Annotated Game #282: If it's the first round, I must lose as Black

This next tournament's first-round game unfortunately comes as no surprise, in terms of the result. I very much dislike ascribing a loss to time trouble, since that is normally a symptom rather than a cause, but in this game it is certainly the most important single factor. Right up to the penultimate move I had at worst an even game, but hallucinated sufficiently under time pressure to make the wrong move. Other than that, it was an interesting Caro-Kann Advance with a number of useful teaching points, including how queenside pawn advances could have been employed to good effect.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class C"] [Black "ChessAdmin"] [Result "1-0"] [ECO "B12"] [Annotator "ChessAdmin/Dragon 3.2"] [PlyCount "77"] [GameId "497288691264"] 1. e4 c6 2. d4 d5 3. e5 c5 4. c3 Nc6 5. Nf3 Bg4 6. Be2 e6 7. Nbd2 {I had not seen this move in the Caro-Kann Advance before - White typically castles - so had to think about what would be best to do.} Qb6 $6 {premature development of the queen.} (7... cxd4 8. Nxd4 Bxe2 9. Qxe2 Nxd4 10. cxd4 {and only now} Qb6 $11) 8. Nb3 {this blocks the b-file pressure and frees the Bc1 up, but does not directly challenge Black. It also leaves the knight potentially open to attack from an advancing Black a-pawn.} (8. dxc5 Bxc5 9. O-O $14 {and Black is going to have to lose time retreating the queen or bishop, due to the threat of b4.}) 8... cxd4 9. cxd4 $6 {this allows Black to pressure d4.} (9. Nfxd4 $11) 9... Bxf3 {another premature move.} (9... a5 {I spotted this idea but thought it was too aggressive. However, it nets the d-pawn, so is worth it. For example} 10. Be3 {and now} Bxf3 11. Bxf3 a4 12. Nd2 (12. Nc1 Bb4+ 13. Bd2 Qxd4 $19) 12... Nxd4 $17 {the pin looks dangerous, but Black can play ...Bc5 next move.}) 10. Bxf3 Bb4+ {I was pleased with this, but did not find the right follow-up idea, namely ...a5 to chase away the knight.} 11. Bd2 $6 (11. Kf1 $11) 11... Nge7 {good enough for equality, but there was more.} (11... a5 $1 12. a4 $2 {blocking the a-pawn does not help, since it leaves the Nb3 unprotected.} (12. O-O a4 $17) 12... Bxd2+ 13. Nxd2 Qxd4 $19) (11... Bxd2+ 12. Qxd2 a5 $15) 12. Bc3 Rc8 13. O-O Nf5 {this is a premature deployment of the knight, there is no rush to determine its square.} (13... O-O) 14. Rc1 O-O 15. Qd2 Bxc3 16. bxc3 Na5 {heading for c4.} 17. Rb1 Nc4 18. Qd3 Qd8 {I thought for a while here. The queen should move away from the discovered attack on the b-file; d8 is as good a square as any, allowing her to be redeployed on the kingisde.} 19. Bg4 Nh4 (19... Nh6 $11 {looks awkward but could have been simpler.}) 20. f4 {I missed this and as a result it caused me great problems, as I was counting on ...Qg5 ideas on the kingside. However, there are multiple good options from here.} b6 {around here I got wise to the possibility of a White knight sac on e6, but analysis shows it does not need to be feared. The text move was intended to keep the White knight out of c5, and is not bad.} (20... g6 {restraining a further f-pawn advance.} 21. Nc5 b6 22. Nxe6 fxe6 23. Bxe6+ Kg7 24. Bxc8 Qxc8 $11 {is equal according to the engine.}) 21. Qh3 Rc6 {defending e6 and also along the third rank in general.} (21... g6 $5 {again is solid.} 22. g3 Nf5 $15) 22. Be2 Nf5 23. Bd3 g6 {long think here and the engine validates the choice, but now I start to get down on time.} 24. Bxf5 exf5 25. Rf3 Kg7 {here I focused on keeping the queen out of h6, which should be all right.} (25... Qd7 {better to clear the square for a rook, to increase flexibility. White cannot break through.} 26. Qh6 Na3 27. Rc1 $6 (27. Re1 Nb5 28. Ree3 Rfc8 29. Rh3 f6 30. exf6 Rxf6 $11) 27... Rfc8 $17) 26. Rg3 Qc8 (26... Na3 {this idea seemed to go nowhere for me, but I did not consider it fully.} 27. Rc1 {always seemed to hold everything, but then ...Qd7 followed by ...Rfc8 is a more dynamic way of playing.}) 27. Re1 {now I decide I need counterplay, so play actively and cast the die.} a5 (27... Rh8 $15 {would be solid.}) 28. Qh4 {expected, with doubling on the h-file to come.} a4 $1 {again, the correct and active way to play, as confirmed by the engine.} 29. Nc1 Nd2 {so far so good.} 30. Rh3 Rh8 {I burned more time trying to decide between this and ...h5, which would have been better.} (30... h5 {I was concerned about the idea of a pin on the g-file and tactics involving taking the h5 or f5 pawns. Although I did not see anything concrete, it was enough to deter me.} 31. Ne2 Ne4 $17 {should be fine, though.}) 31. Qh6+ (31. Nd3 {I was more concerned about this idea, which in fact would cause more problems.} h5 {now this is needed anyway.} 32. Nb4 Ne4 33. Nxc6 Qxc6 $11 {the defensive exchange sac looks like it gives good chances to hold.}) 31... Kg8 32. Ree3 $6 (32. Ne2 $11) 32... Ne4 33. Ne2 g5 {while technically this is still a draw, the loose structure in front of my king now leads to my downfall. I mistakenly thought the active play was necessary.} (33... b5 $17 {leads to a significant advantage, according to the engine. White now is vulnerable to back-rank tactics, among other problems, while Black can target the backwards c-pawn on the open file.}) (33... f6 {is another good idea, opening up the 7th rank for defense.}) (33... Qf8 {this would have been good for a solid defense, but I chose the more aggressive text move, despite not seeing a clear outcome.}) 34. Qh5 g4 35. Rxe4 {I did not consider this, but it works. Low on time, I didn't even consider capturing on h3.} dxe4 $2 36. Rg3 (36. Re3 $1 $18 {leaving h3 open for the knight to move to the attack.}) 36... Kg7 {low on time now, but this should still hold.} 37. d5 Rg6 {I thought for a while and under pressure played the wrong rook move. :(} (37... Rh6 38. Qg5+ Rg6 39. Qe7 $2 (39. Qh5 $11) 39... Qc5+ {I calculated a version of this variation but somehow decided it was bad for Black.} 40. Qxc5 bxc5 $19) 38. Nd4 $1 $18 Rh6 {one move too late.} 39. Nxf5+ 1-0