bug-gnubg
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bug-gnubg] Measuring performance levels 2


From: Douglas Zare
Subject: Re: [Bug-gnubg] Measuring performance levels 2
Date: Sat, 26 Oct 2002 06:05:13 -0400
User-agent: Internet Messaging Program (IMP) 3.1

> How about 1 closed out, extras on the 23, 10, and 7, versus 13 off with a
> blot on the 1? 

In 90,000 trials, Snowie 3 1-ply wins 15.5% (-0.691 +-0.004). 
In  9,000 trials, Snowie 3 2-ply wins 17.3% (-0.655 +-0.004). That's 2-ply huge.

I'm running a 3-ply (huge 33%) rollout now, but if the spares get to good 
places, 3-ply wins more than 21% of the time, if I remember correctly. I doubt 
3-ply will bear in well here, though, and so far it looks like 3-ply will win 
about 19%. I'm not sure if the position ought to be a take; my guess is no, but 
that it is close. Snowie evaluations switch the equities of 13 off and 12 off, 
so the variance reduction isn't great. 

I'll post some Snowie 4 rollouts later.

Douglas Zare





reply via email to

[Prev in Thread] Current Thread [Next in Thread]