Fremeau Efficiency Ratings
College football power ratings and analysis

FEI Final Pre-Bowl Ratings and Connectivity

by Brian Fremeau

This has been a weird college football season. It hasn't been weird like 2007 was weird, with wild results and earth-shattering upsets seemingly every week that turned over the top ten throughout the season. It has been weird for the opposite reason, in fact.

We have four undefeated teams heading into bowl season, including three power programs. No power programs made it to bowl season without a loss last year, and only three total had accomplished that feat in the three years prior. There were five undefeated teams heading into bowl season way back in 2009, but three of those five hailed from non-power conferences. Going undefeated through the regular season is rare. This year, with Alabama, Clemson, Notre Dame, and Central Florida all accomplishing the feat, is especially rare.

The 2018 season has been largely devoid of drama. Alabama, Clemson, Notre Dame, Georgia, Ohio State, Oklahoma, and Washington all finished the pre-bowl season ranked among the top eight teams in the FEI ratings. They're also seven of the top nine teams in the College Football Playoff committee rankings. All seven of those teams were projected as top-eight FEI teams at the beginning of the year. I can't even boast about that prediction -- I didn't really go out on a limb with any of the picks. Six of those teams appeared among the top seven teams in the preseason AP Top 25 poll as well. If it felt like there weren't many surprises this year, you're absolutely right, there weren't.

Back in Week 10, I highlighted another oddity of this season, which is likely related to the lack of drama at the top of the rankings. There seems to be a drop-off in quality after the top tier. Good teams -- the teams that fill out the second half of top-25 polls -- haven't been performing like good teams usually do. Instead, they performed more like the next tier down in quality. That feeds right into the first point – part of the reason why the top teams had staying power this year is that they weren't challenged by the next tier down.

On the subject of the lack of challengers, one of my Twitter followers asked me to look into overall weakness of college football scheduling this year. That kind of inquiry can be approached in a variety of ways. I wanted to see if there was a drop-off in the number of games played between top opponents, suspecting that the lack of drama was also aided by the lack of competition between top teams. My suspicions were confirmed.

I compared the 2018 season with each of the last four years, the College Football Playoff era. Through this past weekend, a total of 733 games between FBS opponents have been played. Only five of those games pitted one top-10 team (according to end-of-year FEI ratings) against another. There were eight such games played prior to bowl season last year, and an average of 7.5 such top-10 matchups from 2014 to 2017. There have been fewer games between top-20 teams played this season than any other playoff year as well. Same goes for games played between top-30 opponents. What is the story here?

The drop-off in games played between top opponents is chiefly tied to non-conference scheduling. The Notre Dame versus Michigan game played all the way back in Week 1 remains the only non-conference game played this year between FEI top-20 opponents. There were seven such games played prior to bowl season last year. There have been only seven non-conference games played between FEI top-30 opponents to date, half as many as last season.

Non-conference scheduling can be fickle. Many teams do make an effort to schedule at least one strong non-conference opponent, and some games that appeared to be matchups between heavyweights prior to the start of the season (Auburn-Washington, Ohio State-TCU, and Notre Dame versus USC, Stanford, Virginia Tech, and Florida State) all turned out to be underwhelming matchups when looked at in retrospect. Scheduling intentions are one thing, but 2018 made many of the non-conference games fizzle when it came time to play those games.

That's disappointing, but it also underscores the tired refrain that "nobody has played anyone" this year. Conference matchups between eight (yes, eight!) SEC teams ranked in the final CFP committee rankings could all look like good games between good teams on paper with ranking numbers assigned to those teams, but appearances can be deceiving. Connectivity between top opponents, and really between all opponents, is more desirable, and not just for the numbers folks like myself who need those inputs to produce reliable outputs. Those games also make college football more interesting. This year wasn't very interesting, and if non-conference scheduling is partially to blame, here's hoping 2018 was an aberration and not the beginning of a trend.

FEI Final Pre-Bowl Ratings

The Fremeau Efficiency Index (FEI) is a college football rating system based on opponent-adjusted possession efficiency. Adjusted Possession Advantage (APA) ratings represent the per-possession scoring advantage a team would be expected to have on a neutral field against an average opponent, calculated as a function of current FEI overall, offense, defense, and special teams ratings.

Strength of Schedule ratings (PSOS) represent the average number of losses an elite team (two standard deviations better than average) would have against the team's schedule to date. Offensive FEI (OFEI) is scoring value generated per drive adjusted for starting field position and opponent defenses faced. Defensive FEI (DFEI) is scoring value generated per opponent drive adjusted for starting field position and opponent offenses faced. Special Teams FEI (SFEI) is scoring value generated per possession by a team's non-offensive and non-defensive units adjusted for opponent special teams units faced. The team's record to date against opponents ranked in the FEI top 10 (v10), top 20 (v20), top 30 (v30), top 40 (v40), and top 50 (v50) are also provided.

Ratings and supporting data are calculated from the results of non-garbage possessions in FBS vs. FBS games.

Click here for ratings for all 130 FBS teams.

Rk Team Rec FEI APA Rk PSOS Rk OFEI Rk DFEI Rk SFEI Rk v10 v20 v30 v40 v50
1 Alabama 12-0 .333 3.15 1 1.46 20 3.96 2 .80 1 .230 4 2-0 4-0 6-0 6-0 6-0
2 Georgia 10-2 .272 2.52 2 2.00 4 3.88 3 1.40 9 .219 5 0-2 2-2 4-2 5-2 6-2
3 Clemson 12-0 .241 2.16 3 .57 101 3.05 9 1.01 3 -.013 67 0-0 0-0 1-0 4-0 6-0
4 Oklahoma 12-1 .222 1.88 5 .97 57 4.26 1 2.60 98 .119 16 0-0 1-0 3-1 4-1 5-1
5 Notre Dame 12-0 .213 1.76 6 .70 85 2.85 19 1.34 8 .033 45 1-0 1-0 2-0 4-0 6-0
6 Central Florida 11-0 .207 1.92 4 .32 120 3.70 4 1.81 29 -.019 71 0-0 0-0 1-0 2-0 3-0
7 Washington 9-3 .203 1.55 10 1.07 45 2.59 36 1.41 11 .006 57 0-0 2-0 4-1 5-1 5-2
8 Ohio State 12-1 .201 1.64 9 1.12 40 3.49 5 2.09 47 .175 8 1-0 3-0 3-0 4-1 4-1
9 Michigan 10-2 .195 1.68 8 1.44 22 2.93 14 1.40 10 .181 7 0-2 2-2 2-2 3-2 4-2
10 LSU 8-3 .179 1.37 13 1.93 5 2.50 41 1.44 12 .294 2 1-1 2-2 3-3 4-3 4-3
11 Mississippi State 7-4 .176 1.75 7 1.70 11 2.79 24 .95 2 .082 24 0-2 0-3 2-4 2-4 2-4
12 Missouri 7-4 .169 1.50 11 1.90 6 2.98 12 1.57 17 .106 19 0-2 1-2 1-3 2-4 3-4
13 Fresno State 10-2 .164 1.41 12 .72 81 2.72 30 1.45 13 .000 60 0-0 1-1 1-1 1-1 2-1
14 Penn State 9-3 .142 1.18 15 1.12 41 2.48 43 1.47 14 .171 10 0-2 1-3 1-3 2-3 4-3
15 Florida 7-3 .140 1.10 18 1.49 17 2.68 31 1.80 28 .215 6 1-1 2-2 2-3 3-3 4-3
Rk Team Rec FEI APA Rk PSOS Rk OFEI Rk DFEI Rk SFEI Rk v10 v20 v30 v40 v50
16 Iowa 7-4 .138 1.07 19 .77 74 2.41 49 1.57 16 .106 18 0-0 0-1 0-1 1-3 1-4
17 Boise State 10-3 .137 1.13 16 .76 75 2.79 25 1.81 30 .010 52 0-0 1-1 2-1 2-1 2-3
18 Michigan State 7-5 .135 .97 24 1.24 31 1.73 106 1.07 4 .010 53 0-2 1-2 2-2 3-4 3-4
19 Utah 8-4 .134 .97 23 1.27 28 2.36 54 1.67 22 .174 9 0-2 0-2 1-3 1-4 2-4
20 West Virginia 7-3 .134 1.07 20 .86 67 3.06 8 2.18 62 .053 35 0-1 0-1 1-1 1-2 1-3
21 Kentucky 8-3 .130 .96 25 1.44 21 2.27 63 1.58 18 .131 13 0-1 3-1 3-2 4-2 5-2
22 Cincinnati 9-2 .130 1.22 14 .79 72 2.45 47 1.22 5 -.072 88 0-1 0-1 1-1 1-2 1-2
23 Washington State 9-2 .128 1.07 21 .71 83 3.36 7 2.43 84 .050 37 0-1 1-1 2-1 2-1 3-1
24 Texas 9-4 .116 .87 28 1.19 35 2.73 29 2.08 43 .009 54 1-1 1-2 1-2 2-2 2-3
25 Utah State 9-2 .109 1.11 17 .59 97 3.02 11 1.82 32 -.037 79 0-0 0-2 0-2 0-2 0-2
26 Ohio 7-4 .105 1.02 22 .56 104 3.40 6 2.35 76 .046 41 0-0 0-0 0-1 0-1 1-2
27 Texas A&M 7-4 .102 .81 30 2.04 3 2.80 22 2.15 58 .323 1 1-2 1-3 2-4 3-4 3-4
28 Army 8-2 .100 .90 26 .73 80 2.93 15 2.08 44 -.160 114 0-1 0-1 0-1 0-1 1-2
29 Auburn 6-5 .095 .82 29 2.39 1 2.37 53 1.62 20 .236 3 1-3 1-4 2-4 2-4 2-4
30 Stanford 7-4 .095 .72 33 1.41 23 2.68 32 2.13 55 .080 25 0-2 0-3 0-4 1-4 3-4

Comments

4 comments, Last at 17 Dec 2018, 11:04am

1 Re: FEI Final Pre-Bowl Ratings and Connectivity

Could this be an issue of the top 6/7 teams beating their in-conference rivals ("2nd"-tier teams) in such a way that those teams' ratings are slightly more depressed than normal? Of the top 6 teams considered for the CFP (the four in the playoff, plus Ohio St. & UCF), they have a total of 2 losses--both to teams below their "rating tier" overall. Georgia has two losses--one to LSU, and one to Alabama--only one to a team below their tier; both of Michigan's losses are to teams in the top 6.
I think what you have this year is that the large majority of teams beat the teams they were supposed to beat, and lost to the teams they were supposed to lose to. I mean, of teams in your top 30 here--how many have bad losses to teams not in the top 30? (I'm not talking about losing in the last seconds, or in a tight game--I'm talking about a Virginia Tech losing to Old Dominion-type loss.)

2 Re: FEI Final Pre-Bowl Ratings and Connectivity

This is a good point, and one I focused on in the Week 10 article that was referenced in this piece. Teams "beat the teams they were supposed to beat and lost to the teams they were supposed to lose to" is unusual as well. Check out my Massey Team Types data for how teams are "supposed" to perform: http://www.bcftoys.com/massey

3 Re: FEI Final Pre-Bowl Ratings and Connectivity

The ultimate weakness of football rating systems is their dependence on preseason subjective opinions of strengths and weaknesses. It's worse for college rankings because teams play so few games against a tiny fraction of potential opponents. College football rankings are for bar arguments, nothing more.

4 Re: FEI Final Pre-Bowl Ratings and Connectivity

None of the data in this article, or in any FEI data posted here since Week 7, includes any preseason data or "subjective opinions of strengths and weaknesses". I'm not aware of any football rating systems (aside from polls) that depend on preseason data at year's end, and I'd dispute the influence of preseason data on end-of-year polls as well.