[Event "St Petersburg Chigorin Memorial"] [Site "St Petersburg"] [Date "1909.02.28"] [Round "10"] [White "Cohn, Erich"] [Black "Rubinstein, Akiba"] [Result "0-1"] [ECO "D21"] [Annotator "Monokroussos,Dennis"] [PlyCount "76"] [EventDate "1909.02.15"] [EventType "tourn"] [EventRounds "19"] [EventCountry "RUS"] [SourceTitle "EXT 2017"] [Source "ChessBase"] [SourceDate "2016.10.25"] [SourceVersion "1"] [SourceVersionDate "2016.10.25"] [SourceQuality "1"] {[%evp 0,76,29,19,23,23,59,47,62,-18,-18,-18,-18,-47,-30,-33,-33,-33,-24,-15,-13,-32,-16,-21,21,0,-4,-5,-5,-11,-11,-11,-6,-30,-27,-29,-21,-21,-23,-19,-13,-16,-13,-20,-20,-20,-20,-22,-19,-68,-68,-68,-71,-75,-77,-77,-77,-77,-77,-77,-76,-242,-242,-364,-364,-1000,-1008,-1006,-1022,-1085,-1105,-1405,-1405,-1413,-1413,-1405,-1427,-1378,-1378]} 1. d4 d5 2. Nf3 c5 3. c4 dxc4 4. dxc5 Qxd1+ 5. Kxd1 Nc6 6. e3 Bg4 7. Bxc4 e6 8. a3 Bxc5 9. b4 Bd6 10. Bb2 Nf6 11. Nbd2 Ke7 12. Ke2 Be5 13. Bxe5 Nxe5 14. Rhc1 Rac8 15. Bb3 Rhd8 16. Nc4 Nxc4 17. Rxc4 Rxc4 18. Bxc4 Ne4 19. Ke1 Bxf3 20. gxf3 Nd6 21. Be2 Rc8 22. Kd2 Nc4+ 23. Bxc4 Rxc4 24. Rc1 $4 {A hideous error, transforming what should be a routinely drawn ending into a lost king and pawn ending. Even if weren't lost, however, it would still be highly unpleasant. That's important: when making decisions it's rare that we can do so based on the ability to find and exhaustively calculate all the relevant variations. What we often can do is figure out what sorts of things could lead us into trouble, and avoid that. The alcoholic should avoid the bar and the compulsive gambler the casino. Neither should say, after visiting those establishments and falling, that they had "miscalculated", believing that they could have been alright had they made one less bet or stayed two minutes less. So too here: it would be ridiculous for Cohn to say afterwards that he saw that Black's king could reach h3 and he'd be completely tied down, but that he missed some subtlety that proved fatal, when he could have reached a position with no need to worry about subleties.} (24. f4 $11 {is fine, with no problems at all. Black's rook has nothing to do on c4, and it can't worm its way any deeper into White's position. White's h-pawn is in no trouble here, and while Rubinstein as the stronger player will no doubt keep at it, there's no reason for White to fear his opponent's prowess in this position.}) 24... Rxc1 25. Kxc1 {[%csl Oh2]} Kf6 $1 $19 {Heading for h3.} 26. Kd2 Kg5 27. Ke2 Kh4 28. Kf1 Kh3 29. Kg1 {Black's achievement is significant, but doesn't automatically win the game. Now what? White will soon have nothing better than to play Kh1-Kg1-Kh1 until and unless Black's king moves away, but while it's not much it still leaves the burden on Black to prove the win.} e5 {This takes f4 off the table.} (29... b5 {was also worth considering, fixing White's queenside pawns as far back as possible. That's often handy in case of an eventual pawn race.}) 30. Kh1 ({Corresponding to the last comment,} 30. a4 {was worth considering.}) 30... b5 {Happiness.} 31. Kg1 f5 32. Kh1 g5 33. Kg1 h5 34. Kh1 {Black has gained loads of space, but what is he going to do with it? Rubinstein's basic idea is simple and strong, though some finesse will be required: he will play for ...g3, so that after the trades his king will gain access to that square. Once that has been achieved it will migrate along the third rank to the queenside, collect White's a- and b-pawns, and then his own b-pawn will promote.} g4 35. e4 $2 {This makes it easy.} (35. fxg4 {is a much better try, as White has a nice resource, one which was overlooked by more than one analyst of the pre-engine era. (The resource, I hasten to add, does not require an engine to find - it's a simple and long-known idea. But it is easy to miss or forget about, which is what makes for such a nice trap.)} hxg4 36. Kg1 f4 37. exf4 exf4 38. Kh1 f3 $3 {This is the only winning move, and there are still more moves to find.} (38... g3 $2 {Obvious, natural, logical, and...a blunder.} 39. hxg3 fxg3 40. fxg3 Kxg3 {Exactly what Black wanted, right? As I said in the note after White's 34th move, the king will migrate along the third rank to take White's remaining pawns, and then he'll promote his b-pawn. Turns out, it's not so simple.} 41. Kg1 Kf3 42. Kf1 Ke3 43. Ke1 Kd3 44. Kd1 Kc3 45. a4 $3 {Whoops! Black either winds up with only a-pawns, which will be useless, or he'll end up with just his b-pawn in a position where White obtains the opposition. Either way, it's a draw.} (45. Kc1 $4 Kb3 46. a4 Kxa4 $1 $19) 45... a6 (45... Kxb4 46. axb5 $11) (45... bxa4 46. Kc1 a3 47. Kb1 $11) 46. axb5 axb5 47. Kc1 Kxb4 48. Kb2 $11) 39. Kg1 Kh4 40. Kf1 Kh5 $1 {Black will use triangulation/corresponding squares to put the white king in zugzwang.} 41. Ke1 Kg5 $1 42. Kf1 (42. Kd2 Kh4 {Triangulation: White's king is too far away.} 43. Ke3 Kh3 44. Kf4 Kxh2 45. Kxg4 Kg2 $19) 42... Kf4 43. Ke1 Ke4 44. Kd2 Kd4 45. Kc2 Kc4 {With a winning zugzwang. Surprisingly, Black wins by just one tempo, but that's enough.} 46. Kd2 (46. Kb2 $6 Kd3 $19) 46... Kb3 47. Ke3 (47. Kd3 Kxa3 48. Kc3 Ka4 $19) 47... Kxa3 48. Kf4 Kxb4 49. Kxg4 a5 50. h4 a4 51. h5 a3 52. h6 a2 53. h7 a1=Q $19) 35... fxe4 $1 (35... gxf3 $4 {is too clever by half.} 36. exf5 e4 {Threatening 37...e3, which wins against 37.f6. Unfortunately for Black, there's a giant hole in this idea.} 37. Kg1 $1 $18 (37. f6 $4 e3 38. Kg1 (38. fxe3 f2 $19) 38... e2 $19)) 36. fxe4 h4 37. Kg1 g3 38. hxg3 hxg3 (38... hxg3 {White can create a passed pawn and even make a new queen; alas for White, "losing" the king is more important than gaining a queen.} 39. f4 (39. f3 g2 $19) (39. fxg3 Kxg3 40. Kf1 Kf3 $19) 39... exf4 40. e5 g2 (40... f3 $19) 41. e6 (41. Kf2 Kh2 $19) 41... Kg3 42. e7 f3 43. e8=Q f2#) 0-1