Writers of Pro Football Prospectus 2008

Most Recent FO Features

BradyTom00-10.jpg

» Week 4 DVOA Ratings

Five different teams from last year's DVOA top eight rank in the bottom half of the league through four weeks of 2014. What can we learn from other teams with similar starts in the past?

14 Jan 2008

Final 2007 FEI Ratings

by Brian Fremeau

The Fremeau Efficiency Index (FEI) principles and methodology can be found here. Like DVOA, FEI rewards playing well against good teams, win or lose, and punishes losing to poor teams more harshly than it rewards defeating poor teams. Unlike DVOA, it is drive-based, not play-by-play based, and it is specifically engineered to measure the college game.

FEI is the opponent-adjusted value of Game Efficiency (GE), a measurement of the success rate of a team scoring and preventing opponent scoring throughout the non-garbage-time possessions of a game. Like DVOA, it represents a team's efficiency value over average.

Only games between FBS (Division 1-A) teams are considered. The Week 14 Ratings represent the results of games played through Monday, January 7 (712 games).


Rank Team W-L FEI Last Wk GE GE Rank
1 LSU (12-2) 0.285 1 0.236 5
2 USC (11-2) 0.262 2 0.238 4
3 Oregon (9-4) 0.260 6 0.173 12
4 West Virginia (11-2) 0.257 4 0.283 1
5 Florida (8-4) 0.235 5 0.180 10
6 Georgia (10-2) 0.227 9 0.142 16
7 Oklahoma (11-3) 0.212 8 0.249 3
8 Ohio State (10-2) 0.204 3 0.232 6
9 Missouri (11-2) 0.191 15 0.175 11
10 South Florida (8-4) 0.190 7 0.120 19
11 Clemson (8-4) 0.183 12 0.186 9
12 Virginia Tech (10-3) 0.182 10 0.116 21
Rank Team W-L FEI Last Wk GE GE Rank
13 Auburn (8-4) 0.180 17 0.084 33
14 BYU (10-2) 0.177 16 0.194 8
15 Cincinnati (9-3) 0.177 11 0.165 13
16 Arizona State (10-3) 0.170 13 0.104 28
17 Tennessee (10-4) 0.164 20 0.070 37
18 Oregon State (8-4) 0.157 21 0.010 53
19 Kansas (11-1) 0.151 25 0.265 2
20 Boston College (10-3) 0.151 19 0.107 25
21 Arkansas (7-5) 0.150 18 0.109 23
22 Michigan (9-3) 0.147 22 0.097 30
23 UCLA (6-7) 0.143 23 -0.002 58
24 California (7-6) 0.139 24 0.041 43
25 Kentucky (7-5) 0.132 26 0.035 46

Offensive Efficiency (OE) represents the raw per-possession scoring rate for each team, and Defensive Efficiency (DE) represents that of its opponents. Adjusted Offensive Efficiency (AOE) and Adjusted Defensive Efficiency (ADE) modifies OE and DE for field position and opponent. Unlike FEI, AOE and ADE recorded against stronger opponents in wins or against weaker opponents in losses do not receive additional weight. Field Position Advantage (FPA) represents the difference between a team's average offensive field position and its opponents'.


FEI Rank Team W-L OE OE Rank AOE AOE Rank DE DE Rank ADE ADE Rank FPA
1 LSU (12-2) 0.403 12 0.444 6 0.192 9 0.188 6 3.6
2 USC (11-2) 0.366 22 0.370 25 0.148 2 0.152 2 4.9
3 Oregon (9-4) 0.380 16 0.444 7 0.223 22 0.228 24 3.6
4 West Virginia (11-2) 0.437 5 0.455 5 0.208 14 0.206 14 4.9
5 Florida (8-4) 0.526 1 0.605 1 0.343 88 0.322 76 3.8
6 Georgia (10-2) 0.377 17 0.416 11 0.240 27 0.208 16 2.1
7 Oklahoma (11-3) 0.447 4 0.433 10 0.232 24 0.202 10 2.2
8 Ohio State (10-2) 0.352 29 0.338 37 0.126 1 0.139 1 7.8
9 Missouri (11-2) 0.414 10 0.478 2 0.275 45 0.238 29 -1.5
10 South Florida (8-4) 0.318 47 0.345 36 0.220 21 0.203 12 3.3
11 Clemson (8-4) 0.361 23 0.391 17 0.178 4 0.203 11 5.4
12 Virginia Tech (10-3) 0.267 75 0.300 54 0.159 3 0.164 3 3.3
FEI Rank Team W-L OE OE Rank AOE AOE Rank DE DE Rank ADE ADE Rank FPA
13 Auburn (8-4) 0.278 68 0.320 43 0.219 17 0.186 5 -0.1
14 BYU (10-2) 0.352 30 0.399 14 0.208 12 0.238 30 -1.6
15 Cincinnati (9-3) 0.345 32 0.345 34 0.201 11 0.214 18 6.3
16 Arizona State (10-3) 0.316 48 0.352 31 0.237 26 0.222 21 0.2
17 Tennessee (10-4) 0.359 26 0.395 15 0.280 48 0.255 38 3.6
18 Oregon State (8-4) 0.227 96 0.260 78 0.219 18 0.199 9 1.0
19 Kansas (11-1) 0.423 8 0.374 23 0.181 5 0.184 4 4.0
20 Boston College (10-3) 0.293 61 0.358 28 0.183 7 0.209 17 -0.6
21 Arkansas (7-5) 0.360 24 0.399 13 0.254 31 0.233 26 1.9
22 Michigan (9-3) 0.309 52 0.298 56 0.220 19 0.192 7 3.7
23 UCLA (6-7) 0.188 115 0.228 95 0.208 13 0.192 8 1.0
24 California (7-6) 0.328 41 0.381 21 0.300 59 0.280 50 -1.6
25 Kentucky (7-5) 0.366 21 0.412 12 0.317 72 0.277 45 2.2

Click here for FEI rankings of all 119 teams. Click here for Offensive/Defensive Efficiency rankings for all 119 teams.

National Champions

LSU claimed the BCS championship in dominating fashion last week, playing exactly like an undisputed No. 1 team should play. And yet, this championship will likely forever be remembered as one of the most disputed in history, with a handful of other strong teams playing some of their best football in their own respective finales, each believing they are every bit as good as the team holding the coveted crystal football in Louisiana. They are wrong.

The LSU Tigers defeated two FEI Top-10 teams, the same number defeated by USC, West Virginia, Georgia, Missouri and Kansas combined. LSU took down a total of five FEI Top-20 teams, two better than anyone else. (Oregon and UCLA each earned three such victories.) On ten occasions this season, an FEI Top-20 team suffered a dominating loss; two of these (Ohio State and Virginia Tech) were administered by LSU. The Tigers defeated eight FEI Top-40 teams; only two other schools in college football even played that many. They did lose twice, however, to good but not great Arkansas and Kentucky teams. While those triple-overtime losses shouldn't actually be considered ties, statistically they are about as close as it gets.

In a season in which no team posted an unblemished resume, LSU should get the credit they deserve. But since two-loss national championships are extremely rare (and will be extremely rare as long as there is no college football playoff) the Tigers must also consider themselves fortunate that no other teams were able to best their season-long effort. In the FEI era (2003-present), LSU posted only the ninth best FEI rating overall, and only the third best rating among two-loss teams.


Top 10 FEI Ratings, 2003-2007
Rank Year Team REC FEI Season Rank
1 2004 USC (13-0) 0.351 1
2 2005 Texas (13-0) 0.349 1
3 2006 Florida (12-1) 0.328 1
4 2005 USC (12-1) 0.325 2
5 2006 USC (11-2) 0.315 2
6 2004 Auburn (12-0) 0.306 2
7 2003 LSU (12-1) 0.295 1
8 2005 Ohio State (10-2) 0.287 3
9 2007 LSU (12-2) 0.285 1
10 2005 Penn State (11-1) 0.283 4

The BCS is the perfect system for everyone when a clear-cut No. 1 and No. 2 play for the title. In the last five years, though, only the 2005 Texas Longhorns earned the BCS title and FEI No. 1 distinction by defeating the second best team in the land. Of course, before the BCS ever existed, the opportunity for No. 1 and No. 2 to meet was even more rare. And each of the last five BCS champions, even if they didn't face the nation's second-best team, did face a formidable foe in the championship game, proving they were No. 1 in both that final game and throughout the season according to FEI. But this metric should be only part of the conversation.

If decades of college football have taught us anything, it is that everything is debatable, perhaps especially the merits of its champions. A college football playoff would certainly afford more opportunities to more teams to prove they are the best, but it would not necessarily create more No. 1 versus No. 2 championship games. Like other blemished champions before them, LSU proved both that they were the best team in the country and that the best can be beaten, one of the enduring hallmarks of college football.

The Best of the Rest

USC finishes No. 2 in the FEI rankings for the fourth time in five years, a run that has both dominated the college football landscape and left Trojan fans disappointed that it hasn't produced more hardware. Oregon and West Virginia both finished the season with big bowl victories, and were it not for injuries to their star quarterbacks, may have contended for more, since each played frequently this season like the nation's best team. Florida remains at No. 5; all four of their losses came against FEI Top-25 teams and were decided late in the fourth quarter.

As in several other instances throughout the FEI final ratings, the Gators rank just above one of the teams that defeated them: Georgia. Why? This system is based on Game Efficiency, and playing several losses close against good teams simply rates more efficiently over the course of the season than losing miserably does (as Georgia did against Tennessee), especially when that miserable loss isn't countered by several outstanding victories. Had the Bulldogs been pitted against a more capable opponent than Hawaii in their bowl game, they may have been able to prove something more -- or they may have been run out of the building like Oklahoma.

And then there's South Florida, the only FEI Top-20 team unranked in the AP poll. The Bulls were FEI darlings all season and for that, the system was frequently and heavily criticized. Of the ten four-loss teams that are ranked in the AP final poll, though, only Oregon defeated more FEI Top-20 teams (three) than USF did. There has been plenty of grumbling about 8-4 Florida topping 8-4 Auburn in the final AP poll rankings, despite the fact that the Tigers won head-to-head against the Gators. Nary a peep has been heard about 8-4 South Florida being left out of the rankings entirely. FEI hasn't forgotten their head-to-head wins over Auburn and West Virginia and gives them their due.

The Games

The 2007 season will be remembered for dozens of heart-stopping thrill rides, half of which seemed to be turned in by LSU alone. With six dramatic overtime periods in two losses and four fourth-quarter comebacks, it would be difficult to keep several Tiger games out of a list of the best of the year. Teams in the final FEI Top 25 played 25 games amongst themselves that were decided by a single possession this season, and they also played in 53 single-score games against non-Top 25 teams, winning 37 of them.

Then there were the dominant performances, several of which were turned in against very good teams. For the following table, I calculated a Calibrated Adjusted Game Efficiency for every team's individual game performances. As explained in the Week 13 FEI ratings article, AGE data modifies raw Game Efficiency data to adjust for the strength of the opponent and relevance of the data point to an individual team. FEI ratings are calculated from multiple-order AGE data. The Calibrated AGE data removes the relevance adjustment in order to directly compare performances of different teams.


Most Dominant Performances of 2007
Rank Winning Team Opponent Opp. FEI Competitive Final Score Garbage Final Score Calibrated AGE
1 Tennessee Georgia 6 35-7 35-14 0.873
2 Oregon South Florida 10 49-14 56-21 0.774
3 LSU Virginia Tech 12 34-7 48-7 0.771
4 USC Arizona State 16 44-17 44-24 0.727
5 Oregon Michigan 22 39-7 39-7 0.713
6 Florida Tennessee 17 49-20 59-20 0.685
7 LSU Ohio State 8 38-17 38-24 0.684
8 Georgia Auburn 13 45-20 45-20 0.679
9 Alabama Tennessee 17 41-17 41-17 0.652
10 Cincinnati Oregon State 18 34-3 34-3 0.651

Tennessee turned in the best single game of the season by this measure, but were also walloped themselves twice. LSU and Oregon each dominated two FEI Top-25 teams, leading the SEC and Pac-10 to the nine most dominant performances of the year. The most forgotten victory on this list has to be Cincinnati's September 6 shellacking of Oregon State, a game that didn't make any headlines in the wake of the LSU-VaTech and Oregon-Michigan beatdowns from the same weekend, but one that contributed mightily to the Big East's resume this season.

Of course, this was also the year of the upset, from Appalachian State's shocker in the Big House all the way through Michigan's takedown of Florida on January 1. At least five different teams suffered an earth-shattering loss this year that not only made a run at "biggest upset ever" recognition, but shook up the ratings world every week. The following table ranks the biggest upsets of the year according to the difference in FEI rating of the two teams involved. Only FBS results are included, so the Wolverines (and the other eight victims of FCS upsets) are off the hook.


Biggest Upsets of 2007
Rank Winning Team FEI Opponent Opp FEI Delta FEI Competitive Final Score Garbage Final Score
1 Louisiana Monroe 94 Alabama 27 -0.258 21-14 21-14
2 Syracuse 110 Louisville 46 -0.247 38-35 38-35
3 Stanford 52 USC 2 -0.231 24-23 24-23
4 UNLV 91 Utah 36 -0.217 20-0 27-0
5 Notre Dame 74 UCLA 23 -0.210 20-6 20-6
6 Pittsburgh 49 West Virginia 4 -0.205 13-7 13-9
7 Wyoming 83 Virginia 31 -0.194 20-3 23-3
8 Colorado 54 Oklahoma 7 -0.183 27-24 27-24
9 Nebraska 82 Wake Forest 34 -0.181 20-17 20-17
10 Kansas State 71 Texas 26 -0.181 41-21 41-21

Only the USC, West Virginia and Oklahoma upsets registered significantly with most fans this season, which is not surprising, since each involved a thwarting of the championship aspirations of a once and future Top-10 team at the hands of an also-ran. Some may not consider the others on this list as worthy of the same recognition; that is understandable, though incorrect. Regardless of what the betting public may have thought would happen in the games or whether the results altered the course of the BCS, every upset -- and non-upset, for that matter -- helped shape the course of a season like no other in college football.

FEI Postscript

The Fremeau Efficiency Index will once again be examined and adjusted in the off-season, with more thorough analysis of more data to provide more effective ratings and unit splits. The adjusted offense and defense efficiency calculations will be a particular focus, along with other metrics. Suggestions and questions related to college football stats here at FO are welcomed here in the comments section or by e-mailing brian-at-footballoutsiders-dot-com.

Posted by: Brian Fremeau on 14 Jan 2008

12 comments, Last at 22 Jan 2008, 9:32pm by Matt

Comments

1
by Sid (not verified) :: Mon, 01/14/2008 - 3:56pm

This is more in line with my opinion than the final polls, that's for sure. I felt Ohio State, Hawaii, and Illinois were far too high in the AP poll.

2
by mm (not verified) :: Mon, 01/14/2008 - 5:29pm

I've been reading all season, though I don't think I've left a comment before. Just wanted to say it's been good reading; nice to get an analytic look at the college game.

3
by noahpoah (not verified) :: Mon, 01/14/2008 - 6:43pm

When you say "it is specifically engineered to measure the college game," what does that mean, exactly?

Also, how do you normalize the scores? I ask only because the #1 team's FEI is .285 and #119's is -.21. Are these proportions above/below the mean (non-normalized) FEI (i.e., LSU's FEI is 1.285 times the average FEI)?

Also (also), this is an odd thing to write: "the Tigers must also consider themselves fortunate that no other teams were able to best their season-long effort. In the FEI era (2003-present), LSU posted only the ninth best FEI rating overall, and only the third best rating among two-loss teams."

First, any team that wins a championship likely considers itself fortunate that no other team outperformed them (to mix and match pronouns a bit). Second, this doesn't have anything to do with where LSU's 2007-2008 FEI ranks in the FEI era. Because FEI is expressed relative to performance in a given year, a .285 this year doesn't bear any known relationship to a .285 (or any other score) in any other year.

Anyway, I like your work even though I don't follow college football too closely. It's good to see more sophisticated staistical football analyses flourishing.

4
by Will B. (not verified) :: Mon, 01/14/2008 - 6:55pm

Why is LSU given credit for going up 38-17 in the final couple minutes, but Ohio State's is classified as garbage time? If you classify LSU's final score also as garbage time, do the rankings change much? The game was decided by that point and both defenses stopped putting up much of a fight.

5
by pm (not verified) :: Mon, 01/14/2008 - 6:56pm

Why do people talk up South Florida as some type of great team. They mention their win over Auburn, but Auburn was overrated at the beginning of the season. They barely beat a weak K State team at home. Then the week after beating USF, they lost to Miss St. at home. No one props up Miss. St. for their win over Auburn, but act like USF is a legit team for beating Auburn.

6
by Pat (not verified) :: Mon, 01/14/2008 - 8:33pm

One thing I'd like to see is an explanation of the difference between AOE/ADE and FEI. It's apparently all in the "weighting of games vs strong/weak opponents," but I really have trouble understanding how that can be determined empirically, rather than ad hoc.

Also, why isn't AOE/ADE weighted just like FEI? If there's some value to weighting like that, why isn't it done for AOE/ADE? As it is, right now, there's no way to really understand why, for instance, Kansas is ranked below Oregon State, Arizona State, Cincinnati, Auburn, or even the much-maligned South Florida. Kansas has a better AOE and ADE than each of those teams.

So why is Kansas ranked worse? No clue.

7
by Roscoe (not verified) :: Mon, 01/14/2008 - 9:35pm

4 - LSU's last touchdown wasn't in "garbage time" because OSU was doing everything it could to stop it. OSU's last touchdown was against a lot of second teamers.

8
by Pat (not verified) :: Mon, 01/14/2008 - 11:36pm

4 - LSU’s last touchdown wasn’t in “garbage time” because OSU was doing everything it could to stop it.

Yes, but I highly doubt Brian watches every game to determine whether or not the teams were trying hard.

9
by Nafe (not verified) :: Tue, 01/15/2008 - 8:13pm

Why no App. State over Michigan in the biggest upset in 07??

10
by vijay (not verified) :: Tue, 01/15/2008 - 8:23pm

Re #9: I believe he only puts this together for D-IA teams (or whatever the new name is). And since App. State is a I-AA school, they don't have an FEI and thus can't calculate a FEI Delta.

11
by kibbles (not verified) :: Wed, 01/16/2008 - 5:53am

In response to #9, here's a quote from the article: Only FBS results are included, so the Wolverines (and the other eight victims of FCS upsets) are off the hook.

Michigan = FBS, Appy State = FCS.

A question for Fremeau- in the years that you have data for, has any AOE even approached Florida's .605 this year? I've been marveling at that number all year, and it just kept getting higher and higher. Scary that Florida's going to be returning almost every key part of that offense next year. If they get even a mediocre defense, they're going to be tough to beat.

12
by Matt (not verified) :: Tue, 01/22/2008 - 9:32pm

Exposing one of the most overrated teams there is. Texas Longhorns.