|
BGonline.org Forums
Variance Reduction
Posted By: Bob Koca In Response To: Variance Reduction (rambiz)
Date: Friday, 18 July 2014, at 12:51 p.m.
It is not really any different than CI in general.
Suppose you have a RO of a game with about 80% win chance without using VR. You get a sequence of numbers like 1 0 1 1 1 1 0 1 1 1 1 1 1 1 1 0 0 1 1 ... 1 One can get the sample mean for your estimate and can get the sample deviation for use in making the CI.
If the RO does use VR the same process is used. The numbers will just be like .75 .82 .74 .91 .78 1.03 .79 .83 ... .81 instead.
Tim Chow might get picky here but a normal distribution approximation is fine in both cases.
|
BGonline.org Forums is maintained by Stick with WebBBS 5.12.