[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: computer-go: Perl Module for next move.
> This was in games between two gnugo versions, which both had a reasonably
> good (and similar!) idea of the ending point of the game. I believe things
> could be much worse if you just have two random neural nets playing, since
> they can "decide" to stop the game at any point. It is next to impossible to
> calculate a valid score on a random position like that, if you play with
> Japanese rules.
--score last or --score [movenum] produces a reasonably
valid score based on Bouzy's algorithm even before the
end of the game, provided the status of all groups on
the board are correctly evaluated. --score end or
--score aftermath generates moves until the end of
the game, then gives a count.
> For the NN experiment, I would warmly recommend using chinese rules, and
> requiring all prisoners to be captured before declaring the game over - that
> is, all stones left on the board are to be counted alive. This sounds like a
> big change in the rules, but actually it can be shown that the optimal play
> is the same under both rulesets for a long way into the game, only in the
> endgame can there be some slight differences...
If this course is adopted I don't think they can use GNU Go
to count the game unless someone writes a Chinese scoring
routine.
Daniel Bump