Basketball average is over?

So in May, the team [Milwaukee Bucks] hired Dan Hill, a facial coding expert who reads the faces of college prospects and N.B.A. players to determine if they have the right emotional attributes to help the Bucks.

The approach may sound to some like palm reading, but the Bucks were so impressed with Hill’s work before the 2014 draft that they have retained him to analyze their players and team chemistry throughout the current season.

There is more here.  How well does this model retrodict various successes and failures, relative to underlying levels of talent?  How about if we apply the model to economists?  Potential graduate students?  Mates?  Where else?  File under speculative.