[Computer-go] Solving semeai

Darren Cook darren at dcook.org
Fri Jun 3 00:17:31 PDT 2011

(I just thought I'd bring together a few recent comments and see if they
inspire anyone.)

Yamato wrote:
> Zen has some code to handle very simple semeais in the playouts. But 
> I am not satisfied with the code. To solve the semeai issue fully,
> we will need a better algorithm to share the tree information with
> the simulation.

Aja wrote:
> it's clear that we need something new and BIG to solve semeai, rather
> than heavy knowledge implementation.

I believe Valkyria solves some semeai in the playouts, and I believe it
does this with a large set of hand-coded patterns (i.e. "heavy
knowledge"). E.g. The Many Faces vs. Aya game that was discussed
recently, where both programs misunderstood it, Valkyria had the needed
knowledge to make it simple.

But Magnus goes on to say:

> Semeai is a good example. Valkyria is good on simple cases close to
> the winning capture but there are an unending amount of complex
> semeai where this knowledge only have an effect deep down in the
> playouts. Some algorithm of online learning of good tactics seems to
> be necessary.


Darren Cook, Software Researcher/Developer

http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)

More information about the Computer-go mailing list