Kevin Dayaratna and Steven Miller develop a theoretical underpinning for the Pythagorean Won-Loss Formula in hockey based on the Weibull Distribution, including an assessment of the independence of GF and GA. This builds on earlier work by Miller for baseball and supplements my 2004 work on a Poisson / Skellam underpinning for the Poisson formula. This work supports a Pythagorean exponent of about 2.15 over three recent seasons.
