clock menu more-arrow no yes mobile

Filed under:

BCS Championship Game 2014, Auburn vs. Florida State: Tracking the Tigers' improvement via FEI

When looking at some advanced statistics, Auburn's numbers don't quite reflect the team we saw on the field in the last two months of the season. Let's fix that.

Nelson Chenault-USA TODAY Sports

In about 60 hours, this won't matter because we'll have an actual result on the field, but I wanted to get it out there because it keeps coming up (at least in the Internet circles in which I run).

What would Auburn's advanced stats look like if the slow start to the season was taken out?

Nearly everyone knows that the Tigers who held on to beat Washington State, came back to beat Mississippi State and lost by two touchdowns to the University of LSU aren't the same Tigers that romped Tennessee, shocked Alabama and broke records against Missouri. Gus Malzahn even said that Nick Marshall only knew about 25 percent of the playbook for the first game, and the coaching staff didn't start to focus on the zone read game until the first bye.

So how can we compare teams with these advanced stats? One way I've seen it done is by charting each teams' offensive and defensive F/+. In those charts, the farther a team is to the right, the better it is defensively, and the farther up a team is, the better it is offensively. It provides a nice visual way to quickly compare teams.

Unfortunately, game-by-game data isn't available for F/+ (as far as I can tell), so the entire season must be included. But Brian Fremeau of BCF Toys has a metric he calls Game Factors. These help make up his FEI statistics, which, in turn, are a major part of the F/+ numbers.

You can read his own explanation of these numbers, but, simply put, these numbers rate how well a team played against each individual opponent with adjustments for things such as the strength of the opponent. For example, looking at the tables, Auburn's OFEI against Missouri was 2.026, good for the 26th-best offensive performance in the country this year. A weight is also provided to represent how much that game should count toward a team's ranking. So Auburn's win over Missouri is worth about 11.4 percent of the season, but the win over Arkansas State was only worth about 4.5 percent.

Taking a weighted average of the each teams' OFEI and DFEI should provide similar numbers to the offensive and defensive F/+. Fremeau has these numbers available back to 2010, so let's look at every FBS team for the last four years and then focus on the BCS bowl and BCS Championship Game participants.


Let me explain the numbers that pop up when you hover over a dot.


In the first line, each dot has it's teams basic bowl designation: non-BCS, BCS or BCS NC. The first number is the GFEI, which is sort of a combination of the OFEI and DFEI. It is listed here but not charted. The second line is the team and year followed by the DFEI and then the OFEI, both of which actually determine the dot's placement.

TheIndefiniteArticle broke down some of this year's F/+ stats in a nice FanPost earlier this week. There he pointed out "some nuggets" like that the 2011 Alabama team had the best defense since 2007. He also noted that Auburn's 2010 team had the best offense. My chart largely agrees. Alabama2011 is the farthest dot to the right, and Auburn2010 is the highest, except for Texax A&M2012.

I also see that Connecticut2010 is possibly the worst team to reach a BCS bowl, Auburn2011 wasn't very good and Auburn2012 was absolutely horrific.

Now, about that improvement. Since the Game Factors are provided on a game-by-game basis, and each game is weighted, portions of the season can be removed. Yes, football already has a sample size issue and this makes it worse, but I think it's worth looking at. By removing the first four games of Auburn's 2013 season and adjusting the remaining weights as if the remaining games made up an entire season, Auburn looks like is a much better team.

Auburn2013 is at (.206, 1.09) and AuburnPostLSU2013 is at (0.20, 1.42).

Not so much defensively, as the actual DFEI fell by the slightest of margins. But offensively, Auburn went from the 14th-best offense to reach a BCS bowl game (just ahead of Wisconsin2010) to the fourth-best (just behind Florida State2013). The OFEI difference is about 0.328. This post-LSU Auburn team is better than the whole-season Auburn team by about as much as Auburn's 2010 offense was better than Arkansas's 2010 offense.

None of this is to say that Florida State doesn't have a statistical edge, as the Seminoles defense still appears to be much better than the Tigers'. But the teams now appear to be evenly matched with the ball in hand.

As TheIndefiniteArticle said, past performance is not always a guaranteed future predictor. We can throw numbers at each other all day long (and we have), but it will all change when Monday night comes along. I'm looking forward to most of these numbers meaning less and less as the game plays out on the field.

Share on Fancred