Wrong Bishop

When my opponent blundered to give me a knight fork, I saw a clear route to a draw.

White to move

46.Nc1+ Kb2 47.Nxa2 Kxa2 48.f4=

The engine suggests that White can play for advantage with 48.Rd1 Bc5 49.Rc1 Bb4 50.Rc7 a3 51.Rxh7 Kb2 52.Ra7 White's active king and three to one pawn majority more than compensates for Black's bishop, although this, too, should be drawn.

My plan was to exchange off Black's g-pawn and assure myself of reaching an ending where Black's bishop cannot guard the promotion square of his remaining pawn.

49.g4 a3 50.f5

Black to move

Only after the game, did I give any thought to how Black might have disrupted my plan.

50...gxf5 

This exchange gave me the draw that I sought. However, Black could try 50...g5 51.f6 a2 52.f7 Bc5 53.h4 again seeking to leave all Black's pawns on the h-file. 53...h6?? 54.hxg5 hxg4 Black's bishop will prove to be overworked after White's king mows down the g-pawn.

51.gxf5

Black to move

My after school chess club saw this position on Tuesday. They were asked how play might proceed and what should be the result. One student identified that the bishop does not control the promotion square for the h-pawn.

51...a2 52.f6 Bxf6 53.h4 a1Q 54.Rxa1 Bxa1 55.Kh1 Bf6 56.h5 h6 57.Kg2 Kc4 58.Kh1 

Post-game analysis with Stockfish set at a depth of 18 plies on chess.com produced a comic evaluation.

Black to move

Posting this position in social media with the comment, "Stockfish is wrong. This is a dead draw", elicited a lot of comments. Some posters revealed that they do not know this elementary endgame. Others sought to explain it to them.

White's pawn is of no consequence. Black cannot evict the White king from h1. The king will shuffle between h1 and any of the adjacent squares until these are taken away. My opponent played it until stalemate.