"Lies, damn lies, statistics."
While true, simple ypg allowed can be misleading. For example, a stat like pass defense ypg is going to be influenced by your own offense. If your offense can't get first downs, the other team will possess the ball more, run more plays, have more opportunities to pass, and rack up more pass yards, even against a good pass defense.
If I'm looking at raw data, I prefer yards per play. This at least mitigates some other factors and gives clearer focus to the defense's performance. Ohio State is still in trouble there, at 96th in the country with 7.8 ypp allowed against the pass, but not close to last.
Next you have things like schedule, game plan, and junk time. Maybe OU is really, really good at throwing the ball. Or maybe they throw it a lot more than average. Either or both of those things will artificially affect raw ypp. Maybe IU was in the bag in the 4th quarter and the Bucks sent the 2s in, who gave up yards. I don't know anything in raw data that can sift through that.
For those of you who don't know about Football Outsider's advanced metrics, S&P+, FEI, and F/+, check 'em out. They're not perfect (no stats are), but they do a great job and I really like their philosophy. Their formulas are proprietary, I think, so you can't get a look at how they're getting what they're getting, but their thought process is detailed on each stat's page. They attempt to throw out garbage time, factor in SOS, efficiency, things like "havoc rate" for defense....stuff like that. You can look at pass defense/offense, rush defense/offense, team defense/offense, and sort by all the categories included. The F/+ is an amalgamation and is basically the power rankings of teams, or how "good" they are, similar to the poll Mr. Pac Ten used to do here. I forget which is which out of the S&P+ and the FEI, but one is more of a per play basis and one is more of a per drive/possession basis. So you get variations in the two rankings on their offensive and defensive outlooks, but probably not wildly different.
They're interesting, even if you don't agree with them. They usually line up decently well with my subjective opinion after watching teams/games even if raw data doesn't seem to back me up. I think there's something to it.
They won't be out until about week 6 or so, but you can look at previous years to get an idea about it.
