Well I was hoping for a static definition which could be assessed with minimal calculation. The problem with a definition relying on a best move metric in my mind is that "best move" is a very difficult thing to define in go since we do not (yet) have an accurate position evaluation function.
--- On Sat, 5/10/08, Evan Daniel <[EMAIL PROTECTED]> wrote: > From: Evan Daniel <[EMAIL PROTECTED]> > Subject: Re: [computer-go] Some beginner's questions concerning go bots > To: "computer-go" <computer-go@computer-go.org> > Date: Saturday, May 10, 2008, 12:55 PM > On Sat, May 10, 2008 at 12:29 PM, Peter Drake > <[EMAIL PROTECTED]> wrote: > > On May 10, 2008, at 7:01 AM, Carter Cheng wrote: > > > > Thanks everyone for the responses. My go skills are > somewhat limited (only > > 6-7kyu on KGS) so hopefully I am not belaboring the > obvious. > > I have a few followup questions- > > 1) What mathematically is a seki? I know this is a > local draw but can it be > > determined statically at some point in all cases using > some sort definition > > without placing stones on the board (i.e. searching). > I only know of three > > cases here- the 1 eye each case with one shared > liberty. two shared > > liberties and the half eye situation. > > > > How about "a situation where there are neutral > points yet the best move is > > to pass"? > > Best move for both players -- in some positions and some > rulesets, > pass can be a ko threat. > > To include Japanese rulesets, where filling dame is > unimportant, you > might want to require that pass be the unique best move. > > Evan Daniel > _______________________________________________ > computer-go mailing list > computer-go@computer-go.org > http://www.computer-go.org/mailman/listinfo/computer-go/ ____________________________________________________________________________________ Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
