Re: [Bug-gnubg] Re: The importance of METs

[Nis Jorgensen]

On Tue, 2 Sep 2003, Joseph Heled wrote:

> Here are the numbers.
> E1 (woolsey wins both) - 31747
> E2 (mec26 wins both)   - 32067
> E3                     - 186186
>
> The verdict is?

I forgot to include it ... even with the variance erduction, the 95%
confidence interval includes 50%. So we need more trials.

One thing: If you still have the matches lying around, it might make sense
to do luck reduction combined with the duplicate dice option (that
is, output LuckAdjustedResult1 + LuckAdjustedResult2. Not sure whether
this is faster than extending with more 0-ply matches though.

If you do this, please make sure you use the same MET for the luck of both games
- to include the exact duplicates (I am not sure how many they are).

Please note that all the normal arguments for Luck Reduction being
unbiased applies, even if we are using one of the involved METs to do it.

Nis