hypatia13 Apr 2004

I think the idea of Least Squares would work here, minimizing the sum of squared differences from something and thereby deriving a best-fit curve. This seems to ramble some, a bit repititious, but still it's a good take on the tug-of-war.