Annotated Game #128: Busted out of the opening

In the post-game review of this fifth-round tournament game, my opponent said he thought he had been busted out of the opening, which is mostly true.  His 9th move was simply bad and his position became further cramped and vulnerable afterwards.  However, he missed multiple opportunities to limit the damage or even regain equality, for example on move 15 when he could have made a very advantageous piece exchange.  I finally am able to pull the tactical trigger on move 19 and Black has a losing game afterwards, quickly going downhill.

The course of this game illustrates the psychological power that trends often have on a player; in other words, once you start down a losing path, it's hard to break that feeling and transform your game for the better.  GM Alex Yermolinsky talked about using "trend-breaking tools" to counter that phenomenon and I previously discussed the idea in "Chess performance and chess skills: not the same thing".  If you've perused other annotated games here, you've no doubt noticed a number of examples where I fell into the same psychological trap and failed to take advantage of opportunities to reverse my fortunes.  In this game, it was good to be on the opposite side of the problem, for once.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class B"] [Result "1-0"] [ECO "A50"] [Annotator "ChessAdmin/Houdini"] [PlyCount "53"] {A50: Queen's Fianchetto Defence (1 d4 Nf6 2 c4 b6)} 1. c4 Nf6 2. Nc3 e6 3. Nf3 b6 4. e4 d6 (4... Bb7 {is normally played here, in which case White can execute the same idea of the bishop development as follows:} 5. Bd3 Be7 6. O-O d6 7. Bc2 {followed by d4.}) 5. d4 Bb7 6. Bd3 Be7 7. O-O Nbd7 8. Re1 {White has a large number of possibilities here, but this is by far the most popular, done in order to overprotect the e4 pawn and free up the Nc3 or Bd3 if necessary.} e5 9. d5 {deciding to gain space rather than leave the center fluid.} (9. h3 {is an alternative way to play.} a5 10. Be3 O-O 11. Qc2 Re8 12. Rad1 Bf8 13. a3 Qb8 14. Nd2 Ba6 15. Bf1 c6 16. Qa4 Qb7 17. d5 c5 18. g4 g6 19. Be2 Red8 20. Kh2 Bg7 21. Rg1 Nf8 22. Rg3 Qd7 23. Qc2 Ne8 24. h4 Nc7 25. a4 Qe7 26. g5 Bc8 27. Bg4 Na6 28. Rdg1 Nb4 29. Qb1 Bxg4 30. Rxg4 Nd7 31. h5 Rf8 32. R4g3 Rfd8 33. Rh3 Nf8 34. Nf1 Rd7 35. Ng3 Re8 36. Kh1 Qd8 37. Rg2 Rb7 38. b3 Qd7 39. Rh4 f6 40. gxf6 Bxf6 41. Rhh2 Kf7 42. hxg6+ hxg6 43. Qd1 Ke7 44. Qf3 Qd8 45. Nf5+ Kd7 46. Nb5 gxf5 47. Qxf5+ Ne6 48. Rg6 {1-0 (48) Bui Vinh (2475) -Bao,Q (2333) Cao Lanh Dong Thap 2006}) 9... Nf8 $6 $146 {Black blocks his own castling. This move suggests that that my opponent had little experience in this type of position. Up until now he had taken a great deal of time on the clock, but had played reasonable moves, if not the most challenging. If followed up with ...Ng6, it might make some sense, but this doesn't occur in the game.} (9... O-O {seems to make the most sense;}) (9... Nc5 {is the other alternative in the database and occurred in a game with GM Uhlmann.} 10. Bc2 Bc8 11. b4 Nb7 12. Ba4+ Bd7 13. Bc6 Rb8 14. Be3 O-O 15. h3 Qe8 16. Bxb7 Rxb7 17. c5 dxc5 18. bxc5 Bxc5 19. Bxc5 bxc5 20. Rb1 Rb6 21. Qc1 Qe7 22. Qa3 Ra8 23. Nd2 Nh5 24. Ne2 Bb5 25. g3 a6 26. Rbc1 Bxe2 27. Rxe2 Rb4 28. Qe3 Rb5 29. Nb3 Nf6 30. Rxc5 Rb4 31. f3 Ne8 32. Rec2 Rab8 33. Kg2 R8b5 34. R5c3 a5 35. Qa7 a4 36. Na5 Rb2 37. Nc6 Rxc2+ 38. Rxc2 Qa3 39. Nxe5 f6 40. Qa6 Rb6 41. Qc8 Qe7 42. Nc4 Rb1 43. d6 cxd6 44. Nxd6 Kf8 45. Rd2 Qe5 46. Nxe8 Rb8 47. Nxf6+ {1-0 (47) Uhlmann,W (2413)-Appelt,G (2046) Dresden 2006}) 10. Nh4 $14 {heading for f5, which would make a wonderful outpost.} g6 {Black decides not to let the knight occupy f5, but at the cost of inflicting a dark-square weakness on his kingside and blocking the square for the Nf8.} (10... Nxd5 {is a tactical possibility, due to the hanging Nh4, which however still leaves White with an edge. I recall seeing the line only after I moved, though, so should have considered it beforehand.} 11. Nf5 Nxc3 12. Nxg7+ Kd7 13. bxc3 $14) 11. Nf3 { The weakness caused in Black's camp more than compensates for the back-and-forth moves by the knight.} Nh5 (11... N8d7 {Black instead should get the knight re-developed, also allowing the dark-squared bishop to redeploy via f8.}) 12. g3 {this further cramps Black's position by taking away f4 and h4, a mirror image of what just occurred with White's knight; however, White is already castled and his pieces are better placed, so it's less weakening.} Nd7 {Black finally admits to himself that the knight is doing nothing on f8 and returns it to the center.} 13. Bh6 {White traps the enemy king in the center, comments Houdini via the Fritz interface.} Ndf6 $6 {now the Nh5 has no safe squares left.} (13... Bf8 {challenging the Bg6 is probably best.}) (13... Nhf6 {does not help due to} 14. Bg7 Rg8 15. Bh6 {and Black has lost the right to castle, without any compensation.}) 14. h4 {here my idea was to create a potential outpost on g5 for the bishop or knight. This is too vague, however.} (14. b4 {Houdini prefers a plan of switching to queenside expansion, now that Black is all tied up on the kingside.} Bf8 15. Bg5 h6 16. Be3 Bg7 17. c5 $16 { is one possible continuation.}) (14. h3 {threatening g4 would be the way to continue with kingside play.}) 14... Ng4 {the knight happily occupies the square left uncovered by the h-pawn advance.} 15. Bg5 $6 {this demonstrates a lack of understanding of how the piece exchange on g5 would affect the game. Black should immediately take, following the principle that the more cramped side should be happy to exchange pieces. It would also get rid of the strong dark-square bishop possessed by White.} (15. Qa4+ {Houdini assesses that it would be better to trade queens and start exploiting White's advantages in piece coordination and space on the queenside.} Qd7 16. Qxd7+ Kxd7 17. Bd2 $14) 15... f6 $6 {with this move, Black forces White to preserve his good bishop and inflicts additional cramping on his own position.} (15... Bxg5 16. Nxg5 Ngf6 $11) 16. Bd2 Bf8 $6 {this looks reasonable at first glance, with the idea of retreating the knight to h6, but is too slow, interferes with castling again, and loses a pawn to a White tactic.} (16... O-O $5) 17. Bf1 {here I play somewhat defensively.} (17. Be2 {could have been played immediately, lining up on the Black knights.} Bc8 18. Nh2 Nxh2 19. Bxh5 Qd7 20. Kxh2 gxh5 21. Qxh5+ $16) 17... Bc8 18. Nh2 Nh6 $2 {a major misstep.} (18... Bh6 $142 $5 $14 {would have held things together, although still at a disadvantage.}) 19. Be2 {I was now able to see that this is obviously good for White. Black cannot protect both his knights adequately.} Qd7 $2 {this loses a piece and the game rather quickly.} (19... Nf7 {would limit Black's material loss to a pawn, although he still has a poor game afterwards.} 20. Bxh5 gxh5 21. Qxh5 $16) 20. Bxh6 {this works, but the alternative bishop capture} (20. Bxh5 $5 {is cleaner, not allowing Black to get in the desperado tactic Nxg3.}) 20... Nxg3 21. Bg4 { this removes the bishop from further desperado tactics involving the Ng3.} ({ inferior would be} 21. fxg3 Bxh6 22. Ng4 Bg7) 21... f5 22. exf5 {a case where the "obvious" move is not best.} (22. Bxf8 {resolves the situation in White's favor.} Rxf8 23. exf5 gxf5 24. Bh3 $18) 22... gxf5 $2 {the pressure is too much, Black crumbles, comments Houdini.} (22... Bxh6 23. fxg3 gxf5 24. Bh3 Qf7 {and Black would at least have some compensation for the material, with the e5/ f5 pawn center and prospects for pressure on White's kingside.}) 23. Bh5+ { again a good response, but not the best.} (23. Qc1 {protecting the bishop} fxg4 24. fxg3 {is Houdini's preference.}) 23... Nxh5 24. Qxh5+ {I had calculated this far in the line on move 20, seeing that White would be able to keep the extra piece.} Kd8 (24... Qf7 {I thought would allow Black to resist longer and Houdini agrees, although it's still lost.} 25. Qxf7+ Kxf7 26. Bxf8 Rxf8 27. f4 $18 {is the line I had considered and the engine validates. The last move is intended to break up Black's pawn duo.}) 25. Bg5+ Be7 26. f4 {as in the variation above, this challenges and weakens the e5/f5 pawn duo.} h6 27. Bxh6 { I decided to take the simplest route to victory, as the Bh6 cannot be effectively pinned due to the possibility of checking the Kd8. Down a full piece and having no real prospects, Black resigned.} 1-0