I have no idea what you're talking about.  I meant to respond to Don's
email, not yours.  Sorry for the confusion.

On Sat, Jun 18, 2011 at 1:23 AM, "Ingo Althöfer" <[email protected]>wrote:

> > Datum: Fri, 17 Jun 2011 18:22:32 -0400
> > Von: Michael Williams <[email protected]>
> >
> > If we are only looking for the shape of the curve, then maybe each
> > datapoint does not need to be as precise as +/- 10 ELO.
>
>
> Sorry, but your comment missed the point.
> My strange result ( MC(4) performs clearly better against MC(1)
> than expected after knowing the quota for MC(4)-vs-(MC(2) and
> MC(2)-vs-MC(1) ) is NOT due to fluctuations of small sample size.
>
> IT IS A PHENOMENON.
>
> Ingo.
>
>
> > > I am not a go programmer, but in experiments with very
> > > simple model games I found the following phenomenon,
> > > when looking at the "basic" Monte Carlo algorithm in
> > > 3 levels: MC(1), MC(2), MC(4).
> > > [Here MC(k) is the version which makes k random games for each
> candidate
> > > move.]
> > >
> > > I found several cases (more than 30 % of the games analysed this way)
> > > with the following phenomenon:
> > > Elo difference between MC(1) and MC(2) = k(1,2)
> > > Elo difference between MC(2) and MC(4) = k(2,4).
> > > Elo difference between MC(1) and MC(4) > k(1,2) + k(2,4)
> > >
> > > The games under investigation were so simple that I was able to run
> > > 100,000 matches in selfplay (or even more), and the inequality above
> > > ( k(1,4) > k(1,2) + k(2,4) ) was statistically significant.
> > >
> > > Ingo
>
>
> --
> NEU: FreePhone - kostenlos mobil telefonieren!
> Jetzt informieren: http://www.gmx.net/de/go/freephone
> _______________________________________________
> Computer-go mailing list
> [email protected]
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
_______________________________________________
Computer-go mailing list
[email protected]
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Reply via email to