On Advanced Metrics, specifically F/+

"Ugly girlfriend means no confidence." -Moneyball

I've read here for quite a while, and on a recent rainy day when my cable was out, decided to finally join in on the fun. I love advanced metrics and all the inferences you can make from them. As one FSU commenter stated in a earlier post, "Your eyes can deceive you, but numbers and data don't lie." That's a strong paraphrase, but one that I believe conveys the point accurately. Several comments made by our friendly FSU fan contingent got me in the mood to look back at historical F/+ of title game participants. For a concise explanation of F/+ (and its component parts FEI & S&P+) go to the glossary at Football Outsiders.

Comparing teams across seasons is a favored hobby for the majority of college football fans which often leads to poor conclusions along the lines of, "Well I just know I'm right." Yuck. I'll concede that I'm not entirely sure the F/+ numbers can be compared across seasons, but if nothing else this was an entertaining exercise.

F/+ with special teams ratings and Offense/Defense breakdown only goes back to 2007 so that's our beginning point. All the archived standings include bowl games , so it's not complete apples to apples, but we'll do the best with what we got. First, a few nuggets.

The best unit (O or D) is the terrifying Alabama defense of 2011. Their defense F/+ rating was 37.1 or 9.3 percentage points higher than tied for #2 2012 Alabama and 2011 LSU. This gap is an outlier, at least at the top. Other offenses and most the defenses are much more bunched together. A natural inference I made--and several of you may make as well--is the 2011 Alabama defense final numbers shot upwards with its dominating performance in the BCS title game. Regardless of how true this is, that defense was really good.

Best offense? Auburn 2010 at 24.9 (2013 Florida State currently sits at 22.6, good for 3rd just slightly behind the 2008 Gators). So maybe they're in the conversation for most prolific offense since 2007, but let's stop short of definite statements here. Auburn's offense sits 8th of 14 at 17.1, just slightly behind the 2011 edition of Alabama's offense.

Top overall rated team? 2011 Alabama at 53.9

2013 Florida State closest comparison is 2008 Florida. FSU- 49.7, UF- 49.4. 2013 Auburn closest comparison is 2007 Ohio State. AU-32.9 OSU-33.4. If you're curious, 2013 Florida State sits 3rd in overall F/+ and 2013 Auburn sits 12th (3rd from the bottom).

Auburn's defense is the lowest rated defense to ever play in the title game, just a tick below our 2010 incarnation (11.3 to 11.0) Using the F/+ metric, Auburn's defense is the 2nd worst unit (offense or defense) to take the field in a title game since 07. The worst? Oregon's 2010 offense at 9.7. A little shocking right? Again, our eyes can deceive us. FSU's defense sits 4th in the overall unit rankings at 26.8.

Silver lining for Auburn fans? Our special teams ranks as the best special teams unit to play in this game since F/+ accounted for special teams. Florida State fans may look at this and say "But, Lou Groza Award," but as many of them have so eloquently told us, numbers don't lie. It'd be great for Steven Clark to continue turning his punts into 120 yard pitching wedges inside the 5. Florida State special teams rank 10th out of the 14 title game participants.

So what does all this mean? Well the numbers aren't in Auburn's favor. I think most of us objectively knew that. FSU's offense is a juggernaut, but they might actually be the lesser juggernaut between their offense and defense.

When I began getting into Bill Connelly's (the originator of one of the components in the F/+ formula and contributor at the excellent Football Study Hall) stuff earlier this year I did it because yards per play didn't seem like enough anymore. There had to be a better metric for determining a team's quality. Lo and behold, F/+.

I love numbers because data doesn't lie. Because data can parse through blithering rhetoric like the quote at the top of this post. But I also know that past performance is not always a guaranteed future predictor. Last year I drafted my fantasy baseball team based on a 3 year average of WAR. WAR is an advanced metric for baseball. I thought I had a leg up on my league as they were using their mish mash of magazines and 'feelings' towards a player. Anyway, my team finished 11th out of 12 and it wasn't because I was a terrible manager. It happened because sometimes past performance won't tell you the future. These F/+ ratings are what led me to believe we'd have extreme difficulty with UGA, and lose to Bama. On one of those counts I was proven correct, the other I was (gladly) wrong. Luckily (pun intended) those games were won on the field.


We're all just trying to have a good time here. Don't be a jerk, and we won't have a problem with you. War Eagle!