|
|
41. g5 Bf5
|
At this point I was doubting this game is drawable unless white blunders.
Apparently Rb5 would have lead to a combination winning the pawn. |

|
|
|
42. Rh6 Rd8
|
Preparing Bd7 to block the rook. |

|
|
|
43. g6 Rd7
|
Bd7 falls to Rh7, where Bf5 (trying to win pawn or rook) loses to Rg7 , etc. to mate since the king can't get to the d file.
I'm thinking that after Rxd7 Bxd7, it's hard to protect the pawn and the rook. Apparently white could have kept the pawn with this move, while Rd6 would have won it. I keep failing to notice the potential pins. |

|
|
|
44. Rc1
|
If Rxd7 Bxd7, White can't safely prevent Kg7, where the bishop can take the pawn. I think Rc6 was the only move that protected against Kg7, as that leads to Rh7 Kg(f)8 Rc8 , with mate to follow. |

|
|
|
44... Kg7
|
At last! |

|
|
|
45. Rc6
|
White makes a final mistake. A better move would have lost the pawn, however, and lead to a drawn endgame. |

|
|
|
45... Kxh6
|
This is still a draw. A rook and bishop vs. rook is usually a draw (and certainly would be with me on offense). |

|
|
|
46. Kf2 Bxg6
|
I offered my opponent a draw rather than trying for a win for 50 turns. An endgame database confirmed that either the 2R vs. R B or R B vs. R are theoretical draws. |

|
|
|