Writers of Pro Football Prospectus 2008

Most Recent FO Features

HarvinPer09.jpg

» Impact of the NFL's Kickoff Rule Change

After three NFL seasons of kicking off from the 35-yard line, what has been the impact on touchbacks, returns, field position, scoring and injuries? Also, is this rule responsible for a record number of big comebacks?

17 Nov 2010

How FEI Watches a Game

by Brian Fremeau

The weekly FEI column has appeared here at Football Outsiders since 2007, and I know many of the regular readers have developed a solid understanding of the elements of my drive-based system. I have been trying to focus this year on revisiting the foundational concepts of those elements to enhance that understanding. I didn't design FEI solely in the pursuit of the "true" value of each college football team as measured against one another. I've also had a parallel goal of continuing to refine the metrics by which I determine those values.

Thus, in the last few weeks I have taken the time to isolate particular elements of the overall drive efficiency data I collect and process. Offensive success is contingent on starting field position expectations and the scoring opportunities that are produced from those opportunities. Field goal success is partly attributed to the special teams unit and the offense that moved the ball into field goal range. Field position following punts is determined by the drive-ending field position of the previous offensive series and the performance of the special teams units on the field. And so on.

All of this seems like a pretty fractured way of examining at the data from a football game. I confess that I sometimes have a hard time wrapping my head around the relationship between these efficiency metrics and the actual points on the scoreboard that determine the winner and loser of a game. I'd like to consider today's FEI column the first step towards clarifying that relationship. It is a work in progress, so please provide feedback and suggestions for future columns.

We'll start with a case study. There have been 556 FBS vs. FBS games this year. Have there been any true gems? It certainly depends on what defines a great football game. If we use game-specific FEI (GFEI) as a guideline, we can isolate a few games in which the winner and loser each had an excellent performance relative to the strength of the opponent faced. LSU's 24-21 victory over Alabama two weekends ago qualifies. The Tigers and Crimson Tide both rank among the Top 5 teams in the current FEI ratings and the game was competitive throughout. The symmetry of game possessions in that contest makes it ideal for our analysis as well -- there were 12 possessions in each half, six for each team, and every drive was a non-garbage possession.

Detailed in the table below is a drive summary account of the LSU-Alabama game. Most of the fields with no shading will look familiar to anyone who has perused a typical box score drive summary. The column labeled "Initiated" refers to the event that initiated the drive: a kickoff, punt, turnover, etc. Plays ("P") and Yards ("Y") provided refer to the field yardage traversed on the drive, including yardage which may have been generated by penalties. The score of the game at the conclusion of each drive is also provided.

First Half Possessions
Drive Possession Initiated Start P Y End Result LSU ALA ODV FGV XPV FPV Seq Ret Auto
1 Alabama Kickoff own 21 3 3 own 24 Punt 0 0 -1.30 - - 1.30 - -.30 1.60
2 LSU Punt opp 43 3 2 opp 41 Punt 0 0 -3.04 - - 3.04 .11 1.07 1.86
3 Alabama Punt own 20 3 2 own 22 Interception 0 0 -1.26 - - 1.26 -.80 .21 1.86
4 LSU Interception opp 35 4 8 opp 27 FG 3 0 -1.77 1.24 - 3.53 1.67 - 1.86
5 Alabama Kickoff own 20 6 10 own 30 Punt 3 0 -1.26 - - 1.26 -.26 -.34 1.86
6 LSU Punt own 36 5 13 own 49 Punt 3 0 -1.93 - - 1.93 -.13 .20 1.86
7 Alabama Punt own 19 11 81 - Touchdown 3 7 5.73 - .04 1.23 -.65 .02 1.86
8 LSU Kickoff own 40 6 18 opp 42 Punt 3 7 -2.12 - - 2.12 -.26 .52 1.86
9 Alabama Punt own 15 3 8 own 23 Punt 3 7 -1.09 - - 1.09 -.79 .02 1.86
10 LSU Punt own 27 3 4 own 31 Punt 3 7 -1.54 - - 1.54 .16 -.48 1.86
11 Alabama Punt own 26 7 27 opp 47 Punt 3 7 -1.50 - - 1.50 -.17 -.19 1.86
12 LSU Punt own 3 9 50 opp 47 Half 3 7 -.72 - - .72 -.72 -.42 1.86
Second Half Possessions
Drive Possession Initiated Start P Y End Result LSU ALA ODV FGV XPV FPV Seq Ret Auto
13 LSU Kickoff own 20 9 52 opp 28 Failed FG 3 7 .41 -1.68 - 1.26 - -.34 1.60
14 Alabama Failed FG own 29 5 27 opp 44 Punt 3 7 -1.62 - - 1.62 -.24 - 1.86
15 LSU Punt own 6 2 94 - Touchdown 10 7 6.15 - .04 .81 -.76 -.29 1.86
16 Alabama Kickoff own 27 10 73 - Touchdown 10 14 5.42 - .04 1.54 -.26 -.07 1.86
17 LSU Kickoff own 29 7 53 opp 18 FG 13 14 .64 .74 - 1.62 -.26 .02 1.86
18 Alabama Kickoff own 37 3 1 own 38 Punt 13 14 -1.98 - - 1.98 -.26 .37 1.86
19 LSU Punt own 23 11 77 - Touchdown 21 14 5.58 - 1.04 1.38 -.39 -.09 1.86
20 Alabama Kickoff own 35 2 -7 own 28 Fumble 21 14 -1.89 - - 1.89 -.26 .28 1.86
21 LSU Fumble opp 28 4 6 opp 22 FG 24 14 -1.91 .92 - 3.99 2.13 - 1.86
22 Alabama Kickoff own 26 9 74 - Touchdown 24 21 5.46 - .04 1.50 -.26 -.11 1.86
23 LSU Kickoff own 23 6 31 opp 46 Punt 24 21 -1.38 - - 1.38 -.26 -.23 1.86
24 Alabama Punt own 15 3 21 own 36 Half 24 21 -1.09 - - 1.09 -.74 -.04 1.86
LSU Total -1.63 1.22 1.08 23.33 1.30 -.04 22.06
Alabama Total 3.63 - .12 17.25 -4.67 -.15 22.06

The yellow columns represent the component values that contribute to each team's scoring, according to my efficiency metrics. In a sense, these columns provide a road map for how FEI "watches" a game. Offensive drive value (ODV) is the scoring value produced (or forfeited) by the offense in the given drive, including the value of driving into field goal range. (ODV is the raw data used to produce Offensive Efficiency ratings, explained here). Field goal value (FGV) is the value produced (or forfeited) by the field goal unit on attempts (introduced and explained here). Extra point value (XPV) is the value produced by the point-after unit touchdown drives, including two-point attempts. Field position value (FPV) is the expected score value of the drive based on starting field position alone.

I have discussed the value of starting field position a number of times, but I don't think I've ever taken the time to explain where it comes from. To be honest, it is difficult to isolate the source with complete certainty based on the drive data I have. I mentioned earlier that this may only be the first step toward better understanding of drive value, and the first step for me is to acknowledge that I can't know everything.

One thing I think I do know is that drives begin with a quantifiable value. This may fly in the face of conventional wisdom -- scores don't appear on the scoreboard at the start of a drive, only at the end. But the data tells me that on average, a little less than two points are granted to the team with the ball on every possession. These points are unearned by the team's offense, defense, or special teams and are awarded simply for participation. An average game may feature 24 total possessions, 12 for each team. Those teams are given 22 points in automatic field position value simply by lining up for those drives. A game of 36 possessions is worth 33 points per team. A game of 16 possessions is worth about 15 points per team.

Above, the automatic field position value of each drive is quantified in the LSU/Alabama game in the gray column labeled "Auto." The drive following the opening kickoff of each half is worth 1.6 points, the average value of starting field position following kickoffs. Subsequent drives in each half are worth 1.86 points, the average value for all remaining possessions. These drives may follow kickoffs, punts, or turnovers, but on average, every team is on equal footing simply due to the alternating possessions of a college football game.

The gray column labeled "Ret" refers to punt or kickoff return value produced from the kick return that initiated the drive. These values and the methodology that produced them were introduced here and here. The final value, labeled "Seq," refers to the field position value produced from a sequence of events that I haven't yet found the best way to isolate and evaluate. I'll examine that issue in a moment. The sum total of the three gray columns equals the total starting field position value of the drive.

Let's take a closer look a few key numbers and drive sequences from this table. LSU's first drive of the game followed a three-and-out forced on Alabama's opening possession. The Crimson Tide punted from their own 24-yard line, a position which on average should have resulted in starting field position for LSU at their own 37-yard line. Instead, the punt traveled only 28 yards, Patrick Peterson returned it another nine yards, and LSU started in Alabama territory. The punt return value (including the punt itself and Peterson's return) represented 1.07 points over the expected value, one-third of the resulting field position value. It was the most valuable return yardage earned by either team all day, but LSU squandered the full value of the drive (-3.04 points) in going three-and-out.

LSU's next possession was a bit more successful. An interception of Greg McElroy by Kelvin Sheppard put the Tigers in striking distance again. LSU gained only eight yards, then successfully kicked a 45-yard field goal to get on the scoreboard first. Who earned those three points? According to this breakdown, the kicker earned 1.24 points on the kick and the starting field position alone accounted for 3.53 points on the drive. Thus, the offense forfeited 1.77 points by going three-and-out.

Note also where the 3.53 points of starting field position came from. The automatic value of 1.86 points represents a little more than half of the total value, leaving 1.67 points of "sequence" field position value. The sequence that produced the additional value includes the following: Alabama's starting field position, the first two plays of the three-down series that netted only two yards, and Sheppard's interception, which was caught 13 yards beyond the line of scrimmage and returned for no gain. Some of that value would be rightfully attributed to the Crimson Tide defense, but unlike the methodology for calculating offensive and defensive efficiency, the baseline for success in forcing this particular sequence is harder to quantify.

Should the interception value be measured against what would have been a punt had the third-down ball been incomplete? If so, what if the interception occurred on first down? Or on a second down 20 yards farther downfield? At this point, it's easier for me to isolate the sequence that produced the extra value than it is to separate all of its elements. This sequence-generated field position isn't a very intuitive metric, but it's a start.

The rest of the table can be evaluated in the same way. LSU's recovered fumble in the second half produced the largest sequence-generated value of the game. The Tigers' two-play, 94-yard drive early in the second half was the most valuable offensive possession of the day. Note that in the summary, however, LSU's total offensive value was actually negative for the game, and Alabama's offense outperformed LSU's by four points. The Crimson Tide lost that value on starting field position, and a big chunk of the difference came in sequence exchanges.

I'll continue digging into these numbers and will provide at least one game summary breakdown in each remaining FEI column this season. As I said before, please let me know if you have further questions or suggestions for this analysis.

Three and Out

Week 3: Three-and-outs, Available Yards, and Explosive Drives
Week 4: Reaching the Red Zone, Methodical Drives, and Late and Close Efficiency
Week 5: Converting 10+ Yard Drives Into Scores, Points Per Possession, and Scoring After Three-and-outs
Week 6: Yards Per TD Drive, Playing With 2- or 3-Score Lead, Third Downs Per First Down Series
Week 7: FEI Team Resumes for South Carolina, Arizona State, and Utah
Week 8: FEI Team Resumes for Oregon, Wisconsin, and Georgia
Week 9: FEI Team Resumes for Iowa, Miami, and Baylor
Week 10: FEI Team Resumes for Arkansas, Michigan State, and Texas

In the tables below, the Game Efficiency, Offensive FEI, Defensive FEI, and "Game" FEI (GFEI) for each team in each game is provided. The ranking of those individual unit and game performances is also provided. Note that there have been 556 FBS vs. FBS game played to date, meaning that there have been 1,112 individual game performances for each category.

The opponent FEI ranking is also provided, as well as a general relevance factor for the particular GFEI, OFEI, and DFEI results for that team in that game. As stated in the FEI principles, my system rewards playing well against good teams, win or lose, and punishes losing to poor teams more harshly than it rewards defeating poor teams. In the formula, the relevance factor is partly a function of the relative ratings of the two teams. Across all games, the least relevant results receive about one-eighth as much weight as the most relevant results. For simplicity, I've generalized the relevance data here into three equally distributed categories, High, Med, and Low.

No. 21 Pittsburgh Panthers (4-4)
Date Wk Opponent Result Opp
FEI
Opp
OFEI
Opp
DFEI
GE GE
Rk
FPA FPA
Rk
OFEI OFEI
Rk
DFEI DFEI
Rk
GFEI GFEI
Rk
Relevance
9/2 1 at Utah L 24-27 49 86 31 -.034 615 .598 122 .279 517 .621 813 .093 424 High
9/23 4 Miami L 3-31 12 26 2 -.261 904 .475 672 .394 442 -.260 270 .074 452 High
10/2 5 Florida International W 44-17 57 57 42 .336 144 .630 59 1.006 142 .413 706 .293 183 Med
10/9 6 at Notre Dame L 17-23 47 44 41 -.082 724 .439 851 .532 355 -.100 364 .098 415 High
10/16 7 at Syracuse W 45-14 34 65 39 .298 180 .579 188 .912 176 -.174 320 .562 24 Med
10/23 8 Rutgers W 41-21 48 106 14 .309 166 .492 588 1.432 46 .201 557 .299 174 Med
10/30 9 Louisville W 20-3 39 63 37 .324 154 .531 409 .615 312 -.650 101 .419 74 Med
11/11 11 at Connecticut L 28-30 73 82 59 -.023 590 .407 971 .388 446 .212 563 -.024 600 High

The Panthers are the best four-loss team in the FEI ratings and didn't suffer much of a blow after their lackluster loss to Connecticut. What does FEI see in this team that our intuition and eyeballs do not? With the exception of the Miami game, the other close losses aren't particularly damaging, and the Hurricanes strong rating helps as well. The biggest thing that jumps out to me are the moderately strong ratings of the Big East foes they've faced. There aren't any complete albatrosses on Pittsburgh's schedule according to FEI, and the respective strong showings against Syracuse and Louisville are rewarded. Florida International is another team that will earn the Panthers no credit in the court of public opinion, but FEI respects. Their games against South Florida, West Virginia, and Cincinnati are all winnable down the stretch, and the Panthers are still in the driver's seat for the league's BCS bowl berth.

No. 28 Texas A&M Aggies (6-3)
Date Wk Opponent Result Opp
FEI
Opp
OFEI
Opp
DFEI
GE GE
Rk
FPA FPA
Rk
OFEI OFEI
Rk
DFEI DFEI
Rk
GFEI GFEI
Rk
Relevance
9/11 2 Louisiana Tech W 48-16 86 93 88 .267 204 .476 663 .426 420 -.013 430 .028 517 Low
9/18 3 Florida International W 27-20 57 57 42 .059 435 .374 1048 .146 594 -.718 76 .015 535 Med
9/30 5 at Oklahoma State L 35-38 17 15 28 -.029 603 .429 894 .947 161 -.625 108 .405 83 Med
10/9 6 vs. Arkansas L 17-24 8 3 15 -.065 687 .462 745 .410 426 -1.404 6 .369 111 Med
10/16 7 Missouri L 9-30 15 23 5 -.386 998 .467 720 .133 603 .272 608 -.072 661 Med
10/23 8 at Kansas W 45-10 109 105 97 .381 122 .582 177 .114 624 .336 660 .051 485 Low
10/30 9 Texas Tech W 45-27 70 59 63 .275 190 .459 765 1.050 127 -.183 315 .134 372 Med
11/6 10 Oklahoma W 33-19 18 14 20 .138 331 .420 926 .992 149 -1.239 12 .420 73 High
11/13 11 at Baylor W 42-30 66 25 110 .137 334 .576 201 -.341 873 .146 523 .168 328 Med

The Aggies also defeated Florida International, though not as dominantly as Pittsburgh. Field position disadvantage might have had something to do with it -- only three teams overcame a greater FPA deficit and won this season. That game doesn't paint the entire picture for Texas A&M, but they are the second-best team the Aggies have beaten this year, according to FEI. The big win over Oklahoma on November 6 and the defensive effort in a loss to Arkansas on October 9 are also notable. Nebraska and Texas await. My numbers give the Aggies a 27 percent chance of winning both. A victory over the Cornhuskers will boost their rating significantly.

No. 56 San Diego State Aztecs (6-3)
Date Wk Opponent Result Opp
FEI
Opp
OFEI
Opp
DFEI
GE GE
Rk
FPA FPA
Rk
OFEI OFEI
Rk
DFEI DFEI
Rk
GFEI GFEI
Rk
Relevance
9/11 2 at New Mexico State W 41-21 120 118 115 .197 275 .587 165 -.525 953 1.299 1050 -.354 980 Low
9/18 3 at Missouri L 24-27 15 23 5 -.026 594 .537 372 .786 224 -.520 141 .433 68 Med
9/25 4 Utah State W 41-7 101 70 103 .421 101 .380 1034 1.153 95 -.526 136 .050 488 Med
10/9 6 at BYU L 21-24 81 99 64 -.041 637 .467 719 .244 530 .796 907 -.114 721 High
10/16 7 Air Force W 27-25 44 36 56 .026 520 .551 307 .409 427 .277 611 .066 468 High
10/23 8 at New Mexico W 30-20 118 117 108 .167 302 .597 128 -.892 1035 .695 855 -.355 982 Low
10/30 9 at Wyoming W 48-38 112 104 118 .102 368 .474 685 -.355 887 .972 972 -.273 907 Low
11/6 10 Colorado State W 24-19 116 112 114 .060 434 .475 669 -1.046 1065 .655 827 -.585 1081 Low
11/13 11 at TCU L 35-40 10 19 8 -.045 647 .474 676 1.053 125 -.370 210 .443 64 Med

Our preseason predictions zeroed in on Air Force to join the Mountain West Conference upper crust, but the Aztecs of San Diego State nabbed that distinction by knocking off the Falcons on October 16. That's the only victory SDSU can claim against teams not ranked in the 100s in FEI. But their best resume bullet points are narrow defeats at the hands of Missouri and TCU. San Diego State might be able to whip Utah this weekend if the Utes are still stinging from back-to-back losses, and a conference runner-up finish is within reach.

Thanks to reader "mm" for the team resume suggestions this week. If you have a suggestion for an FEI team resume you'd like to see, drop me a line on Twitter or in the comment section here. I'm happy to answer data inquiries or provide team resume tables for bloggers interested in investigating the data themselves.

FEI Week 11 Top 25

The principles of the Fremeau Efficiency Index (FEI) can be found here. 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 to date. SOS listed here does not include 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 RW) represent the average expected team wins for games scheduled but not yet played.

Offensive FEI (OFEI) and Defensive FEI (DFEI) are the opponent-adjusted ratings of all non-garbage-time drives from scrimmage. Field Position Advantage (FPA) is the share of the value of total starting field position for the season earned by each team against its opponents. Field Goal Efficiency (FGE) is the point value per field goal attempt earned by the field goal unit.

Only games between FBS teams are considered in the FEI calculations. The FEI ratings published here are a function of the results of games played through November 13.

FEI ratings for all 120 FBS teams are listed in the stats page section of FootballOutsiders.com. Click here for current ratings; the pull-down menu in the stats section directs you to 2007 through 2009 ratings. There are also now separate pages for offensive and defensive FEI ratings for 2010.

Rk Team FBS
Rec
FEI LW
Rk
GE GE
Rk
SOS SOS
Rk
FBS
MW
FBS
RW
OFEI OFEI
Rk
DFEI DFEI
Rk
FPA FPA
Rk
FGE FGE
Rk
1 Auburn 10-0 .301 1 .169 14 .206 26 9.3 0.5 .809 1 -.354 21 .520 37 .134 46
2 Oregon 9-0 .266 2 .278 4 .353 62 9.6 1.7 .330 18 -.442 10 .551 8 .088 55
3 Alabama 8-2 .258 8 .266 5 .159 15 8.7 0.5 .551 5 -.382 16 .547 13 .236 38
4 Stanford 8-1 .254 6 .251 6 .256 34 9.4 1.8 .377 12 -.330 26 .541 15 .185 41
5 LSU 8-1 .246 3 .122 21 .125 11 8.4 1.5 .210 34 -.380 17 .596 1 .702 8
6 Virginia Tech 8-1 .239 10 .247 7 .296 46 9.1 1.5 .434 9 -.369 19 .554 6 .580 12
7 Boise State 9-0 .231 7 .466 1 .535 92 10.9 2.7 .335 17 -.442 9 .547 12 .243 36
8 Arkansas 7-2 .220 11 .164 16 .104 5 7.7 1.2 .613 3 -.382 15 .502 64 .835 2
9 Nebraska 8-1 .214 5 .212 10 .417 76 9.4 1.6 .373 13 -.427 11 .529 25 .757 6
10 TCU 10-0 .211 4 .329 3 .730 112 10.3 1.0 .320 19 -.447 8 .562 3 .119 49
11 Wisconsin 8-1 .209 16 .225 9 .303 48 9.0 1.6 .568 4 -.180 35 .548 10 .438 22
12 Miami 6-3 .207 14 .065 43 .199 23 7.9 1.3 .265 26 -.595 2 .491 70 .256 34
Rk Team FBS
Rec
FEI LW
Rk
GE GE
Rk
SOS SOS
Rk
FBS
MW
FBS
RW
OFEI OFEI
Rk
DFEI DFEI
Rk
FPA FPA
Rk
FGE FGE
Rk
13 Ohio State 9-1 .204 12 .339 2 .351 61 9.7 1.3 .342 16 -.377 18 .559 4 -.042 76
14 Iowa 6-3 .197 9 .174 13 .382 70 8.7 1.5 .255 29 -.426 13 .525 31 .054 58
15 Missouri 7-2 .196 13 .125 19 .310 50 8.9 1.9 .279 23 -.478 5 .517 43 .419 25
16 South Carolina 6-3 .187 23 .091 31 .106 6 7.4 1.5 .489 6 -.344 23 .509 50 -.043 77
17 Oklahoma State 9-1 .183 17 .199 11 .542 94 10.1 1.6 .361 15 -.306 28 .514 45 .745 7
18 Oklahoma 8-2 .180 18 .162 17 .363 64 9.3 1.2 .363 14 -.362 20 .529 24 .068 57
19 North Carolina State 6-3 .172 15 .106 25 .325 52 7.7 1.1 .088 45 -.427 12 .531 22 -.106 84
20 USC 7-3 .172 25 .118 22 .218 27 9.7 2.3 .423 10 .088 73 .558 5 -.791 113
21 Pittsburgh 4-4 .163 19 .089 32 .450 82 7.8 2.0 .276 24 -.229 32 .517 42 .008 69
22 Michigan State 8-1 .162 22 .112 23 .334 53 8.3 1.7 .243 31 -.513 3 .484 84 .588 11
23 Clemson 4-5 .159 20 .039 51 .111 9 6.9 1.5 .117 41 -.509 4 .516 44 -.816 114
24 West Virginia 5-3 .136 35 .124 20 .399 72 7.5 1.7 -.041 68 -.604 1 .507 54 -.069 81
25 Florida State 6-3 .136 30 .073 39 .199 24 6.7 1.2 .392 11 -.114 36 .513 47 .112 51

Posted by: Brian Fremeau on 17 Nov 2010

9 comments, Last at 18 Nov 2010, 2:02pm by Brian Fremeau

Comments

1
by mm (not verified) :: Wed, 11/17/2010 - 3:10pm

So the SEC West has half of the top 8. Bet these are their favorite rankings.

Thanks for taking my suggestions this week. If you don't get enough next, week, how about Sunbelt-leading Florida International (if they don't get beat)? It's interesting to look at how some of these small conference schools do in the system.

Also, maybe at some point a feature on who's been most erratic? The GFEI rankings are the one thing I look at more than anything else in these ratings. Who's been bouncing up and down the most from game to game? Cal seems like a good bet to be one of the schools, but who else?

2
by Kal :: Wed, 11/17/2010 - 3:40pm

Huh.

I don't think anyone would have been able to reasonably say that Oregon has a better defense than they do offense, but that's certainly the feeling I've had some times this year.

Also, Boise and Oregon are almost identical in their values; is it only SOS that ranks Oregon higher?

3
by Brian Fremeau :: Wed, 11/17/2010 - 4:51pm

Here are the Oregon and Boise State GFEI numbers, best to worst (national GFEI rank in parentheses).

OREGON
.618 (16) vs. No. 20 USC (Medium)
.577 (19) vs. No. 4 Stanford (High)
.498 (43) vs. No. 74 UCLA (Low)
.310 (162) vs. No. 40 Arizona State (Med)
.302 (170) vs. No. 75 Tennessee (Low)
.193 (296) vs. No. 69 Washington (Low)
.135 (371) vs. No. 118 New Mexico (Low)
.089 (427) vs. No. 95 Washington State (Low)
.048 (494) vs. No. 67 California (Low)

BOISE STATE
.508 (38) vs. No. 58 Hawaii (Low)
.506 (40) vs. No. 6 Virginia Tech (High)
.369 (110) vs. No. 114 San Jose State (Low)
.291 (186) vs. No. 105 Idaho (Low)
.286 (193) vs. No. 35 Oregon State (Med)
.256 (231) vs. No. 103 Toledo (Low)
.206 (280) vs. No. 120 New Mexico State (Low)
.197 (289) vs. No. 112 Wyoming (Low)
.121 (390) vs. No. 86 Louisiana Tech (Low)

Overall strength of schedule is a factor here, but very strong performances against USC and Stanford are also better than Boise State's best efforts.

4
by bsharp :: Wed, 11/17/2010 - 6:14pm

Interesting. That helps explain how LSU has managed to win 9 games and look completely disfunctional for a large part of the season Having the top FPA in the country at .596 has everything to do with that. All season long, the story has been punt returns, kickoff returns, bad kicks to avoid returns, interceptions, fumbles, downing punts inside the 10 yard line, etc. When the game is over and we've scored 51 points on just 251 total offensive yards (ULM), it leaves a lot of fans dumbfounded.

On a related note, I'm going to miss Patrick Peterson after this year. Man is he good.

5
by TomTom (not verified) :: Thu, 11/18/2010 - 12:50am

I think if one of the undefeated teams loses this next week (well, not Oregon, bye week), then you could pull off looking at those 3 teams. I would imagine you would want to do that by the end of the season, but there is no certainty on when there will be 3 undefeated teams left.

P.S. I would love to see how the garbage time is done. If some team makes a huge comeback this weekend (like Penn St. did against Northwestern a bit back), that would be a good choice for a game to look at.

6
by D :: Thu, 11/18/2010 - 1:14am

I'm confused about something. How can Ohio State have a better game efficiency rating and a tougher schedule than TCU and still be ranked behind them in FEI? Isn't FEI just the opponent adjusted version of GE?

7
by Brian Fremeau :: Thu, 11/18/2010 - 9:39am

FEI is opponent-adjusted GE, but there isn't a linear relationship between GE, SOS, and FEI for two reasons. One, SOS is calculated to mean something very specific: the likelihood that an elite team would go undefeated against the given team's schedule. Second, the "relevance" weight given to each game is a critical factor in the relationship between GE and FEI. Here are the two team's resumes, best to worst, according to GFEI:

OHIO STATE
.467 (58) vs. No. 12 Miami (High)
.388 (94) vs. No. 36 Illinois (Med)
.341 (140) vs. No. 97 Purdue (Low)
.339 (142) vs. No. 85 Minnesota (Low)
.292 (184) vs. No. 89 Indiana (Low)
.251 (235) vs. No. 11 Wisconsin (High)
.242 (242) vs. No. 84 Ohio (Low)
.221 (259) vs. No. 50 Penn State (Med)
.133 (374) vs. No. 104 Marshall (Low)
.015 (534) vs. No. 111 Eastern Michigan (Low)

TCU
.684 (8) vs. No. 49 Utah (Med)
.530 (32) vs. No. 44 Air Force (Med)
.385 (96) vs. No. 112 Wyoming (Low)
.356 (127) vs. No. 66 Baylor (Low)
.237 (246) vs. No. 35 Oregon State (Med)
.212 (271) vs. No. 71 SMU (Low)
.055 (481) vs. No. 81 BYU (Low)
.005 (546) vs. No. 56 San Diego State (Med)
-.034 (611) vs. No. 117 UNLV (Low)
-.214 (847) vs. No. 116 Colorado State (Low)

The very top of TCU's resume trumps Ohio State's, even if the depth of the schedule is better for the Buckeyes.

8
by NGD (not verified) :: Thu, 11/18/2010 - 1:32pm

Brian, you said, "A game of 36 possessions is worth 33 points per team. A game of 16 possessions is worth about 15 points per team."

I understand the point--please don't pummel me with math--but isn't this backwards in some sense? Doesn't fewer possessions generally indicate that they are marching down the field for points? Wouldn't more possessions probably mean a lot of 3-and-outs?

9
by Brian Fremeau :: Thu, 11/18/2010 - 2:02pm

In terms of the value of the automatic field position itself, the pace of the game isn't relevant. In your example, a game of fewer possessions might indicate a higher scoring rate than one of many possessions, but this isn't always the case. And it might only be the case for one of the two teams in the game.