Hugo Pfoertner wrote: If you look at applications, I would say: A good random sequence is something that passes G. Marsaglias "DIEHARD" tests: A description can be found at: http://stat.fsu.edu/pub/diehard/cdrom/source/tests.txt Hugo