It's not that easy. The statistics I used were obtained from 9x9 games, and there was quite some manual tuning based on visual inspection of problematic positions (using my own subjective judgement as a Go player).
Erik On Sun, Jan 2, 2011 at 5:05 PM, Olivier Teytaud <olivier.teyt...@lri.fr> wrote: > > Ok. > If you can send your binary (linux ?) to me, I'll do the comparison on 13x13 > and 19x19 - I think the advantage of Yizao's pattern is mainly on big > boards. > (if it's under windows I'm a bit too lazy for that ...) > Best regards, > Olivier > > >> >> The playout policy I used in the 2007 version of Steenvreter was >> >> developed independently of the Mogo policy. However, some time after >> >> the olympiad I also implemented Mogo's version for comparison. IIRC >> >> the policy with Yizao's patterns was of comparable quality at equal >> >> numbers of simulations but a bit faster so therefore over-all better. >> >> >> > >> > Thanks for this interesting comment. Is your policy described somewhere, >> > and >> > was this comparison in 9x9 or 13x13 or 19x19 ? >> >> No, I never published anything on it. The comparison was only for 9x9. >> > > _______________________________________________ > Computer-go mailing list > Computer-go@dvandva.org > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go > _______________________________________________ Computer-go mailing list Computer-go@dvandva.org http://dvandva.org/cgi-bin/mailman/listinfo/computer-go