxb5 Nc5 20. b4 Nd3 21. Rd1 Ba6 22. Qxa5 Qxc3 23. Bd2 Qc2) 18. Nh4 $1 {Comme prévu par Viswanathan Anand, le Russe se tourne vers l'aile-Roi noire. Viswanathan Anand : « Je n'aime pas 17...Tb8. Il y a des lignes où le Cavalier en a6 n'est pas défendu. De plus, les coups des Blancs comme 18.Ch4 sont faciles. »} Bf8 $5 {En 13 minutes.} ({Bien que laid,} 18... Be6 $5 {pour se débarrasser du Fou blanc de cases blanches, était sans doute supérieur.} 19. Bxe6 fxe6 20. Nh5) 19. Qf3 {[%csl Ga2,Gc1,Gf3,Gg3,Yg8,Gh4] En moins de 2 minutes. Toujours le pragmatisme de Ian. La menace est simplement de prendre en h6.} (19. Ng6 $5 Nc5 20. Nxf8 Kxf8) 19... bxa4 {Après 15 minutes. Cinq pièces blanches visent le Roi noir ; Ding Liren est en grand danger.} 20. Bxh6 Nc5 {[%csl Rb2,Gb8,Gd8,Yf3,Yh6][%CAl Yb8b2,Yd8d3,Yd3f3] Menace de prendre en b2 et aussi de faire irruption en d3 avec l'autre Tour.} 21. Ng6 $5 ({Stockfish 15.1 préférait} 21. Bg5 $5 Rd3 22. Qe2 Nh7 $1 23. Be3 Rxe3 $5 24. Qxe3 Rxb2) 21... Rxb2 $5 {Avec la menace de prendre en a2 puis en g6, sans oublier l'irruption de l'autre Tour en d3..} (21... Rd3 $5 22. Re3 Rxe3 23. Qxe3 Qd6 $1 24. Nxe5 Be6) (21... gxh6 $2 22. Qxf6 Bg7 23. Ne7+ Kh7 24. Qxf7 $18) 22. Nxf8 {Difficile d'envisager autre chose que cette prise en f8, affaiblissant les cases noires autour du Roi adverse.} Rxf8 {0h50-0h48} ({Surtout pas} 22... Kxf8 $4 {à cause de} 23. Qxf6 {et 1-0.}) 23. Bg5 $6 {C'est avec cette manière de jouer, quelque peu nonchalante et rapide (2 minutes) que ce trouve sans doute le principal défaut de Ian Nepomniachtchi.} (23. Bc4 $1 {était le coup de la machine, avec avantage blanc.}) 23... Nh7 $1 {Simple et bon.} 24. Bc1 Rb5 $6 {A tempo, mais imprécis !} ({Ding Liren rate} 24... Rc2 $1 {qui maintenait l'égalité.}) 25. Ba3 $1 Re8 {Ian Nepomniachtchi a maintenant l'occasion de gagner une qualité.} 26. Bc4 {Après 15 minutes de réflexion.} (26. Bxc5 $5 Rxc5 27. c4 {[%csl Yc5] et si la Tour n'a pas de case, on ne voit pas comment elle pourrait être capturée.}) (26. Rad1 {était une alternative plus ou moins équivalente du point de vue de l'évaluation informatique.}) 26... Be6 $1 {0h32-0h39} 27. Bxe6 $5 {Inattendu !} ({Les Blancs n'ont sans doute pas aimé} 27. Bxb5 cxb5 {et la qualité a coûté le Fou de cases blanches, et les Blancs doivent se méfier des pions noirs de l'aile-Dame.}) ({La machine envisageait aussi} 27. Qe2 $5 Bxc4 28. Qxc4 Qe7 29. Nf5 Qe6 30. Qe2) 27... Nxe6 28. Nf5 c5 29. Qe2 {Et Sesse.net affiche son 0.00 !} (29. Bc1 $5) 29... Rb3 30. Qc4 Qc6 {0h27-0h28} 31. Bc1 Nf6 32. Qxa4 {Ian Nepomniachtchi récupère son pion, mais il n'est plus question d'un avantage pour les Blancs.} Qxa4 33. Rxa4 Rxc3 34. Bb2 Rb3 35. Bxe5 Rb4 36. Rxa5 Rxe4 {La partie prend le chemin de la nulle.} 37. Rxe4 Nxe4 38. Ra4 Nd4 {Ding Liren a pris son temps pour jouer ce coup liquidateur au prix d'un pion.} 39. Bxd4 cxd4 40. Rxd4 {Et le Chinois a raison, cette finale est nulle.} g6 {Le contrôle du 40e coup est atteint et les joueurs se voient attribuer une heure supplémentaire pour les 20 prochains coups.} 41. Ne3 Kg7 42. Rb4 Ng3 $5 {Fabiano Caruana : « C'est juste un coup amusant, parce que vous pouviez déplacer le Cavalier n'importe où et vous le mettez en g3. Cela ressemble presque à une blague, une taquinerie. Mais il n'y a aucune raison particulière de le mettre en g3, ça n'aide pas à la tâche défensive. »} 43. Rb7 Nf5 44. Ng4 {Évidemment, Ian Nepomniachtchi, avec un pion de plus, ne prend aucun risque à continuer de jouer.} Re7 $5 {Alors que l'échange des Cavaliers menait directement au demi-point, proposer l'échange des Tours exige de savoir exactement où l'on met les pieds.} 45. Rb5 {Ian Nepomniachtchi refuse le troc.} Re1+ 46. Kh2 Re2 47. Rb7 Nd6 {Avec tous les pions sur la même aile, cette position doit se terminer par la nulle. Toutefois, il est indispensable de jouer avec précision.} 48. Ra7 Kf8 49. Kg3 f5 50. Kf3 Re7 51. Ra8+ {Apparemment Ian Nepomniachtchi veut faire payer à son adversaire le fait d'avoir voulu jouer au malin avec son 42.Cg3!?.} Re8 {Force l'échange des Tours.} 52. Rxe8+ Kxe8 {Et nous obtenons ainsi la finale de Cavalier avec un pion de plus pour le Russe. Le Chinois va maintenant devoir démontrer que sa vision était correcte.} 53. Ne5 g5 $1 {Seul coup évidemment. Fabiano Caruana a souligné que Magnus Carlsen a réussi à battre Vishy Anand dans une finale de Cavalier à 3 contre 2 lors du Tata Steel Masters 2019.} 54. h4 $5 {Même si les échanges de pions facilitent la défense. Cependant, Ian Nepomniachtchi a pris 20 minutes pour ce coup, il a donc forcément une idée.} gxh4 {A tempo, bien entendu.} 55. Kf4 {On pouvait lire de l'inquiétude sur le visage de Ding Liren et nous commencions à craindre qu'une gaffe vienne mettre un terme à cette partie. Et à ce moment-là, Daniil Dubov : « Ding semble bouleversé. Son langage corporel est comme s'il se sentait perdu. »} h3 $1 {0h37-0h16} (55... Ne4 $2 {Perdait sur} 56. Nd3 Nd6 57. Ke5 Ke7 58. Nf4 Ne4 59. f3 Ng3 60. Ng6+ Kf7 61. Nxh4 $18) 56. gxh3 Ke7 {Viswanathan Anand : « Ayant perdu une finale de Cavalier avec 2 pions contre 3, je ne me serais pas précipité pour échanger les Tours. Ding connaît peut-être simplement ses finales. »} 57. Nc6+ Kf6 58. Nd4 Ne4 59. f3 Nf2 60. h4 Nd3+ {Avec un ajout de 15 minutes pour le reste de la partie, et maintenant un incrément de 30 secondes par coup.} 61. Kg3 Kg6 {C'est la première fois du match que les joueurs dépassent les 60 coups.} 62. Ne6 Kf6 63. Nf4 Nb4 {Daniil Dubov : « Il suffit de mettre le Roi noir en h6 et il n'y a plus rien à craindre. »} 64. Kf2 Ke5 65. Ke3 Nc2+ 66. Kd2 Nd4 67. Nd3+ Kf6 68. Ke3 Nc2+ 69. Kf4 Nd4 70. Kg3 Ne2+ 71. Kf2 Nd4 72. Nf4 Ke5 73. Ne2 Ne6 74. Kg3 Kf6 75. Kg2 Kg7 76. Kf2 f4 {La fin approche, les joueurs ne vont pas tarder à se séparer bons amis.} 77. Kg1 Kg6 78. Kg2 Kh6 79. Nc1 Kh5 80. Kh3 Nd4 81. Nd3 Nxf3 82. Nxf4+ {Mal embarqué dans cette Espagnole variante berlinoise, le Chinois a néanmoins obtenu le demi-point avec les pièces noires. Mené par le Russe Ian Nepomniachtchi 5 points à 4 avec encore cinq parties à jouer, Ding Liren aura trois fois les Blancs et tout reste donc possible dans ce match. Repos samedi 22 avril, rendez-vous dimanche à partir de 11h00.} 1/2-1/2 [Event "Nepo - Ding, WCC 2023: Game 9 - Nepo v L"] [Site "lichess.org"] [Date "2023.04.21"] [Round "?"] [White "Nepomniachtchi"] [Black "Ding Liren"] [Result "1/2-1/2"] [ECO "C65"] [Annotator "https://lichess.org/@/rowrulz"] [PlyCount "163"] [EventDate "2023.??.??"] {[%evp 0,163,29,16,16,25,26,13,14,14,26,-1,-1,-1,25,22,21,-3,0,-5,0,0,0,-13,14,0,16,15,16,15,3,24,14,-1,-21,-5,-4,-9,29,8,31,-31,19,-16,-16,17,17,24,21,37,60,60,52,54,56,-13,-13,-14,-17,-17,-31,-17,0,0,17,14,25,25,32,26,26,31,23,28,28,28,28,39,95,100,100,95,95,97,98,90,85,86,139,81,94,73,81,83,89,73,85,92,98,97,110,98,98,100,119,113,125,132,130,97,104,102,148,143,144,127,121,102,104,101,101,89,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,101,120,101,101,39,101,0,0,0,0,0,0,0,0,0,0,0,0,0]} 1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 {Ding switches to the Berlin - arguably THE top choice amongst professionals} 4. d3 {The most fighting approach keeping all the tension in the position. This shows to me that Nepo is fighting for a win in this game and not yet looking to draw his way to the title} (4. O-O Nxe4 5. Re1 {I was sort of expecting this approach from Nepo leading to quiet positions where white isn't carrying any risks (though the draw ration is high)}) 4... Bc5 5. c3 (5. Bxc6 {This has become a major option}) 5... O-O 6. O-O d5 (6... d6 7. Nbd2 (7. Bg5)) 7. Nbd2 (7. exd5) 7... dxe4 8. dxe4 a5 $5 {Preventing queenside expansion from black} (8... Qe7 9. b4 Bd6 10. a3) 9. a4 (9. Bxc6 bxc6 10. Nxe5 {Grabbing the pawn is of course the critical test} Re8 11. Nb3 Qxd1 12. Rxd1 Bb6 13. Nc4 {This line has been tested a few times at high level} (13. Nxc6 Bb7 14. Ncd4 Bxe4 {With the usual 2 bishop compensation for the pawn being sufficient}) 13... Nxe4 14. Nxb6 cxb6 15. f3 Nc5 16. Nxc5 bxc5 {With some small advantage for white but black should make a draw}) 9... Qe7 10. Qc2 {The players are following a game Radjabov - Nakamura from the 2022 Candidates (at which both Nepo and Ding were present)} Nb8 {The most played move. This is the 'Breyer' manoeuvre} (10... Na7 {Radjabov, T. - Nakamura, Hi, 1-0, FIDE Candidates 2022, https://lichess.org/payNb8D6}) 11. Re1 Rd8 (11... c6 12. Bf1 {Gives white extra option of Bf1 which he doesn't have after Nf1}) 12. h3 h6 (12... Nbd7 {Anand, V. - Vidit, S., 0-1, Tata Steel India Blitz, 2019, https://lichess.org/xj4iUoeW}) 13. Nf1 c6 14. Bc4 (14. Bd3) 14... Na6 $5 {Usually the knight goes to d7 after b8.} (14... Nbd7) 15. Ng3 {Some ideas for white include Nh4 (aiming at f5 or even g6 if the bishop goes to f8), Qe2/Be3} Qc7 {Ding was playing quickly and seemingly following the top choice of the engine and some correspondence games. ...Qc7 hints at ...Bf8...Nc5. another factor here is that by defending the e-pawn ...Be6 becomes an option in some lines} (15... Nc7 16. Qe2 {and it 's slightly annoying for black that he can't play ...Be6 yet} (16. Nh4 Be6 17. Nhf5 Qf8 $10) 16... Be6) 16. Ba2 (16. Nh4 Bf8 (16... Re8 $5 {Top choice of the engine hinting at ...Be6}) 17. Be3 Nc5 18. Ng6 Bd6 19. Nxe5 {is something black should avoid}) 16... b5 17. Qe2 Rb8 $6 {Not liked by Anand or the engine} (17... Bf8 18. axb5 cxb5 19. Qxb5 Nc5 {With good activity for black compensating the pawn deficit. Ding was worried here about b4} 20. b4 Nd3 21. Rd1 Ba6 22. Qxa5 Qxc3 23. Bd2 Qc2) 18. Nh4 $1 Bf8 19. Qf3 (19. Ng6 $5) 19... bxa4 20. Bxh6 (20. Ng6 $5) 20... Nc5 21. Ng6 (21. Bg5 $5 Rd3 22. Qe2 Nh7 23. Be3) 21... Rxb2 {Ding bravely grabs the pawn!} 22. Nxf8 Rxf8 (22... Kxf8 $4 23. Qxf6 $18) (22... Rd3 $5 {This interim move avoids the Qe3 lines - but it seems very computerish and messy} 23. Be3 Nb3 24. Bxb3 axb3 25. Nh5 Qd6 $1 {with strange complications}) 23. Bg5 $6 {White chases the knight to h7 before playing Bc1, which seems like a nice subtlety but in a critical line the knight is actually better on h7 than f6 !} (23. Qe3 $5 {[%CAl Ge3g5]} gxh6 24. Qxc5 (24. Qxh6 Ne8 {[%CAl Gc7d6] and now white should take the perpeptual}) 24... Qb6 25. Qxb6 Rxb6 26. Bc4 $14 {This looks like a nice endgame for white with some pawn weaknesses}) (23. Bc1 {This is stronger immediately} Rc2 24. Ba3 Nb3 25. Bxf8 Kxf8 {In this line the king has to take on f8 which makes a difference} 26. Bxb3 axb3 27. Qe3 $1 {[%CAl Ge3c5]} Be6 28. c4 Bxc4 29. Nf5 {With tactical issues for black on the kingside}) 23... Nh7 24. Bc1 Rb5 $6 (24... Rc2 $1 {Weird looking move but it works!} 25. Ba3 $2 {This natural move is a mistake} (25. Bb1 $2 Rxc1 26. Rxc1 Nb3) 25... Nb3 $1 26. Bxf8 Nxf8 {The knight taking on f8 makes a difference} 27. Bxb3 axb3 $17 {Black is on top with the connected passed pawns} 28. Qe3 Be6 29. c4 Bxc4 30. Nf5 {Here with the knight on f8 this is an empty shot compared to 23.Bc1 line}) 25. Ba3 Re8 26. Bc4 Be6 27. Bxe6 (27. Bxb5 cxb5 {Looks unclear and perhaps Nepo didn't want to allow any counter-chances. Still this was perhaps the way to go to press for a win}) (27. Qe2 Bxc4 28. Qxc4 Ne6 29. Nf5 Rd8 30. Qxa4 $14 {Was perhaps a better way to try and press}) 27... Nxe6 28. Nf5 c5 29. Qe2 Rb3 30. Qc4 Qc6 {The position is now pretty even} 31. Bc1 Nf6 32. Qxa4 Qxa4 33. Rxa4 Rxc3 34. Bb2 Rb3 35. Bxe5 Rb4 36. Rxa5 Rxe4 37. Rxe4 Nxe4 38. Ra4 Nd4 $1 {Ding decides to give up a pawn to simplify the position. i think this is a good practical decision as the other lines were complicated and white may have ended up with better winning chances there} (38... Nd2 {Is playable but looks a bit loose and there is certainly stuff to calculate here} 39. Bxg7 $4 (39. Rg4 Nf4 $1 (39... g6 40. Bc3 {Looks a bit tricky for black} (40. Rh4 $2 f6 $1 41. Bc3 Nb1 42. Nd6 Rd8 43. Ne4 Nxc3 44. Nxc3 $15 {[%csl Gc5]}) 40... Rd8 41. Rh4 f6 {forced} 42. Bxf6 Nf3+ $1 43. gxf3 Rd1+ 44. Kg2 gxf5 {maybe white can still press here}) 40. Nh6+ (40. Rxg7+ Kh8 {this is the most accurate way for black but looks scary at least from afar} 41. Bxf4 Re1+ 42. Kh2 Nf1+ $10) (40. Bxf4 Re1+ 41. Kh2 Nf1+ $10) 40... Kh7 41. Nxf7 Ng6 42. f4 Nxe5 43. Nxe5 (43. fxe5 Kg8 $10) 43... c4 {with counterplay}) 39... Nxg7 40. Rg4 Re1+ 41. Kh2 Nf1+ 42. Kg1 Ne3+ $19) (38... N4g5 39. f4 (39. h4 f6 40. hxg5 fxe5 41. g6 {Black has some problems to solve here as well}) 39... f6 40. fxg5 fxe5 41. g6 {with problems for black to solve}) (38... f6 39. Rxe4 Ng5 40. Nxg7 Nxe4 41. Nxe8 fxe5 42. Nc7 c4 43. Nd5 c3 44. Nb4 {could be a draw - but it is clear all these lines are more dangerous than 38...Nd4!}) 39. Bxd4 cxd4 40. Rxd4 g6 {White has won a pawn but black's defensive task is relatively straightforward and both players are obviously aware that this should be a draw. However a long defence could be in store for black} 41. Ne3 Kg7 42. Rb4 Ng3 43. Rb7 (43. fxg3 Rxe3 {is a draw}) 43... Nf5 $5 44. Ng4 {No hesitation from Nepo} (44. Nxf5+ gxf5 {Ding is confident that even this rook endgame is an easy draw}) 44... Re7 45. Rb5 {Again no hesitation} (45. Rxe7 Nxe7 {In this endgame there are more winning chances than the Rook endgame. Dvoretsky mentions that 4v3 is usually winning but there are far more defensive resources in 3 v 2 knight endgames. Still some white players may choose to go this way}) 45... Re1+ 46. Kh2 Re2 47. Rb7 {Prevents ...Nh6. White wants to keep all the pieces on the board, for now black is making it harder for white to move the knight back into the game} (47. g3 Nh6 $10) 47... Nd6 48. Ra7 Kf8 (48... Ne4 {was possible again trying to exchange knights} 49. f3 (49. Re7 Nc3) 49... Nf2 $10) 49. Kg3 (49. Rd7 f5 $4 50. Rxd6 fxg4 51. h4 $1 {white wins} Rxf2 (51... Kf7 52. Kg3 Re4 53. Rd5 {[%CAl Gd5g5,Gg5g4]}) 52. Rxg6 $18 {[%csl Rg4]}) 49... f5 50. Kf3 Re7 51. Ra8+ Re8 52. Rxe8+ Kxe8 {Ding gets his rook exchange but looks like he did it in quite a risky way. Nepo's facial expressions indicated he was surprised though he often does this} 53. Ne5 g5 {If the white king were on e3 here then f4 would be winning which perhaps indicates that this was a risky choice from Ding - perhaps he had calculated everything in advance?} 54. h4 {Nepo goes for one of the direct choices} gxh4 {only move} 55. Kf4 {Now black only has 2 moves which hold} h3 $1 {The most practical move} (55... Ke7 {This is the most natural of the two} 56. Ng6+ Kf6 57. Nxh4 Ne4 58. f3 Nd6 59. g3 Ke6 60. Kg5 Nf7+ 61. Kg6 Ne5+ 62. Kg7 {This was the position being discussed in commentary with the view that it looks quite scary for black if you can't force an immediate draw. Here tablebase indicates two ways to draw - but they are both quite difficult I think} Nd3 (62... Nc4 $2 63. f4 {black will be zugzwanged} Ne3 (63... Kd5 64. Nxf5) 64. Kg6 $18) (62... Nc6 63. f4 Ne7 64. Nf3 Nc6 65. Kg6 Ne7+ 66. Kg5 Nc6 {and black holds}) 63. f4 Kd5 $1 {only drawing move} 64. Kf6 (64. Nxf5 Nxf4 65. gxf4 Ke4 $10) 64... Ke4 65. Kg5 {and black still needs to be accurate here!} Nf2 (65... Nc5 {even draws in a more unique way} 66. Nxf5 Kf3 {with a study-like draw!}) 66. Nxf5 Nh3+ 67. Kg4 Nf2+ $10) (55... Ne4 $2 {This natural move loses} 56. Nd3 $1 {the black knight has to go back} (56. f3 $2 Nf2 57. Kxf5 (57. Ke3 h3 $10 58. Kxf2 $4 h2 $19) 57... h3 $10) 56... Nd6 (56... Kf7 57. Kxf5 h3 58. gxh3 {is a win for white according to tablebases}) 57. Ke5 Ke7 58. Nf4 Ne4 (58... Kd7 59. Ng6 Ne4 60. f4 $1 $18 (60. f3 $4 h3 $1 61. gxh3 Ng5 $10)) 59. f3 Nd6 (59... Ng3 60. Ng6+ Kf7 61. Nxh4 $18) (59... h3 60. Nxh3 Ng3 61. Nf2 Kf7 62. Kf4 Nf1 63. Nd1 $1 (63. Kxf5 $4 Ne3+ $10) 63... Kf6 64. Ne3 $18) 60. Ng6+ Kd7 61. Nxh4 $18) 56. gxh3 Ke7 57. Nc6+ (57. Kg5 Ne4+ $10) 57... Kf6 (57... Ke6 $4 58. Nd4+ Kf6 59. h4 (59. f3 $2 Nf7 $1 60. Nxf5 Ng5 61. h4 Nxf3 $10) 59... Kg6 60. f3 Kf6 61. h5 $18 {zugzwang}) 58. Nd4 Ne4 59. f3 Nf2 60. h4 Nd3+ 61. Kg3 Kg6 {Ding has survived the hardest phase} 62. Ne6 Kf6 63. Nf4 Nb4 64. Kf2 Ke5 65. Ke3 Nc2+ 66. Kd2 Nd4 (66... Kxf4 67. Kxc2 Kxf3 68. h5 Kg2 69. h6 f4 70. h7 f3 71. h8=Q f2 {is a theoretical draw but understandable that ding didn't want to check this as he is holding easily as it is}) 67. Nd3+ Kf6 68. Ke3 Nc2+ 69. Kf4 Nd4 70. Kg3 Ne2+ 71. Kf2 Nd4 72. Nf4 Ke5 73. Ne2 Ne6 74. Kg3 Kf6 75. Kg2 Kg7 76. Kf2 f4 77. Kg1 Kg6 78. Kg2 Kh6 79. Nc1 Kh5 80. Kh3 Nd4 81. Nd3 Nxf3 82. Nxf4+ {draw agreed} 1/2-1/2 [Event "WCh 2023"] [Site "Astana KAZ"] [Date "2023.04.21"] [Round "9.1"] [White "Nepomniachtchi, Ian"] [Black "Ding Liren"] [Result "1/2-1/2"] [ECO "C65"] [Annotator "Crowther,Mark"] [PlyCount "163"] [EventDate "2023.04.09"] {[%evp 0,163,19,31,34,42,20,27,32,10,7,22,10,-24,-8,6,28,1,0,-15,-8,-8,15,-8,-7,-19,-6,-20,-5,-5,0,16,23,0,-1,-3,9,-16,34,15,41,-17,-14,-67,-50,-46,-3,21,-17,3,53,40,67,56,59,-11,-43,-19,-15,-45,-31,-21,-5,-5,-5,0,0,7,62,49,39,42,32,42,34,45,27,32,107,101,100,100,97,89,103,89,86,85,140,70,87,68,78,74,74,78,85,89,89,98,116,103,98,104,116,115,113,112,118,129,113,100,142,149,162,127,127,126,127,95,109,105,109,105,110,109,103,108,108,91,90,122,121,122,108,108,108,108,108,104,104,104,109,106,104,113,122,104,120,104,104,55,104,0,0,0,0,0,0,0,0,0,0,0,0,0]} 1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. d3 Bc5 5. c3 O-O 6. O-O d5 (6... d6) (6... Re8) ({Relevant:} 6... Re8 7. Bg5 a6 8. Bxc6 dxc6 9. Bh4 Be7 10. Bg3 Bd6 11. Nbd2 Nh5 12. Bh4 Qd7 13. Nc4 Qg4 14. Bg5 Nf4 15. Ne3 Qh5 16. Bxf4 exf4 17. Nc4 Bg4 18. h3 Bxh3 19. gxh3 Qxh3 20. Nxd6 cxd6 21. Ng5 Qh4 22. Nf3 Qh3 23. Ng5 Qh4 24. Nf3 Qh3 {?-? (24) Firouzja,A (2778)-Aronian,L (2775) Saint Louis 2022}) 7. Nbd2 dxe4 8. dxe4 a5 9. a4 {A slightly unusual move but still played at the very top level.} (9. Qe2) 9... Qe7 10. Qc2 Nb8 {The top scoring move.} 11. Re1 Rd8 12. h3 h6 13. Nf1 c6 {This has been played before but perhaps surprised Nepomniachtchi.} 14. Bc4 Na6 $146 (14... Nbd7 15. Ng3 Nb6 16. Bf1 Re8 17. Nh4 Nh7 18. Nhf5 Bxf5 19. Nxf5 Qf6 20. Be3 Bxe3 21. Rxe3 Qe6 22. Rg3 g6 23. Nxh6+ Kh8 24. Re1 Kg7 25. Nf5+ Kh8 26. Ne3 Rad8 27. b3 Rd6 28. c4 Nd7 29. c5 Rd4 30. Nc4 Nxc5 31. Nxa5 Qe7 {Morkunas,G (1905)-Krivonosov,O (2378) Panevezys op-A 2022 (9) ?-?}) 15. Ng3 Qc7 {[#] Probably still prep from Ding.} 16. Ba2 $146 {A very good waiting move.} ({Predecessor:} 16. Qe2 Bf8 17. Be3 Nc5 18. b4 Ncxe4 19. Nxe4 Nxe4 20. Bb6 Qxb6 21. Qxe4 Re8 22. Nxe5 Be6 23. Bd3 f5 24. Qf3 axb4 25. Bxf5 Bxf5 26. Qxf5 Qc7 27. Nd7 Rxe1+ 28. Rxe1 bxc3 29. Qe6+ Kh7 30. Qe4+ Kh8 31. Ne5 Bd6 {?-? (31) De la Rosa Solorzano,J (2426)-Genga,S (2424) ICCF email 2021}) 16... b5 (16... Bf8 17. Be3) 17. Qe2 Rb8 $6 {White seems a bit better after this.} (17... Bf8 18. axb5 cxb5 19. Qxb5 Nc5 {and black has full compensation for the pawn.}) 18. Nh4 $1 {Immediately putting black under pressure.} Bf8 19. Qf3 (19. Ng6 $5) 19... bxa4 {Probably had to be tried but black's pawn structure is very ugly.} 20. Bxh6 Nc5 21. Ng6 (21. Bg5) 21... Rxb2 (21... Rd3 $5 {Is interesting.}) 22. Nxf8 Rxf8 (22... Rd3 {is a try.} 23. Be3 Nb3 24. Nh5 Qe7 25. Bxb3 axb3 26. Nxg7 Rxe3 27. Qxe3) 23. Bg5 (23. Bc4 $5 Be6 24. Nf5 Bxf5 (24... Bxc4 25. Bxg7 {isn't playable.}) 25. exf5) 23... Nh7 24. Bc1 Rb5 (24... Rc2 {is best according the the engines. But the rook looks like it might get trapped.}) 25. Ba3 Re8 26. Bc4 Be6 {This offer of the exchange is strictly necessary.} 27. Bxe6 (27. Bxb5 {is obviously critical.} cxb5 28. Qe3) 27... Nxe6 28. Nf5 c5 29. Qe2 (29. Bc1) 29... Rb3 30. Qc4 Qc6 31. Bc1 Nf6 32. Qxa4 {This liquidates to an endgame a pawn up.} Qxa4 33. Rxa4 Rxc3 34. Bb2 Rb3 35. Bxe5 Rb4 36. Rxa5 Rxe4 37. Rxe4 Nxe4 38. Ra4 Nd4 {Black goes into an endgame which he should hold but he's signing up for a bit of suffering.} (38... Nd2 {The engines make this work but it's not surprising Ding couldn't make it work.} 39. Rg4 Nf4 $11) 39. Bxd4 cxd4 40. Rxd4 g6 41. Ne3 Kg7 42. Rb4 Ng3 43. Rb7 Nf5 44. Ng4 (44. Nxf5+ gxf5 {looks ugly but probably is holdable.}) 44... Re7 45. Rb5 Re1+ 46. Kh2 Re2 47. Rb7 Nd6 {Ding spent quite a bit of time on this.} 48. Ra7 Kf8 49. Kg3 f5 {This starts to get tricky now.} (49... Rb2) 50. Kf3 Re7 51. Ra8+ Re8 52. Rxe8+ Kxe8 53. Ne5 g5 54. h4 gxh4 55. Kf4 h3 $1 {Ding thought this was the only way to draw.} (55... Ne4 56. Nd3 {and white wins.}) 56. gxh3 Ke7 57. Nc6+ Kf6 58. Nd4 Ne4 59. f3 Nf2 60. h4 Nd3+ 61. Kg3 Kg6 62. Ne6 Kf6 63. Nf4 Nb4 64. Kf2 Ke5 65. Ke3 Nc2+ 66. Kd2 Nd4 67. Nd3+ Kf6 68. Ke3 Nc2+ 69. Kf4 Nd4 70. Kg3 Ne2+ {The worst is over for Ding now.} 71. Kf2 Nd4 72. Nf4 Ke5 73. Ne2 Ne6 74. Kg3 Kf6 75. Kg2 Kg7 76. Kf2 f4 77. Kg1 Kg6 78. Kg2 Kh6 79. Nc1 Kh5 80. Kh3 Nd4 81. Nd3 Nxf3 82. Nxf4+ 1/2-1/2 [Event "FIDE World Chess Championship 2023"] [Site "Chess.com"] [Date "2023.04.21"] [Round "9"] [White "Nepomniachtchi, Ian"] [Black "Ding, Liren"] [Result "1/2-1/2"] [ECO "C65"] [WhiteElo "2795"] [BlackElo "2788"] [Annotator "Rafael"] [PlyCount "163"] [EventDate "2023.??.??"] [TimeControl "7200+30"] {[%evp 0,163,32,23,23,23,23,22,19,13,17,7,-11,6,8,30,25,6,12,12,12,0,5,-6,-14,-8,0,12,16,9,15,24,24,-2,0,-5,-5,2,29,17,15,-11,26,0,0,0,0,11,11,-13,24,48,64,75,56,-16,-4,-15,-17,-30,-21,0,17,-9,0,0,0,0,47,31,46,36,36,39,33,33,33,33,103,105,105,97,93,99,99,86,89,89,137,81,94,67,81,78,84,84,90,82,86,87,101,97,97,92,118,124,119,129,137,103,107,105,153,146,146,120,130,121,118,104,103,104,104,108,108,108,107,108,103,103,92,104,103,104,103,103,108,108,108,103,108,108,108,108,108,108,108,103,112,108,108,38,108,38,0,0,0,0,0,0,0,0,0,0,0,0]} 1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 {The Berlin Defense, widely considered Black's most reliable option against 1.e4, makes its first appearance in this duel. This is impressive, as a feature of world championship matches is that the openings are usually repeated. This is yet another feature that makes this match unique.} 4. d3 Bc5 5. c3 O-O 6. O-O d5 7. Nbd2 dxe4 8. dxe4 a5 {All of this is a critical variation of the Anti-Berlin and was played in an important game of the Candidates Tournament. The idea of this move is to gain space on the queenside and prevent b2-b4.} (8... Qe7 {was played by Ding in a blitz game.} 9. b4 Bd6 10. Bb2 a5 11. a3 Be6 12. Qc2 h6 {Nakamura-Ding, Speed Chess (blitz) 2021.}) 9. a4 {A good option, fixing the pawn structure on the queenside.} (9. Bxc6 {is a critical move, but this has been worked out by the computers, as it often happens in forcing lines in the opening.} bxc6 10. Nxe5 Re8 {and some games by elite players prove that Black manages to equalize.}) 9... Qe7 10. Qc2 Nb8 $1 {[%c_effect b8;square;b8;type;GreatFind;persistent;true] This is a typical maneuver, improving the knight's position. The pawn will go to c6, and the knight can choose whether to go to a6 or d7.} (10... Na7 {was played by Nakamura in the Candidates Tournament, but it's less accurate.} 11. Be2 Nc6 12. Nb3 Ba7 $6 {[%c_effect a7;square;a7;type;Inaccuracy;persistent;true]} 13. Bb5 {with an advantage for White in Radjabov-Nakamura, Candidates Tournament Madrid 2022.}) 11. Re1 Rd8 {Not the easiest move to play since Black has natural alternatives, but Ding surely was still following his preparation.} 12. h3 {A typical move, avoiding problems with Bg4 or Ng4.} h6 13. Nf1 c6 14. Bc4 Na6 $5 {[%c_effect a6;square;a6;type;Interesting;persistent;true] A surprising route for the knight, as Nd7-b6 is a simple alternative. But Ding's plan is more active and better.} (14... Nbd7) 15. Ng3 Qc7 $1 {[%c_effect c7;square;c7;type;GreatFind;persistent;true] Black's plan becomes clear. He wants to retreat the bishop to f8, protect the king, and bring the knight to c5.} 16. Ba2 {Nepo thought for a long time before playing this mysterious and not very convincing move.} (16. Qe2 {was tried in a recent correspondence game. The plan is to bring the queen to the attack with Nh4-Qf3, but Black is in time.} Bf8 17. Be3 (17. Nh4 Nc5 18. Qf3 Nh7 $5 {[%c_effect h7;square;h7;type;Interesting;persistent;true] Other moves are possible as well.}) 17... Nc5 18. b4 Ncxe4 19. Nxe4 Nxe4 20. Bb6 Qxb6 21. Qxe4 Re8 22. Nxe5 Be6 23. Bd3 f5 {De la Rosa Solorzano-Genga, email 2021.}) (16. Nh4 Bf8 17. Ng6 Bd6 {is not dangerous for Black, despite the intruding knight.}) 16... b5 (16... Bf8 {is also natural and good, following the plan.} 17. Be3 Nc5 18. Rad1 Rxd1 (18... Be6 $2 {[%c_effect e6;square;e6;type;Mistake;persistent;true] as Caruana showed during the broadcast, here we can see a possible idea of the Ba2 move.} 19. Bxc5 $1 {[%c_effect c5;square;c5;type;GreatFind;persistent;true]} Bxc5 (19... Bxa2 $2 {[%c_effect a2;square;a2;type;Mistake;persistent;true]} 20. Bxf8 Kxf8 21. b3 {the bishop is trapped.}) 20. Bxe6 fxe6 {Black's pawn structure is spoiled.}) 19. Rxd1 b5 20. axb5 cxb5 {Black has a good position.}) 17. Qe2 $1 {[%c_effect e2;square;e2;type;GreatFind;persistent;true] White attacks the b5-pawn while preparing the attack with Nh4-Qf3. This is a critical moment in the game.} Rb8 $6 {[%c_effect b8;square;b8;type;Inaccuracy;persistent;true] An inaccuracy that leads to a difficult position for Black. This move is too slow.} (17... b4 $6 {[%c_effect b4;square;b4;type;Inaccuracy;persistent;true] is a positional mistake, giving up control of the important c4-square.} 18. Bc4) (17... bxa4 {is possible, although it's a move one usually wants to avoid, as it may weaken the queenside.} 18. Bc4 Bf8 (18... a3 19. b3 $1 {[%c_effect b3;square;b3;type;GreatFind;persistent;true]} Qe7 20. Nh4 {with the initiative for White.}) 19. Rxa4 Nc5 20. Ra2 {White's position is more comfortable due to the better pawn structure.}) (17... Bf8 $1 {[%c_effect f8;square;f8;type;GreatFind;persistent;true] is the best move. A dynamic pawn sacrifice to activate the pieces.} 18. Be3 {It's better not to accept the pawn.} (18. axb5 cxb5 19. Qxb5 Nc5 {with excellent compensation, as the knight is going to d3 and the bishop to a6.}) 18... Nc5 (18... bxa4 19. Bc4 Nc5 20. Bxc5 Bxc5 21. Rxa4 {with a roughly equal game.}) 19. axb5 cxb5 20. Bxc5 Bxc5 21. b4 $1 {[%c_effect b4;square;b4;type;GreatFind;persistent;true]} Ba7 $1 {[%c_effect a7;square;a7;type;GreatFind;persistent;true]} (21... axb4 $2 {[%c_effect b4;square;b4;type;Mistake;persistent;true]} 22. Bxf7+ {loses material.}) 22. Qxb5 Qxc3 {with enough counterplay for a balanced game.}) 18. Nh4 $1 {[%c_effect h4;square;h4;type;GreatFind;persistent;true] Suddenly, the attack on the kingside gets serious.} Bf8 19. Qf3 (19. axb5 {This capture is a move White generally wants to avoid, as now it's easier for Black to defend the kingside through the sixth rank.} cxb5 20. Qf3 Rb6 $5 {[%c_effect b6;square;b6;type;Interesting;persistent;true]} (20... b4 {The computer prefers this move, defying my positional explanation for not capturing on b5.}) 21. Be3 Bc5 {with an acceptable position.}) 19... bxa4 {It's necessary to go for some counterplay, as defensive moves are not enough.} (19... Nh7 $6 {[%c_effect h7;square;h7;type;Inaccuracy;persistent;true]} 20. Ng6 {with a dangerous attack.}) (19... Kh7 $6 {[%c_effect h7;square;h7;type;Inaccuracy;persistent;true]} 20. Ngf5 Ng8 21. Qh5 $1 {[%c_effect h5;square;h5;type;GreatFind;persistent;true]} g6 22. Nxg6 $1 {[%c_effect g6;square;g6;type;GreatFind;persistent;true]} fxg6 23. Bxg8+ Kxg8 24. Qxg6+ Kh8 25. Nxh6 {with a decisive attack.}) (19... Nc5 {is also possible, likely transposing to the game after} 20. Bxh6 bxa4) 20. Bxh6 $1 {[%c_effect h6;square;h6;type;GreatFind;persistent;true]} Nc5 (20... gxh6 $2 {[%c_effect h6;square;h6;type;Mistake;persistent;true]} 21. Qxf6) (20... Rxb2 $2 {[%c_effect b2;square;b2;type;Mistake;persistent;true]} 21. Bg5) 21. Ng6 $5 {[%c_effect g6;square;g6;type;Interesting;persistent;true]} (21. Bg5 {is very tempting and enough for an advantage.} Rd3 $1 {[%c_effect d3;square;d3;type;GreatFind;persistent;true] This is the only possible defense.} 22. Qe2 Nh7 23. Be3 {Black is suffering in this position.}) 21... Rxb2 22. Nxf8 Rxf8 {White has many interesting options now.} 23. Bg5 $6 {[%c_effect g5;square;g5;type;Inaccuracy;persistent;true] Not the best one.} (23. Bxg7 {is a direct attempt to mate, but it backfires.} Kxg7 24. Qe3 Ne6 $1 {[%c_effect e6;square;e6;type;GreatFind;persistent;true]} 25. Nf5+ Kh8 26. Qh6+ Nh7 27. Bxe6 fxe6 {Everything is defended by miracle.}) (23. Bc1 Rb5 24. Ba3 {is similar to the game, but it's an improved version for Black with the knight on f6.} Rd8 25. Bc4 Nb3 $1 {[%c_effect b3;square;b3;type;GreatFind;persistent;true]} 26. Bxb5 cxb5 {with compensation for the exchange.}) (23. Qe3 $5 {[%c_effect e3;square;e3;type;Interesting;persistent;true] is natural and good.} gxh6 (23... Ne6 24. Bxe6 (24. Nf5 {is also very dangerous.}) 24... fxe6 (24... Bxe6 $2 {[%c_effect e6;square;e6;type;Mistake;persistent;true]} 25. Bxg7 Kxg7 26. Qg5+ {wins.}) 25. Qc1 $1 {[%c_effect c1;square;c1;type;GreatFind;persistent;true]} Rb3 26. Be3 {with a much better position for White due to Black's horrible pawn structure.}) 24. Qxc5 (24. Qxh6 {is enough only for a draw.} Ne8 25. Qg6+ (25. Re3 Qd6 $1 {[%c_effect d6;square;d6;type;GreatFind;persistent;true]}) 25... Kh8 26. Qh6+ {with perpetual check.}) 24... Qb6 $1 {[%c_effect b6;square;b6;type;GreatFind;persistent;true] Black manages to exchange queens.} 25. Qxb6 Rxb6 26. Bc4 Be6 27. Bxe6 fxe6 28. Rxa4 Ra8 {Black has good drawing chances, despite the bad pawn structure.}) (23. Bc4 $1 {[%c_effect c4;square;c4;type;GreatFind;persistent;true] This is an unpleasant move to face. White avoids the threat of Nb3 and maintains all the advantages of his position.} Be6 24. Nf5 Bxf5 25. Qxf5 {with a strong initiative.}) 23... Nh7 24. Bc1 Rb5 $6 {[%c_effect b5;square;b5;type;Inaccuracy;persistent;true]} (24... Rc2 $1 {[%c_effect c2;square;c2;type;GreatFind;persistent;true] is the best move, although it's really difficult to place the rook on such a strange square.} 25. Bc4 (25. Ba3 $6 {[%c_effect a3;square;a3;type;Inaccuracy;persistent;true]} Nb3 $1 {[%c_effect b3;square;b3;type;GreatFind;persistent;true]}) 25... Nb3 26. Rxa4 Nxc1 27. Qe3 Be6 28. Rxc1 Rxc1+ 29. Qxc1 Bxc4 30. Rxc4 Ra8 {with a likely draw.}) 25. Ba3 Re8 {Ding prepares an exchange sacrifice.} 26. Bc4 Be6 {Another critical moment of the game.} 27. Bxe6 $6 {[%c_effect e6;square;e6;type;Inaccuracy;persistent;true] This makes things easier for Ding.} (27. Bxb5 cxb5 28. Qe3 Nd7 {Black has some compensation for the exchange, but the position is still dangerous.}) (27. Qe2 $1 {[%c_effect e2;square;e2;type;GreatFind;persistent;true] is an interesting way to continue, playing for a positional advantage instead of material.} Bxc4 28. Qxc4 {The queenside pawns are weak, and White's knight is ready to jump to f5. Black has a tough defense ahead.}) 27... Nxe6 28. Nf5 c5 {Necessary to protect the d6-square.} 29. Qe2 (29. Bc1 {is more accurate, preparing to take on a4 immediately.}) 29... Rb3 30. Qc4 Qc6 31. Bc1 Nf6 32. Qxa4 Qxa4 33. Rxa4 Rxc3 34. Bb2 Rb3 35. Bxe5 Rb4 36. Rxa5 Rxe4 37. Rxe4 Nxe4 38. Ra4 {The game simplified a lot and it's clear that a draw is on the way.} Nd4 $1 {[%c_effect d4;square;d4;type;GreatFind;persistent;true] A practical and safe option, leading to an ending that is easily defensible for a player of Ding's stature.} (38... Nf6 $2 {[%c_effect f6;square;f6;type;Mistake;persistent;true] loses to} 39. Nxg7 Kxg7 40. Rg4+ {winning a pawn.}) (38... N4g5 39. h4 f6 $1 {[%c_effect f6;square;f6;type;GreatFind;persistent;true]} (39... Nh7 40. Bxg7 $1 {[%c_effect g7;square;g7;type;GreatFind;persistent;true]} Nxg7 41. Rg4 {loses a pawn.}) 40. hxg5 fxe5 41. g6 {Black should make a draw, but it's not so easy.}) (38... Nd2 {is objectively the best, according to the machine, but it's necessary to foresee some brilliant and inhuman variations.} 39. Rg4 Nf4 $3 {[%c_effect f4;square;f4;type;Brilliant;persistent;true] This wonderful resource leads to a draw. No human can play this and calculate the following variations.} (39... g6 $6 {[%c_effect g6;square;g6;type;Inaccuracy;persistent;true]} 40. Bc3 {is dangerous.}) 40. Nh6+ (40. Rxg7+ Kh8 $1 {[%c_effect h8;square;h8;type;GreatFind;persistent;true] No fear $1} 41. Bxf4 Re1+ 42. Kh2 Nf1+ {with perpetual check.}) 40... Kh7 41. Nxf7 Ne2+ 42. Kh2 Kg8 43. Nh6+ Kh7 44. Bxg7 Nf1+ 45. Kh1 Nd4 $1 {[%c_effect d4;square;d4;type;GreatFind;persistent;true]} 46. Bxd4 Re1 {The computer claims a draw is inevitable.}) 39. Bxd4 cxd4 40. Rxd4 g6 41. Ne3 {The presence of the knights gives some winning chances, but it's still a safe draw since there are not many pawns remaining.} Kg7 42. Rb4 Ng3 $5 {[%c_effect g3;square;g3;type;Interesting;persistent;true]} 43. Rb7 Nf5 44. Ng4 (44. Nxf5+ gxf5 {leads to an easy draw in the rook endgame.}) 44... Re7 45. Rb5 (45. Rxe7 Nxe7 {If this was a four-against-three-pawns endgame, the winning chances would be considerable. But one pawn less makes the task much easier.}) 45... Re1+ 46. Kh2 Re2 47. Rb7 Nd6 48. Ra7 Kf8 49. Kg3 f5 50. Kf3 Re7 51. Ra8+ Re8 52. Rxe8+ Kxe8 {Finally we get to the knight endgame.} 53. Ne5 g5 54. h4 gxh4 55. Kf4 h3 $1 {[%c_effect h3;square;h3;type;GreatFind;persistent;true]} 56. gxh3 Ke7 57. Nc6+ Kf6 58. Nd4 Ne4 59. f3 Nf2 60. h4 Nd3+ 61. Kg3 Kg6 {Now the draw is simple, since White cannot make progress.} 62. Ne6 Kf6 63. Nf4 Nb4 64. Kf2 Ke5 65. Ke3 Nc2+ 66. Kd2 Nd4 67. Nd3+ Kf6 68. Ke3 Nc2+ 69. Kf4 Nd4 70. Kg3 Ne2+ 71. Kf2 Nd4 72. Nf4 Ke5 73. Ne2 Ne6 74. Kg3 Kf6 75. Kg2 Kg7 76. Kf2 f4 77. Kg1 Kg6 78. Kg2 Kh6 79. Nc1 Kh5 80. Kh3 Nd4 81. Nd3 Nxf3 82. Nxf4+ {A well-played game with interesting ideas in all its phases, leading to a fighting draw. This match is as good as it gets.} 1/2-1/2