Re: [Bug-gnubg] A new variance reduction technique

[Christian Anthon]

2009/3/31 Massimiliano Maini <address@hidden>:
> I also think that Variance Reduction (Dahl's) should not be applied on
> perfectly stratified rolls since it can just introduce errors: average
> luck over the 36 (1296) rolls is zero.
>

If the Varred is correctly implemented it adds nothing on perfect
stratification, otherwise it should be fixed. And we can't know when a
user stops a rollout so we cannot leave out stratification on, e.g.,
the first and second rolls in a 1296 trial rollout.

Christian.