Yes. And I just follow your context that you want F1 has higher probability to 
be played. In this way, you can add up the probability of C1 to F1. 

I haven't such a rule in Erica right now, so I would expect your result.

Aja
  ----- Original Message ----- 
  From: Go Fast 
  To: Aja ; [email protected] 
  Sent: Saturday, December 04, 2010 1:25 PM
  Subject: Re: [Computer-go] Tactical misevaluations in Fuego MCTS


  Are you proposing that in the playout mode, if it happens to pick up a 
"meaningless" self atari, try to find a more meaningful position around the 
nearby strings?

  I simply ignore such move, and pick next random move. I may try that approach 
and see if it improves its overall strength.


  2010/12/3 Aja <[email protected]>

    Hi Go Fast,
    Â 
    A simple solution to this sitaution is: 
    Â 
    When the forbidden selfatari (also an atari) of the bottom-down 3 stones 
is detected, then try to atari White's string from its the other liberty 
point, which is the one you want.
    Â 
    Aja
      I have one test case similar as following diagram. All other places are 
mostly settled, except the bottom three string.

      O O X X X O O O
      X X O O O X . .
      . X . O O . X .
      - - - - - - - -

      In most cases, StoneGrid's playout ends up with the left bottom corner 
being killed. With self-atari analysis, either side will not play self atari 
moves, which is right behavior. However, without analysis on the 3-lib string, 
the chance for the black to find the right move in the random playout is very 
small. It is far less than 1/2. As when black plays randomly at one of its two 
outside libs, it will bring the white to play around it, and end up with black 
being killed.


    _______________________________________________
    Computer-go mailing list
    [email protected]
    http://dvandva.org/cgi-bin/mailman/listinfo/computer-go


_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Reply via email to