[Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.14"] [Round "1.1"] [White "Firouzja, Alireza"] [Black "Caruana, Fabiano"] [Result "1-0"] [Annotator "Monokroussos,Dennis"] [ECO "B35"] [WhiteElo "2762"] [BlackElo "2795"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "63"] [GameId "2257725097442885"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "Triveni Continental Kings"] [BlackTeam "Alpine SG Pipers"] {[%evp 0,63,15,19,17,21,27,21,36,27,40,23,27,36,27,11,4,7,20,16,11,32767,-3,0,2,9,-2,0,9,3,3,-12,-14,-30,24,-24,-15,-37,64,-31,46,0,-95,29,54,70,19,16,103,32767,120,188,74,298,175,441,802,32767,29991,29992,29993,32767,29995,29996,29997,29998]} 1. e4 c5 2. Nf3 g6 3. d4 cxd4 4. Nxd4 Bg7 {Not a normal part of Caruana's repertoire, even if there is *one* other game in the database where he was Black. (Last year, vs. Nepomniachtchi.)} 5. Nc3 (5. c4) 5... Nc6 6. Be3 {Transposing to a few more Caruana games.} Nf6 7. Bc4 O-O 8. Bb3 d6 9. f3 Bd7 (9... Qa5 10. Qd3 Ne5 11. Qd2 Bd7 12. O-O-O Rfc8 13. Kb1 Nc4 14. Bxc4 Rxc4 15. g4 Be6 16. Nb3 Qa6 17. Bd4 Rac8 18. g5 Nxe4 19. fxe4 Bxd4 20. Nxd4 Rxc3 21. a3 Qc4 22. b3 Qc5 23. Rc1 Qxa3 24. h4 R8c5 25. Rcf1 Bxb3 {0-1 Nepomniachtchi,I (2767)-Caruana,F (2793) Sinquefield Cup 11th Saint Louis 2024 (6)}) 10. Qd3 {Very, very unusual - weird, even, as Black often wants to play ...Ne5 (followed by ...Nc4 with the help of a rook on c8), so why is White inviting this with a loss of time? They're playing on physical boards, in person, so it's not a mouse slip.} (10. Qd2) ({and} 10. h4 {are usual.}) 10... Ne5 11. Qd2 Rc8 12. h4 h5 13. O-O-O Nc4 14. Bxc4 Rxc4 {I still have no clue whatsoever about what Firouzja is hoping for. This is essentially a theoretical position with an extra tempo for Black. White's position is playable, but not more playable than it would have been.} 15. Kb1 Qc7 16. Nde2 $6 (16. Ncb5 $142 $11 {That Firouzja's play is inaccurate here suggests that this wasn't some sort of brilliant prep--but then the mystery deepens, as the standard 10th move options aren't obscure; they've been played thousands of times.}) 16... b5 $6 (16... Rc8 $15 {/? is even better according to the engine.}) 17. Nf4 (17. Bh6 $142) 17... Bc6 (17... Rc8 $142 $17) 18. g4 $1 $11 Rc8 $2 {Too slow. It was time to do something, not continue building.} (18... d5 $1 19. gxh5 dxe4 20. hxg6 exf3 21. gxf7+ Rxf7 22. Ne6 Qd7 23. Qxd7 Bxd7 24. Ng5 b4 25. Rxd7 Nxd7 26. Nxf7 Kxf7 27. Nd1 $11) (18... Rxc3 $1 19. Qxc3 Nxg4 20. Bd4 e5 21. Nd5 exd4 22. Qxc6 Qxc6 23. Ne7+ Kh7 24. Nxc6 Nf2 25. Nxa7 $11 {are a couple of bonkers options.}) 19. gxh5 $16 {/+- Firouzja's play has been iffy for a while, but now he turns into a beast.} b4 20. hxg6 $1 $18 bxc3 21. gxf7+ Kxf7 22. Qg2 e5 23. Rhg1 $1 exf4 24. Qxg7+ Ke6 25. Qg6 $2 (25. Qg5 $1 $18 {was best, so that if} Bd7 {, which is best after the text though not here, White is winning after} 26. Qxf4 $18) 25... fxe3 $2 {Caruana gets no further chances.} (25... Bd7 $16) 26. Qf5+ Kf7 27. Rg6 Qe7 28. Rdg1 $1 Bxe4 29. Rg7+ Ke8 30. Rxe7+ (30. Rg8+ {mates a move faster, not that it matters.}) 30... Kxe7 31. Rg7+ Ke8 32. Qe6+ 1-0 [Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.15"] [Round "2.1"] [White "Mamedyarov, Shakhriyar"] [Black "Rapport, Richard"] [Result "1-0"] [Annotator "Monokroussos,Dennis"] [ECO "E26"] [WhiteElo "2730"] [BlackElo "2741"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "65"] [GameId "2257725097496153"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "Upgrad Mumba Masters"] [BlackTeam "Fyers American Gambits"] {[%evp 0,65,20,14,24,23,13,19,13,6,12,12,3,5,35,32767,24,-14,11,-4,34,53,31,26,32,23,49,32767,102,87,87,88,164,153,182,32767,117,47,53,66,48,91,93,112,114,80,76,88,102,59,60,69,68,118,122,181,151,140,211,146,192,137,166,78,145,132,1175,1079]} 1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. a3 {It's safe to say that the Saemisch is back after a generations-long hiatus.} Bxc3+ 5. bxc3 c5 6. e3 b6 7. Bd3 Bb7 8. f3 Nc6 9. Ne2 Qe7 {Very rare. Black generally castles short and then combines prophylaxis with moves aimed at winning the c4-pawn. Rapport asks why and castles long instead.} 10. O-O $146 O-O-O 11. e4 d6 12. Be3 $146 Kc7 $6 {The idea is presumably to clear the way for the rook to return to b8 if necessary, e.g. in a line like 13.e5 dxe5 14.dxc5 bxc5 with White's rook coming to b1 in the near future.} (12... Na5 13. Rb1 Nd7 $11) 13. Nc1 $14 {/?} Na5 14. Nb3 Nxb3 15. Qxb3 e5 (15... Nd7) 16. a4 $16 a5 17. Bg5 $1 {White would like to put a rook on b1 without allowing Black to play ...Nd7.} Bc6 {Allowing the rook to defend b6 from b8.} 18. f4 h6 19. Bh4 $1 exf4 20. Rxf4 g5 21. Raf1 $1 {This had to be foreseen when playing 19.Bh4, and was almost certainly intended on move 18 or even move 17 (if not earlier).} gxf4 22. Rxf4 Qd7 23. Bxf6 Bxa4 24. Qb1 Rhe8 25. Bxd8+ Qxd8 26. e5 $1 $18 dxe5 27. Rxf7+ Re7 28. Rf6 {The b6 pawn is a gift that keeps giving.} Kb7 (28... Qb8 29. d5 Rg7 30. d6+ Kb7 31. d7 Kc7 32. Rxb6 $1 $18) 29. Be4+ Ka7 30. dxc5 bxc5 31. Bd5 $1 {Keeping Black's queen out frees White's queen for attacking purposes.} e4 32. Rxh6 e3 33. Qg6 {An impressive win by Mamedyarov. Black's plan of castling long didn't make a sterling impression, but we should wait to see how it does when Black chooses better options on move 12.} 1-0 [Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.16"] [Round "3.2"] [White "Vidit, Santosh Gujrathi"] [Black "Rapport, Richard"] [Result "0-1"] [Annotator "Monokroussos,Dennis"] [ECO "A40"] [WhiteElo "2708"] [BlackElo "2741"] [WhiteFideId "5029465"] [BlackFideId "-1"] [PlyCount "54"] [GameId "2257725097553521"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "Triveni Continental Kings"] [BlackTeam "Fyers American Gambits"] {[%evp 0,54,22,15,37,15,34,42,32767,49,43,33,53,46,32767,32767,57,56,70,72,58,32767,32767,40,43,51,33,8,32767,32767,0,18,-5,-9,12,-2,32767,-13,3,-48,-63,-42,-19,32767,-23,-43,-24,-86,-85,32767,32767,-172,-29990,-29991,-29992,-29993,-29994] Probably the most spectacular game of the tournament, at least so far.} 1. d4 e6 2. c4 c5 3. d5 g6 4. e4 Bg7 5. Bd3 d6 6. Nf3 exd5 7. exd5 Qe7+ (7... Ne7 {is more common, but the text is better.}) 8. Be2 Nh6 $146 {This would be silly if the knight were staying on h6, but it's headed to f5 or f7.} 9. O-O O-O 10. Bg5 f6 11. Bd2 Nf7 12. Re1 Nd7 {Say what you will about Black's position, he at least controls the e5 square. More importantly, while White's position is fine it's hard to see what he can do. Black will put a knight on e5 and push his kingside pawns; for White, the only obvious plan is to play b4 and try to break in on the queenside. He eventually settles on that idea, but only after an excessive delay.} 13. Bf1 Nde5 14. Bc3 $6 (14. Nxe5 fxe5 15. Nc3 $11) 14... Bg4 15. Nbd2 f5 16. h3 Bh5 $1 (16... Bxf3 17. Nxf3 Rae8 18. Nh2 $1 Qh4 {and with the bishop on f1 White can play} 19. g3 $11 {with equality.}) 17. Be2 Bxf3 $1 {Now it's good.} 18. Nxf3 Rae8 19. Nh2 Qh4 $17 {There's no g3 here with the loose h-pawn.} 20. Bf1 f4 {Both creating a possible battering ram with ...f3 and again preventing g3.} 21. b4 {Finally - but too late.} Ng5 $1 22. bxc5 f3 $1 23. cxd6 fxg2 {Threatening ...Qxf2#.} 24. Bxg2 Rxf2 25. d7 $2 {Black is winning with the prosaic 25...Ref8, but there's even better.} (25. Bxe5 {had to be tried. Black is in great shape, close to winning, after 25...Rxe5 or 25...Nxh3+ 26.Bxh3 Bxe5, but the game goes on.}) 25... Qg3 $3 {There's no stopping ...Qxg2#, even with two queens.} 26. dxe8=Q+ Bf8 {White is a queen and a rook ahead, and helpless all the same.} 27. Qxf8+ Kxf8 0-1 [Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.17"] [Round "4.3"] [White "Wei, Yi"] [Black "Keymer, Vincent"] [Result "1-0"] [Annotator "Monokroussos,Dennis"] [ECO "B12"] [WhiteElo "2754"] [BlackElo "2776"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "67"] [GameId "2257725097606792"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "Triveni Continental Kings"] [BlackTeam "Ganges Grandmasters"] {[%evp 0,65,18,26,36,24,33,25,24,29,26,35,23,19,27,22,33,20,15,11,39,42,52,36,36,53,56,19,29,7,16,13,13,21,15,1,5,2,37,24,33,40,65,23,46,33,67,67,96,89,178,162,268,263,325,348,348,301,323,252,365,239,474,306,343,322,409,453]} 1. e4 c6 2. d4 d5 3. e5 Bf5 4. Nd2 e6 5. Nb3 Nd7 6. Nf3 Qc7 7. Be2 c5 8. c3 c4 9. Nbd2 b5 $6 {The engine doesn't believe it, but if the idea is new the question is how easily a strong player can work out the resulting position.} (9... h6 {is usual here, precisely because of White's next move.}) 10. Nh4 $146 {?} Bg6 11. O-O Be7 12. Nxg6 hxg6 13. f4 (13. g3 $16 {followed by h4 was better.}) 13... a5 14. b3 $5 $14 (14. Nf3 $142) 14... a4 (14... Nb6) 15. bxc4 bxc4 16. Qc2 a3 17. g4 Nh6 18. h3 Nb6 ({It looks like Black's kingside setup is all about stopping White from playing f5; surprisingly, Black's best may be going for ...f5 himself.} 18... f5 $1 19. exf6 gxf6 20. Qxg6+ Nf7 $44) 19. Rf3 (19. Rb1 $142 $14) 19... Qc6 (19... f5 $1) 20. Nf1 $14 Kd7 21. Ne3 (21. Rb1 $142) 21... f5 (21... Qa4 $1) 22. Rb1 (22. Rg3 $142 $14) 22... Rab8 (22... Rhb8 $142) 23. Bd1 Kc7 24. Rg3 g5 $2 {Trying to force matters on the kingside, but he hasn't calculated everything correctly.} (24... Na4 $3 25. Rxb8 Rxb8 26. gxf5 Nxf5 27. Nxf5 exf5 28. Bxa3 (28. Qxa4 Qxa4 29. Bxa4 Rb1 30. Rxg6 Rxc1+ 31. Kh2 Ra1 32. Rg2 g5 33. fxg5 Bxg5 $1 $11) 28... Bxa3 29. Qxa4 Qxa4 30. Bxa4 Rb6 $11) 25. fxg5 f4 26. gxh6 $1 fxg3 (26... Rxh6 27. g5 Rhh8 28. Rf3 fxe3 29. Bxe3 $18) 27. hxg7 Rhg8 28. Qg6 $18 {Materially Black is fine, but White's kingside passers will quickly decide the game.} Qd7 29. Ng2 (29. Nf5 $18 {is nice but wholly unnecessary.}) 29... Na4 30. Rxb8 Rxb8 31. Bxa4 Qxa4 32. Ne3 $1 {Black "owns" the queenside, but can't do anything with it as White's queen covers b1 and the knight prevents the queen's invasion on c2 or d1. Meanwhile, g8Q is threatened, so Black must return the heavy pieces to the kingside.} Rg8 33. Qf7 Qe8 34. Qxe6 {Black's d-, c-, and g-pawns are all dropping, and without a shred of compensation.} 1-0 [Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.18"] [Round "5.3"] [White "Rapport, Richard"] [Black "Sindarov, Javokhir"] [Result "0-1"] [Annotator "Monokroussos,Dennis"] [ECO "D02"] [WhiteElo "2741"] [BlackElo "2726"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "44"] [GameId "2257725097651867"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "Fyers American Gambits"] [BlackTeam "Ganges Grandmasters"] {[%evp 0,42,17,20,20,8,15,5,10,-3,6,4,32767,32767,14,11,20,18,11,32767,15,-4,11,24,3,15,-6,-6,-14,-10,-18,32767,-6,-94,-50,32767,32767,-103,-118,-272,-250,-251,-280,-388,-364]} 1. Nf3 Nf6 2. d4 d5 3. Bf4 c5 4. e3 Nc6 5. Nbd2 Nh5 6. Bg5 (6. dxc5) 6... h6 7. Bh4 g5 8. Ne5 Nxe5 9. dxe5 Ng7 10. Bg3 Nf5 11. e4 {All theory. Black has done very well in this line despite the engine's insistence that White is slightly better (or at least pushing for a minimal edge).} dxe4 12. Nxe4 Qb6 13. Bc4 Qb4+ 14. Nd2 Bg7 15. c3 (15. O-O $142 O-O 16. Bb3) 15... Qxb2 $1 16. Rb1 $2 Qxc3 $19 {White will prevent Black from castling with Bb5+, either now or on the next move, but it doesn't make up for the two missing pawns.} 17. Rb3 Qa5 18. Bb5+ Kf8 19. Bd3 Nxg3 20. hxg3 Be6 $1 21. Rxb7 c4 {Now Black wins even more material.} 22. Be2 c3 (22... c3 23. Nb3 $2 Bxb3 {followed by ...c2+ wins a queen, no matter which way White recaptures.}) 0-1 [Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.19"] [Round "6.2"] [White "Giri, Anish"] [Black "Wei, Yi"] [Result "1-0"] [Annotator "Monokroussos,Dennis"] [ECO "D02"] [WhiteElo "2760"] [BlackElo "2754"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "59"] [GameId "2257725097676454"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "Alpine SG Pipers"] [BlackTeam "Triveni Continental Kings"] {[%evp 0,59,22,15,15,6,20,6,11,4,11,20,11,1,18,13,22,-25,-32,22,15,-21,-22,-13,-31,-33,-29,-3,22,23,40,15,0,32767,-33,-6,40,-78,1,-54,32767,78,405,409,512,614,572,532,609,520,493,474,458,503,524,463,765,523,526,500,465,472]} 1. d4 Nf6 2. Nf3 d5 3. c3 c5 4. dxc5 {The most principled response, asking Black to prove his compensation.} e6 5. Be3 Nc6 6. b4 Be7 7. g3 Ng4 8. Bd4 e5 9. h3 exd4 10. hxg4 dxc3 11. Nxc3 Nxb4 12. Qa4+ Nc6 13. Rd1 Bxc5 14. Nxd5 Be6 15. Rh5 Rc8 16. Bh3 g6 17. Qe4 $1 {My recollection is that both players were still banging out their moves - great prep so far.} Kf8 (17... gxh5 18. g5 $1 Bb4+ 19. Nxb4 Qe7 20. Nd5 Bxd5 21. Qxe7+ Nxe7 22. Bxc8 h6 $1 23. g6 $1 fxg6 24. Ne5 O-O 25. Rxd5 Nxd5 26. Be6+ Kg7 27. Bxd5 h4 $1 28. gxh4 Rf5 29. f4 Rxf4 30. Nf3 $11 {was probably also on one or both players' laptops as well.}) 18. Qf4 Qa5+ $2 {At this point Wei Yi was on his own, and this natural move is a mistake. As difficult as it is to maintain the queen opposite White's rook on d1, that's what should have been done.} (18... h6 $1 19. Rxh6 Rxh6 20. Qxh6+ Kg8 $11 {The position is still sharp and I'm sure both players' analyses - certainly Giri's - continued a while longer. Still, the engine gives 0.00 after both 21.Qg5 and 21.e4.}) 19. Kf1 $18 gxh5 20. Ng5 Nd8 21. Qe5 $1 Bxd5 (21... Rg8 22. Nxh7+ Ke8 23. Ndf6+ Ke7 24. Rd7# {is an attractive mate.}) 22. Qxh8+ Ke7 23. Rxd5 Qc7 24. gxh5 Ne6 25. Nxe6 $1 Qc6 (25... Rxh8 26. Nxc7 $18) 26. Qe5 fxe6 27. Qg7+ Ke8 28. Qg8+ Bf8 29. Bxe6 Rc7 30. Rd8+ $1 (30. Rd8+ Ke7 31. Qf7+ Kxd8 32. Qxf8+ Qe8 33. Qd6+ Rd7 34. Qxd7+ Qxd7 35. Bxd7 Kxd7 36. g4 $18 {is a trivially winning pawn ending for White. Great prep and great play by Giri!}) 1-0 [Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.19"] [Round "6.3"] [White "So, Wesley"] [Black "Keymer, Vincent"] [Result "1-0"] [Annotator "Monokroussos,Dennis"] [ECO "C26"] [WhiteElo "2753"] [BlackElo "2776"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "63"] [GameId "2257725097692844"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "Upgrad Mumba Masters"] [BlackTeam "Ganges Grandmasters"] {[%evp 0,63,22,13,28,18,0,-6,8,10,0,-1,11,10,-3,1,6,2,12,6,-1,32767,0,5,0,4,10,-8,11,32767,63,52,77,32767,62,62,81,74,88,76,134,154,167,160,188,32767,207,180,180,162,326,427,397,32767,439,475,420,337,439,408,862,936,2542,2636,2328,2468]} 1. e4 e5 2. Nc3 {The Vienna is becoming increasingly popular. Could it be that players are finally getting sick of the Italian?} Nf6 3. g3 Bc5 4. Bg2 O-O 5. Nge2 d6 6. O-O a5 7. d4 exd4 8. Nxd4 Re8 9. h3 Nbd7 10. Kh2 a4 11. b3 axb3 12. cxb3 $14 Bb4 13. Nde2 h6 14. Qc2 Ne5 15. f4 $16 Ng6 16. a3 Bd7 17. Bb2 Bc5 18. b4 Ba7 19. Rad1 h5 20. Nd5 $18 {White has seemingly made purely natural moves, and his reward - against the world's 4th highest-rated player - is a winning position in the early middlegame.} Nxd5 21. Rxd5 h4 22. Rg5 Ne5 23. gxh4 $1 {It's not as if Black can exploit the weakening of White's kingside.} f6 24. Rg3 Nf7 25. e5 $1 {Qg6 is a huge threat.} f5 26. exd6 (26. e6 {is even stronger (and fairly obvious), threatening both minor pieces and a capture on g7. (And maybe Qc3 as well.) Still, So's move is good enough.}) 26... Qxh4 27. Rxg7+ (27. Qc3) 27... Kf8 28. Qxc7 Rad8 29. Ng3 Nh6 30. Rxd7 (30. Qc3) 30... Rc8 31. Bg7+ Kg8 32. Bd5+ {An absolute massacre. 2.Nc3, anyone?} 1-0 [Event "TechM GCL 2025"] [Site "Mumbai IND"] [Date "2025.12.20"] [Round "8.2"] [White "Erigaisi, Arjun"] [Black "Keymer, Vincent"] [Result "1-0"] [Annotator "Monokroussos,Dennis"] [ECO "C58"] [WhiteElo "2775"] [BlackElo "2776"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "71"] [GameId "2257725097762506"] [EventDate "2025.12.14"] [EventType "team-tourn"] [SourceVersionDate "2025.01.05"] [WhiteTeam "PBG Alaskan Knights"] [BlackTeam "Ganges Grandmasters"] 1. e4 e5 2. Nf3 Nc6 3. Bc4 {Remember what I said, all of one game ago, about people getting sick of the Italian? Never mind.} Nf6 4. Ng5 {The "duffer's move", to quote Tarrasch. (Or rather, Fischer quoting Tarrasch.)} d5 5. exd5 Na5 6. Bb5+ c6 7. dxc6 bxc6 8. Qf3 (8. Bd3 {is by far the main move nowadays (it used to be 8.Be2), but the text is an important option Black must be ready for.}) 8... Rb8 (8... cxb5 9. Qxa8 Qc7 10. Qf3 Nc6 11. c3 Bg4 12. Qg3 Nd4 {is the start of a topical, forcing line. Keymer wants to keep things relatively simple.}) 9. Bd3 (9. Bxc6+ $6 {is somewhat popular at the amateur level but it's not very good. Black has a serious advantage after} Nxc6 10. Qxc6+ Nd7 $17) 9... Bd6 (9... Be7) 10. Nc3 $2 {The engine really dislikes this move, but it's hard to believe Erigaisi didn't know this. Is this a bluff for a rapid time control?} (10. Bf5) 10... Bg4 $17 11. Qg3 e4 12. Qe3 O-O 13. Ngxe4 Nxe4 $2 (13... Re8 $1 14. f3 (14. h3) 14... Nd5 $17 {/-+ is clearly better for Black according to the engine, and White's score is abysmal, too (two draws, 12 losses, no wins), so it's not the sort of position that only an engine can handle with Black.}) 14. Nxe4 $11 Re8 15. f3 Bh5 (15... Bc8 $142 $11 16. g4 f5 17. gxf5 Bxf5 18. b3 Be5 19. c3 c5 20. Ba3 c4 21. Bc2 cxb3 22. axb3 Kh8 23. d4 Bc7 24. Rb1 Rb6 25. Ke2 Qh4 26. Rbg1 h6 27. Kd1 Nxb3 28. Re1 Na5 29. Qe2 Qd8 30. Rhg1 Qd5 31. Bc1 Nb3 32. Qg2 Rg6 33. Bxb3 Qxb3+ 34. Qc2 Qd5 35. Rxg6 Bxg6 36. Qd3 Rf8 37. c4 Qh5 38. Rf1 Qh3 39. Ke2 Bxh2 40. d5 Bg3 41. Be3 Qg2+ 42. Bf2 Bh5 43. Rg1 Bxf3+ 44. Kd2 Qh2 45. Qa3 Rf7 46. Qa4 Bxf2 47. Qe8+ Kh7 {0-1 Yemelin,V (2573)-Tomashevsky,E (2646) EU-Cup 24th Kallithea 2008 (3)}) 16. g4 Bg6 17. b3 $146 Be5 18. c3 Bf4 $6 (18... h5 $11 {was better, working speedily to prevent White from consolidating.}) 19. Qe2 (19. Qd4 $1 $14 {/?}) 19... h5 20. Rg1 (20. O-O) 20... Bxh2 $5 21. Rg2 Bf4 22. gxh5 Qh4+ $2 (22... Bxh5) (22... Bh7 $1 $44) 23. Kd1 $16 {/+-} (23. Qf2 Qxh5 24. Be2 $18) 23... Qxh5 24. Kc2 $2 (24. Rg4 $1) 24... f5 25. Bc4+ (25. Nf6+ $5 gxf6 26. Qf2 $44) 25... Kh7 26. d3 fxe4 27. dxe4 {Now it's Black who's up material but worrying about his unsafe king.} Bc7 $2 (27... Be5 $1 $11) 28. Rg5 (28. Bd3 $1 $16 {/+-}) 28... Qh3 $8 29. Bd3 Rbd8 30. Bd2 (30. Be3 $142) 30... Rxd3 $2 (30... Rf8 31. Rag1 Rxf3 32. Ba6 Rf6 $11) 31. Qxd3 Rf8 $2 {Black's clever idea works if White now plays 32.Rf1 or 32.f4 to stop ...Rxf3, but he has missed something.} 32. Rag1 $1 (32. Rf1 $4 Rd8 $1 33. Qe2 Rxd2+ $1 34. Qxd2 (34. Kxd2 Bf4+ $19) 34... Qxf1 $19) (32. f4 $2 Qxd3+ 33. Kxd3 Rd8+ 34. Ke3 Bb6+ 35. Ke2 Bxe4 $11) 32... Rf6 (32... Rxf3 33. Rxg6 $1 Rxd3 (33... Be5 34. Qe2 $18) 34. Rxg7+ Kh8 35. Rg8+ Kh7 36. R1g7#) 33. Rxg6 {Simple and strong.} Rxg6 34. e5 Qh5 35. Rg5 $1 Qh6 36. Qf5 {Black has no way to save his queen without losing the rook on g6.} 1-0