4 O O O O . . . . . 4
3 . . . . O . . . . 3
2 # # # # O O O O O 2
1 O . O # O . . . . 1
A B C D E F G H J
On Wed, Jun 27, 2001 at 01:22:53AM +0200, Grajdeanu, Adrian wrote:
> > No program is strong enough to make intelligent use of
> > suicide moves, or to
> > play triple kos correctly. When that is no longer true then
> > the issue may
> > be a problem. At the moment it seems that some programs are making
> > unreasonable restrictions on the choices their opponents
> > make. (Don't get
> > me started on the handicap stones).
>
> Yeah, you guys got me curious: when would a suicide move be desirable? Or a
> triple ko (other then asking for a draw)?
A suicide move can be useful as a ko threat. E.g. white B1 here:
4 O O O O . . . . . 4
3 . . . . O . . . . 3
2 # # # # O O O O O 2
1 O . O # O . . . . 1
A B C D E F G H J
This is pretty academic at the level of ko understanding that most
(all?) current programs have. Judging from my experience, it's more
likely that you'll kill your opponent's program by playing a suicide
move it can't process. That's not a strategy which would have occurred
to me, but it did actually happen to my program in an AGA Congress
tournament a few years ago. (The suicide move was due to a bug in my
program, not due to my devious plan, and certainly not due to my
program's deep understanding of ko!)
Probably triple ko *isn't* a problem if you're willing to treat some
games as draws/jigos. But I prefer rulesets which always give a pure
win or loss.
--
William Harold Newman <william.newman@xxxxxxxxxxxxxxxxx>
As usual, this being a 1.3.x release, I haven't even compiled this
kernel yet. So if it works, you should be doubly impressed.
-- Linus Torvalds, announcing kernel 1.3.3 on the linux-kernel mailing list.
PGP key fingerprint 85 CE 1C BA 79 8D 51 8C B9 25 FB EE E0 C3 E5 7C