Estimating run times

I have replied to this StackOverflow question.

The user wanted to know how to estimate the runtime of its O(N!) function in base to some meassurements performed for small values of N.

There were several errors in the way he was trying to fit a curve to the experimental data as not using the right curve or having too few data samples.

But the interesting thing about this node is that he was trying to predict the biggest N he would be able to run over the weekend and the response is that he doesn't need to!