[computer-go] use for Monte Carlo on 19X19?

[forrestc at adnc.com]

Lars said:
> I had build an Monte-Carlo GO-Engine (GOMonCy) wich uses the Japanese
> scoring system. It reached a win  rate against GnuGO 3.6 level 10 of
> stable 50%-52%. I used territorry-statistics about the Monte-Carlo
> outcomes. You get a probability for every field telling you who is the
> owner. It works quite good, but I thougt  that nearly everyone is using
> such statistics, isnt't it?

I don't know what dragongoserver.net uses to score human-human games, but
I almost always have to tell it which chains are dead. Does this mean that
their scoring routine should be using Monte Carlo, or are those
territory-statistics pretty questionable?

Forrest Curo



-----------------------------------------
This email was sent using AIS WebMail.
http://www.americanis.net/