Yagiz Kaan Erdogmus vs Murali Karthikeyan: game review

27…Nxh4! The only good move here: every alternative is clearly worse.

• Takes the knight on h4.
• The next best move, 27…Nf4, would drop Black's winning chances from 63% to 50% (0.00).
• 27…Nxh4 trades off the attacked knight instead of losing it to Nxg6.
• White was threatening Nxg6, winning a knight and a pawn. The knight takes something with it instead.
• White answers 28.Nxh4, winning a pawn.
• After 28.Nxh4 a3 29.bxa3 Nc3 30.Re1 Nd5, Black is a pawn down (−1.57).

30.Rd7 is a mistake: White's winning chances go from 31% to 16%.

• Black answers 30…Nc3: moving away uncovers the rook on b4, which now hits the knight on h4.
• After Nc3, Black wins 1 point of material within 6 moves.
• Better was 30.Nf3 (−2.14), which attacks the e5 pawn, which is undefended.

Better: 30.Nf3 wins a pawn, picking it up with 31.Nxe5. After 30…c4 31.Nxe5 c3 32.Rd7 Rb5 33.Rf7, White comes out a pawn up (−2.14).

31…Ne2+ lets the win slip: 31…Qe8 kept Black winning. Black's winning chances go from 85% to 70%.

• White answers 32.Kh2: the king steps out of check to h2.
• Better was 31…Qe8 (−4.76): the queen forks the rook and the knight: both can't be saved at once.

Better: 31…Qe8 wins a rook for a piece and a pawn, picking it up with 33…Qxd7, and forks the rook and the knight. After 32.Ne7+ Kh8 33.Qxc3 Qxd7 34.Qxc5 Qc7, Black comes out a pawn up (−4.76).

37…Qxd1! The only good move here: every alternative is clearly worse.

• Takes the queen on d1.
• Protects the knight on e2, which was hanging.
• The next best move, 37…Qc7, would drop Black's winning chances from 68% to 50% (−0.05).
• 37…Qxd1 trades off the attacked queen instead of losing it to Qxd7 and trades queens.
• White was threatening Qxd7, winning the queen. The queen takes something with it instead.
• Trades queens: the queens come off while ahead, which makes the extra material easier to convert.

38…Bc3! The only good move here: every alternative is clearly worse.

• Protects the a3 pawn, which was hanging.
• The next best move, 38…e4, would drop Black's winning chances from 68% to 58% (−0.84).
• 38…Bc3 stops Rd3.
• White was threatening Rd3, winning a pawn, picking it up with 39.Nxe5. Bc3 takes care of it.
• White answers 39.h4, pushing the pawn storm towards the king, leaving Black with a king with pawns missing from its shelter.
• Engine line: 39.h4 h5 40.Bd6 Rd8 41.Ne7+ Kh7 (−2.14).

39…Bb2! The only good move here: every alternative is clearly worse.

• Shores up the a3 pawn, which was under as much attack as it had defence (2 against 1).
• The next best move, 39…Bd4, would drop Black's winning chances from 73% to 56% (−0.69).
• 39…Bb2 shores up the a3 pawn.
• White answers 40.Rd2, attacking the knight on e2, which is undefended.
• Engine line: 40.Rd2 Nc1 41.Ne7+ Kh7 42.Rc2 e4 (−2.85).

41.h4 is a mistake: White's winning chances go from 23% to 13%.

• Black answers 41…Nc1: the knight forks the rook and the pawn: both can't be saved at once.
• Moves a pawn from in front of the king, loosening its cover.
• Better was 41.Rd2 (−3.22), which attacks the knight on e2, which is undefended.

Better: 41.Rd2 threatens Rxe2, winning a knight. Engine line: 41…Nc1 42.Rc2 e4 43.Kg1 Rd8 44.Nc6 (−3.22).

• Both sides followed opening theory up to 12…Ng6: 24 moves in all, 12 by White and 12 by Black.
• Theory runs out after 12…Ng6: too few games between strong players reached this position to go further.