Not to mention math so bad he wouldn't pass a high school stats class. Sampling from a highly cross correlated population expecting to see population-wide gausians; amateur hour before you even got to the phrenology and the misapplication of personality theory.

I feel like 90% of people didn't read the drivel he wrote; I'd have fired him purely for the fact anyone that bad at stats (and proud of it) shouldn't be anywhere near a machine learning team.