2016-03-26 2:48 GMT+00:00 Petr Baudis <pa...@ucw.cz>: > > The word covered by the speaker's head is "self". Bot results in > self-play are always(?) massively exaggerated. It's not uncommon to see > a 75% self-play winrate in selfplay to translate to 52% winrate against > a third-party reference opponent. c.f. fig 7&8 in > http://pasky.or.cz/go/pachi-tr.pdf . Intuitively, I'd expect the effect > to be less pronounced with very strong programs, but we don't know > anything precise about the mechanics here and experiments are difficult. >
Note that recently for Crazy Stone and Zen improvements in self-play also transferred to playing strength against human players. According to Remi and Hideki, Crazy Stone and Zen are both >=80% stronger with a policy net and they both reach 7d on KGS (1 stone stronger). But generally I agree that we should be cautious about self-play results. Aja It's no doubt today's AlphaGo is much stronger than the Nature version. > But how much? We'll have a better idea when they pit it in more matches > with humans, and ideally when other programs catch up further. Without > knowing more (like the rest of the slides or a statement by someone from > Deepmind), I wouldn't personally read much into this graph. > > -- > Petr Baudis > If you have good ideas, good data and fast computers, > you can do almost anything. -- Geoffrey Hinton > _______________________________________________ > Computer-go mailing list > Computer-go@computer-go.org > http://computer-go.org/mailman/listinfo/computer-go >
_______________________________________________ Computer-go mailing list Computer-go@computer-go.org http://computer-go.org/mailman/listinfo/computer-go