>With X to move, Many Faces immediately gives about a 1% win rate, and after >a few seconds, resigns, showing 23742 playouts with 0.5% win rate. 10 ply >principal variation, staring with A7. I don't have any special code to >detect superko or give-2, get-1 in playouts, but the playouts don't generate >self- atari moves in a seki, so I think it never tries J6 for either side.
How does MF recognize that it is a seki without analyzing what happens in the ko? The eye on H5 is false if X can fill J6, so it is premature to use the "both sides have an eye with only shared liberties" rule. > 1 2 3 4 5 6 7 8 9 > A - O - O X - - X - > B - X O - O X - O O > C - O O - O - X O - > D X X X O O O O O O > E - O X X X O X O O > F - - O X O X X X X > G - - X X O O X X - > H - O X O - O O X X > J - O X O O - X - X > X to play. _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
