|
|
1. d4
|
This was part of a Semi-Slav thematic mini. The starting position can be found at black's fourth move. |

|
|
|
1... d5 2. c4 c6 3. Nc3 Nf6 4. Nf3 e6
|
The starting position. |

|
|
|
5. e3
|
White heads for a probable Meran. |

|
|
|
5... Nbd7 6. Bd3 dxc4 7. Bxc4 b5
|
The Meran Variation. This is a very heavily analysed variation, and both sides are following mainline theory until about move 14. |

|
|
|
8. Bd3 Bb7 9. O-O
|
The current main line. Alternatives are the aggressive e4 or the quiet a3. |

|
|
|
9... a6 10. e4 c5 11. d5 c4 12. Bc2 Qc7 13. dxe6 fxe6 14. Ng5
|
A major choice point; the main alternative is Nd4. |

|
|
|
14... Nc5 15. e5 Qc6
|
Black departs from theory (15 ...Qxe5, usually leading quickly to an endgame). The text is an idea of Shirov's, virtually forcing f3 and clogging up white's lines of attack. |

|
|
|
16. f3 Nfd7 17. Qe2 Nd3
|
Delaying white's attack down the e-file, and freeing c5 for a bishop or queen check. Be7 is another interesting possibility here. |

|
|
|
18. Bxd3 cxd3 19. Qxd3 Nxe5 20. Qe2 h6 21. Re1
|
I was expecting 21 Qxe5, whereupon Qc4 is a way of easing the pressure. |

|
|
|
21... hxg5
|
The beginning of a very risky plan which in the end did not succeed. Black tries to break white's king's position open by marching this loose pawn down the king's side, while white's major pieces queue up to attack his uncastled king! |

|
|
|
22. Qxe5 g4 23. Bg5
|
The problem is that taking the e5 pawn does not get white far, e.g. Qxe5 Qxe5, Rxe5 Kd7, and white is at risk of a strong attack from black's bishops and h8 rook. The move played stops black castling and makes the threats to his king more serious. |

|
|
|
23... gxf3 24. gxf3
|
White thought about this move for some time, but it seems to be correct. 24 Qxe6 Qxe6, 25 Rxe6 Kd7, the rook must move and Bc5 seems to be winning. Alternatively 24 Rd1 Bc5 and again black has fg with check. |

|
|
|
24... Qc5+
|
Bc5 is answered by 25 Be3. The text is intended to deprive white of this possibility. |

|
|
|
25. Kg2
|
25 Qxc5 Bxc5 and black has time to castle, and then it is white's king's position that looks precarious. |

|
|
|
25... Qxe5 26. Rxe5 Bd6
|
Obviously aiming at h2, but there are tactical risks to black as well. |

|
|
|
27. Rxe6+ Kd7 28. Rae1
|
Allowing Rxh2 . |

|
|
|
28... Rxh2+ 29. Kf1 Bxf3 30. Rxd6+
|
I had seen this tactic earlier but taken my eye off the ball. |

|
|
|
30... Kxd6 31. Bf4+ Kd7 32. Bxh2 g5 33. Ne4 Rh8
|
33´...Rg8 fails to 34 Nf6 . The idea here was 34 Nxg5 Rxh2, 35 Nxf3 Rxb2 and white's king is cut off on the second rank. Liquidating the pawns gives black a chance of holding a R v R N (or R v R B) pawnless ending. |

|
|
|
34. Nc5+
|
...but white doesn't want any of that. |

|
|
|
34... Kc6 35. Bg1 a5 36. Nb3 Bd5
|
Of course not 36 ...a4?, 37 Nd4 ! |

|
|
|