Writers of Pro Football Prospectus 2008

Most Recent FO Features


» 2018 Free Agency Cost-Benefit Analysis

Is Kirk Cousins the best free-agent quarterback in recent memory? Should Trumaine Johnson or Malcolm Butler have gotten the larger contract? And what makes a free-agent contract good or bad, anyway?

23 Sep 2009

Week 3 FEI Ratings

by Brian Fremeau

In the aftermath of Florida’s less-than-epic 23-13 victory over Tennessee Saturday, head coach Urban Meyer was asked about why his vaunted Gators squad didn’t obliterate the Volunteers. Meyer offered two explanations: flu symptoms kept key Florida players off the field, and Tennessee didn’t play to win; rather, they "shortened the game" and played not to lose too badly. Tennessee head coach Lane Kiffin responded a day later with a sarcastic comment about using the flu as an excuse the next time his team didn’t meet expectations. And darn it if we won’t be hearing those sound bites ad nauseam during the build-up to next year’s game.

Cattiness aside, is there something to be said for the "shorten the game" hypothesis? Do fewer game possessions help weaker teams compete more effectively against stronger teams, or are these teams just playing to lose more modestly?

Over the last six seasons, the average FBS vs. FBS game has been contested over the course of approximately 26 total game possessions, 13 for each team. This number includes all offensive versus defensive series, including clock-kill drives as well as special teams and defensive scores and turnovers. Even though plays from scrimmage were not run on such possessions, the given team did possess an opportunity to score.

Figure 1 illustrates the distribution of total possessions per game in the more than 4000 FBS games played since 2003, plotted in blue. The red plot represents games in which an inferior team defeated a superior team, according to end-of-year FEI ratings (21.8 percent of all games played). The green plot represents "upset" games in which the winning team overcame an FEI rating deficit of more than 0.075 -- approximately 0.5 standard deviations of FEI, or a ranking deficit of around 20 to 25 spots (8.5 percent of all games). The consistent bell curve distributions suggest that pace of game does not appear to have a statistically significant relationship with pulling off an upset.

"Shortening the game" by limiting game possessions can seem to make a game more competitive if margin of victory is the only measure considered. The average margin of victory in games with 20 or fewer total game possessions is only about 11 points per game, nearly a full touchdown less than average margin of victory in other games. It stands to reason that fewer possessions provide fewer opportunities for winning teams to run up the score. But in terms of Game Efficiency (GE) -- the tempo-free measure of a team’s possession-by-possession success, and the foundation of FEI -- shorter games aren’t more competitive than longer ones. In fact, they are less competitive. The average GE in games with 20 or fewer possessions is 0.21, nearly three percent greater than the average GE in other games.

As for the teams themselves, shortening a game can potentially disrupt the opponent’s comfort and rhythm on offense, but certainly only if that team doesn’t actually prefer a shortened game. Since 2003, 13 percent of Tennessee’s games have had 20 or fewer total game possessions; Florida has played 12 percent of its games at that pace. The Volunteers and Gators each rank among the 20 FBS teams that most often play such "shortened" games. Their records in low possession games -- 6-4 for Tennessee, 7-2 for Florida -- are similar to their overall records in the same span. Against one another, four of the last five games have featured 20 or fewer game possessions, and Florida has won each with varying Game Efficiency results:

Year Winning Team Losing Team Home Team PF PA Poss GE
2003 Tennessee Florida Florida 24 10 27 0.154
2004 Tennessee Florida Tennessee 30 28 21 0.027
2005 Florida Tennessee Florida 16 7 20 0.129
2006 Florida Tennessee Tennessee 21 20 20 0.014
2007 Florida Tennessee Florida 59 20 28 0.377
2008 Florida Tennessee Tennessee 30 6 19 0.457
2009 Florida Tennessee Florida 23 13 18 0.179

The Game Efficiency difference between the 2008 and 2009 Florida victories is significant despite the tempo of each game being similar. Had Tim Tebow scored instead of fumbling at the Tennessee 2-yard line early in the fourth quarter, the GE may have been nearly identical. As it was, the 10-point victory wasn’t flashy, but the outcome was never in doubt, and the Gators produced the third most-efficient victory in the series since 2003.

Week 3 FEI Top 25

The principles of the Fremeau Efficiency Index (FEI) can be found here. Like DVOA, FEI rewards playing well against good teams, win or lose, and punishes losing to poor teams more harshly than it rewards defeating poor teams. Unlike DVOA, it 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. Like DVOA, it represents a team's efficiency value over average. Strength of Schedule (SOS) is calculated as the likelihood that an elite team would win every game on the given team's schedule.

Only games between FBS teams are considered in the FEI calculations. Since limited data is available at the beginning of the season, the ratings to date are a function of both actual games played and projected outcomes based on the 2009 Projected FEI Ratings. The weight given to projected outcomes will be reduced each week until mid-October, at which point the projections will be eliminated entirely.

Rank Team FBS Record FEI Last Week GE GE Rank SOS SOS Rank
1 Florida 2-0 0.264 1 0.363 1 0.326 58
2 Texas 3-0 0.228 3 0.305 2 0.334 59
3 USC 2-1 0.211 2 0.205 13 0.202 15
4 Ohio State 2-1 0.205 4 0.224 9 0.283 46
5 Auburn 3-0 0.199 16 0.232 7 0.260 35
6 West Virginia 1-1 0.196 6 0.201 14 0.366 68
7 Miami 2-0 0.185 28 0.068 40 0.162 5
8 Florida State 1-1 0.183 30 0.098 30 0.094 1
9 Oklahoma 1-1 0.182 7 0.227 8 0.132 3
10 Alabama 3-0 0.177 5 0.253 5 0.289 49
11 Clemson 2-1 0.171 14 0.158 24 0.244 25
12 Virginia Tech 2-1 0.169 8 0.168 23 0.248 28
Rank Team FBS Record FEI Last Week GE GE Rank SOS SOS Rank
13 LSU 3-0 0.155 9 0.172 21 0.211 19
14 Boise State 3-0 0.155 27 0.236 6 0.671 118
15 Iowa 2-0 0.154 19 0.183 16 0.256 32
16 BYU 2-1 0.145 10 0.209 10 0.371 71
17 Michigan 3-0 0.140 12 0.207 11 0.288 48
18 Notre Dame 2-1 0.137 15 0.180 18 0.290 50
19 Georgia Tech 1-1 0.136 11 0.036 49 0.192 11
20 Texas Tech 1-1 0.130 25 0.206 12 0.263 38
21 Georgia 2-1 0.126 17 0.072 38 0.127 2
22 Penn State 3-0 0.126 13 0.262 3 0.356 65
23 California 2-0 0.124 21 0.141 26 0.351 64
24 Pittsburgh 2-0 0.118 33 0.105 28 0.373 72
25 Utah 2-1 0.112 31 0.173 20 0.408 77

Ratings for all 120 FBS teams can be found here.

Posted by: Brian Fremeau on 23 Sep 2009

42 comments, Last at 25 Sep 2009, 3:15pm by beargoggles


by bird jam :: Wed, 09/23/2009 - 3:33pm

Re: Urban Cryer's post-game whining, his suggestion that Tennessee wasn't playing to win the game is preposterous and simply sour grapes from a coach whose team didn't win by as much as he wanted them to.

The sad and surprising thing is that the national media seems to be lapping it up. People are criticizing Kiffin for not going no-huddle or "opening the offense up." What these people are ignoring is that Jonathan Crompton is Tennessee's quarterback. The Vols were moving the ball fine with the gameplan they were using. On the couple of plays where they abandoned that strategy and threw downfield, Crompton was picked off.

It is easy with the benefit of hindsight to say that Tennessee shoul dhave done something different. But had they gone pass-wacky int he second half, that game would have turned into the bloodletting that everyone predicted going in. Kiffin's gameplan gave the Vols their best chance to win the game. It didn't result in a victory, but the Vols acquitted themselves well and surprised everyone.

by Kibbles :: Wed, 09/23/2009 - 4:32pm

Meyer was right, Kiffin wasn't playing for the win. Not because Kiffin wasn't passing, but because Kiffin wasn't HURRYING. On Tennessee's opening drive, down 3-0, they averaged 35 seconds off the clock per play run. On Tennessee's penultimate drive, down 23-6 (three scores), they averaged... 32 seconds off the clock per play run. On Tennessee's last drive, down 23-13 with 6 minutes left in the game, they averaged... 33 seconds off the clock per play run before Crompton's incompletion and interception. Even if Crompton had converted the 4th down, Tennessee would have been at around the Florida 40 yard line, down two scores, with under two minutes to go. Tennessee burned 4 minutes off the clock travelling 30 yards on that last drive while down 2 scores. Please, explain to me how that gives Tennessee the best chance of winning? It might give Tennessee the best chance of scoring once, but in order to win they have to score TWICE, and it seems to me that running out the entire clock en route to a single score is not a good way to score twice. It's a pretty good way to keep the margin of defeat down, though.

I have no problem with the suggestion that Kiffin thought running gave his team the best chance of winning the game. Urban Meyer wasn't criticizing Kiffin for calling runs, though... he criticized Kiffin for not speeding up the game. If time is against you and you're going to call a bunch of runs, then tell your kids to hurry up to the line and stop milking the entire play clock for every play. Tennessee played at the exact same pace down 3 scores with 8 minutes left in the game as they did down 3 points with 55 minutes left in the game. That's not playing to win, that's playing to minimize the beatdown.

When the Miami Dolphins milked the entire playclock while trying to mount a comeback against Indy, they got *LAMBASTED* for terrible clock management... and rightly so. Yet when Urban Meyer points out that Lane Kiffin was doing the exact same thing, everyone is quick to rush to Kiffin's defense? Huh? If anything, Kiffen's offense was worse. Miami let the playclock run down while down a score with 3 and a half minutes to play. Kiffin let 35 seconds run off the clock while down TWO scores with UNDER THREE MINUTES TO PLAY. "Playing to win" my hindquarters.

by bird jam :: Wed, 09/23/2009 - 4:40pm

"Meyer wasn't criticizing Kiffin for calling runs, though... "

Oh, really?

"When I saw them start handing the ball off, I didn't feel like they were going after the win," Meyer said. (http://sports.espn.go.com/ncf/news/story?id=4492076)

And criticizing the Vols for taking too much time between plays is valid. But Tennessee has a long way to go before they have their act together on offense. As an astute Vol fan commented on another message board (I don't want to take credit for it because it's not mine although I agree 100%):

Because they don’t have a hurry-up offense ready to deploy against a high-caliber defense on the road. Even the best offensive schemes rarely get installed over the first offseason (Auburn is a very curious exception so far), and getting your base offense down is more important than the situational material. We still have problems in the passing game that need to be worked out before we can nail down the hurry-up.

Remember that Kiffin was still in the huddle calling plays through much of our fall scrimmages. Do you honestly think that the last two weeks of fall camp plus the first three weeks of the season is enough to get from Kiffin in the huddle to an effective no-huddle? 8 interceptions into the season, I’m guessing the answer is no.

Short answer: Tennessee didn’t go no-huddle because Tennessee can’t go no-huddle. A slow, effective offense is still a better idea than a fast train wreck. As regards the article, the main point was that the David could try to increase the potential for gain at cost of increasing the potential for cost. So now it becomes a risk-reward analysis. How plausible were the rewards for going no-huddle, pass-wacky for Tennessee when receivers are messing up routes and the qb is a Florida-proclaimed catfish? How does that compare to the risk?

I get the concept, but it involves a few assumptions about the offense that weren’t necessarily valid. After all, if Tennessee’s supposed best receiver can’t keep his feet in bounds when open in the end zone, who can be trusted to run their routes right in a no-huddle that’s probably not even implemented yet?

(That comment and more here: http://www.rockytoptalk.com/2009/9/23/1051371/killing-the-playing-not-to... )

As I posted there, I believe I will defer to Kiffin and his staff on decisions involving the personnel and playcalling on his team. Jonathan Crompton scares the crap out of me when he has the full play clock to wrap his brain around what the Vols are trying to do. Going no-huddle and “opening up the offense” with more passing sound like a recipe for disaster.

by Kibbles :: Wed, 09/23/2009 - 11:00pm

So you're saying that Lane Kiffin didn't have enough time this offseason to tell his players "hey, when we need two scores and time is running down, snap the ball with more than 3 seconds left on the play clock!". Lane Kiffin didn't have enough time to install runs to the outside? Lane Kiffin didn't have enough time to tell his players to try to get out of bounds? Sounds to me like Lane Kiffin is the worst coach ever.

Look, I get that Tennessee couldn't go no huddle, and they couldn't just let Crompton loose... but you can't tell me that a coach was doing everything in his power to win the game when he is literally milking every second off the play clock when down two scores late. By the time the 4th quarter rolled around, Tennessee wasn't making the least effort to win the game. Period. And if Kiffin was content to just leave town with a less embarrassing loss, then why shouldn't Urban Meyer oblige and grant him his wish?

by vcn (not verified) :: Thu, 09/24/2009 - 12:16am

I didn't get to watch much of UT-Florida, but if the last week's game was any indication, receivers were still having problems running the correct routes (or Crompton was looking in the wrong place for them). Who knows what chaos could come out of a no-huddle or even a rushed offense.

Is that a coaching problem? Sure. I'm waiting to see whether it gets better with time. (Incidentally, the Miami example is a bad one--Miami was certainly NOT playing to reduce the eventual margin of the Colts victory; they were playing to win.)

by ChaosOnion (not verified) :: Thu, 09/24/2009 - 11:31am

No huddle and running the play clock down before the snap are to ends of a continuum. There are any number of things they could have done to speed up their possessions. Snapping the ball with 12 seconds left on the play clock instead of 2 does not require the installation of a complicated, no-huddle scheme. Shave 7-8 seconds off per possession and you add a minute back to the game clock on an 8 play drive.

I can understand shortening the game and lengthening offensive possessions to keep the Florida offense off the field. It is a good strategy and good clock management. But when you are down by two scores, you want to keep some time on the clock and get between the huddle and the line efficiently. That is also good clock management. Down by 2 scores, running the play clock down is /bad/ clock management. I do not think he was playing not to lose, I think his overall strategy was to keep it close and expect Florida to make mistakes under the pressure. He underestimated Florida's ability to hold a lead in a close game. In the fourth quarter when TENN was down by two scores, his play calling should have remained the same, but he should not have been running out the play clock. That was poor clock management.

by bird jam :: Thu, 09/24/2009 - 12:44pm

Poor clock management is one thing. Deliberately losing the game is another. I am fine admitting the former for that one drive. I take offense (on behalf of my coach and team) to accusations of the latter.

by Fourth :: Wed, 09/23/2009 - 4:45pm

Watching the game live, I was actually stunned by Tennessee's lack of urgency in the 4th quarter. I am not suggesting that anyone should go pass-wacky with Crompton at the helm, but I sat there and watched as the Vols, down 10 with 6 minutes to play in the game, took about 3 minutes to move the ball 30 yards. This wasn't just because they were running every down. They weren't bothering to try and run to the sidelines to get out of bounds, and there was NO hurry-up whatsoever. They simply huddled after every play and snapped the ball with single digits on the playclock while the gameclock kept on rolling. The gameplan to that point had actually worked: you've got a chance in the 4th quarter, however small, to pull off a miracle by beating Florida. But the lack of any semblance of a hurry-up offense was a shocking display of poor coaching on Kiffin's part, something that hopefully he can improve upon in the future.

By the way, never heard anyone use Urban Cryer before. Clever. For more clever, http://www.everydayshouldbesaturday.com/wp-content/uploads/2009/01/hello...

by Fourth :: Wed, 09/23/2009 - 4:47pm

Nevermind, Kibbles said it.

by bird jam :: Wed, 09/23/2009 - 4:48pm

And my response to Kibbles' comment addresses yours as well.

by Fourth :: Wed, 09/23/2009 - 5:29pm

Well I read your reply to Kibbles, and my only problem with it is that you seem to be assuming that no huddle = pass wacky. That is not the case. Why not call the next 2 plays (both runs!) in the huddle, and only huddle up half the time? Further, why not make that second call (at least the majority of the time) a sweep or a toss out into the flat so the runner can get out of bounds and you can huddle with the clock stopped? Why not at least HUSTLE from the huddle to the line of scrimmage and save 3 or 4 seconds every play? Why not call a play at the line with your senior qb? Does his lack of arm strength, accuracy, and decision-making in the pocket also mean he can't make a read pre-snap? (That last question is only half-rhetorical).

I'm sure you will have a defense for most of these questions as you seemed very passionate in your defense of your team before, and that's fine because really that's what makes college football fun. But to convince a neutral party or heaven forbid a Florida fan is another thing entirely. In 10 or 15 years of watching college and pro football I'm pretty sure I've never seen anything like the -stunning- lack of urgency by Tennessee in the 4th quarter last week. I emphasize stunning because it's not rhetoric. My friends and I were just pointing at the field looking at each other, not finding the words to express how surreal it was to see what was happening with the Tennessee offense. Yes, Kiffin knows more football than I do, and he certainly knows his team better than I do. But I've never seen anything like that before.

by bird jam :: Wed, 09/23/2009 - 5:43pm

Crompton's deficiencies are not physical at all. He has a pretty good arm. His problems are all mental. So to answer your questions, no - he can't make a pre-snap read. He can't be trusted to call plays at the line. Before the season began, Crompton praised Kiffin's (and Chaney's) system because it didn't require him to think at all. People didn't really pay much attention to that comment, but it spoke volumes about what Crompton is capable (and incapable) of.

And any play on which Crompton's primary receiver is not wide open (so basically every passing play after the Western Kentucky game) is going to be an adventure.

I stand by my position that Kiffin was making chicken salad from chicken *&^$^%. I agree that there should have been more urgency between plays with the clock running. But I don't have a problem with any of the rest of what the Vols did in the second half on Saturday. And to suggest that they weren't trying to win the game is insulting. And ridiculous. And frankly, it doesn't even make sense. On that last drive, they were guaranteed to cover the huge spread - what's the difference between losing by 10 and losing by 17? What would the point even be of not trying to win the game in that scenario?

by Fourth :: Wed, 09/23/2009 - 6:12pm

Your last question touches on the original subject, which is the thing we are never going to agree on. My answer as a Florida fan is that you weren't trying to win because "winning" was never in the gameplan, and Urban was (mostly) right. It's interesting that you would phrase the question in such a way, because if we were having this conversation in person my response would probably be, "Exactly!" The logic goes like this:

A. A good, rational coach prepares his team for close games and has a plan for every late-game situation that maximizes his team's chances of winning the game.

B. On Saturday, Tennessee did not implement a strategy in the 4th quarter that maximized its chances of winning the game (regardless of how bad or stupid the quarterback is).

C. Tennessee's coach is either not good enough to coach a major program (because he can't or won't teach the simple concept of hurrying to the line of scrimmage, calling two plays in the huddle, etc), or he was being irrational (not trying to win). Honestly the latter is more plausible. It's almost as if he himself didn't expect it to be close late, and didn't devote any significant practice time for such an event in the week leading up to the game.

Anyway that's just the way I see it through orange and blue glasses. At least we know the rivalry won't die as long as Kiffin is around.

by bird jam :: Wed, 09/23/2009 - 6:22pm

Ha ha - it's true. Orange and blue glasses versus orange and white.

I have to admit that I do really enjoy having a coach that inspires such hatred (after only 9 months on the job!) that Florida fans would honestly believe that he would deliberately lose a game in which a win would have made him an instant and eternal hero. This must be how you guys felt when you had Spurrier.

by DEW (not verified) :: Wed, 09/23/2009 - 10:09pm

Speaking as a fan of neither team...nor anyone in the SEC, for that matter...I'd have to say you folks are talking around the telling point, which is this:

Tennessee doesn't have the personnel to beat Florida. Period.

Without some extremely fluky random bounces, there was no chance for Tennessee to win the game. It doesn't matter what strategy Kiffin picked; by the time the drives you're discussing rolled around, the game was all but settled because those flukes hadn't occurred.

Kiffin may be a mouth, but he knows this. You can bet that he knew his quarterback lacked the ability to run an effective downfield passing game and an effective no-huddle office. He may be a good coach or he may be an awful coach, but regardless of his skill level as a coach, he can't overcome the personnel difference in the guys on the field.

By coaching to keep the game close on the scoreboard, he appeals to the random Tennessee alum--the guys who kick in cash to support the football program (and who, if they're not happy, push for him to get canned). He appeals to the random Tennessee fan. He might even appeal to the team itself ('cause losing by 10 feels a lot better than making a lot of embarrassing mistakes while losing by 30).

Mind you, he dug the hole himself by running his mouth before the game. But lambasting him for "not coaching to win the game" assumes that "winning the game" is a possibility in the first place, which by the second half, it basically wasn't.

by Kibbles :: Wed, 09/23/2009 - 11:06pm

I have no problem with the fact that Lane Kiffin wasn't trying to win the game. I think that, in his situation, a 75% chance at a less-embarrassing-than-expected loss would do a lot more good for his program than a 5% chance at a program-defining upset. I think that playing for the merciful loss was a perfectly acceptable coaching strategy.

I have a problem with people getting upset at Urban Meyer for calling a spade a spade. Kiffin was perfectly content leaving Gainesville with a "moral victory" rather than trying for an actual victory. Don't hate Urban Meyer for recognizing that fact and granting him his wish.

by bird jam :: Thu, 09/24/2009 - 8:27am

So I guess when Kiffin called that timeout on Florida's last drive to stop the clock on third and long (with just over a minute remaining,) he was... um, trying to give Florida more time to design a good play?

How does that timeout (which was Tennessee's final timeout) fit into the whole “Kiffin wasn’t trying to win” argument? Wouldn’t he have let the clock keep running there if he were really just trying to get out of town with a 10-point loss?

by hrudey (not verified) :: Thu, 09/24/2009 - 1:51pm

Perhaps he hoped to get a stop and a punt return for a score to lose even more impressively.

I don't think anyone's saying Kiffin was trying to lose, but they certainly were looking to avoid losing big. If you watch their fourth quarter offense and didn't know the score, you'd be much more likely to believe that they were up 10 rather than down 10.

But I'll give credit to Kiffin for putting together a gameplan that reduced the magnitude of their eventual and inevitable loss. I for one hope he continues to delight Tennessee fans with more respectable losses for many years to come.

by jayinalaska :: Wed, 09/23/2009 - 3:37pm

Was the weighting factor for each week of the season ever published? It seems a little strange when the FEI number 2 team (USC) loses to the FEI number 91 team (UW) that USC falls only to number 3. (UW rose all the way to 70. Go Dawgs!) This would suggest to me that the projection is really holding USC up right now. Not that I think USC's ranking should have fallen off the face of the Earth, it just seems strange their ranking isn't closer to, say, 10. --Jay

by Will :: Wed, 09/23/2009 - 7:36pm

Don't put too much faith into these (or any computer) rankings until fairly late in the season.


by jayinalaska :: Wed, 09/23/2009 - 8:23pm

I have no faith invested in the ratings at all, now or at the end of the season.

I was just trying to reconcile in my mind why USC only fell one spot from 2 to 3 after losing to a UW team that I can only conclude FEI currently considers "poor". Right above the comment section each week is the following text: "FEI rewards playing well against good teams, win or lose, and punishes losing to poor teams more harshly than it rewards defeating poor teams." I went back and looked at USC's schedule so far: crushed cream-puff San Jose State (a "poor" team), narrowly beat on the road a "good" tOSU team, narrowly lost on the road to a "poor" UW team.

As I said before, it must be the projected rankings that are keeping USC so high in this week's rankings. It will be interesting to see how the USC ranking changes as the projected ratings weight diminishes. And, I am curious what those weightings are week-to-week until they've gone to zero. Over the next 4 weeks, USC hosts WSU (a "poor" team), goes to Cal (currently a "good" team), has a bye, then goes to ND (currently a "good" team). Of course, much depends on how SJSU, tOSU, and UW continue to play as well as the play of USC's upcoming opponents. --Jay

by Brian Fremeau :: Wed, 09/23/2009 - 9:32pm

Since every team hasn't played the same number of FBS games, the weighting factor of projected data and actual data isn't rigid. The projections carry approximately 50% of the weight for each team's rating, though, which is the primary reason USC hasn't fallen far.

Their rating did take a significant dip, even if their ordinal ranking did not. If Washington keeps winning, the Trojans probably won't drop too swiftly in the coming weeks, but I would expect at least somewhat of a weekly dip going forward.

by Muldrake (not verified) :: Wed, 09/23/2009 - 11:59pm

I agree that the formula will probably balance out in the long run but I think it's also pretty clear that the projected FEI is taken too much into account at this part of the season. According to the predictions that were published in last week's 7th Day Adventure the formula went a brutal 1-9 this last week. It would seem to me that the only point in including the FEI projections in the formula would be to allow some predictions to be made about the upcoming week...otherwise what would it matter whether or not all the teams played the same number of FBS games? Sure the poll would fluctuate wildly but that makes as much sense as USC being ranked 60+ spots ahead of a team that it just lost to.

by jayinalaska :: Thu, 09/24/2009 - 8:07pm


Thanks for taking the time to respond. Your point about looking at the actual FEI value changes versus the ranking changes is taken. --Jay

by The PAC (not verified) :: Wed, 09/23/2009 - 3:43pm

And FBS means? I assume it means Football Stats. You might want to include a link that explains that otherwise. This is my first time here and its hard enough to figure out your acronyms. DVOA, FIA, GA. I want to learn, but give a brother a break.

The PAC!

by bird jam :: Wed, 09/23/2009 - 3:45pm

It's not a Football Outsiders acronym - it's Football Bowl Subdivision, the NCAA's term for what is really just Division I.

by sam :: Wed, 09/23/2009 - 4:21pm

Not quite. It's the old Division I-A (Division I-AA became the FCS--Football Championship Subdivision--and has a playoff system I assume like Division II and Division III).

Division I still has two subdivisions , they just have names instead of letters.

sam! or the original sam from the old FO

by bird jam :: Wed, 09/23/2009 - 4:32pm

Well, although Division I technically included I-AA, most people referred to major college football (i.e., Division I-A) as "Division I" in common parlance.

by zlionsfan :: Wed, 09/23/2009 - 11:44pm

Prior to 1978, there was no split between I-A and I-AA, just a single Division I. I didn't know it had occurred that recently (I was 10 at the time), I thought it had been split for longer than that, but it does mean that there was a time when there was no I-AA, and thus referring to Division I football would have made more sense (although it would not have been accurate to mean major-college football then either, because the teams that became part of the FCS were still Division I teams at the time).

Saying DI and meaning I-A was/is not accurate, regardless of the number of people who did it, and if we're talking about a glossary for people who are new/newer to college football, we should be referring to the splits correctly.

And yes, sam, I-AA/FCS has postseason playoffs like DII and DIII. Sixteen-team single-elimination tournament (although next year it seems to be expanding to twenty teams), held at campus sites up until the finals.

by YourNationalChampionHuskies (not verified) :: Wed, 09/23/2009 - 4:03pm

At least Pitt is ranked somewhere.

by Tom Gower :: Wed, 09/23/2009 - 4:16pm

Does Game Efficiency correlate better to score in games with more possessions? The normal explanation for why an underdog wants a shorter game is it's easier for a lucky or fluke occurrence to affect the game in their favor. For example, Florida in 50 red zone possessions may score 35 TDs, 10 FGs, and 5 times come away with 0 points (say, by turnover). In a single game, though, UF may come away with a TD, an FG, and a turnover (say, a fumble like the one Tebow had) in 3 red zone possessions. If, say, Tebow's fumble had been returned for a touchdown, you could be looking at a 23-20 game with very little difference in Game Efficiency and the Vols could have been driving for a game-tying or -winning score. That would seem to validate the "shorten the game" underdog strategy.

by peachy (not verified) :: Wed, 09/23/2009 - 7:21pm

There was an excellent post a while back at Smart Football which covered underdog strategies - I believe the gist was that a running-heavy 'shorten the game' approach can work, but it relies on the favourite making a sufficiency of serious mistakes to offset their superiority. If those don't occur, then a true underdog (ie, a team that really is inferior on the field) has little hope of winning a straight-up contest. (The other approach is to play aggressively, and try to force mistakes, accepting a higher risk of being blown out for a higher chance of winning.)

by Brian Fremeau :: Wed, 09/23/2009 - 9:36pm

Agreed on the opportunity for a break or two turning a low-possession 10 point game into an upset alert. But reciprocally, the same game can turn into "dominant" efficiency game, too -- the UF/UT 2008 game outcome is the opposite of your example, and without running data, I would expect that outcome to occur more frequently than not.

by peachy (not verified) :: Thu, 09/24/2009 - 1:05am

I should add that the actual fumble occurred at 23-6 - Tennessee scored on the ensuing drive, so a return would just have sped things up a bit. Tennessee had possession once at 23-13... that would be the drive that is being discussed up-thread.

by The Ninjalectual :: Wed, 09/23/2009 - 4:48pm

Colorado State is not the 103rd best team in college football. This is the stupidest rating system I have seen, if that is what it says. What's the point? If all you can do well are the top 25, why even list all 120 teams?

"Just look at that pumpkin."
-John Madden, looking at the moon.

by Fourth :: Wed, 09/23/2009 - 6:17pm

I can't decide if you are being funny or serious, because your comment looks a lot like this...

(team) is clearly ranked too high/too low because (reason unrelated to DVOA). (subjective ranking system( is way better than this. (unrelated team-supporting or -denigrating comment, preferably with poor spelling and/or chat-acceptable spelling)

...but not quite.

by Will :: Wed, 09/23/2009 - 8:01pm

Don't put too much faith into these (or any computer) rankings until fairly late in the season.


by zlionsfan :: Wed, 09/23/2009 - 11:55pm

Well, they've beaten Colorado by six on the road, FCS Weber State by one at home, and winless Nevada at home. They may not turn out as low as they are displayed now, but I don't know that they've necessarily done much to stake a claim any higher based on what they've done this season.

by ChiTown11111 (not verified) :: Wed, 09/23/2009 - 8:29pm

I know that in basketball (High School) shortening the game is done alot. There is no shot clock and teams facing better opponents just hold the ball and pass until they get an excellent opportunity. This is different than football of course because you don't get infinite plays to achieve the goal in football.

by Mr Shush :: Fri, 09/25/2009 - 12:00pm

Certainly in soccer the equivalent behaviour is absolutely standard practice. Inferior teams nearly always adopt an extremely negative strategy (10 men behind the ball, attacking almost exclusively on the counter). Jose Mourinho once referred to it as "parking the team bus in front of the goal". Obviously it doesn't shorten the game, but it is analogously an attempt to keep the score low and hope for a freak event or two to give them a win or draw. In soccer, I think almost no-one seriously doubts that it's the optimal strategy for a severely outclassed team to adopt. That doesn't necessarily mean it is in football, of course.

by ninerbeliever (not verified) :: Thu, 09/24/2009 - 2:05pm

alright I will be one of those guys, but I got ot ask why has Cal fallen from their preseason ranking? The preseason ranking was .154 while they have fallen to .124, so actually their preseason rankings is the only thing holding them in the top 25. I won't go bat shit crazy till after the SC game but still very strange.

by beargoggles :: Fri, 09/25/2009 - 3:15pm

FEI didn't make any sense last year, so I don't expect it to make sense after 2-3 games this year.