Re: [gnugo-devel] reinstating value(100) owl defense moves?

[SP Lee]

> Thinking about the owl code, it occurred to me that there could
> perhaps be significant gains to be had by reinstating and judicious
> use of the special case of value(100) for owl defense moves.  That is,
> if an owl defense pattern matches and has value(100), owl_defend
> declares victory without further reading.
>
> Clearly, bad patterns here could cause significant problems, so
> caution would be needed.  However, I believe the potential gains are
> also quite significant.  Currently, for example, owl reads out the
> life and death of the tripod group every time it appears in a game.
> It gets it right, and considers all important variations, but it takes
> over 400 owl nodes to do it.  That means that there could well be
> cases in deep reading where owl would arrive at a knowbly alive tripod
> group, but be unable to read it correctly due to depth or node count
> limits.  A handful of patterns could at least partially fix this:
>
> |.????
> |.????
> |.O???
> |*.O??
> |..O..
> |.....
> +-----
>
> :8,-,value(100)
>
> This pattern reduces the node count on ld_owl:151 from 435 to 363.  A
> few related patterns could easily drop it much further.
>
> After writing a small handful of patterns, I was able to reduce owl
> nodes on ld_owl.tst from 22636 to 21420 (~4%); more gains are clearly
> possible without much work.
>
> So, before I go spend much more time on this, are there any objections
> to the idea?
>
>
> Evan Daniel
>
>

My question is, are we going to make a gnugo which is just 1 or 2 stones
stronger than the current version or are we aiming at a much stronger
gnugo? I would believe that this patch can improve the performance of
gnugo, but it must be reworked in the future versions. In my opinion,
there are too many options to consider during defense (even in the above
example) and we should be careful for emphasizing any one option.


SP Lee