The reason they didn't find anyone breaking the model was that they had a very high standard: players had to really outscore the model
over multiple seasons in order to produce a
sample size large enough for them to pay attention, and individual players have rarely done that.