I was able to replicate the success (and with more iterations, failure) of the all-as-first heuristic. But I have not been able to see an improvement when I prohibit multi-stone suicides (I always prohibit single-stone suicides). Forgive me, but I am only interested in your response if you have actually tried allowing multi-stone suicides while preventing single-stone suicides in your engine, because feelings and intuition don't trump what I have witnessed in my two independent engines.
Also, I have been unable to reproduce the benefit of the proximity heuristic (play adjacent or diagonal to the last move 25% of the time) in the one engine that I have tried it. Has anyone besides Peter Drake (it was his paper, right?) reproduced this benefit? _______________________________________________ computer-go mailing list [email protected] http://www.computer-go.org/mailman/listinfo/computer-go/
