Good game!
Unusual Opposite Colored Bishop Endgame Where Side Down a Pawn Comes Back to Win!
This is an interesting endgame, thank you for sharing.
9 Bd1? loses. 9 Be2 is more active as the bishop attacks b5 while the king defends pawn b3.
15...bxa4? is a missed win: 15...b4+ a knight pawn is stronger than a rook pawn.
17 Kc2? loses 17 Kc4 is much more active defence.

This is an interesting endgame, thank you for sharing.
9 Bd1? loses. 9 Be2 is more active as the bishop attacks b5 while the king defends pawn b3.
15...bxa4? is a missed win: 15...b4+ a knight pawn is stronger than a rook pawn.
17 Kc2? loses 17 Kc4 is much more active defence.
Thanks. Glad you found it interesting.
And I agree with all your points. As far as 15...bxa4 is concerned I erroneously thought that having the colored bishop that controls the promotion square was correct but obviously looking back my king then has a hard time participating in the pawn's advancement because my king can be easily cut off.
I do not know exactly how often the scenario referenced in the title occurs but being that opposite colored bishop (ocb) endgames are usually difficult to win even if up a pawn (or sometimes even up two pawns) it probably does not occur in a very high percentage of ocb endings. And at the point where this got down to an ocb endgame it was pretty drawish so it is not as if though it was an atypical ocb ending where one side had a clear edge.
Also by no means was this played perfectly by either side (for example 15...bxa4 rather than 15...Kxa4 throws away the winning advantage!) but this game just goes to show that the title scenario can occur without egregious blunders (hanging a bishop, erroneously allowing a pawn to queen, etc.) by either side.