So far I played these MC programs at it seems they are doing well against humans mostly because the moves they play are bizarre and some times throw unreasonable contact fight challenges. They win more often than they deserve just because many weak players (like KGS 4k level players are) quite often panic and start making ever funnier moves than they do usually.
Point is that being tactically challenged is rather common amongst weaker players but as get to higher ranks it gets rather rare quality. And MC programs simply suck tactically and adding more simulations gives very little gain for amount of CPU spent. The full board usage may getter better faster. Even top MC programs fail to see that a group with 3 liberties with no eyes is dead. So by doubling the simulations it would now fail on 4 liberties instead of 3? How many fold the simulation count would have to understand 2-eye concept? I am pretty sure that 32 fold increase will not do that, but obviously there is amount of doublings that will do the trick. Question is will that happen on our lifetimes. Or will the next generation be some sort of hybrid approach? Petri -- Petri Pitkänen e-mail: [EMAIL PROTECTED] _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
