Writers of Pro Football Prospectus 2008

Most Recent FO Features

ReplayIns14.jpg

» Scramble for the Ball: Getting it Right?

Instant replay review is one of the cornerstones of the modern NFL. The process and its myriad special rules have been internalized and constantly debated. Mike Kurtz wonders: is it worth it?

07 Sep 2011

FEI: The Most Valuable Play

by Brian Fremeau

I had a bit of time to kill this past weekend as I waited through two separate severe weather delays in the concourse of Notre Dame Stadium. The Fighting Irish were shooting themselves in the foot repeatedly throughout the first three quarters against South Florida. I was with family members at the time, attempting to explain the specific value lost on each turnover according to my new offense, defense and special teams unit value splits. The exercise was equal parts therapeutic and aggravating.

The discussion also brought back memories of a Notre Dame game from 2002, a stinging 14-7 upset loss to Boston College that ruined what had been a perfect season to date and ultimately exposed the Irish that year as a team built on unsustainable non-offensive success. Oddly enough, that game actually has a special place in my heart. The turnovers and drive failures that miserable afternoon were the inspiration for my work today with college football statistics. When I left the stadium that day, I wanted to know the value of each possession and momentum-changing event, and I began to collect data to do so soon thereafter. Nine years later, there I was, evaluating another Irish defeat according to what I had learned since then.

Exhilarating and maddening pretty much sums up the last decade-and-a-half of Notre Dame football for me. And if it weren't both, I might not have been moved to dig into numbers the way I do, and I might not be writing this today. So thank you, Boston College. And damn you.

The South Florida game this past weekend was similar in many ways to the Boston College game in 2002. Notre Dame coughed up five turnovers against each team (though the Irish also chipped in two more turnovers on downs against BC). Multiple drives into scoring range were repeatedly wasted in both games. The Irish offense outperformed its opponent in both games, but couldn't capitalize. And in both games, a first half turnover deep in opponent territory was returned the length of the field for a touchdown.

The value of that kind of turnover was the heart of our discussion during the storm delay. After a strong drive by Notre Dame to open the game against South Florida, running back Jonas Gray had the ball stripped at the 1-yard line, and it was scooped up and run back for a touchdown by the Bulls' Kayvon Webster. My data ultimately credits South Florida's defense with creating 6.3 points of scoring margin value on the play, but that's only in terms of net value of the two-drive sequence. In other words, South Florida had lost significant value by allowing Notre Dame to drive to the 1-yard-line, but erased that deficit by creating the turnover and score.

The Gray fumble occured on third-and-goal, but let's imagine the same play occuring on a drive that starts with first-and-goal from the 1-yard-line. In this case, the expected scoring value of the offense's drive is more than six points. The same play in this scenario is then worth the value of killing a six-point drive plus the value of the touchdown return. We might like to call it a 14-point play, but according to my unit value splits, it would officially be credited with a defensive value of 11.1 points. Remember that there is unearned value on every possession, so the defense doesn't get the full credit for the 14-point swing, but a defensive score following a goal-to-go turnover is the most valuable single play in football.

South Florida's special teams created another valuable single-play possession-change sequence by recovering a muffed punt in the second quarter. The turnover by Notre Dame's Theo Riddick was worth 1.7 points in lost possession value, and the resulting field position for South Florida at the Irish 20-yard-line was worth an additional 3.4 points generated by the special teams play. The total value of the sequence (5.1 points) wasn't quite as strong as the total value of the drive-turnover-return sequence that opened the game (6.3 points), but it was awfully close.

In the end, special teams account for the scoring margin of the game. South Florida earned a 12-point advantage through punt exchanges, turnovers, and place kicking success. Notre Dame's second half offense actually erased the entire deficit generated by its red zone miscues by moving the ball and creating enough other scoring opportunities to win. And the defense held South Florida in check throughout the day, surrendering only one touchdown drive. The complete possession-by-possession unit values for the game are provided below.

-

How FEI Watches a Game: South Florida vs. Notre Dame
Drive Possession Start End P Y Result USF Off USF Def USF ST USF Ex Score Poss Mar
1 Notre Dame own 20 opp 1 8 79 Fumble 0.0 1.3 0.4 -1.6 0-0 -
2 South Florida - - - TD (+XP) 0.0 5.0 0.1 1.9 7-0 +7
3 Notre Dame own 25 50 5 25 Punt 0.2 1.5 0.2 -1.9 7-0 -
4 South Florida own 27 opp 32 8 41 FG -0.3 -0.7 2.1 1.9 10-0 +3
5 Notre Dame own 33 own 37 3 4 Punt 0.2 1.8 -0.2 -1.9 10-0 -
6 South Florida opp 40 opp 1 5 39 FG -0.4 -0.4 1.9 1.9 13-0 +3
7 Notre Dame own 32 opp 7 11 61 Int 0.3 1.8 -0.2 -1.9 13-0 -
8 South Florida own 20 opp 35 12 45 Failed FG -0.4 -0.6 -0.9 1.9 13-0 -
9 Notre Dame own 35 own 40 3 5 Punt 0.0 1.9 0.0 -1.9 13-0 -
10 South Florida own 23 own 32 3 9 Punt -1.4 -0.5 0.0 1.9 13-0 -
11 Notre Dame - - - - Fumble 0.2 0.0 1.7 -1.9 13-0 -
12 South Florida opp 20 opp 19 4 1 FG -2.3 0.0 3.4 1.9 16-0 +3
13 Notre Dame own 26 own 25 3 -1 Punt 0.0 1.5 0.4 -1.9 16-0 -
14 South Florida own 42 50 3 8 Punt -2.2 0.0 0.3 1.9 16-0 -
15 Notre Dame own 14 own 15 3 1 Punt 0.7 1.1 0.1 -1.9 16-0 -
16 South Florida own 44 opp 42 4 14 Half -2.3 0.5 -0.1 1.9 16-0 -
Drive Possession Start End P Y Result USF Off USF Def USF ST USF Ex Score Poss Mar
17 South Florida own 19 own 21 3 2 Punt -1.2 0.0 -0.4 1.6 16-0 -
18 Notre Dame opp 49 opp 5 7 44 Int -0.1 2.7 -0.7 -1.9 16-0 -
19 South Florida own 4 own 21 5 17 Punt -0.8 -1.2 0.0 1.9 16-0 -
20 Notre Dame own 34 - 5 66 TD (+XP) -0.2 -5.1 0.2 -1.9 16-7 -7
21 South Florida own 19 own 30 6 11 Punt -1.2 -0.2 -0.4 1.9 16-7 -
22 Notre Dame own 33 opp 13 7 54 Failed FG 0.2 -0.6 2.3 -1.9 16-7 -
23 South Florida own 20 - 14 80 TD (+XP) 5.7 -0.6 0.0 1.9 23-7 +7
24 Notre Dame own 24 - 12 76 TD (failed 2Pt) 0.2 -5.5 1.2 -1.9 23-13 -6
25 South Florida own 32 opp 40 5 28 Punt -1.8 -0.3 0.1 1.9 23-13 -
26 Notre Dame own 9 own 13 2 4 Int 0.8 0.9 0.2 -1.9 23-13 -
27 South Florida opp 30 opp 32 3 -2 Punt -3.9 2.0 0.0 1.9 23-13 -
28 Notre Dame own 1 - 10 99 TD (+XP) 0.9 -6.3 0.3 -1.9 23-20 -7
USF Total -8.8 Off -0.3 Def 12.1 ST 0.0 Ex

Notre Dame wasn't the only team to surrender a double-digit scoring margin special teams deficit this weekend. Naturally, each of the others lost as well (Troy -11.3 ST points vs. Clemson, Oregon -10.5 ST points vs. LSU, Florida Atlantic -10.5 ST points vs. Florida, and San Jose State -10.5 ST points vs. Stanford). In each of the other cases, however, the total scoring margin for the game was greater than the deficit created by special teams.

Four other teams in addition to Notre Dame would have won if they had played an even game on special teams. If the game were entirely decided in terms of offensive versus defensive play, the outcome for these teams would have reversed.

Team Opponent Result Off Def ST Ex
Boston College Northwestern L 17-24 2.4 -2.2 -7.1 0.0
Middle Tennessee Purdue L 24-27 4.4 -3.2 -4.2 0.0
Notre Dame South Florida L 20-23 0.3 8.8 -12.1 0.0
UCLA Houston L 34-38 25.0 -23.7 -5.3 0.0
Utah State Auburn L 38-42 23.2 -18.3 -8.9 0.0

The complete unit value breakdown for non-garbage possessions in every FBS game this past weekend can be found here.

FEI Week 1 Top 25

The Fremeau Efficiency Index (FEI) rewards playing well against good teams, win or lose, and punishes losing to poor teams more harshly than it rewards defeating poor teams. FEI 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. FEI represents a team's efficiency value over average. 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 future games scheduled.

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 expected team wins for games scheduled but not yet played.

Only games between FBS teams are considered in the FEI calculations. Since limited data is available in the early part of the season, preseason projections are factored into the current ratings. The weight given to projected data will be reduced each week until Week 7, when it will be eliminated entirely. Offensive and defensive FEI ratings will also debut in Week 7.

These FEI ratings are a function of results of games played through September 5. The ratings for all FBS teams can be found here.

Rank Team FBS
W-L
FEI Last
Wk
GE GE
Rk
SOS SOS
Rk
FBS
MW
FBS
RMW
1 LSU 1-0 .296 2 .212 18 .082 4 9.1 8.4
2 Alabama 1-0 .283 1 .295 15 .125 12 9.2 8.2
3 Oklahoma 1-0 .241 3 .503 7 .317 62 10.6 9.6
4 Florida State 1-0 .239 7 .429 9 .166 23 8.8 7.9
5 Boise State 1-0 .235 11 .154 23 .613 102 10.8 10.3
6 Ohio State 1-0 .231 9 .556 3 .250 42 10.1 9.1
7 Virginia Tech 0-0 .221 4 n/a n/a .512 92 9.8 9.8
8 West Virginia 1-0 .217 15 .333 12 .133 15 8.2 7.3
9 Auburn 1-0 .202 5 .052 33 .043 1 7.2 6.2
10 Clemson 1-0 .198 16 .157 22 .098 8 7.4 6.4
11 Wisconsin 1-0 .194 24 .901 1 .287 50 8.7 7.7
12 Stanford 1-0 .178 8 .524 5 .402 77 10.1 9.1
Rank Team FBS
W-L
FEI Last
Wk
GE GE
Rk
SOS SOS
Rk
FBS
MW
FBS
RMW
13 Arkansas 0-0 .176 12 n/a n/a .091 7 7.4 7.4
14 Florida 1-0 .176 10 .533 4 .055 2 7.0 6.0
15 Oregon 0-1 .171 6 -.212 63 .206 33 8.4 8.2
16 Pittsburgh 1-0 .170 13 .143 24 .204 31 8.9 7.9
17 North Carolina State 0-0 .153 18 n/a n/a .218 35 7.3 7.3
18 Notre Dame 0-1 .148 22 -.031 45 .160 20 7.8 7.2
19 USC 1-0 .144 14 .03 37 .282 48 8.8 8.0
20 South Florida 1-0 .137 37 .031 35 .167 24 6.9 6.6
21 Georgia 0-1 .135 17 -.154 58 .325 65 7.4 6.9
22 South Carolina 1-0 .133 21 .166 21 .204 32 6.7 5.9
23 Michigan 1-0 .131 35 .457 8 .252 43 8.2 7.3
24 Missouri 1-0 .124 19 .137 26 .301 54 8.0 7.1
25 Arizona State 0-0 .116 26 n/a n/a .392 75 7.6 7.6

Posted by: Brian Fremeau on 07 Sep 2011

4 comments, Last at 07 Sep 2011, 10:28pm by Brian Fremeau

Comments

1
by brandond03 :: Wed, 09/07/2011 - 2:31pm

Does the strength of schedule ranking change during the course of the season based on how opponents perform, both before and after playing? Same question for Mean Wins.

Go Gators and Go Navy

2
by Brian Fremeau :: Wed, 09/07/2011 - 2:35pm

Yes, strength of schedule will be recalculated each week. By my methodology, you play against the team in its totality, now how strong it was perceived to be at the time of the game.

3
by GoNiners :: Wed, 09/07/2011 - 10:23pm

Are you providing ratings for all teams this year or just the top 25? Didn't you have a link last year to look at all teams?

4
by Brian Fremeau :: Wed, 09/07/2011 - 10:28pm

Just added the link. Sorry for the oversight!