On 7/12/07, Robert Jasiek <[EMAIL PROTECTED]> wrote:
Jason House wrote: > I personally find territory scoring more interesting. 90% of my > reason > for that is because the game ends sooner... I don't have to go filling dame > (open spaces between chains of opposing colors). This is for most part an illusion. In real world games, the dame are filled anyway. In go server games, it depends on the server whether filling is meaningful. E.g., on KGS it is essential because KGS makes scoring mistakes otherwise. E.g., on IGS it is undefined whether it would be meaningful or not.
IMHO, KGS does a fine job with unfilled dame. It only finds lost points that would be because dame filling causes a chain to be in atari. Honestly, a more complex attack could be missed by the opponent and it seems reasonable for KGS to miss it too. Occasionally, I'll fill dame to force the opponent to acknowledge a more involved vulnerability and fill it. I checked my recent personal game record to see how this applies "in the real world". For this month, I had 5 relevant games (vs. human, no resignation). On average, 4 dame remain unfilled in one of my games. It seems like most games that number of unfilled dame are low (0,3,4,4), but occasionally there can be a lot (10). 4 dame unfilled: http://files.gokgs.com/games/2007/7/10/Touch-jhouse.sgf 0 dame unfilled: http://files.gokgs.com/games/2007/7/9/teee-jhouse.sgf 10 dame unfilled: http://files.gokgs.com/games/2007/7/3/bearkid-jhouse.sgf 4 dame unfilled: http://files.gokgs.com/games/2007/7/3/jhouse-GoTooom.sgf 3 dame unfilled: http://files.gokgs.com/games/2007/7/3/bitchesbru-jhouse.sgf
-- robert _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
_______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
