Experiments in Computer Go Endgames

Martin Müller
Ralph Gasser
Informatik, ETH Zürich

Recently, the mathematical theory of games has been applied to late stage Go endgames [Berlekamp 91], [Wolfe 91a]. Based upon this theory, we developed a tool to solve local Go endgames. We verified all exact game values in [Wolfe 91a] and analyzed some more complex positions. We extended our method to calculate bounds for positions where optimal play depends on Ko.

Our program Explorer uses this tool to play full board endgames. It plays a non-trivial class of endgame positions perfectly. In a last section we discuss heuristic play for a wider range of endgames.