[Event "Superbet Chess Classic Romania"] [Site "Bucharest, Romania"] [Date "2023.05.06"] [Round "1.5"] [White "Firouzja, Alireza"] [Black "So, Wesley"] [Result "0-1"] [WhiteElo "2785"] [BlackElo "2760"] [Annotator "Lang, JJ"] [UTCDate "2023.05.11"] [UTCTime "13:01:41"] [Variant "Standard"] [ECO "C50"] [Opening "Italian Game: Giuoco Pianissimo"] [Source "https://lichess.org/study/gLTaYxXi/fk4K8bfo"] [Orientation "white"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 Bc5 4. d3 d6 5. c3 Bb6 6. O-O h6 { A precise move order which So has successfully employed before. } 7. Nbd2 (7. a4 Nf6 8. b4 a5 9. b5 Ne7 10. Be3 Bxe3 11. fxe3 O-O 12. Nbd2 Ng6 13. Qe1 { gave White nothing after } 13... c6 14. Nh4 d5 15. exd5 cxd5 16. Bb3 Nxh4 17. Qxh4 Be6 18. d4 Rc8 19. Rac1 Ng4! 20. Qxd8 Rfxd8 21. Rfe1 Rc7 22. h3 Nf6 23. dxe5 Nd7 { when Black controlled too many key squares and eventually won in Adhiban – So, Chess. com, 2021. }) 7... Nf6 8. Bb3 (8. Re1 { Another idea of the early ... h7-h6 was demonstrated in 1997 by the inventive Hort. } 8... g5!? 9. Nf1 Be6 10. Bb5 g4 11. N3d2 Nd7 12. Nc4 a6 13. Nxb6 Nxb6 14. Bxc6+ bxc6 15. f4 gxf3 16. Qxf3 Qh4 17. Ng3 h5 18. Rf1 Qg4 19. Qf2 Rg8 20. Be3 Nd7 21. Rad1 h4 { and Black went on to win in 34 moves in Renner – Hort, Germany, 1997. }) 8... a5! { This precise move order is a novelty, but the game will shortly transpose back to known territory. That said, the specific line they end up in is one where Black (particularly Nakamura) has had fantastic results, so this seems like some excellent preparation from So. } 9. Re1 O-O 10. h3 Be6 { White has tried a number of moves here, and Nakamura has beaten most of them. } 11. Bxe6 $146 { On the face of it, this is a counterintuitive move. The doubled e-pawns are by no means bad for Black: they help control the center, attack on the half-open f-file, and cover the key f5-square. } (11. Bc2 Re8 12. Nf1 d5 13. exd5 Qxd5 14. Ng3 Qd6 15. Nd2 Rad8 16. Qf3 Qd7 17. a4 Nh7 18. Nc4 Ba7 19. Be3 Bxe3 20. Rxe3 f6 21. Rae1 Ng5 22. Qe2 Bf7 23. h4 Ne6 { when Black's knights are monstrous in this fixed structure. Dominguez Perez – Nakamura, Chess.com, 2020. }) (11. Ba4 Nd7 (11... Qb8 { actually gave So good chances as White after } 12. Nf1 Qa7 13. Re2 Ne7 14. Ng3 Rad8 15. d4 exd4 16. cxd4 d5 17. e5 Ne4 18. Be3 Nxg3 19. fxg3 Nf5 20. Bf2 c5 21. g4 Nxd4 22. Nxd4 cxd4 23. Qd3 Rc8 24. Rd1 Rc4 25. Bc2 { and White went on to win on move 62 in So – Xiong, Chess.com, 2021. }) 12. Nf1 f5 13. exf5 Bxf5 14. d4 exd4 15. cxd4 Nf6 16. Ng3 Bh7 17. Bb3+ Kh8 18. a3 Ne7 19. Be6 c6 20. Bd2 Nfd5 21. a4 Nb4 22. Bxb4 axb4 23. Qb3 Ba5 24. Re3 Ng6 25. Bf5 Bg8 26. Qc2 Nh4 27. Nxh4 Qxh4 28. Qd3 Bb6 29. Re4 Qf6 { with a serious initiative for Black in Caruana – Nakamura, Chess.com, 2020. }) (11. Nc4 { is probably White's most promising try. } 11... Ba7 12. Ne3 Ne7 13. Bc2 c5 14. d4 cxd4 15. cxd4 Ng6 16. Nf5 Re8 17. Be3 d5 18. dxe5 Nxe4 19. Bxa7 Bxf5 20. Bd4 Nf4 21. Be3 Ne6 22. Bb3 { White has successfully created pressure against Black's weakness, although Black went on to win in Vachier Lagrave – Nakamura, Chess. com, 2020. }) 11... fxe6 12. Nc4 Ba7 13. Be3 { A typical idea in this opening. The bishop is actually well-placed on c1, as it eyes an important diagonal for kingside attacks. But with a knight unlikely to hop in on the f5-square, attacks on the kingside dark squares are less likely. Thus, the bishop is more of a "piece of wood" in the way of White's queenside rook. } 13... Bxe3 14. Rxe3 a4 15. d4 exd4 16. cxd4 d5! { A precise decision. White's isolated d-pawn is by no means bad here, with a rook already active on the third rank and knights ready to occupy the e5-square. Mostly, I'm charmed by the decision to give up control of the f5-square just after White elected to trade the dark-squared bishop. In other words, the attack White will need to execute to justify the isolated d-pawn will be much harder to pull off without the bishop, so now (and only now) it is reasonable to push ... d6-d5 and play ... e6xd5. } 17. exd5 exd5 18. Ncd2 Qd6 19. a3 Qf4 20. Nb1 Ne4 21. Nc3 Nxc3 22. Rxc3 (22. bxc3?! { would be playable, but it only "solves" White's isolated pawn problem by creating an even more vulnerable target on a3. }) 22... Qd6 23. Rac1 Rf7 24. R1c2 Rf4 25. Re3 Qf6 26. Rd3 Re8 27. Rc5! { Another classy move. In "fixed" centers like this, the squares each side's center pawn protects often end up being key squares from which to outpost and improve pieces. } 27... Qd6 28. Rcc3 Rfe4 29. g3 Re2 30. Rd2 Rxd2 31. Qxd2 Qg6 32. Kg2 Qb1 33. Rc1 Qf5 34. Rc3 Qb1 35. h4 { White declines Black's offer to repeat moves. } 35... Re7! { An instructive idea. This is not merely a defensive move, but rather a response to White's h-pawn push and the "hook" it creates! } 36. Rc1 Qf5 37. Qd1 g5! 38. hxg5 hxg5 39. Qd2 g4 40. Nh4 Qe4+ { It isn't just that Black is winning the d-pawn, but that the space advantage and coordination of the pieces have all drastically turned in Black's favor over the past five moves. } 41. Kh2? (41. Kg1! { was the most stubborn. The point is that now, after } 41... Qxd4 (41... Nxd4 42. Rd1 Ne2+ 43. Kh2 Re5 44. Qh6 { The computer insists that White has more compensation for the pawn here, presumably because the knight is worse off on e2. }) 42. Qg5+! Rg7 43. Qh5 { , White is able to push for a perpetual, as the f2-pawn is not hanging with check. }) 41... Qxd4 42. Qg5+ Rg7 43. Qf5 { This move is forced if White wants to threaten to check Black's king, but, unlike from the h5-square, it does not threaten a perpetual from here. Black is better. } 43... Rf7 44. Qg5+ Kf8 45. Kh1 Qe4+ 46. Kg1 Ke8 47. Qd2 Re7 48. Rd1 d4 { Now, it's fair to say Black is up a full passed pawn, and also has the compensation in the form of White's weakened kingside and passive knight. } 49. Qg5 Ne5 50. Qd2 d3 51. Qf4 Qxf4!? { Black's attacking chances were substantial after more or less any move that kept queens on. Then again, White's chances to swindle would be higher, too. So trusts his own technique enough to play the long game. This is probably a reasonable decision against anyone, but especially against a notoriously crafty blitz and bullet player like Firouzja. } (51... Qe2! { sacricing the a-pawn is worth it: } 52. Qxa4+ c6 53. Qa8+ Kf7 54. Qa4 c5 55. Qf4+ Ke6 { and Black's pawns advance in front of a surprisingly sheltered king. }) (51... Qc6 { was probably the most "correct" choice, as it does not require calculations after sacrificing the a-pawn, but still keeps the heat on against White's king. } 52. Re1 { What else? } 52... Qb5 53. Rd1 Qb3 54. Rd2 b5 { and Black is in great shape, as White has no serious threats of a perpetual. For instance, after } 55. Qf5 Qe6 { Black's king is snug. }) 52. gxf4 Nc4 53. Rxd3 Nxb2 { A lot of "mortal" players might have struggled psychologically to part with the passed d-pawn. But there are only two types of passed pawns: those who become queens and those who get captured! So reasons that the d-pawns chances of promotion were smaller than the b- and c-pawns after this trade. } 54. Rc3 Rd7 55. Kg2 Nd3 56. Kg3 c5 57. Kxg4 b5 { Black is completely winning here. The connected outside passed pawns pose far more threats than White's pair of isolated f-pawns. } 58. Nf5 Kf7 59. Kg5 Rd5 60. Rc2 Ne1 (60... b4 { would be the cleanest. } 61. axb4 cxb4 62. Rc7+ Ke8! 63. Kf6 Rxf5+ 64. Kxf5 b3 { is lights out. } 65. Ke6 Nxf4+ 66. Ke5 Nd3+ 67. Ke6 Kd8 68. Rd7+ Kc8 69. Rxd3 b2 70. Rd1 a3 { and the passed pawns are advanced enough to dominate the rook. }) 61. Rc3 (61. Ra2! { was the most stubborn defense, going against every principle of keeping pieces active in the endgame. The point is simply that now ... b5-b4 will hang the a-pawn, and ... c5-c4 "fixes" the b-pawn as backwards. } 61... Nd3 62. Kg4 Rd8 { and now Black asks White how his pieces can untangle. But there is much more work to do. }) 61... Nd3 62. Rc2 b4 { and now it's only a matter of time. } 63. axb4 cxb4 64. Rc7+ Ke8 65. Kf6 Kd8 66. Ra7 b3 67. Ne3 Rd6+ 68. Kf5 b2 69. Rb7 Rd4 70. Nc2 Rxf4+ 71. Kg5 Rxf2 72. Na3 Rf1 { A fantastic game from So. Firouzja's mistakes were neither drastic nor obvious, but they were enough to give So a clear initiative in the late middlegame, and he is far too accurate of a player to give it back. } 0-1