no one knows the exact formula for wages, there is some discussion about it in the GI forums though

I assume all salaries will get updated in the off-season?

Steve
SD Blitz

pstimpel wrote: Xiong Yoshi wrote: ...lot of math... I ran a regression using the Squares of the Football skills as independent variables - and the P-Values are all well below 1%, with the exception of Kicking and Punting. Admittedly my sample size was very small (using only my own team), and I was using Excel (so the maximum variables is quite limited), But the results come out pretty good. great work, thank you!

Bah, Don't really look to that as all that good.. I've been toying with it some - I wish I had my Stata on this computer, it'd be a cinch to just come up with a few dozen different models and run them and probably even get the exact formula. I'm dead certain that AGE will be important, but as the export tool doesn't include it - I haven't been using it.

My newest model is actually just using a multiplier on each of the squares separately, to calculate an estimated wage, then calculating the %difference between the estimated and actual, then running covariance checks on all of the variables.... This atleast tells me which attributes are calculated more heavily than others in the final Wage formula... Then tweaking my input multipliers for each one.. running the Covariance again. Right now my Covariances are very low for each of the Football Skills, which is great.. but my Difference between calculated and Actual is still not good.. but when I look at the particular Players who have too high or too low estimated wages - It's pretty clear that Age is important and I'm not using it in any of my calculations.

So... I'll probably put the ages into my data and see how close I get then.

Summary: I'm still working on it, this time with individual multipliers. I believe Age is a big factor that I haven't inputted yet.