[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gnugo-devel] minor cleaning

From: Gunnar Farneback
Subject: Re: [gnugo-devel] minor cleaning
Date: Sat, 21 Sep 2002 20:07:04 +0200
User-agent: EMH/1.14.1 SEMI/1.14.3 (Ushinoya) FLIM/1.14.2 (Yagi-Nishiguchi) APEL/10.3 Emacs/20.7 (sparc-sun-solaris2.7) (with unibyte mode)

Paul wrote:
> actually, i wanted to know what you think about this:
> owl always consider only one vital point per eyespace because
> compute_eyes() and compute_eyes_pessimistic() return only one
> attack/defense point. maybe it would be useful to consider several
> vital points if we have single eyespace, like in the position we
> discussed on nngs some days ago: 
>   !.XX..!
> here white has no chance to survive if he picks up left marginal
> vertice, but has such a chance if he tries the right one.

Sometimes it can be useful to try more than one vital point for the
same eyespace, at other times it's just a waste of time and more
importantly owl nodes.

> should i try this?

You could do that but I'd guess it would pay off better to try to
improve the heuristics for choosing what margin to try playing when
the graph matching considers the eye value as settled.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]