[computer-go] Re: Explanation to MoGo paper wanted.

[steve uurtamo]

> If an improvement is worth 100 Elo, there is no need for extensive testing. 
> One sees this immediatly. In fact also smaller improvements are in the end 
> chosen by intuition/feeling.

if the win rate is close to 50% (and you can tweak things so that this is
the case), you can get away with a smaller number of experiments, using
a good significance test to make sure that it's actually an improvement.
it only gets really ugly if you are either close to 0% win rate, or close to
100% win rate, in which case it's nearly impossible to measure progress.

s.





       
____________________________________________________________________________________
Pinpoint customers who are looking for what you sell. 
http://searchmarketing.yahoo.com/