<<They do a "percentage based random drug test" every so many months.
  It seems to be based on the last digit of the SSN...If they want 70%
they just pick seven of the 10 digits 0,1,2,3,4,5,6,7,8,9 and everyone
who's SSN ends in that number gets to be tested! Oh Joy of Joys hehehe
:)>>


Which is generally accepted, because
1) Any arbitrary segment of the population is going to have a pretty even
distribution of last-digit values (and the CG is pretty arbitrary),
2) Perhaps more important, the "picker" has no influence over the
assignment of SSANs, so they can't readily be accused of trying to skew
the selection process.

<<I think Steve is right about just using the RAND() function.... tho I
think that it can produce "reproducable and therefore predictable"
results ??>>


Yeah, that's the downside; "computer" and "random" are essentially
incompatible concepts. Wasn't it Von Neumann who said that anyone
expecting to produce random sequences from arithmetic methods (which is
definitely the case for computers) is in "a state of sin"?

To avoid offending of the purists in the crowd: I previously referred to
a truly random sequence of numbers as requiring that any particular value
have the same probability of appearance as any other. This is correct for
most normal-person uses of the term, but "uniform distribution" random
sequences are not the only kind. Other distributions, such as Gaussian
(the "drop balls down the pegged wall" thing) or Boltzmann (energy-level
distributions in physical systems) are definitely not uniform, or, in the
case of the Boltzmann distribution, even symmetrical.

Steve