I played in the MAC. We beat Purdue on the road, we lost to IU close on the road... both early in the non-conference season. If we would have been in the B1G we would have finished in the bottom third. Welcome to the advantage of being the best mid major, you get up for the two P5 games and they don't. They beat a bad Fitzgerald NW barely and smartly scheduled Illinois who, like Purdue would finish 3rd or worse in the MAC. They are definitely outliers.
WMU is not one of the best 25 teams in the country. They would finish 6th or worse in all the P5 conferences. If you wanna geek out with numbers and fantasy boi stats, fine but when you have to line up week in and week out against 4-5 star guys that out everything you, the law of averages takes it's toll. No one is for the mid major more than I but the thing these stats don't take into account is the cumulative effect of playing bigger, faster players every week and they rely on the transitive property to explain the NIU, WMU, Toledo success in limited sample sizes. I didn't need an algorithm to figure that out, I did it with busted teeth, bones & ligaments. What does F/+ say about WMU's 2nd/3rd string OL & DL who will all be playing by the end of the season in the B1G or SEC?