Estimated market value from age, position and six attribute ratings.
A ridge regression fitted to 14,106 outfield players from the FIFA player database, using age, playing position and the six summary attribute ratings It predicts the logarithm of market value rather than the value itself, because the difference between €50k and €100k matters as much proportionally as €50m and €100m.
I refined the data in two 2 ways. Age enters the model as a quadratic, since player value peaks in the mid-twenties, which a straight line cannot capture. Also, each attribute contributes seperately depending on position, since defending ability is worth a great deal to a centre back and very little to a striker. Together these lift the fit from an R² of 0.77 to 0.94.