[Event "Firecracker Open"]
[Site "San Francisco"]
[Date "2001.06.30"]
[Round "2"]
[White "Aigner, Michael"]
[WhiteElo "2191"]
[Black "GM Wojtkiewicz, Aleks"]
[BlackElo "2692"]
[Result "1/2-1/2"]

© 2001 Michael Aigner (annotations for September/October 2001 issue of CalChess Journal)

1. e4

One of my goals for this tournament was to see how I matched up versus a strong opponent. I didn't anticipated being paired against the #4 rated player in the country, and I certainly didn't expect the result of this game.

1... c5 2. Nc3 g6 3. g3 Bg7 4. Bg2 Nc6 5. d3 d6 6. Be3

The game begins along one of the main lines of the closed Sicilian. Although most common at this point is 6... e6, I have encountered Rb8, e5, and the text in previous games.

6... Nf6 7. h3 e5

By playing his pawn to e5 instead of e6, black controls the light squares on the kingside for future attacking chances. Such a choice signals the obvious: rated 500 points higher than his opponent, black is playing to win. A less aggressive alternative is simply castling.

8. Nge2 Be6!?

This seemingly logical move is almost a theoretical novelty, as I found morely one game in ChessBase with this position. If my GM opponent sought to confuse me by leaving "book", he certainly succeeded.

9. O-O

Another plan here is 9. f4 with the thematic idea of destroying black's kingside with 10. f5. However, black hasn't castled yet, so I chose a less commital move.

9... Qd7 10. Kh2 h5 11. Qd2?!

This thematic move runs straight into the main point of black's tactics: the weakness of the light squares on the kingside. More precise would have been 11. Nd5 O-O-O 12. c3 after which white controls the center.

11... Nd4

One of the key concepts of the closed Sicilian is posting a knight on d5 (for white) and d4 (for black). The weakness of the f3 square presents an additional reason for white to be unhappy with his opening play. Fortunately, Bxh3 is not a real threat yet: 12. Nd5 Bxh3? 13. Bxh3 Nf3+ 14. Kg2 Nxd2 15. Bxd7+ Kxd7 16. Nxf6+ Bxf6 17. Bxd2 and white's up a piece.

12. Bg5

With black threatening Ng4+ to open up the h-file in some variations, this move anticipates trading off minor pieces.

12... Ng4+!

And black declines the trade! The piece is immune, because after 13. hxg4 hxg4+ 14. Kg1 Nf3+ 15. Bxf3 gxf3 white must return the piece to avoid worse, thereby leaving his kingside shattered.

13. Kg1 f6! 14. Nxd4?

White should have tried 14. hxg4 hxg4 15. Nxd4 cxd4 16. Nd5 fxg5 and although the open f- and h-files look mighty promising, black has no immediate breakthrough.

14... cxd4 15. Nd5 fxg5 16. hxg4 Bxd5 17. exd5 Qxg4

Instead, a series of trades leaves black up a pawn with a promising kingside attack. Although Fritz suggests the computer defense 18. Rae1 h4 19. Re4 Qh5 20. g4 Qh7, this leaves white with too many holes for black to probe. Remembering that bishops of opposite color favor the attacker, I decided to burn all my bridges. Objectively, this plan nearly loses by force, but only if my opponent is prepared to find the precise move order!

18. Qb4!? O-O-O 19. c4

White's plan is to pry open both the c-file and the long diagonal in order to destroy the safety of black's king.

19... h4 20. c5 dxc5!?

Black seems to wins with 20... hxg3, but looks can be deceiving. 21. c6 b6 22. f3 Qh5 23. Rfd1 Qh2+ 24. Kf1 g4 25. Qb5 Rdf8 26. Qa6+ Kd8 27. c7+ Ke7 and white quickly runs out of ways to delay checkmate. But 22. Qc4! Rh1+ 23. Bxh1 gxf2+ 24. Kxf2 Rf8+ 25. Ke1 Qg3+ 26. Kd2 Qh2+ 27. Ke1 Rxf1+ 28. Kxf1 Qxh1+ 29. Kf2 is spectacular yet drawn. It is understandable that black wanted to avoid any possibly drawish continuations.

21. Qxc5+ Kb8 22. Rac1 Qd7

For the first time in the game, I felt like I wasn't going to perish in a miniature. It is always a good sign when a GM believes your attack and retreats to defend against it! But after 22... hxg3 23. Qc7+ Ka8 24. d6 e4 25. d7 (25. Qxg7 Qh4 26. Rfd1 Qh2+ 27. Kf1 gxf2 and black wins) a6 26. Qxg3 white has weathered the storm and has good chances to hold the bishops of opposite color endgame.

23. d6

This sacrifice to open up the long diagonal is the only logical move. If black has time to play Qd6 to blockade the pawn, then white's attack is doomed.

23... Qxd6 24. Qb5 Qb6?!

Black forces a queen trade by indirectly attacking b2. This ends white's initiative and forces an endgame up two pawns, but with bishops of opposite color. Perhaps 24... Qd7 preserving the pawn structure was more precise.

25. Qxb6 axb6 26. Be4 Rd6

At this point of the game, I assessed the position as giving white very real drawing chances because of his much superior bishop and more active rooks. If a white rook ever penetrates to the 7th rank, then b7 is indefensible. But black's still up two pawns.

27. Kg2 Bf8

The bishop is heading to d6 to guard the c7 penetration square.

28. Rc4?!

Better was Rc2, defending the kingside while preparing to double on the c-file.

28... Rf6 29. gxh4!

After 29. Rfc1 h3, white's king has no choice but to sit passively on h2 in front of the black passed pawn. This violates the rule of activating all your pieces in an endgame. After the text, black's kingside pawns are far weaker and the half open g-file provides attacking opportunities.

29... gxh4 30. Kh3 g5 31. Rg1!

Forced! After 31. f3 g4+ 32. Kxg4 h3 33. Kg3 Rg8+ 34. Kh2 Bh6 (34... Rg2+ 35. Kh1 Rxb2 also wins) 35. Rg1 Bf4+ 36. Kh1 Rxg1+ 37. Kxg1 h2+ 38. Kh1 Rf7 and white has no choice but to play 39. Rc2 Rg7 40. Rxh2 Bxh2 41. Kxh2. Up an exchange, black should have no difficulties converting the endgame.

31... Rg8 32. f3 Be7 33. Rgc1

Around here, I was confident that I had a draw, and wanted to see if my GM opponent would overextend in an attempt to win.

33... Bd6 34. a4

Instead of creating weaknesses by pushing pawns, white should have maintained the status quo with a move like 34. Rg1 or Bd5.

34... Rgf8 35. Kg2 Rf4

Black has no other way to make progress other than by opening up the kingside with g4 and hoping that white blunders. White now must look out for exchange sacrifices where black eliminates white's powerful bishop and advances his own dangerous passed pawn(s).

36. R4c2 g4 37. fxg4 Rxg4+ 38. Kh3 Rg3+ 39. Kh2

39. Kxh4?? Rg7 would have been tragic.

39... Ka7 40. Rg2 Rxg2+ 41. Kxg2

By trading one set of rooks, white's king is no longer in imminent danger of checkmate.

41... Ka6 42. b3 Bc5 43. Kh3 Rf4 44. Bd5

White avoids any tricks whereby black sacrifices the exchange in an attempt to promote the d-pawn.

44... Ka5 45. Bxb7 Kb4 46. Bd5 e4!? 47. Bxe4 Kxb3

After the game, GM Wojtkiewicz told me that he played 46... e4 with the intent of following up with 47... Rxe4 48. dxe4 d3 <see diagram>. All variations win for black except for one, which wins for white!

<A> 49. Kxh4? d2 50. Rd1 Kc3 51. e5 Kc2 0-1

<B> 49. Kg2? d2 50. Rd1 Kc3 51. Kf3 Kd3 52. e5 h3 53. e6 h2 54. Ra1 Kc2 55. Ke2 Bg1 56. e7 h1Q 57. e8Q Qg2# 0-1

<C1> 49. a5!! Kxa5 50. Rd1 1-0

<C2> 49. a5!! bxa5 50. Rc4+ Kxb3 51. Rxc5 a4 52. e5 a3 53. Rb5+ Kc2 54. e6 d2 55. e7 d1Q 56. e8Q Qf3+ 57. Kxh4 and eventually white blocks check with his rook and queen to win the endgame 1-0

<C3> 49. a5!! d2 50. Rd1 Kc3 51. a6 b5 52. e5 Kc2 53. Rxd2+ Kxd2 54. e6 and the bishop can't stop both passed pawns 1-0

48. a5!

Now the game is a dead draw. This my first half point ever versus a GM in a slow tournament game.

48... Kb4 49. axb6 Bxb6 50. Rc6 Bc5 51. Rc7 Bb6 52. Rc6 1/2-1/2