Writers of Pro Football Prospectus 2008

Most Recent FO Features

ProctorSpe15.jpg

» VN: Hits, Misses, and the Year Zero Effect

Bill Connelly looks at the college offenses, defenses, and overall teams that have improved (Air Force!) or regressed (North Texas!) the most in 2014. Year Zero is a real thing (sometimes).

24 Nov 2010

FEI: Head to Head

by Brian Fremeau

The college ratings we publish here at Football Outsiders have been fairly fluid week to week, and as Bill Connelly pointed out yesterday, it's just as important to focus on the big picture as it is the individual team rankings. With 120 teams, there are bound to be outliers and teams that move up or down a handful of ranking positions without much change in their actual rating.

That said, we'd like to think that most of our team comparisons are intuitive for those equipped with a basic understanding of our approach. But inevitably, a few ratings stump us. And then there are weeks when a half dozen examples of head-to-head results don't seem to fit. In the new FEI ratings, Stanford ranks ahead of Oregon, Alabama is ahead of LSU, Virginia Tech is ahead of Boise State, Ohio State is ahead of Wisconsin, etc. What's really going on here?

Ranking violations -- teams ranked incongruently with head-to-head results -- are impossible to avoid entirely, of course. There are "minimum violations" rating systems out there that are specifically engineered to maintain the transitive property as much as possible, but even these systems result in a ranking violation of around 8 percent or so. FEI has a current ranking violation percentage of about 23 percent -- that is, 23 percent of actual head-to-head results on the field are flipped in the rankings. Most computer rating systems hover between 15 percent and 20 percent.

The reasons for individual ranking violations of head-to-head results are many. In the case of FEI, all of the foundational elements are at least partly responsible -- drive data, opponent adjustments, game relevance weights, garbage time, FBS vs. FBS games only, etc. Why might Stanford be ranked ahead of Oregon even though the Ducks handed the Cardinal a 52-31 defeat on October 2? How much weight should that one result receive in comparison with the other 17 FBS games played by those two teams?

I re-ran the FEI ratings under four different scenarios, eliminating one of the head-to-head games listed above in each instance -- Oregon-Stanford, Boise State-Virginia Tech, LSU-Alabama, and Wisconsin-Ohio State. These games alone had little impact on the overall rankings, but they did affect those of the two teams involved in each case. In Stanford's case, its other nine games would rank the Cardinal No. 2 in the nation behind Auburn, and the Ducks dropped all the way to No. 10. The other games had similar results. Boise State would rank No. 13 without the win over Virginia Tech. LSU would rank No. 12 without the win over Alabama. Wisconsin would rank No. 17 without the victory over Ohio State. Virginia Tech, Alabama, and Ohio State would have a slightly better FEI rating in each case, but not significantly so.

Is it possible for both teams to receive an FEI ratings boost simply by playing one another? In isolation, it doesn't appear to be so, but this weekend could put that theory to the test. The FEI ratings are infatuated with the SEC West's best four teams, and they face off in two matchups this weekend. If the games are competitive, I fully expect Auburn, Alabama, LSU, and Arkansas will all remain among the top eight teams according to FEI. Depending on other results, all four might move up. Stay tuned.

FEI Drive Summary Breakdown

Last week, I introduced a new feature in the weekly column -- the FEI Drive Summary breakdown. The FEI system "watches" each game as a series of alternating possession and breaks down the scoreboard value of each possession to the offensive, defensive, and special teams units that contributed to that value. In the final weeks this season, I'll feature at least one drive summary table from the previous weekend.

In the tables below, the white fields represent basic box score drive data. The yellow fields represent the component values that contribute to each team's scoring, according to my efficiency metrics. 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. Field goal value (FGV) is the value produced (or forfeited) by the field goal unit on attempts. 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. The gray fields represent the component values of FPV. Automatic field position value (Auto) is unearned drive value based on national average score expectations from possession alone. Punt or kickoff return value (Ret) is the value above or below average produced from the kick return that initiated the drive. The remaining value (Seq) is produced from the sequence of defensive and special teams events that preceded the drive.

The total score values include only non-garbage possessions and are the only data that is included in the weekly FEI ratings. Garbage possession data is provided as a reference, including first half clock kill possessions. The start of garbage time at the end of a game is calculated retroactively from the game's conclusion as a function of the score margin and the remaining possessions to be played.

First Half Non-Garbage Possessions
Drive Possession Initiated Start P Y End Result Utah SDSU ODV FGV XPV FPV Seq Ret Auto
1 San Diego State Kickoff own 26 8 74 - Touchdown 0 7 5.46 - .04 1.50 - -.10 1.60
2 Utah Kickoff own 19 8 66 opp 15 FG 3 7 1.14 .64 - 1.23 -.26 -.37 1.86
3 San Diego State Kickoff own 25 3 2 own 27 Punt 3 7 -1.46 - - 1.46 -.26 -.14 1.86
4 Utah Punt own 23 4 37 opp 40 Punt 3 7 -1.38 - - 1.38 -.01 -.47 1.86
5 San Diego State Punt own 2 11 98 - Touchdown 3 14 6.26 - .04 .70 -.81 -.35 1.86
6 Utah Kickoff own 47 3 -8 own 39 Punt 3 14 -2.48 - - 2.48 -.26 .88 1.86
7 San Diego State Punt own 21 4 79 - Touchdown 3 20 5.66 - -.96 1.30 -.42 -.14 1.86
8 Utah Kickoff own 37 3 63 - Touchdown 10 20 4.98 - .04 1.98 -.26 .38 1.86
9 San Diego State Kickoff own 14 14 86 - Touchdown 10 27 5.91 - .04 1.05 -.26 -.55 1.86
10 Utah Kickoff own 32 3 68 - Touchdown 17 27 5.21 - .04 1.75 -.26 .15 1.86
11 San Diego State Kickoff own 29 10 48 opp 23 Interception 17 27 -1.62 - - 1.62 -.26 .02 1.86
12 Utah Interception own 15 5 85 - Touchdown 24 27 5.87 - .04 1.09 -.77 - 1.86
Second Half Non-Garbage Possessions
Drive Possession Initiated Start P Y End Result Utah SDSU ODV FGV XPV FPV Seq Ret Auto
13 Utah Kickoff own 20 7 40 opp 40 Punt 24 27 -1.26 - - 1.26 - -.34 1.60
14 San Diego State Punt own 10 3 90 - Touchdown 24 34 6.03 - .04 .93 -.81 -.12 1.86
15 Utah Kickoff own 25 10 25 50 Punt 24 34 -1.46 - - 1.46 -.26 -.14 1.86
16 San Diego State Punt own 14 4 23 own 37 Punt 24 34 -1.05 - - 1.05 -.67 -.14 1.86
17 Utah Punt own 21 14 79 - Touchdown 31 34 5.66 - .04 1.30 -.37 -.19 1.86
18 San Diego State Kickoff own 20 3 -5 own 15 Punt 31 34 -1.26 - - 1.26 -.26 -.34 1.86
19 Utah Punt opp 3 4 3 - Touchdown 38 34 1.08 - .04 5.88 .58 3.44 1.86
20 San Diego State Kickoff own 24 9 46 opp 30 Interception 38 34 -1.42 - - 1.42 -.26 -.18 1.86
21 Utah Interception own 12 6 42 opp 46 Punt 38 34 -.99 - - .99 -.87 - 1.86
22 San Diego State Punt own 15 6 47 opp 38 Interception 38 34 -1.09 - - 1.09 -.74 -.04 1.86
Utah Total 38 16.36 .64 .20 20.80 -2.74 3.34 20.20
San Diego State Total 34 21.42 - -.80 13.38 -4.74 -2.08 20.20
Garbage Time Possessions
Drive Possession Initiated Start P Y End Result Utah SDSU ODV FGV XPV FPV Seq Ret Auto
23 Utah Interception own 20 3 -10 own 10 Half 38 34 -1.26 - - 1.26 -.60 - 1.86

The Utah Utes pulled off the biggest comeback of last weekend, erasing a 17-point first-half deficit against San Diego State to win 38-34. Utah's most valuable (and critical) drive was a five-play, 85-yard touchdown drive at the end of the first half that followed an Aztecs interception, but overall, San Diego State's offense produced more drive value than Utah. The Utes' big advantage in the game came from starting field position, keyed by a punt block early in the fourth quarter. Note that the value listed in the return column (3.44 points earned on the punt block) was three times more valuable than the offensive series that punched in the three-yard touchdown drive. San Diego State also had a failed extra point on its third touchdown drive (-.96 points lost in extra point value) which, if successful, would have allowed the Aztecs to drive for a game-tying field goal instead of an attempted game-winning touchdown on their final possession.

First Half Non-Garbage Possessions
Drive Possession Initiated Start P Y End Result Stan Cal ODV FGV XPV FPV Seq Ret Auto
1 California Kickoff own 20 3 6 own 26 Fumble 0 0 -1.26 - - 1.26 - -.34 1.60
2 Stanford Fumble opp 26 4 15 opp 11 FG 3 0 -1.65 .52 - 4.13 2.27 - 1.86
3 California FG own 34 9 41 opp 25 Interception 3 0 -1.84 - - 1.84 -.26 .24 1.86
4 Stanford Interception own 5 7 95 - Touchdown 10 0 6.18 - .04 .78 -1.08 - 1.86
5 California Kickoff own 22 8 21 own 43 Punt 10 0 -1.34 - - 1.34 -.26 -.26 1.86
6 Stanford Punt own 14 9 86 - Touchdown 17 0 5.91 - .04 1.05 -.53 -.28 1.86
7 California Kickoff own 26 6 15 own 41 Punt 17 0 -1.50 - - 1.50 -.26 -.10 1.86
8 Stanford Punt own 10 9 90 - Touchdown 24 0 6.03 - .04 .93 -.48 -.46 1.86
9 California Kickoff own 22 5 27 own 49 Interception 24 0 -1.34 - - 1.34 -.26 -.26 1.86
10 Stanford Interception own 39 6 61 - Touchdown 31 0 4.89 - .04 2.07 .21 - 1.86
Second Half Non-Garbage Possessions
Drive Possession Initiated Start P Y End Result Stan Cal ODV FGV XPV FPV Seq Ret Auto
12 Stanford Kickoff own 36 8 64 - Touchdown 38 0 5.03 - .04 1.93 - .33 1.60
13 California Kickoff own 20 3 -3 own 17 Punt 38 0 -1.26 - - 1.26 -.26 -.34 1.86
Stanford Total 38 26.39 .52 .20 10.89 .40 -.41 10.90
California Total 0 -8.54 - - 8.54 -1.30 -1.06 10.90
Garbage Time Possessions
Drive Possession Initiated Start P Y End Result Stan Cal ODV FGV XPV FPV Seq Ret Auto
11 California Kickoff own 34 1 -1 own 33 Half 31 0 -1.84 - - 1.84 -.26 .24 1.86
14 Stanford Punt own 44 11 56 - Touchdown 45 0 4.64 - .04 2.32 .47 .00 1.86
15 California Kickoff own 21 12 79 - Touchdown 45 7 5.66 - .04 1.30 -.26 -.30 1.86
16 Stanford Kickoff own 29 13 48 opp 23 FG 48 7 .41 .97 - 1.62 -.26 .02 1.86
17 California Kickoff own 20 6 33 opp 47 Downs 48 7 -1.26 - - 1.26 -.26 -.34 1.86
18 Stanford Downs own 47 3 -1 own 46 Punt 48 7 -2.48 - - 2.48 .62 - 1.86
19 California Punt own 19 13 81 - Touchdown 48 14 5.73 - .04 1.23 -.59 -.04 1.86
20 Stanford Kickoff own 25 1 3 own 28 Half 48 14 -1.46 - - 1.46 -.26 -.14 1.86

By GFEI, the single-game opponent-adjusted measure of team performance, Stanford's rout of California was the second most dominant effort of the season by any team in the nation. The Cardinal offense settled for a field goal on a short field in its first drive (-1.65 ODV), but had touchdown drives of 95, 86, 90, 61, and 64 on its next five possessions. Cal managed to cross the 50-yard line only once on its six offensive possessions. Compare this performance to Oregon's effort against California a week earlier. In that game, the Ducks had 26.89 points of field position value but had a negative offensive drive value of -9.36 points the game. The difference in the games against Cal is a big reason why Stanford leapt Oregon in this week's FEI ratings.

Three and Out

Previous topics:

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 609 FBS vs. FBS game played to date, meaning that there have been 1,218 individual game performances for each category.

The opponent FEI ranking is also provided, as well as a general relevance factor (Rel) 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. 33 Mississippi State Bulldogs (6-4)
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
Rel
9/4 1 Memphis W 49-7 117 117 115 .6 34 .500 610 .393 476 .235 645 -.056 691 Low
9/9 2 Auburn L 14-17 1 2 18 -.036 675 .525 478 .406 464 -1.602 2 .495 48 Low
9/18 3 at LSU L 7-29 8 30 24 -.419 1106 .328 1204 .342 525 -.308 275 .118 420 Med
9/25 4 Georgia W 24-12 32 31 68 .321 171 .626 69 .324 533 -.903 36 .465 61 High
10/9 6 at Houston W 47-24 74 19 96 .299 195 .640 54 .042 722 -.606 122 .275 240 Med
10/16 7 at Florida W 10-7 29 47 25 .045 513 .585 184 .326 531 -.255 295 .346 150 High
10/23 8 UAB W 29-24 86 58 72 .057 482 .574 230 -.144 849 .307 694 -.198 896 Low
10/30 9 Kentucky W 24-17 35 16 83 .074 450 .458 844 -.172 868 -.969 33 .189 323 High
11/13 11 at Alabama L 10-30 2 4 15 -.386 1097 .562 281 -.035 781 -.073 409 .212 304 Med
11/20 12 Arkansas L 31-38 6 3 21 -.077 778 .502 594 .904 198 -.804 57 .330 167 Med

The Bulldogs have played the toughest schedule in the country according to FEI, having played four division mates that currently rank among the top eight teams in the country. They lost all four games, but played Auburn and Arkansas close. Those efforts, along with strong performances against Georgia and Kentucky, all rank among the Top 100 opponent-adjusted defensive games of the year.

No. 46 Nevada Wolf Pack (9-1)
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
Rel
9/11 2 Colorado State W 51-6 119 116 114 .835 5 .560 290 1.382 62 .306 693 .166 355 Low
9/17 3 California W 52-31 87 100 78 .286 203 .551 331 1.462 50 1.237 1133 .028 558 Med
9/25 4 at BYU W 27-13 82 98 62 .222 274 .395 1104 .888 202 .182 607 .150 380 Med
10/2 5 at UNLV W 44-26 116 115 109 .271 219 .436 944 .585 356 .988 1066 -.213 912 Low
10/9 6 San Jose State W 35-13 115 104 112 .286 207 .503 591 .408 462 .390 752 -.254 960 Low
10/16 7 at Hawaii L 21-27 60 55 85 -.071 764 .529 455 -.564 1050 .160 593 -.018 646 High
10/30 9 Utah State W 56-42 105 86 108 .261 229 .640 53 .674 306 1.936 1205 -.166 857 Low
11/6 10 at Idaho W 63-17 104 81 102 .416 114 .595 145 1.363 68 .589 871 .168 351 Low
11/13 11 at Fresno State W 35-34 79 84 82 .012 593 .619 87 -.017 772 1.339 1154 -.054 689 Med
11/20 12 New Mexico State W 52-6 120 119 117 .435 106 .476 735 .125 658 .460 792 -.249 956 Low

On the complete opposite end of the schedule perspective, Nevada's best opponent to date, Hawaii, ranks 60th nationally. The Wolf Pack's overall SOS ranks dead last. Oh yeah, and Nevada lost the game against the Warriors. There simply aren't any impressive opponent-adjusted victories or defeats here, even though Nevada's raw offensive numbers have been eye-popping. The test this weekend against Boise State will be unlike anything Nevada has seen, offensively or defensively.

No. 59 Florida International Golden Panthers (5-5)
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
Rel
9/11 2 Rutgers L 14-19 57 97 37 -.045 702 .46 828 -.178 871 .070 516 -.110 768 High
9/18 3 at Texas A&M L 20-27 25 41 3 -.059 739 .626 70 .419 456 -.315 272 .299 210 Med
9/25 4 at Maryland L 28-42 36 53 27 -.138 853 .508 575 .640 333 .844 1013 .116 427 Med
10/2 5 at Pittsburgh L 17-44 20 24 32 -.336 1061 .370 1155 .936 177 .289 677 .069 492 Med
10/9 6 Western Kentucky W 28-21 91 60 92 .087 422 .452 868 .067 697 .004 460 -.193 894 Med
10/16 7 at North Texas W 34-10 90 87 86 .443 101 .535 417 .459 427 -.366 239 .308 192 Med
10/30 9 at Florida Atlantic L 9-21 92 83 51 -.143 860 .465 807 -.573 1051 .433 776 -.305 1021 High
11/6 10 Louisiana Monroe W 42-35 103 93 49 .077 443 .562 282 .535 387 .673 920 -.290 1006 Med
11/13 11 at Troy W 52-35 83 56 80 .157 345 .432 962 .834 227 .010 463 .065 504 Med
11/20 12 at Louisiana Lafayette W 38-17 100 73 73 .190 311 .517 525 .255 575 -.006 455 -.009 627 Med

Per a reader suggestion from "mm," we take a look at Florida International, the top-rated team from the Sun Belt conference. We didn't project much success for the Golden Panthers at the start of the year, but a few close losses in non-conference play set the stage for a strong conference run down the stretch. Florida International hasn't defeated anyone in the top two-thirds of the college football world, but they haven't had to in the Sun Belt. It has been a breakthrough year for a program that had won only 13 games in five seasons of FBS. They can wrap up a conference championship with a victory over Arkansas State this weekend.

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 12 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 20.

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 .306 1 .169 15 .189 32 9.3 0.5 .787 2 -.371 18 .520 41 .134 49
2 Alabama 8-2 .266 3 .266 6 .143 17 8.7 0.5 .544 4 -.408 15 .547 10 .236 38
3 Virginia Tech 9-1 .264 6 .236 8 .180 26 9.3 1.0 .481 6 -.389 17 .553 5 .657 6
4 Stanford 9-1 .259 4 .293 4 .284 52 9.6 0.9 .430 9 -.305 26 .542 13 .208 42
5 Oregon 9-0 .253 2 .277 5 .398 82 9.5 1.6 .325 15 -.412 14 .551 8 .088 59
6 Arkansas 8-2 .243 8 .153 18 .067 3 8.0 0.6 .661 3 -.369 21 .501 60 .595 10
7 Boise State 10-0 .237 7 .466 1 .466 90 10.9 1.8 .333 14 -.458 9 .546 12 .337 30
8 LSU 9-1 .236 5 .116 22 .110 10 8.1 0.4 .227 30 -.331 24 .594 1 .674 5
9 Ohio State 10-1 .236 13 .315 3 .205 36 10.1 0.9 .387 12 -.404 16 .553 4 .050 65
10 Wisconsin 9-1 .231 11 .229 9 .251 44 9.3 1.0 .508 5 -.271 29 .552 7 .467 21
11 TCU 10-0 .215 10 .329 2 .701 113 10.3 1.0 .321 18 -.463 7 .562 3 .119 55
12 Nebraska 8-2 .208 9 .187 13 .312 59 9.2 1.0 .249 27 -.487 6 .534 20 .804 3
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 South Carolina 7-3 .206 16 .149 19 .088 6 7.6 0.6 .480 7 -.370 20 .541 14 -.043 78
14 Missouri 8-2 .205 15 .130 20 .267 48 8.9 1.0 .241 29 -.510 4 .521 40 .419 24
15 Oklahoma 9-2 .202 18 .188 12 .306 57 9.6 0.5 .322 17 -.423 11 .546 11 .143 48
16 North Carolina State 7-3 .192 19 .099 26 .229 40 8.1 0.7 .109 43 -.461 8 .528 24 .006 70
17 Miami 6-4 .192 12 .042 47 .117 11 7.4 0.8 .259 26 -.545 2 .488 77 .148 47
18 Iowa 6-4 .191 14 .154 17 .273 51 8.5 0.9 .275 21 -.422 12 .524 33 -.070 80
19 Oklahoma State 10-1 .173 17 .212 10 .520 98 9.8 0.5 .356 13 -.276 28 .519 42 .552 16
20 Pittsburgh 5-4 .169 21 .089 32 .374 74 7.8 1.4 .263 24 -.230 32 .514 47 -.048 79
21 Florida State 7-3 .161 25 .082 36 .156 21 7.3 0.7 .394 11 -.136 38 .528 25 .022 67
22 Michigan State 9-1 .156 22 .105 25 .305 55 8.1 0.7 .294 20 -.439 10 .488 76 .588 11
23 West Virginia 6-3 .154 24 .118 21 .350 67 7.9 1.2 -.025 66 -.634 1 .500 63 -.126 84
24 Clemson 5-5 .152 23 .060 41 .101 8 6.7 0.4 .091 46 -.495 5 .517 45 -.589 109
25 Texas A&M 7-3 .145 28 .082 35 .186 30 7.2 0.8 .135 41 -.537 3 .460 107 .349 28

Posted by: Brian Fremeau on 24 Nov 2010

9 comments, Last at 08 Dec 2012, 3:16am by john123456

Comments

1
by Damien (not verified) :: Wed, 11/24/2010 - 3:34pm

Hi Brian,

I noticed something that on first glance appears counter intuitive, and I was hoping you could provide some insight. I was looking at the SOS rankings and I noticed that Iowa has a lower SOS ranking than Wisconsin. Being a Wisconsin fan, this seemed odd to me because I thought the Hawkeyes had played a harder schedule. So I looked through their opponents FEI rankings.

On first glance, Iowa's non conference schedule of Arizona(#34), Iowa State(#85), and Ball State(#108) seems a lot tougher than Wisconsin's Arizona Sate(#41), San Jose State(#115), and UNLV(#116).

The same thing would also seem to be true of their respective conference schedules. The only difference between the two teams has been that Iowa has played Penn State(#50) and Northwestern(#55), instead of Minnesota(#81) and Purdue(#88) who Wisconsin played.

So I was pretty surprised to see that Iowa's SOS(#51) is lower than Wisconsin's(#44). The only thing I can think of that would cause this to be true would be a pretty severe home/away adjustment. Iowa played all three of Wisconsin, Michigan State and Ohio State at home, instead of Wisconsin playing Michigan State and Iowa on the road. Is this what's causing the difference? Sorry if it seems like I'm nitpicking, it just seemed odd to me, and I want to understand the difference

5
by Brian Fremeau :: Wed, 11/24/2010 - 9:58pm

Thanks for the question. The home/away factor is definitely a part of it. The following games are a wash since home/away was the same for both teams:

at Michigan
vs. Ohio State

The following games are effectively a wash because an elite team would have a near-100% win likelihood home or away:

Iowa at Indiana / Wisconsin vs. Indiana
Iowa at Northwestern / Wisconsin vs. Northwestern
Iowa at Minnesota / Wisconsin vs. Minnesota

These non-conference games are a wash as well for the same reason:
Iowa vs. Iowa State and vs. Ball State
Wisconsin at UNLV and vs. San Jose State

Here are the individual game win likelihoods for an elite team against the schedules for the remaining Iowa and Wisconsin games:

IOWA
.809 at Arizona
.977 vs. Penn State
.732 vs. Wisconsin
.853 vs. Michigan State

WISCONSIN
.952 vs. Arizona State
.712 at Michigan State
.646 at Iowa
.985 at Purdue

2
by Will :: Wed, 11/24/2010 - 5:50pm

Virginia Tech, #3. I believe this is the year that throwing out FCS results breaks FEI.

Will

4
by Kal :: Wed, 11/24/2010 - 9:52pm

Ya know, I would have said the same. The fact of the matter is that VTech has recovered nicely from that abysmal game and played well the rest of the season - and consistently so. They lost to BSU by 3 then proceeded to do well the rest of their way. FEI rarely cares about who won or lost (that's the point of it after all) so it kind of makes sense to ignore it.

That being said, I'd be happy to play VTech over oregon or Stanford if I were going for a NC bid :)

8
by Will :: Thu, 11/25/2010 - 4:13am

James Madison isn't even a good FCS school - they are 3-5 in their own conference.

I agree that Virginia Tech is a very good team, but no elite team would ever lose to James Madison. FEI is ranking VT as an elite team, and it wouldn't if this game were taken in to account properly.

Total side note on just how remarkable of a game that was, James Madison only managed to score more than 20 points three times this season - against Morehead State, William & Mary, and Virginia Tech.

Will

6
by Brian Fremeau :: Wed, 11/24/2010 - 10:06pm

I'll definitely take a look at it in the off season, but I'm not sure what conclusions I'll reach. As it stands right now, I'm less inclined to believe that Virginia Tech would be more "accurately" ranked in the 15-25 range than I am to believe they are a legitimate top-10 team that somehow lost a game they had no business losing. The circumstances were unique (they played JMU less than five days after the Boise State loss), and Virginia Tech has been a slow starter for much of the year. But still.

7
by Kal :: Wed, 11/24/2010 - 10:19pm

What I'd worry about is the times when that would give an otherwise accurate depiction of the team. In this case it worked out, but in other cases I suspect it wouldn't have; playing a close FCS game might be a bad indicator (or at least back up other information). Plus it affects all the teams that VTech has played, making them all look artificially better than they were.

As you said, 5 days rest + a tough game against BSU are both mitigating factors, but as it stands they look exactly the same as if VTech had won 100-0. It might be the case that simply looking at really skewed results of any flavor is not going to give any predictive value, so throwing out games where the ratings are absurdly imbalanced might give better predictive value. For instance, Boise State looks like a juggernaut - but they've had a stupidly easy schedule and tromped all over it. In their competitive games they've looked just ok. Does that mean they're awesome because they stomped bad teams?

Or does it mean that playing a FCS team or a bad team from FBS have equally non-predictive value?

3
by mm (not verified) :: Wed, 11/24/2010 - 9:28pm

So MSU's close loss to Auburn is given a 'low' relevance because Auburn is rated so much higher than MSU?

The fun thing about this week is we can see the odds FEI gives each school winning this week (for those teams ending their season this week). Auburn and Alabama each have about a 50% chance of winning, Arkansas has about 60% chance of defeating LSU, while Virginia Tech is around 100% chance of beating Virginia.

9
by john123456 (not verified) :: Sat, 12/08/2012 - 3:16am

Examsmafia.com provides its customers genuine and quality study material. The study material for the certification 70-498 exam is thoroughly researched and prepared by IT industry experts who have deep knowledge and exposure of the candidates’ needs and problems. The 70-498 study material is also being constantly updated and amendments are being made whenever required in order to keep the content valid and relevant. Thus there is no question of authenticity with the 70-498 exam material provided by http://www.examsmafia.com/70-498.html. Our 70-498 study material is to the point, interactive and very easy in learning. Thus candidates from varying academic backgrounds find it the best and easy way of preparation for their certification 70-498 exams.