My read is he's less preficting winning but assessing WAR for teams where his goal is just "who is best relative to the middle".I'm surprised Pomeroy hasn't changed that formula, since he had a long article showing that offensive efficiency had about 2/3 of the control over scoring. Maybe it's baked in somewhere with the calculation of AdjOE and AdjDE (versus raw OE and DE). But if not, a more accurate measure of strength against an average D-I team would weight the offense more heavily. Something like:
(2*AdjOE + D1avg(AdjOE)) / 3
-
( AdjDE + 2*D1avg(AdjDE)) / 3
would seem more appropriate.
Actually, let's just check out what that gives. Math is easy in Excel.
I give you, the "Adjusted Adjusted Efficiency Margin" Rankings. The first 4 columns are KenPom #s. The 3rd from right is the new formula, with rankings next to it - and the last column is the difference between the new ranking and KenPom. I didn't subtract the EM numbers because I'm not sure they really represent the same thing...
View attachment 31619
Re: the methodology difference (essentially it being additive is the difference) he explains here: