[Computer-go] Seki in playouts

55 views
Skip to first unread message

Michael Williams

unread,
Aug 7, 2012, 4:34:39 PM8/7/12
to compu...@dvandva.org
Could anyone describe the basic implementation for detecting some seki
in a fast enough way to be used in playouts? I understand that it's
not practical to detect them all.

Thanks.
_______________________________________________
Computer-go mailing list
Compu...@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Peter Drake

unread,
Aug 7, 2012, 4:39:18 PM8/7/12
to compu...@dvandva.org
Orego doesn't do this directly. Instead, we massing passing a possible
move, but with a very low prior success rate (1 win in 10 runs). It
will be chosen if all other moves lead to losses.

--
Peter Drake
https://sites.google.com/a/lclark.edu/drake/

Erik van der Werf

unread,
Aug 7, 2012, 6:07:43 PM8/7/12
to compu...@dvandva.org
On Tue, Aug 7, 2012 at 10:34 PM, Michael Williams <michaelw...@gmail.com> wrote:
Could anyone describe the basic implementation for detecting some seki
in a fast enough way to be used in playouts?  I understand that it's
not practical to detect them all.

Just prevent self-atari unless it makes a basic nakade shape and the surrounding group doesn't have eye potential elsewhere.

Erik

Petr Baudis

unread,
Aug 8, 2012, 7:31:08 AM8/8/12
to compu...@dvandva.org
That sounds simple, but "just" testing if a move is self-atari is not so
straightforward, with possible connections to neighboring groups etc.
Also, you must allow throw-ins.

But yes, proper selfatari testing is the way to prevent breaking seki.

Petr "Pasky" Baudis
Reply all
Reply to author
Forward
0 new messages