Christian Anthon wrote:
>     > can you change the 'supernatural' etc setting in the match statistics?
> 
>     No, not without recompiling.
> 
>     (I believe it is a good idea to have these thresholds fixed, isn't it? A
>     discussion around this may be fruitful.)
> 
> 
> I think it is important to keep the error rates consistent, but not
> necessarily the translation to supernatural etc.
> 
> In fact the current translation is plain stupid as it uses the same
> numbers for move, cube and totals. In my experience the following ratios
> are fairly independent of match length:
> cube-rates:move-rates:total-rates:snowie-rate =  1:2:2:4.

Perhaps another approach is to ignore "easy" chequer plays, i.e. similar
to cube decisions, only count the moves that can potentially be
misplayed (perhaps were the worst move is more than a certain equity
from the best move)?  This would automatically include forced plays.

This would push up the total chequer error - so perhaps then all the
ratings would need to be rethought?

I'm far from an expert in this area, so this might be a stupid idea :o)

Jon

Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
Bug-gnubg mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/bug-gnubg

Reply via email to