On Thu, 2007-03-15 at 19:38 -0400, Chris Fant wrote: > > If you allow multi-stone suicide, it will probably avoid a test > > that may be expensive in your program, and so it may turn out to > > be a net improvement in strength per second - especially if your > > testing proves that it doesn't hurt in any measurable way. > > Doing that test does make my C# engine about 18% slower. But even if > fix the number of playouts and not the amount of time, allowing > multi-stone suicide is still a winner.
I thought you were saying it doesn't play worse, but now you seem to be saying it plays better. I suppose there is a chance it would play better, perhaps the suicide helps resolve the fact that these groups are weak anyway? Just a half baked theory. - Don > _______________________________________________ > computer-go mailing list > [email protected] > http://www.computer-go.org/mailman/listinfo/computer-go/ _______________________________________________ computer-go mailing list [email protected] http://www.computer-go.org/mailman/listinfo/computer-go/
