TalkChess.com

[H.G.Muller]

I am currently writing an engine for playing large Chess variants (upto 36 x 36 board). Because of the big size and powerful pieces invites a fresh look onto the problem. Perhaps some of the techniques I am developing for it can be useful for 8x8 Chess as well (e.g. for building a more efficient Quiescence Search). But in any case it is a lot of fun. Today I had it play its first game on a 12 x 12 board. I beat it easily (although it was only my second game in that variant), because it only searched 3-ply plus a limited QS (of at most 2 ply). But there still is a very long way to go until I have implemented everything I have in mind (like incrementally updated attack map / mobility eval).