Writers of Pro Football Prospectus 2008

04 Nov 2015

FEI Week 9: 90th Percentile Update

by Brian Fremeau

Before the season began, I discussed the 90th percentile as a way to evaluate the teams that produce elite opponent-adjusted single-game results at a high rate. The postseason championship contenders first need to navigate their regular seasons with as few losses as possible to even merit consideration for the playoff, but which teams showcase their potential to play at an elite level most often?

Bill Connelly uses percentile performances based on S&P+ ratings in his Varsity Numbers columns, and I have my own that are based on my FEI ratings. There are similarities between our two approaches, and similar results. Clemson is the top-ranked team in each of our respective rating systems by a wide margin. Bill credits the Tigers with 90th percentile performances against FBS opponents Appalachian State, Georgia Tech, and Miami. My numbers agree with those games and add in the victory over Notre Dame as a 90th percentile performance as well.

Clemson's dominating win over Miami in Week 8 actually ranks as the No. 2 opponent-adjusted game efficiency result of the season. The best GFEI rating of the season to date is Utah's 62-20 victory over Oregon back in Week 4. In neither case does the opponent rank very highly in FEI -- Oregon has slumped down to No. 47 and Miami sits at No. 53. Both wins were extremely efficient, of course, and that combination of efficiency and opponent strength is the key to the GFEI results and therefore the overall FEI ratings.

Efficiency against a strong opponent, even in a losing effort, is rewarded in the FEI formula. Notre Dame has the third best GFEI rating of the season in its 22-24 loss on the road against Clemson in Week 5. Few Irish fans would likely view the loss to the Tigers as Notre Dame's best performance of the year, but my formula is designed to give more credit to a narrow road loss to an elite opponent than to a dominant win over an overmatched opponent. Notre Dame's 38-3 win over Texas (FEI No. 73) was very good, but it only ranks in the 85th percentile in my GFEI ratings.

Clemson and Notre Dame are two of the seven teams this season that have recorded four 90th percentile GFEI results to date. As one would expect, the teams at the top of this list are also at the top of the FEI ratings. All teams with at least two 90th percentile performances to date are included.

90th Percentile GFEI Single-Game Results To Date
FEI
Rk
Team W-L FEI 90th
Pctl
1 Clemson 7-0 .322 4
2 Notre Dame 7-1 .262 4
3 Alabama 7-1 .255 4
4 LSU 7-0 .254 4
6 USC 5-3 .228 4
10 Michigan 6-2 .209 4
15 Tennessee 3-4 .184 4
5 Stanford 7-1 .251 3
26 California 4-3 .129 3
7 Michigan State 8-0 .215 2
8 Ohio State 8-0 .213 2
FEI
Rk
Team W-L FEI 90th
Pctl
9 Oklahoma 7-1 .213 2
11 Florida 7-1 .207 2
12 TCU 7-0 .199 2
14 Utah 7-1 .189 2
20 Texas A&M 6-2 .151 2
22 Mississippi 6-2 .137 2
36 Northwestern 5-2 .093 2
42 Arkansas 3-4 .068 2
47 Oregon 4-3 .061 2
48 Arizona State 3-4 .056 2
54 Georgia Tech 2-6 .044 2

Tennessee has a losing FBS record but still ranks among several of the playoff challengers in terms of 90th percentile performances to date. Two of their 90th percentile performances came in road losses -- a one-point defeat at Florida and a five-point defeat at Alabama. USC has three losses on the season, but four 90th percentile performances (including three straight since Steve Sarkisian was fired). Undefeated Baylor, Iowa, and Memphis have one 90th percentile GFEI result apiece. Oklahoma State and Houston have none so far.

Every team has an opportunity to post a 90th percentile GFEI result in any given week, regardless of the strength of the opponent, but the GFEI ratings make clear that it is much tougher to do so against weak opposition than against strong opposition. As the final month of the season plays out, some playoff contenders will have their first opportunities to deliver 90th percentile results -- to impress selection committee members and to underscore their capacity to play like one of the elite teams in the nation.

FEI Degree of Difficulty through Week 9

FEI Degree of Difficulty (DOD) ratings are based on current FEI ratings, but instead of measuring efficiency against schedule, DOD measures record against schedule. What is the likelihood that an elite team (two standard deviations better than average) would play a given team's schedule to date and achieve that team's record?

My hypothesis is that the College Football Playoff selection committee is likely to value and reward something akin to DOD through their process and deliberations. As of this week, the following teams have the best records relative to the schedule each has faced to date. Each team's current FEI rating, remaining strength of schedule (RSOS) rating, and current College Football Playoff ranking (CFP) are also provided. DOD ratings for all teams can be found here.

FEI Degree of Difficulty through Week 9
Rk Team W-L DOD FEI Rk RSOS Rk CFP
1 Clemson 7-0 .510 .322 1 .745 74 1
2 LSU 7-0 .541 .254 4 .214 8 2
3 Notre Dame 7-1 .586 .262 2 .347 26 5
4 Michigan State 8-0 .647 .215 7 .433 35 7
5 Memphis 7-0 .732 .169 18 .485 40 13
6 Utah 7-1 .734 .189 14 .596 54 12
7 Florida 7-1 .735 .207 11 .742 73 10
8 Stanford 7-1 .786 .251 5 .276 16 11
9 Iowa 7-0 .770 .172 17 .751 75 9
10 Alabama 7-1 .786 .255 3 .272 13 4

FEI Week 9 Ratings

The Fremeau Efficiency Index (FEI) is a college football rating system based on opponent-adjusted drive efficiency. Approximately 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. Unadjusted game efficiency (GE) is a measure of net success on non-garbage possessions, and opponent adjustments are calculated with special emphasis placed on quality performances against good teams, win or lose. Other definitions:

  • SOS: Strength of Schedule, measured as the likelihood of an elite team going undefeated against the given team's regular season schedule. Schedule strength data based on FEI ratings and calculated across other dimensions can be explored in this interactive visualization.
  • FBS MW: Mean Wins, the average number of games a team with the given FEI rating would be expected to win against its regular season schedule of FBS opponents.
  • FBS RMW: Remaining Mean Wins, the average number of games a team with the given FEI rating would be expected to win against the remainder of its regular season schedule of FBS opponents.
  • OFEI: Opponent-adjusted Offensive Efficiency value generated per possession.
  • DFEI: Opponent-adjusted Defensive Efficiency value generated per opponent possession.
  • STE: Special Teams Efficiency value generated per game possession.
  • FVE: Field Value Efficiency value generated per game possession.

Preseason projection data receives no weight in this week's ratings. Ratings for all FBS teams can be found here.

Rk Team FBS
Rec
FEI LW GE Rk SOS Rk FBS
MW
FBS
RMW
OFEI Rk DFEI Rk STE Rk FVE Rk
1 Clemson 7-0 .322 1 .250 5 .380 68 10.5 3.9 .77 14 1.01 5 .00 69 -.06 89
2 Notre Dame 7-1 .262 3 .166 23 .078 3 9.8 3.2 1.15 5 .42 32 .06 28 .10 26
3 Alabama 7-1 .255 4 .204 11 .137 16 9.2 2.2 .44 26 1.09 1 -.07 105 .12 20
4 LSU 7-0 .254 5 .216 9 .116 12 9.1 2.9 .95 9 .54 23 -.03 88 .00 67
5 Stanford 7-1 .251 2 .226 7 .134 14 10.2 3.4 1.10 7 .17 51 .13 5 .16 11
6 USC 5-3 .228 8 .188 15 .166 26 9.5 3.7 1.09 8 .46 29 -.01 78 .12 17
7 Michigan State 8-0 .215 14 .150 26 .280 52 10.1 3.1 .79 12 .40 31 -.13 123 .01 57
8 Ohio State 8-0 .213 9 .267 3 .267 50 10.4 2.9 .48 25 .73 15 .07 24 .23 1
9 Oklahoma 7-1 .213 6 .261 4 .195 34 9.9 2.7 .70 17 .96 6 -.04 92 .06 40
10 Michigan 6-2 .209 7 .217 8 .226 42 9.9 3.2 .03 58 1.09 2 .16 1 .12 19
11 Florida 7-1 .207 10 .154 25 .305 59 10.0 3.7 .27 37 .91 7 -.01 79 .22 2
12 TCU 7-0 .199 20 .201 13 .230 43 9.0 2.5 1.12 6 -.12 73 .02 50 .08 29
13 Baylor 6-0 .191 13 .376 1 .213 38 9.0 3.3 1.39 2 .25 44 .02 55 .02 55
14 Utah 7-1 .189 12 .147 28 .243 46 9.4 3.3 .27 39 .77 13 .12 7 .09 27
15 Tennessee 3-4 .184 18 .079 42 .266 49 8.2 3.8 .69 18 .38 35 .11 10 .18 8
Rk Team FBS
Rec
FEI LW GE Rk SOS Rk FBS
MW
FBS
RMW
OFEI Rk DFEI Rk STE Rk FVE Rk
16 Mississippi State 5-2 .183 16 .144 30 .140 17 8.1 2.7 .52 24 .71 17 .08 21 .08 30
17 Iowa 7-0 .172 11 .169 21 .579 98 9.3 3.4 .24 41 .62 19 .03 44 .07 36
18 Memphis 7-0 .169 15 .149 27 .355 64 8.7 2.8 .71 16 .01 66 .10 15 .07 37
19 Oklahoma State 7-0 .160 19 .181 18 .190 33 8.3 2.3 .24 40 .38 34 .10 14 .14 13
20 Texas A&M 6-2 .151 17 .098 38 .160 22 7.5 1.9 .00 62 .81 10 .11 13 .07 38
21 Washington 3-4 .150 28 .036 50 .173 29 7.1 2.9 -.32 87 1.01 4 .08 22 .07 39
22 Mississippi 6-2 .137 21 .112 33 .083 6 6.2 1.3 .35 35 .52 25 .00 72 -.02 74
23 North Carolina 5-1 .136 22 .184 17 .526 91 7.8 2.8 .61 19 -.06 70 .04 36 -.01 70
24 Navy 5-1 .134 30 .166 24 .303 58 8.1 3.5 .89 11 .06 64 -.03 84 -.12 110
25 Houston 7-0 .130 27 .317 2 .403 70 9.0 2.6 .59 22 .49 26 .01 62 .21 4
26 California 4-3 .129 25 .010 59 .145 19 6.3 2.5 .40 30 .21 47 -.05 99 -.06 88
27 Florida State 7-1 .125 26 .203 12 .115 11 7.7 1.0 .90 10 .26 46 -.02 82 .00 65
28 Temple 7-1 .109 38 .144 29 .327 61 8.9 2.9 -.22 78 .52 24 .11 9 .17 9
29 Pittsburgh 5-2 .102 23 .048 48 .184 31 6.5 2.1 .10 49 .11 58 .09 16 .17 10
30 BYU 5-2 .101 35 .033 52 .453 80 7.2 3.2 .40 29 .12 61 .03 46 .05 45

Posted by: Brian Fremeau on 04 Nov 2015

0 comments