Forums

Mate with king+rook vs. king+bishop

Sort:
SacrifycedStoat
r+k vs k involves ditching the king to move. If they have a bishop, it’s a draw via 50 moves because they never have to move towards the corner.
MARattigan
Knight_king1014 wrote:

K and R vs K and B is a draw assuming the one with the bishop doesn't make a mistake.

Really endgames don't work like that. If the side with the bishop converts into the ending more than a third of the possible resulting positions are theoretical wins for his opponent. If the side with the rook makes the conversion then only around 3¼% of the possible resulting positions are theoretical wins for him. (But most positions, 81.6%, are drawn.)

The mismatch suggests that most of the wins with the rook to move are immediate captures, forks and skewers. The remaining wins can take up to 29 moves.

If you get the hang of this game, you should have cracked it. (It's maximal DTZ - the KRK after conversion in a longer mate is not enlightening.) When how to handle alternative Black responses is not obvious try defending them against Syzygy and see what it does to you.

 
MARattigan
SacrifycedStoat wrote:
r+k vs k involves ditching the king to move. If they have a bishop, it’s a draw via 50 moves because they never have to move towards the corner.

If you play the rook in either endgame, you don't necessarily need the opposing king to move into a corner, but neither is it true that he never has to move towards the corner if he has a bishop - see the example above (e.g. Black's move 9).

MARattigan
ChessNetwork wrote:

No, it's not always possible. It's theoretically a draw. In practice however, many continue to not hold the position to a draw.

How can it be theoretically a draw with >4M theoretically winning positions?