I'm saying xG is a poor system.....

We know with a large data set the superior penalty taker is better than 75% yet we blindly give him that anyway....

How simple can I make it.

Messi has outperformed xG his whole career yet we blindly give him lower than correct probability of chance in the face of a large database of evidence.

Posted By: usacanary, Sep 1, 14:14:25