Start with the "low hanging fruit". You haven't implemented alpha-beta pruning yet, but you should, that is an enormous speedup. After implementing alpha beta pruning you should spend a lot of time on move ordering - doing a lot of experiments to see what is best as this will give you additional enormous speedups.
But even with random move ordering, alpha beta pruning is an enormous speedup and it's absolutely free - no degradation in move quality. With excellent move ordering it can speed-up up many times faster yet. - Don On Mon, 2006-11-13 at 01:20 +0100, Wodzu wrote: > Greetings guys. > > Ive implemented standard minimax algorithm but is so enormous slow even on > 9x9 board. For example when depth = 3 (player1 move, player2 move, player1 > move) computing time at the begining is like 1 minute. > I know that I should implement alpha-beta pruning variation to speed up it > but what else can I do? > I am eveluating position by using floodfill metod so for every single > evaluaton i need to copy entire board. Ive heard that there is a > possibility to store the board and not using floodfill to often by > remembering what has changed and to just undo such changes but i cant > figure it out by myself. Any faster way than floodfill to check the teritory > and catched stones? > Thanks for any tips, > > Regards. > > _______________________________________________ > computer-go mailing list > computer-go@computer-go.org > http://www.computer-go.org/mailman/listinfo/computer-go/ _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
