26 June 2015


In a piece in the Sports section, of all places, we get this:
In a 2005 academic paper on the N.F.L. draft, the economists Cade Massey and Richard Thaler concluded that lower picks were more economically valuable than higher ones, partly because of the salary difference. "In all my years of studying market efficiency," Thaler, a University of Chicago professor, wrote in his recent memoir, "Misbehaving," "this is the most blatant violation I have ever seen."

Which mostly serves to demonstrate that "analytics", especially in the sports world, is just simplistic descriptives. No GLMs or Random Forests or even simple regression. And these sports metrics guys get paid about the same as The London Whale. At least their musings, mostly, only affect gamblers.

No comments: