I know it was a subtle point but I had to keep it simple.....

If you publish an xG table and other simple stats correlate better to actual league position then what is the point.

We already know for a fact better teams/players outperform xG so what value do you give to this?

As I mentioned below I wouldn't use GD as a predictive model just as I wouldn't use xG
I think you see more outliers with xG than with GD.

Posted By: usacanary, Jan 15, 16:24:40