Am Dienstag, den 06.03.2012, 15:35 +0000 schrieb Matthew Woodcraft: > Stefan Kaitschick wrote: > > Bots love to throw in useless ko threat type moves now and then. > > Sometimes its just a wasted threat, at other times it loses a useful > > liberty too. Why do even strong bots show this "stupid" behaviour? > > > Ko threat type moves have one thing in common: a big payoff if the > > opponent answers incorrectly. If the first response tested is wrong, > > then the ko threat will start off with a very good winrate. The > > correct response will be found fairly soon, but by then the original > > move will have a winrate bonus. > > > This bonus is small, but it is there. The problem is, that the winrate > > degradation, that a useless ko threat causes, can also be small. So if > > the (erroneous) bonus outweighs the (real) loss, the bot will conclude > > that throwing in the ko threat is good. > > > I think it's a little more complicated than that. > > Suppose that X and Y are a ko threat and its only response, and A is the > 'best' move (the move that the computer would play if the ko threat > wasn't an option). > > Then X is competing against A. > > X has a (large) advantage for each time the 'opponent' fails to choose > Y, but it also has a disadvantage for each time the 'player' chooses a > move other than A in the position after X-Y.
The problem is even worse: if the playout does not handle using ko threads, X would (most of the time) not have any disadvantage at all... > > -M- > _______________________________________________ > Computer-go mailing list > [email protected] > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go > _______________________________________________ Computer-go mailing list [email protected] http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
