"Tom Mort" wrote...
I did mean perumutions. I realize that the same numbers in a different
order would have the same average, but, I'm interested in what
percentage of the total permutations would meet a criteria.
....

And what do you mean by 'percentage of the total permutations would meet a
criteria'? If these criteria were any of the typical descriptive statistics,
then either 0% or 100% of *permutations* would satisfy them. That is, *all*
permutations of the same combination of elements would have the same
descriptive statistics. Or are you interested in order-dependent criteria?

A typical array for this would be <100 datapoints with the number of
items in a set 5 or less

I should asked this direct question befo what are you trying to do?

Note that SUMPRODUCT(COMBIN(90,{5;4;3;2;1})) returns 46,626,033, which is
rather a lot of individual combinations to analyze using brute force
techniques. Unless you intend to provide these users with a run-over-night
or run-over-the-weekend application, you need to return the initial design
stage and rethink what you're trying to do. It's a dead certainty your
application as you've described it can't be interactive.