[Computer-go] pachi questions

Michael Williams michaelwilliams75 at gmail.com
Fri Jul 15 15:02:09 PDT 2011


Of course it affects mirror go.  But I wouldn't call the lack of a tengen a
"technical reason" that even-size boards are not popular.  It's a subjective
reason.  People like the way the board looks when there is a tengen.

On Fri, Jul 15, 2011 at 1:36 PM, steve uurtamo <uurtamo at gmail.com> wrote:

> there is a technical reason why even sized boards never show up except
> for artificial situations -- there is no tengen. i believe (although i
> haven't thought about it carefully) that this affects mirror go, so
> makes it a technically different game (more so than just size). so for
> instance, 20x20 boards shouldn't behave "almost" like 19x19 boards,
> but 17x17 should (except for being much smaller). i'm not sure if
> there's a reason to believe that prime sizes are better than composite
> sizes, but could easily be convinced of this.
>
> s.
>
> On Fri, Jul 15, 2011 at 4:58 AM, Petr Baudis <pasky at ucw.cz> wrote:
> > On Thu, Jul 14, 2011 at 07:33:54PM -0400, Michael Williams wrote:
> >> How does Pachi handle integer komi?
> >
> > Naively, but reasonably well. It reached several jigos in the integer
> > komi 9x9 KGS tournaments.
> >
> >
> > On Thu, Jul 14, 2011 at 09:25:11PM -0400, Michael Williams wrote:
> >> I noticed some unusual beharior from Pachi on 6x6 with komi 4.  I
> understand
> >> that this is not a format that you are targeting or have tested
> extensively,
> >> but I thought I would mention it in case it is symptomatic of something
> that
> >> may also come into play at other sizes.
> >
> > Thanks for reporting this. It is caused by the symmetry code, the
> > symmetry is not breaked properly on even-sized boards (I almost never
> > tested Pachi on these). Adjusting the symmetry code to handle even-sized
> > boards is possible, but it would require a serious thought I cannot
> > spare right now, so I applied a band-aid and disabled symmetry
> > compensation for even-sized boards altogehter for the time being.
> > Patches for a real solution are welcome. :-)
> >
> >
> > On Fri, Jul 15, 2011 at 12:05:28AM -0400, Michael Williams wrote:
> >> A feature request related to this:  In uct_evaluate, don't process moves
> >> that are symmetrically-equivalent to moves that have already been
> processed.
> >
> > I have done that now, thanks for the suggestion. But as explained above,
> > it will have no effect on even-sized boards.
> >
> > --
> >                                Petr "Pasky" Baudis
> > UNIX is user friendly, it's just picky about who its friends are.
> > _______________________________________________
> > Computer-go mailing list
> > Computer-go at dvandva.org
> > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
> >
> _______________________________________________
> Computer-go mailing list
> Computer-go at dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://computer-go.org/pipermail/computer-go/attachments/20110715/291ec5d4/attachment.html>


More information about the Computer-go mailing list