Annotated Game #124: Crazy attacking interlude

Before resuming analysis of the tournament begun in Annotated Game #123, I could not resist looking at a recent Slow Chess League game that featured an unsound attack for Black (me) and surprising resources for both sides.  Having been caught off guard by an unpleasant opening sequence, I decide to sacrifice a piece for two pawns and an attack.  This is in fact a typical recipe for disaster at the novice level and I fail to do any better with it.  However, I was able to see a number of attacking ideas and thought it would be much more fun to go out fighting than be squeezed to death out of an inferior opening.  The attacking idea on move 19 for Black is especially noteworthy and the analysis shows how I could have legitimately obtained a dominant kingside attack by finding a way of employing all of my limited resources, despite being the equivalent of two pieces down.  In the end, at least it was a fun game with a worthy opponent.

[Event "DHLC Slow Swiss #13"] [Site "Chess.com"] [Date "2014.05.17"] [Round "6"] [White "Constantine73"] [Black "ChessAdmin_01"] [Result "1-0"] [ECO "D11"] [WhiteElo "1650"] [BlackElo "1451"] [Annotator "ChessAdmin/Houdini"] [PlyCount "55"] [EventDate "2014.??.??"] [TimeControl "45"] {D11: Slav Defence: 3 Nf3 sidelines and 3...Nf6 4 e3 Bg4} 1. d4 d5 2. c4 c6 3. g3 {the fianchetto setup aganst the Slav is rarely played, as the bishop is normally not nearly as useful on g2 as it can be elsewhere.} Nf6 4. Bg2 Bf5 { I thought for a bit about the best square for the bishop, eventually deciding that the traditional f5 placement was best, since it dominated the h7-b1 diagonal without being able to be challenged by its White counterpart now.} 5. Nf3 e6 6. Nc3 Nbd7 (6... h6 {scores close to 52 percent for Black and gives the Bf5 a place to retreat to. It could also have been played on the next move. }) 7. O-O Be7 $6 (7... dxc4 {looks good here, taking advantage of the bishop not being available to recapture on c4.} 8. Nh4 Bg4 9. Qc2 Be7 10. h3 Bh5 11. g4 Nd5 $15) 8. Nh4 {a strong, logical move that looks to chase Black's bishop and gain space. White scores 61 percent with it.} Bg4 9. h3 Bh5 10. g4 Nxg4 $146 {rather than face the awkward bishop retreat and come under lasting pressure from White, I decide to sacrifice the piece for an attack.} (10... Bg6 {would be the best move here, but I was worried about the follow-on threat of g5, which I had not initially considered when entering the sequence, and lurking attacking possibilities for White. Play could have continued} 11. Qb3 Qb6 12. g5 Nh5 13. Nxg6 hxg6 14. cxd5 exd5 15. e4 $14) (10... Bxg4 {has actually been tried before (unsuccessfully).} 11. hxg4 Nxg4 12. Nf3 Qc7 13. Qd3 dxc4 14. Qxc4 Nb6 15. Qd3 h5 16. Qe4 g5 17. Bd2 Nc4 18. Rfc1 Nxd2 19. Nxd2 Qh2+ 20. Kf1 Qh4 21. Nd1 Bb4 22. Nf3 Nh2+ 23. Nxh2 Qxh2 24. d5 h4 25. dxc6 h3 26. Bf3 Rh4 27. cxb7 Rd8 28. Qc6+ Kf8 29. Rc4 Rxc4 30. Qxc4 Qd6 31. Qc8 Kg7 32. Ne3 g4 33. Bxg4 h2 34. Kg2 Ba5 35. Rd1 {1-0 (35) Barberi,A (2160)-Cantore,A (2195) Asti 1998}) 11. hxg4 $16 Bxg4 12. Nf3 h5 {a logical continuation, seeking to work the open flank, but not decisive enough.} (12... dxc4 {would at least grab another pawn, although Houdini still shows White with close to a pawn equivalent advantage here; the open nature of the position should allow White's extra piece to be an effective attacker. Although I did notice the capture idea earlier, I effectively dismissed it in favor of an all-out (desperate) kingside effort.}) 13. Bf4 (13. Qb3 Qc7 14. cxd5 exd5 $16) 13... g5 (13... dxc4) 14. Bh2 {presumably the original idea behind the bishop move. White adds a defender in front of his king but still dominates the full diagonal.} f5 (14... h4 {would be a better way to continue with Black's quasi-attack, not committing the f-pawn and making the h-pawn more of a direct annoyance.}) 15. Ne5 f4 16. Nxg4 (16. Ng6 {is what I was expecting, disrupting Black's ability to castle and exchanging off a piece.} Rh6 17. Nxe7 Qxe7) 16... hxg4 17. e4 (17. cxd5 {is the correct idea here and could also have been played before or afterwards by White. With an extra piece and Black's king in the center, blowing up Black's protective pawn shell is the indicated strategy. } exd5 18. e4 f3 19. Bh1) 17... Nf6 {this turned out to have been an excellent practical choice, as the move played in response - the obvious reaction - is actually rather dangerous for White.} (17... dxc4 18. Qxg4 Rh6 $14) (17... f3 { would have had better practical chances.}) 18. e5 (18. cxd5 f3 19. dxe6 fxg2 20. Kxg2 $18) 18... f3 19. exf6 Bd6 $4 {I thought for a long time at this point, considering the rook sac on h2, but I didn't see how I could make it work in the end. Houdini however does not have that limitation.} (19... Rxh2 20. Bxf3 (20. Kxh2 Bd6+ 21. Kg1 Qxf6 $1 (21... fxg2 $2 {was the only move I considered.} 22. Kxg2 $18) 22. Re1 O-O-O $1 {and now the rook has been brought into play, which will allow Black to have local superiority of forces on the kingside despite being down material. White's king cannot easily run away and Black should get at least a draw out of it, although the position remains complicated.}) 20... gxf3 21. Qxf3 Bd6 22. cxd5 cxd5 $11) 20. Be5 $18 {White finds the best answer and from this point has a fully won game.} (20. Bxd6 $4 Qxd6 {and mate is unavoidable.}) 20... Bxe5 21. dxe5 Kf7 22. cxd5 cxd5 23. Bxf3 $1 {a beautiful, forced end to the game says Houdini via the Fritz interface. Not that hard to find, in reality, as White is up so much material that simplifying down in this manner is an easy win.} gxf3 24. Qxf3 Rh4 {trying to stay as active as possible on the kingside in hopes of a swindle.} 25. Nb5 { now the knight decisively enters the fray, ready to move to the d6 outpost and kick the blockading Kf7.} g4 26. Nd6+ Qxd6 {I was rather proud to find this - indeed the best move - but my opponent finds the obvious flaw.} 27. Qg3 Rh3 28. exd6 {and there is no point in continuing a full rook down.} (28. exd6 Rxg3+ 29. fxg3 $18) 1-0