Re: [gnugo-devel] Connection point not seen

[Evan Berggren Daniel]

I haven't looked at the position in detail; from your description, it
sounds like the solution in the current framework is to add a pattern.
However, I'm guessing Arend's escape rewrite fixes it (arend_3_20.5).  If
so, that is probably the way to go; in either case, a test case would be
useful.  If you're interested in learning about the pattern databases,
this would probably be a relatively straightforward one to add; just be
careful not to make it too general, or it will have negative performance
impacts.

Evan Daniel

On Tue, 3 Jun 2003, SP LEE wrote:

>
> > Attached is a game I just played with the latest GNUGO (from the cvs). There
> > was an obvious blunder on move 158, where GNUGO thought its group L9 is
> > dead, while this group can be simply connected with outside on O14. In GTP
> > mode, owl_attack L9 showed that O14 is an attack point. However, it didn't
> > even consider the move O14 in "owl_defend L9" and thought the group L9 is
> > dead without any defense. O14 is generated by the move generator, but almost
> > without any value.
>
> When I tried owl_does_defend O14 L9, the result is "successfully escaped". So 
> my question is, does it seem like that a pattern should be added? To simplify 
> the situation, refer to the following graph (a=O14):
>
> O O .
> X a X
> . O O
>
> I'm not so familiar with the pattern data base. However, I would think GNUGO 
> should be able to find that 'a' is a connection point for black.
>
> Your comments are highly appreciated.
>
> SP Lee
>
>
>
> _______________________________________________
> gnugo-devel mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/gnugo-devel
>