> > and which, i agree, gives far too much credit to MC players, > > since they roughly emulate this line of play "purely" by accident. > I don't agree with that. There is no accident here, MC really only > cares about winning and has no ego about winning big.
It was intended as a little play on words, since they're roughly random draws. I'm not as funny as I think I am, apparently. > > if you can prove* that you can stop an invasion with no > > risk to yourself, people will generally play to stop it, > but of course MC programs don't care. They might stop it by accident > but not on purpose. Is this what you mean by unbalanced? Monte Carlo > programs don't have a specific policy to lose stones on purpose just to > keep the score "balanced" or close. They simply don't care either way. No, but as a technique for very slightly reducing the error in your measure of the win probability, once you have secured all of your territory in a provable way (likely testing for this is time consuming), and the rest of the moves are dame or new invasions (i.e. the first off-color stone dropped into secure territory), those invasions can be stopped at will by a strong enough player, and probably should be. i can't think of any counterexamples off hand, which doesn't mean that i'm right, of course. similarly, connecting not-yet live stones to provably live groups to prevent them from being captured should often very slightly reduce the error in the measure of the win probability. there are times when doing so is not as valuable as reducing an opponent's territory, so this would only be an interesting mode to switch to if there were no more reduction to be had anywhere on the board for the player involved. again, this is perhaps painful to test for. one thing i wonder is how close to the end of a 19x19 game any particular MC machine (with some number of draws at some cpu strength, etc.) plays perfectly, in the sense of not being able to lose a won game. at moves near and before this point, some heuristics like this should help. i think that there are likely some moves during which a player has provably won the game in an efficiently computable way but isn't yet aware of the fact and still has an opportunity to screw it up. s. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
