Ah I was afraid someone would use the word "approximations"!... Oh well, I'll see what I can do for my particular problem.
Thanks very much for the reply! :D
It's been a long time since I've worried about this - but could someone help me make Pr(x) the subject (I can't remember if it's possible, if it's not, I'd love a brief explanation):
T = S [(1-Pr(x))^N] + Pr(x)
Thanks in advance!