Quiescensce Search

I've changed my mind. I think the most important next step is to implement a quiescensce search algorithm. This should help prevent boneheaded moves like Qxa2 on move 3 of the center counter. What the quiescensce search does is goes to a fixed ply, but then continues on for any position where there is a capture until no more captures are available. So, it will take each line until it's "quiet". Then it can get a much "fairer" evaluation. I think this will help tremendously - though is should still play weak chess since the eval function is currently only a piece value count. But at least it should stop playing the clearly terrible blunders that even a non-chessplayer would generally avoid.

Since my eval function is a recursive algorithm, this should be easy to implement. I think all I'll need to do is add a check to see if at max ply we have any captures, and if so continue with those captures...

Ed.