> And still no one will give me a straight answer. DO all the top program
> now in existence have exploitable flaws far (lets say at least a few
> stones in rank) below their claimed rank or server-calculated rating?

That was my question too, but I can give you some pointers. MCTS
programs are weak in these areas:
  seki/nakade (especially ones bigger than three-in-a-row)
  positions where precise play is required by one side but not the other

As an example of the latter a semeai or L&D problem where black has 3 or
4 moves at each ply but white has just 1 move. What will happen is the
programs (as either black or white) will think it is settled and ignore it.

If you can find a joseki that creates such a situation, *and* that the
program agrees to play with you, then you have your exploit.

Darren

-- 
Darren Cook, Software Researcher/Developer

http://dcook.org/gobet/  (Shodan Go Bet - who will win?)
http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Reply via email to