Unfortunately, it's just not that simple, because it depends far more on _how_ the playout is improved, rather than, say, the ELO numbers that measure that improvement. For example, programs exist that are good (in part) because they entirely disregard some lines of play. They may be correct to disregard these lines in almost every case, which generally makes the playout program stronger. However, for the few cases where this heuristic pruning is not correct, the calling program will suffer greatly, because these lines of play become completely invisible to the random playouts, no matter how many playouts are performed. As an extreme example, consider programs that play completely deterministic strategies. These are obviously useless as random players, yet it is in principle possible to construct ones that play arbitrarily well.
Weston On 2/7/07, Ephrim Khong <[EMAIL PROTECTED]> wrote:
Is there any known (by theory or tests) function of how much a increase in the strength of the simulation policy increases the strength of the MC/UCT Program as a whole?
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
