I did say "extrapolate" didn't I. Good eye Michael. In either case, it should be clear to him by now that he needs to use a look up table and not try to fit a curve. Did you notice that HE was the one who brought up breaking the curve into segments and doing separate fits?

And yes Kai, you are correct. He needs to calculate weight from counts. And first he needs to get the corrected number of counts based on the measured number of counts. So the curve he needs to fit is the number of error counts as a function of measured counts. The first time I played with the data I was thinking of this consciously, and I still started using the weight column as a domian out of nothing but force of habit.