nddulac
Doh! mer
It may have coincided with his inclusion of his ELO-CHESS method that theBCS wanted when they mandated that the ratings systems remove margin of victory from their formulae.I think I read he did use a least squares method for the older data but for current years (after he really started ranking current teams) he went to a Bayesian method. Not sure when he went to Bayesian models but I do think you're correct about his least squares for really older data.