Hi,

Just a comment on strength. It does not really show anything other then
GNU Go strength against GNU Go. But I played 100 games between GNU Go
3.8 lvl 10 and GNU Go 3.8 lvl 12 and actually lvl 10 won 52 against lvl
12 48. I have not analyzed any of the games. Both played black 50 times.

Jens Yllman

On 2010-07-09 13:05, Darren Cook wrote:
> At http://www.gnu.org/software/gnugo/gnugo_3.html#SEC35 it says "The
> default values of these parameters may be found by running gnugo --help."
> 
> That is confusing (maybe it applies to a previous version of gnugo?), it
> should say:
>  ... by running "gnugo --help debug" or "gnugo --print-levels".
> 
> BTW, has anyone done any studies of win rates of version 10 against
> version 12 (or any versions about 10), and/or its success on problem
> sets? The level differences seem to mostly apply to tactical search, so
> I'm thinking it should at least do better on problem sets?
> 
> Darren

_______________________________________________
gnugo-devel mailing list
gnugo-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/gnugo-devel

Reply via email to