If SEE was never wrong, what would be the point of a quiescence search ?OliverBr wrote:I see. This seems to be a common pattern. but I think it is dangerous. I can create situations where SEE is just wrong (due to pinned pieces), and so you leave out good moves in your QS. This yields to different results...Thomas Gaksch wrote:Hello Oliver,
you can reduce a lot of nodes in QS if you only take capture moves where SEE >= 0.
Tony