Re: computer-go: question

[Tim Klinger <klinger@xxxxxxxxxx>]

At 06:30 PM 12/11/99 -0800, you wrote:
>
>Efficiently using your hypothetical oracle might involve invoking it
>thousands of times per move, at deeper levels of the move tree. One
>of the characteristics you presumed for it is that it is fast. Such
>an oracle would greatly speed evaluation of positions. One would
>still need to estimate territory but this is fast (using Bouzy's 5/21
>algorithm, for example) compared with the overhead of determining
>which groups are alive and dead. With such an oracle, brute force
>searching of the move tree (like chess programs) would not seem
>so impossible.
>
>But such a use could not be mimicked with the human expert.

That's true, but personally I still think that brute-force is an
unpromising approach, even with a much better evaluation function.  For one
thing, as I mentioned in my last post, as good as the evaluation function
is using the oracle, it is still not great.  It doesn't take into
interactions between life and death strategies.  And, regardless of how
much better it is than current evaluations, you'll still have an awful lot
of nodes to search...

Tim