I don't think you can ignore superko in the UCT search - you will lose games. You will also lose games even if you only check the move you will actually play in the game.
But in the play-outs, it has almost no value whatsoever, even if it was free. (It probably has SOME value, but miniscule and clearly would weaken the program because you would do less play-outs in a given amount of time.) If you keep superko tests in the UCT tree, you have a scalable program. - Don On Fri, 2007-05-18 at 09:33 -0700, Peter Drake wrote: > True... > > > My experience has been that (largely) ignoring the extremely rare case > of superko is a better use of the finite resources we have. > > > Have others found the same thing? > > Peter Drake > http://www.lclark.edu/~drake/ > > > > > > On May 18, 2007, at 9:19 AM, Chris Fant wrote: > > > > After search, when actually making a move: > > > 1) Make a copy of the board > > > 2) Compute the Zobrist hash of the current position from scratch > > > 3) Check for superko violations (against a stack of previous > > > Zobrist hashes > > > for positions in the real game,) > > > 4) If there is a violation, go back to the copy and try the next > > > best move > > > > > > But then you have an engine which does not converge to perfect play > > given infinite resources. > > _______________________________________________ > > 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/ _______________________________________________ computer-go mailing list [email protected] http://www.computer-go.org/mailman/listinfo/computer-go/
