Re: [Bug-gnubg] Proposal - Luck rating

On Mon, Jul 6, 2009 at 10:10 PM, David Levy<dml.public@xxxxxxxxxxxxx> wrote:
> Hello,
>
> I have a question or two about how luck is reported, followed by a comment
> on Massimiliano's suggestion.
>
> Both players begin play at 50% to win a match. One of them wins, going to
> 100%, the other loses going to 0%. It seems to me that as a matter of
> definition, the ânet luckâ plus the ânet skillâ (difference in MWC of the
> two playersâ errors) should be exactly 50%. Correct?
>

With a perfect luck-evaluator that would be true, but suppose we
constructed a neuralnet like this:

the evaluation is a function of the position, but a random number.

If you let this neuralnet play against itself skill would be perfect
but luck could be evaluated as just about anything. Luckily for us the
gnubg net is close enough to perfect :)

Christian.


_______________________________________________
Bug-gnubg mailing list
Bug-gnubg@xxxxxxx
http://lists.gnu.org/mailman/listinfo/bug-gnubg