Silly idiotic rule: K+N vs. K+p is 1-0 when timeout

@Toadofsky

> 1. Flood-fill whether either player can capture a piece; if so, it is very likely the attacker can checkmate

Example: both sides can capture, but neither can checkmate (if I understand correctly)

That last position isn't a draw. White can play Be6 and black can take the bishop.

true; I understood

here is another try; position seems to be possible (4 black pawns moved right 1 step), although not practical

That's not a draw either. Both players can win.

ok, next try lichess.org/dVQxAco6 although it seems to be impractical in standard chess, but I managed to get similar positions (as white!) in 960 chess, he-he

nvm

#87 (edited; pawn on h3) is a draw and makes me realize that flood-fill should include both promotions & captures.

Where flood-fill (i.e. null-move search where only one color's pieces can move) might fail is if a knight has just enough freedom to hang itself to a pawn capture and not enough to make a capture, but currently isn't en prise.

@lovlas a variation on the same theme with space 8*64*64*64 (approximately) to search:

I see, it is indeed simpler and more consistent. Summarized in this little book:

goo.gl/images/C3C8XT

PS: Please, keep it simple! It‘s sound.