On 02/20/11 20:48, Erik van der Werf wrote:
AFAICS this (japanese-test-7.sgf) is simply not a final position. If
white plays first at least there's a ko and he will probably capture a
lot of black stones.
I agree that it's unsettled. I think the T18/T17 ko is won by white
even if black starts but with white going first some more stones must
be given up.
Moreover, I think officially the group would be
in seki if the dame are not filled (so then none of the intersections
around S6 are territory).
Neither would a lot of white's territory be, but that official
nonsense really is of very little interest.
[Next comes a quoting nightmare, but I think I've only left Rémi's
parts in.]
On Sun, Feb 20, 2011 at 8:06 PM, Rémi Coulom<[email protected]> wrote:
Crazy Stone is probably wrong, too, because there are some
complicated things happening around S6. I would say S6 is dame,
not S7 and S5. B12 and P1 look like really weird dame, though. So
do K15 and N12. The right-side of the board is complicated, but
the rest should be obvious for GNU. Unless I am mistaken...
The weirdness comes from GNU Go starting to play ko and ko threats in
the middle of the final status determination. To see more exactly
what's going on run the command line
gnugo -l japanese-test-7.sgf --score aftermath -o analysis.sgf
After Black connects all the false eyes in the top right, and White
fills dame, then White could kill by atari at T6. So I believe
Black must play S6 to prevent T6, so S6 is dame.
It's not uncommon that one player needs to reinforce with another move
inside what looks like territory after dame has been filled, but that
reinforcement is not necessarily unique. Thus it's not always possible
to say exactly which points are dame, although their total number is
well determined.
Also when GNU Go starts playing ko threats in the determination of
final status the list of dame points may be completely off but the
score should be correct if it managed to resolve the ko fight
correctly (big if there).
/Gunnar
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
