Your cited article gives the method that MS claims to use in 2007, but what
they actually use remains a mystery, since as I previously noted, McCullough
recently showed (Computational Statistics & Data Analaysis 52(10):4587-4593,
2008) that the sequence produced by RAND is not consistent with these
formulas. A preprint can be downloaded from
http://citeseerx.ist.psu.edu/viewdoc...10.1.1.93.9671

Jerry

"JP" wrote:

There's a short description he

http://support.microsoft.com/kb/828795

--JP

On Sep 21, 8:26 pm, Jerry W. Lewis wrote:
Consequently, no one knows how good or bad the 2007 random number generator
is because no on knows how MS is actually producing its random numbers.

Jerry