On Tue, Jul 5, 2011 at 1:48 AM, Leon Matoh <[email protected]> wrote: > > I know that MCTS winrate does not give any usufull information about quality > of the move except winning chance > > I said that normalized winrates with random komi window give much more > information. > > So to show the concept I played two games on 13x13 between > > 1) standard pachi > > 2) and modified pachi with random komi from fixed window [-17.5,32.5] for > all game > board.c 1404 > return (32.5 - rand() % 51) + board->handicap + > scores[S_WHITE] - scores[S_BLACK]; > meaning that for EACH quick count I aplliyed komi as a random number from > window > > Games were close and very informative. > > First one standard pachi won for 0.5 and lost all points in yose. They > played until end. > > Second was misread semeai and nakade so while modified was winnig it > played correct moves and when it lost semeai played nice endgame and lost > for 3.5
Wait. The evidence you present that your change is an improvement is only two games? And the modified version LOST BOTH? > Please take a look at the games. I don't think I will... Álvaro. _______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
