SEE

Discussion of chess software programming and technical issues.

Moderators: hgm, Rebel, chrisw

abulmo
Posts: 151
Joined: Thu Nov 12, 2009 6:31 pm

Re: SEE

Post by abulmo »

Yes, after Qd1, Ra8 is mate. I tried to make an existing position simpler to concentrate on the Bd1 move, the other possible moves like Qxa1, or the moves following Qd1 like Rxd6 or Ra8 does not matter here. I am sorry to get a too artificial and somewhat meaningless position, but my problem is just that SEE scores differently just based on the order of two rook moves are played.
Richard
User avatar
hgm
Posts: 27808
Joined: Fri Mar 10, 2006 10:06 am
Location: Amsterdam
Full name: H G Muller

Re: SEE

Post by hgm »

The point is that Ra1# is just coincidental, and should indeed be ignored for this discussion. But Rxd6 is intrinsic for the problem you are addressing. If there is an X-ray attacker of the SEE square that X-rays an opponent attacker of that square, these two pieces attack each other, and the outcome of the SEE will be totally meaningless.

The best solution is thus to never cut moves where this is the case, no matter what their SEE is.

Similarly, it is pretty pointless to NOT cut 'good' captures (say SEE = +3) if the opponent has an RxQ attack on you. OK, so you will win a piece if he recaptures you on that square. But of course he will recapture your Queen on that other square instead...