Monday’s puzzle was taken from a game I played with the white pieces against Roger in last week’s RTCC lichess arena. I’ll start by taking you back a couple of moves.

After various misadventures in the middlegame we reached this position where I had to make a decision.

I should be winning here by pushing my pawns, but I completely misjudged the position, deciding that, because his rook was actively placed on an open file I should trade it off.

So we continued 33. Rd1?? Rxd1 34. Bxd1, reaching Monday’s position: here it is again.

I’d assessed this position as favourable for me, and, judging from his comments after the game, Roger did as well. I have four pawns to his three, they’re further advanced than his, and I have a bishop against a knight in a position with play on both sides of the board.

I was astounded when Stockfish immediately told me that Black could win here by playing either 34… a5 or 34… b5. The decisive factor seems to be that Roger has outside passed pawns, while my pawns are on the same side of the board as both kings. Play it out for yourself and see what happens.

The engine tells me Black has quite a few drawing moves as well. Roger chose the very natural 34… Ke7, which isn’t one of them. Now I do indeed have a winning advantage. I’ll tell you what happened next another time.

Extraordinary – what an amazingly difficult and deep game chess can be. Roger and I have, in total, been playing chess for 130+ years, but neither of us has ever encountered an ending like this before.

So if you gave Black a win on adjudication in Monday’s position (the second diagram above) you were correct. Well judged!