Fusilli: 37.Qxa5 has its risks. Rather than 37...d4 I am looking at 37...Rxg2 38.Kxg2 d4+ (diagram) click for larger viewWith the following continuations:
A) 39. Kf1 Qxh3+ wins, for example: 40.Ke2 Qf3+ 41. Ke1 Qh1+ 42.Kd2 c3+ 43.Kc2 Be4+ winning the rook. B) 39. f3 Bxf3+! 40.Kxf3 Qxh3+ 41.Ke2 41...Qxe3+ (41...Qg2+ 42.Kd1 c3 43.Qd8+ Kh7 44.Qxd4 White wins) 42.Kd1 c3 (diagram)  click for larger view...but here White has the saviour 43.Qa8+! (43.Rc2 Qd3+ 44.Ke1 Qxc2 45.Qd8+ Kh7 46.Qxd4 Qb1+ and 47...c2 winning) Kh7 44.Qg2 and black has to give perpetual with 44...Qd3+ 45.Ke1 Qe3+ 45.Kd1. C) 39. Kh2 dxe3 40.fxe3 Qf5 41.Qd2 Qe5+ 42.Kg1 Qg3+ 43.Kf1 Qxh3+ (diagram)  click for larger viewand now 44.Kg1 draw, or 44.Ke1 Qh1+ 45.Kf2 (45.Ke2 is suicidal because of 45...Bf3+ 46.Kf2 Qg2+ 47.Ke1 Qg1#) Qg2+ draw. I am moving the pieces around but I'm not checking with the computer, so the analysis may well be flawed (feel free to refute!) For example, in the second line, I suppose White might not take the bishop (but why wouldn't he? His king is exposed anyway and it looks like a good idea to eliminate an attacking piece). |