Quote Originally Posted by BearDownMU View Post
I get what you are saying but not sure I can agree with it. Holding Auburn to 24 and forcing 5 turnovers certainly seems relevant to evaluating the defense. Also, Jacksonville is D1 FCS.

But, hey, it's your math.
In math terms, the way you play against FBS is "noisy" more than it is a signal, just because of how different an atmosphere it is. Citadel lost to Chatt but beat Carolina, in 2008 Wofford played USC closer than App State and so on. All I'm saying is that if we're evaluating overall performance, it's better to look at conference competition because those are the most consistent variables. Everyone plays everyone. Jacksonville may be Division one, but they're also not full-scholarship. The same goes for Davidson. And Colgate (I think?). But if we're comparing OOC performances as if they are all equal, you'd think Mercer's offense is a lot more effective than it is. Heck, you'd say the same about the Citadel (The citadel is about +15 ppg better on offense when you factor in OOC).

So, when doing the math and trying to figure out a good predictive model, take OOC out.