Writers of Pro Football Prospectus 2008

29 Oct 2014

FEI Week 9: USC Holds Steady

by Brian Fremeau

The FEI ratings had a strong weekend predicting game results, forecasting the straight-up winners in 83.7 percent of Week 9 games and 54.5 percent of games against the spread. As a result, the FEI ratings didn't shift very dramatically from last week. Teams pretty much held steady, and FEI is still enamored with the Pac-12.

Seven of the FEI top-25 teams hail from the Pac-12. Six of those teams are ranked among the top 12 teams in FEI this week, five in the top 10, four in the top 5. Of all of those teams, only USC (5-3) appears to stick out like a sore thumb at No. 4 overall.

We've seen this story play out before in the FEI ratings and I'm not alarmed by the results. The system solves for mean wins each week and USC's rating is no more unusual than many other teams in the top 25 according to the distribution of results. A team with USC's FEI rating against USC's schedule should have 5.7 FBS wins to date. No. 8 Florida State (6-0) should have only 5.2 FBS wins to date against its schedule. All but one team in the FEI top 25 currently falls within one game of its expected mean wins to date. The exception is No. 17 Mississippi State (7-0), 1.1 mean wins above expectation through seven games.

FEI measures opponent-adjusted game efficiency, and the Trojans have a few key factors working in their favor. They've played the No. 1 strength of schedule to date. They've dominated the weakest opponents they've faced; 52-13 in non-garbage time against No. 91 Fresno State, 56-21 in non-garbage time against No. 61 Colorado, both top-50 unadjusted game efficiency performances. And when they have lost, they have been competitive; three losses by a combined 13 points, two of the three on the road, two of the three on a touchdown given up with less than 10 seconds left in the game. In total, USC has eight results in the books that rank among the top 200 single-game opponent-adjusted performances (GFEI) of the year (77th percentile). No other team has more than six such GFEI results. Only nine other teams have at least four such performances.

We're at the point of the season when the FEI ratings are expected to be relatively stable. I posted the following chart a few weeks ago, and it is a good reminder of how the tangled results of the first half of the season are expected to smooth out beginning around Week 9. Week 9 ratings had a .952 correlation with end-of-year FEI ratings in 2013.

FEI 2013 Team Ratings By Week

The majority of teams are in close proximity of where they'll end up at season's end in the FEI ratings, but there will obviously still be big jumps and big drops in a few instances. Florida State and Auburn ranked No. 4 and No. 14 respectively in the FEI ratings at this time last year. Miami and Ohio State were both undefeated and ranked among the top seven in the BCS standings at this time last year, though FEI had both outside of the top 10 at this point and at season's end. FEI had a four-loss Utah team at No. 9 at this time last year, for many of the same reasons USC ranks so highly this year. Utah eventually crept down to No. 30 overall at the end of the year, the highest-ranked team with a losing record.

USC isn't destined for the same fate unless it has sub-standard GFEI performances down the stretch. FEI expects the Trojans to go 3-1 against Washington State, California, UCLA, and Notre Dame over the next five weeks. But the Trojans may see a drift downward in their rating and expectations if any of their opponent-adjusted results to date begin to slip. If Boston College nosedives against Virginia Tech, Louisville, Florida State, and/or Syracuse, that will drag USC down.

We've also reached the point in the season when the College Football Playoff selection committee has begun to produce weekly top 25 rankings of their own, a new era for college football and, based on the first results posted last night, a set of results that don't necessarily follow the same poll logic we've grown accustomed to over the years. Ole Miss fell to No. 7 in the Associated Press poll after losing to LSU over the weekend, but the committee didn't treat that loss as particularly significant compared to their head-to-head victory over Alabama from several weeks ago. That's refreshing. I'm particularly intrigued by how the committee will handle the results of games between top teams going forward.

Auburn and Ole Miss are ranked No. 3 and No. 4 respectively in the first CFP rankings and they'll face off this Saturday night in Oxford. When two top teams in FEI meet and play a close game, the ratings of both teams are usually bolstered. (Utah and USC are a good example from this past weekend). How will the committee treat the winner and loser of this weekend's top game? If the committee defies poll logic and doesn't drop the loser, will stability reign supreme over the final month of the season in their system as well?

I'm not sure the scrutiny over the process does the committee or the playoff any favors. But as a data analyst, I love that we'll have weekly results to pore over because it will help us build better playoff projection models going forward.

Degree of Difficulty

The degree of difficulty rankings (DOD) are based on current FEI ratings, but instead of measuring efficiency against schedule, DOD measures record against schedule. How difficult would it be for an elite team (two standard deviations better than average) to play a given team's schedule to date and achieve that team's record?

My hypothesis is that the inaugural College Football Playoff selection committee is likely to value and reward something akin to DOD through their process and deliberations. Updated DOD rankings for all teams with two or fewer losses are provided each week, and the current selection committee rankings are provided in the table for comparison.

Degree of Difficulty: Record Against Schedule To Date
DOD
Rank
Team Record DOD FEI FEI
Rank
SOS
Pvs
Rank SOS
Fut
Rank Win
Out
CFP
1 Mississippi State 7-0 .475 .216 17 .475 68 .293 13 .105 1
2 Florida State 6-0 .534 .246 8 .534 81 .371 24 .257 2
3 Oregon 6-1 .609 .294 1 .229 18 .388 28 .410 5
4 Arizona 6-1 .644 .253 3 .192 13 .215 7 .133 12
5 Arizona State 5-1 .689 .226 10 .258 24 .247 10 .082 14
6 Auburn 6-1 .743 .249 6 .292 31 .154 4 .078 3
7 Alabama 7-1 .783 .229 9 .329 40 .389 29 .205 6
8 Michigan State 6-1 .793 .176 23 .265 27 .627 60 .253 8
9 Mississippi 7-1 .797 .246 7 .408 59 .470 35 .325 4
10 Duke 5-1 .800 .184 22 .362 47 .664 64 .313 24
11 Georgia 6-1 .803 .258 2 .409 60 .505 36 .403 11
12 Utah 5-1 .819 .250 5 .336 42 .144 3 .073 17
13 UCLA 6-2 .824 .221 12 .134 4 .334 20 .124 22
14 Colorado State 6-1 .828 .152 27 .427 62 .901 94 .634 -
15 Kansas State 5-1 .833 .196 18 .375 51 .281 11 .081 9
DOD
Rank
Team Record DOD FEI FEI
Rank
SOS
Pvs
Rank SOS
Fut
Rank Win
Out
CFP
16 Notre Dame 6-1 .852 .193 19 .376 52 .233 8 .041 10
17 Clemson 5-2 .852 .218 16 .152 6 .540 45 .338 21
18 Nebraska 6-1 .887 .218 13 .464 65 .640 61 .431 15
19 Baylor 5-1 .892 .186 21 .514 77 .449 33 .153 13
20 Marshall 7-0 .900 .131 31 .900 126 .938 111 .777 -
21 TCU 5-1 .903 .225 11 .459 64 .576 53 .395 7
22 LSU 6-2 .908 .143 28 .221 16 .646 62 .200 19
23 East Carolina 5-1 .919 .111 37 .555 84 .920 101 .467 23
24 Georgia Tech 5-2 .949 .218 15 .353 45 .348 22 .131 -
25 West Virginia 5-2 .958 .128 32 .343 43 .550 47 .083 20
26 Ohio State 6-1 .960 .192 20 .652 96 .576 51 .326 16
27 Boise State 6-2 .965 .111 36 .366 48 .935 109 .561 -
28 Oklahoma 5-2 .970 .218 14 .323 39 .730 73 .546 18
29 Louisville 5-2 .972 .174 24 .358 46 .331 19 .071 25
30 Missouri 5-2 .994 .115 34 .497 73 .824 85 .308 -
DOD
Rank
Team Record DOD FEI FEI
Rank
SOS
Pvs
Rank SOS
Fut
Rank Win
Out
CFP
31 Central Florida 4-2 .996 .027 55 .632 92 .747 76 .120 -
32 Minnesota 5-2 .996 .042 53 .515 78 .355 23 .002 -
33 Georgia Southern 5-2 .998 .039 54 .543 83 .916 98 .508 -
34 South Alabama 5-2 .998 -.139 101 .685 107 .777 81 .001 -
35 Iowa 4-2 .999 .076 44 .738 112 .584 54 .030 -
36 Northern Illinois 5-2 .999 -.130 100 .724 111 .909 95 .009 -
37 Wisconsin 4-2 .999 .115 35 .667 102 .587 56 .073 -
38 Air Force 4-2 .999 .002 60 .788 118 .745 74 .030 -
DOD - Likelihood that an elite team would have the given team's record against the given team's schedule
SOS Pvs - Strength of Schedule of all games played to date
SOS Fut - Strength of Schedule of all games remaining to be played
Win Out - Likelihood given team will win all of its remaining games
CFP - College Football Playoff selection committee ranking

FEI 2014 Week 9 Ratings

The Fremeau Efficiency Index (FEI) is a college football rating system based on opponent-adjusted drive efficiency. Nearly 20,000 possessions are contested annually in FBS vs. FBS games. First-half clock-kills and end-of-game garbage drives and scores are filtered out. Game Efficiency (GE) is a function of the starting field position and outcome of non-garbage possessions. Opponent adjustments are calculated with special emphasis placed on quality performances against good teams, win or lose.

Strength of Schedule (SOS) is calculated as the likelihood that an "elite team" (two standard deviations above average) would win every game on the given team's schedule. SOS listed here includes all regular season games scheduled. A multifaceted approach to measuring schedule strength is available here and visualized here.

Mean Wins (FBS MW) represent the average total games a team with the given FEI rating should expect to win against its complete schedule of FBS opponents. Remaining Mean Wins (FBS RMW) represent the average total games a team with the given FEI rating should expect to win against the remaining opponents on its schedule.

Offensive FEI (OFEI) is opponent-adjusted offensive efficiency. Defensive FEI (DFEI) is opponent-adjusted defensive efficiency. Special Teams Efficiency (STE) is the composite efficiency of field goal kickoff, punt, and return units. Field Position Advantage (FPA) is the average share of field position value generated by offense, defense, and special teams efficiency.

These ratings are exclusively produced from 2014 game data and are not influenced by preseason projections. Complete ratings and ratings splits for all 128 FBS teams are available here. Supplemental data including points per drive, game splits, and game factors are available as well.

Rk Team FBS
Rec
FEI LW GE GE
Rk
SOS Rk
FBS
MW
FBS
RMW
OFEI Rk DFEI Rk STE Rk FPA Rk
1 Oregon 6-1 .294 1 .193 14 .089 23 9.1 3.3 .762 2 -.327 32 1.334 26 .536 22
2 Georgia 6-1 .258 2 .279 2 .207 53 9.2 3.2 .490 15 -.257 40 2.801 9 .608 1
3 Arizona 6-1 .253 4 .123 22 .041 5 9.0 3.5 .488 16 -.353 28 .930 36 .557 9
4 USC 5-3 .253 3 .127 21 .038 4 8.7 3.0 .526 13 -.291 36 .027 69 .487 84
5 Utah 5-1 .250 8 .112 26 .048 8 7.8 3.0 -.132 78 -.736 5 3.156 3 .572 5
6 Auburn 6-1 .249 11 .180 15 .045 7 8.0 2.3 .665 5 -.543 13 .477 50 .523 29
7 Mississippi 7-1 .246 6 .259 4 .192 51 9.0 2.1 .256 35 -.743 3 1.360 25 .567 6
8 Florida State 6-0 .246 7 .163 18 .198 52 9.1 3.9 .622 7 -.429 20 .547 44 .487 85
9 Alabama 7-1 .229 14 .250 5 .128 30 8.4 1.8 .503 14 -.502 15 -1.091 96 .474 100
10 Arizona State 5-1 .226 16 .071 35 .064 11 7.5 3.2 .592 10 -.331 31 1.030 31 .508 44
11 TCU 5-1 .225 5 .249 6 .264 60 9.2 4.2 .383 22 -.560 11 2.586 10 .576 4
12 UCLA 6-2 .221 9 .042 46 .045 6 7.8 2.4 .648 6 -.240 42 .360 59 .498 63
Rk Team FBS
Rec
FEI LW GE GE
Rk
SOS Rk
FBS
MW
FBS
RMW
OFEI Rk DFEI Rk STE Rk FPA Rk
13 Nebraska 6-1 .218 20 .242 7 .297 65 9.2 3.3 .250 37 -.573 9 2.414 12 .556 10
14 Oklahoma 5-2 .218 10 .195 13 .236 56 10.0 4.5 .615 8 -.265 38 2.818 8 .543 14
15 Georgia Tech 5-2 .218 22 .103 30 .123 29 8.0 2.6 .965 1 .186 85 1.566 20 .548 12
16 Clemson 5-2 .218 12 .078 33 .082 19 8.0 3.2 .235 40 -.740 4 -.882 91 .513 38
17 Mississippi State 7-0 .216 18 .229 8 .139 34 8.5 2.6 .377 23 -.551 12 .924 37 .540 16
18 Kansas State 5-1 .196 19 .203 12 .106 24 7.9 3.3 .351 26 -.526 14 1.040 30 .535 23
19 Notre Dame 6-1 .193 21 .168 17 .088 22 8.4 3.0 .266 32 -.494 16 1.088 29 .519 35
20 Ohio State 6-1 .192 13 .263 3 .375 77 10.2 4.1 .450 18 -.283 37 2.902 7 .581 3
21 Baylor 5-1 .186 17 .204 11 .231 55 8.5 3.7 .321 28 -.389 24 .453 52 .551 11
22 Duke 5-1 .184 24 .157 19 .240 57 8.5 4.0 .074 56 -.654 7 2.948 5 .558 8
23 Michigan State 6-1 .176 25 .227 9 .166 45 8.2 2.9 .204 43 -.481 17 -.238 75 .540 18
24 Louisville 5-2 .174 29 .113 25 .119 27 7.4 2.2 -.100 72 -.867 1 .511 46 .468 109
25 Stanford 4-3 .169 28 .106 29 .027 2 5.7 1.4 -.071 68 -.705 6 -.570 84 .525 28

Posted by: Brian Fremeau on 29 Oct 2014

0 comments