Re: [computer-go] Gnugo results

[Don Dailey <drd@xxxxxxx>]

I'm testing 3.3.21 vs  3.4, so now I know the code  is almost the same
and it makes sense.

This is  the second test where  3.3.21's level 10 appears  to place in
the middle  of the pack.  It doesn't  appear to be an  issue with 3.4.
Are you  aware of any bugs  that might have been  corrected that could
cause this?

- Don



   Hi Don,

   > Here are the current results of my gnugo testing.  I'm testing version
   > 3.3 at all even levels and the same with version 3.4

   > I don't  see any strong evidence that  gungo 3.4 is better  in head to
   > head games  against 3.3,  although I suppose  it's possible  that it's
   > much better against other programs or humans.

   Which exact version do you mean by 3.3? If you are comparing 3.3.21
   with 3.4 then that is hardly a surprise as there wasn't that much change
   in between (some consolidation of the then newly introduced  break-in
   code plus minor tuning).

   Arend

_______________________________________________
computer-go mailing list
computer-go@xxxxxxxxxxxxxxxxx
http://computer-go.org/mailman/listinfo/computer-go