Statistical Preview: Mississippi State #48

If Mississippi State taught us nothing else in 2012, we learned that schedule matters. The Bulldogs started 7-0. Two wins in the next three games against Alabama, Texas A&M and LSU would have thrust them into national title talks. Instead, they scored 37 in the three games combined while allowing at least 37 in each. They won one of their final six games. The amazing thing is that, knowing what we know now, every outcome was completely predictable - they beat the bad teams and lost to the good ones.

The Mississippi State offense was outstandingly reliable. The Bulldogs were 47th and 57th in EPA (expected points added) per pass and rush, respectively, and about 10 spots better when we adjust for competition. Tyler Russell lived between 3 and 8 defense-independent EPA (that is to say, he was 3 to 8 points better than the average quarterback), scoring in that range in 9 of 12 games against FBS opponents. Only against Northwestern did he have a truly bad performance (-6.6 EPA). The offense was surprisingly bad on third downs - 35.8% conversion, 86th nationally.

On the other side of the ball they were 39th and 66th in EPA per pass and rush allowed, respectively. Adjusting for competition, the run defense was fairly good: 25th in EPA per rush and 13th in both yards per rush and explosive plays per rush. The pass defense was not good. They allowed completions on 61.3% of passes and allowed 25+ yards on 6.3% of passes. But the fault lies with the front seven, not the secondary. They sacked the quarterback on only 4.2% of passes and only 4.0% of passes on passing downs (112th nationally); they picked off 4.4% of passes, so they were more likely to intercept a pass than sack the quarterback.

Thanks mostly to those interceptions, Mississippi State turned over opponents twice as often (19.5% of possessions) as they were turned over themselves (9.7% of possessions).

Projection:
If they can find someone to fill Chad Bumphis' shoes, the offense should tick on in 2013 just as it did in 2012 - good, well-rounded, not spectacular, hopefully improved on 3rd downs. Defensively, the talk is that the line will be a strength and the DC will be more aggressive with the linebackers. If they get to the quarterback more often, Mississippi State has room to make a step forward. I'm not holding my breath. At least this year they don't play Bama, A&M and LSU in succession. That's something.
-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Preview: Michigan State #49

The Michigan State defense was every bit as good as we expected. By opponent adjusted EPA per rush (EPA/rush+), the best available metric for measuring run defense, Michigan State's was second to none. By the same measure their pass defense was 11th. After Michigan State, only Alabama, Florida, Stanford, and LSU finished in the top 11 in both categories. They were also excellent in the red zone, allowing only 3.6 points per try; only Utah State and Notre Dame were better.

On the other hand, the Michigan State defense fell short in a few key areas. They didn't force enough turnovers (12.7% of possessions), didn't get enough out of the turnovers they did force, and they didn't get to the quarterback enough (sacks on 4.5% of pass attempts). They were good but not exceptional on third downs and this was probably directly tied to the sacks per pass - Michigan State was 106th in the country in sacks per pass on passing downs.

The offense was uninspiring: 1.5 points per possession, 52.5% completions, 5.3 yards per pass.

Projection:
Michigan State was as close to 12-2 as any 7-6 team in history. They lost five games by 13 total points. The Spartans were a handful of timely turnovers from being grouped with Stanford and Notre Dame this off-season. Looking forward, where do they make up those 13 points in 2013?

There's no reason to expect anything less (or all that different) from the defense in 2013 than what we saw in 2012, and that's fine. The defense was good enough in 2012 to carry a national title contender. But there's little room for them to get better. And so the onus falls on the offense, and that's a problem. The offensive line will be better, but the skill positions are a mess. As with 2012, the 2013 season will depend on a series of coin flips; you never want to go in against a Sicilian when death is on the line, but I'll bet against Dantonio when a football game is on the line every time.
-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Miami #50

Miami did two things well on offense - they got a lot of big plays and not a lot of really bad plays. They knocked off 25 or more yards on 5.7% of plays and were one of 12 teams in the top 30 nationally in both explosive plays per rush and per pass (opponent adjusted; other teams on the list: Georgia, Alabama, Texas A&M, Michigan, Florida State; teams that did not make the list: Oregon, Baylor, Clemson.)

Miami turned the ball over on 1 in 10 possessions, which is good but short of spectacular, but those turnovers cost them only 38.9 points (just over 3 per game). Morris threw interceptions on only 1.67% of passes (fewer than Teddy Bridgewater, David Fales, Johnny Manziel, Marcus Mariotta, Ryan Nassib, and Landry Jones to name a few) while averaging 13.7 yards per completion (more than Teddy Bridgewater, David Fales, Johnny Manziel, Marcus Mariotta, Ryan Nassib, and Landry Jones to name a few).

A quick strike offense is complemented by a defense that gets off the field in a hurry one way or another. That was not the case with Miami. The Miami defense was 120th in plays per possession, 119th in yards per possession and 115th in average time per possession. They allowed "only" 2.4 points per possession because the offense handed them great field position - 74.2 yards from the goal line, 5th best nationally - and they turned over their opponents on 15% of possessions. Otherwise, the Miami defense was pretty useless.

Projection:
Miami is expecting a better pass rush in 2013. If they could double their sacks/pass from 2.9% to 5.8% (league average) without sacrificing coverage in other areas, that should result in a few more third down stops and shorter drives. That could be huge. The defensive front has added beef, which should help with stopping the run. The good news is they couldn't get much worse.

At a glance, we should have a lot of confidence in the offense. They scored a lot of points in 2012 and most of the same folks are coming back to do it again. But the hype doesn't match reality. Miami was only 46th in points per possession in 2012. They were entirely dependent on big plays and avoiding costly turnovers, both of which are high standard error events (in other words, they can fluctuate a lot from season to season). Otherwise, they were league average or worse.

Most folks see Miami as a borderline top 25 team for 2013. From my perspective, they finished 2012 as a borderline top 50 team. I don't see enough improvement on the defense and too much risk of regression on offense to bump them up 25 spots. In this case, don't be seduced by the explosive offense.
-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: San Diego State #51

San Diego State ran off 2 min 40 sec per possession (14th best nationally), scored 5.6 points per possession in the red zone (9th best) and forced turnovers worth 95.4 points (9th best nationally). The last two helped the nation's 78th best offense by EPA per pass score 2.5 points per possession (31st best in the country). The long average time of possession is a completely meaningless stat, except it kept San Diego State scores low by reducing the number of possessions per team (12.4 possessions per game).

The Aztecs were 6th worst in the country in keeping their quarterback upright. They allowed sacks on 10.2% of pass attempts. To make matters worse, they didn't accompany this with high yards per completion - a high sack rate means the quarterback is holding the ball too long but often in exchange for some big plays downfield. The reason a sack rate is high with good line play is that the quarterback is a legitimate threat to scramble; A&M's sack rate was a "weakness" of the offense, especially considering the quality of the offensive line and Manziel's escapability, because Manziel risked sacks for the opportunity to break a long run. But Katz and Dingwell averaged 4.2 and 2.4 yard per carry, not exactly Manziel, Mariota or Lynch.

The story changes when we split Katz and Dingwell. Katz had a 6 point lead on Dingwell in sacks per pass (14% to 8%, respectively); Katz was sacked only slightly less often than sack specialists Jeff Driskel and Kiehl Frazier (Frazier was sacked twice for every seven completions!). And Katz gained 25+ yards on 10.4% of pass attempts, better than all but 12 FBS quarterbacks. Only Tajh Boyd, AJ McCarron and Aaron Murray attempted more passes and gained 25 or more yards at a higher rate.
-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: West Virginia #52

People didn't play a lot of pass defense when West Virginia was on the field in 2012. Four of the eight most productive performances happened in West Virginia games - West Virginia against Baylor and Baylor, Texas Tech and Oklahoma against West Virginia. The defense was torn apart by explosive plays, allowing 25 or more yards on 10% of pass plays. The offense was unimpressive in yards and explosive plays per pass, but was 3rd nationally in completion percentage.

The truth of the matter is that, outside of big performances against Marshall, Baylor and Kansas, West Virginia was pretty unexceptional on offense. Geno Smith and crew managed more than 4.8 points per possession in three games, but less than 2 points per possession in five others (Oregon and Alabama were held under 2 points per possession once each all season). Fourth in the conference in points per possession, third in EP3, fourth in EPA per pass. The unit's best quality was avoiding turnovers, and they ranked second in the Big 12 in turnovers per possession.

And the defense was legitimately bad. They finished the season 116th in points per possession allowed, which improves to 76th when we adjust for context and competition. Only Colorado and Army allowed a higher EPA per pass. As noted earlier, they were particularly susceptible to long pass plays. Without the 45 pass plays that went for 25 or more yards, West Virginia would have reduced its EPA per pass from .33 to .05 (which would place them just below the national average). This kind of defense generally leads to bad 3rd down defense as well, where teams converted 45% of the time against the Mountaineers.

Take home message is if you want to get your quarterback in the Heisman discussion, schedule Baylor and Marshall early and often.
-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Ball State #53

Ball State did face some offensive juggernauts, but with directional Michigan and Army on the schedule there's no excuse for allowing 2.61 points per possession. The rest of the team did everything they could to help out. Opponents started 74.3 yards from the end zone (3rd best nationally). And the defense turned over opponents on only 7.7% of possessions.

Ball State had the least explosive competent offense in the country. The necessary corollary of that is that they were among the nation's best at not getting caught in the backfield - if you're not going to be explosive you have to make sure you keep moving forward.
-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Louisiana-Lafayette #54

Louisiana-Lafayette has the unique distinction of being the only team in the bottom 15 in plays per possession (5.4) while also being in the top half nationally in points per possession (25th, 2.61 points per possession). And they did this despite being 97th in average starting field position (71.3 yards from the end zone). In other words, they started a long way from the end zone, they didn't keep the ball very long, and yet they scored more often than most. They did this, in part, with explosive plays: 14.1 yards per completion and 25+ yards on 8.1% of pass plays. More importantly, Louisiana-Lafayette scored 5.79 points per possession in the red zone, second only to Oregon. This proficiency over average added a full 1/3 of a point per possession and was the difference between them and Texas State in points per possession.

The Ragin' Cajuns struggled to stay on the field in part because they converted only 38% of 3rd and 4th down attempts. This number is remarkable for a team that was 7th nationally in yards per pass. The second worst team in the top 10 in yards per pass in terms of third down conversions is 30 spots better (36th versus 66th). One part of the explanation may be that Louisiana-Lafayette did not have faith in their running game in short yardage situations; they were in the top 40 nationally in pass attempts in short yardage situations and they were stopped for a loss on almost 1/5 of rush attempts. Only three teams in the top 25 in yards per rush were caught in the backfield more often - USC, Wisconsin and Utah State.

Defensively, the Cajuns were unexceptional, except that opponents attempted field goals on only 14% of possessions. Their games were fairly high scoring so teams may have played past field goal opportunities. Teams may have been intimidating by the kick blocking potential. Or, more likely, it was just random chance that opponent's didn't kick a lot of field goals.

-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Washington #55

What do Keith Price and Geno Smith have in common? They both used Baylor to trick people into thinking they were a lot better than they really are. Price didn't quite match Smith's 48 EPA performance against Baylor in 2012, but he did roll up 33.6 expected points added in the 2011 Alamo Bowl, the 5th most productive performance that season. He followed that up with negative EPA performances in eight of 11 games in 2013 and a combined -30 against LSU and Stanford. Instead of the explosive offense we saw against Baylor in 2011, Washington in 2012 was not explosive, they turned the ball over too often, and they didn't excel in any single area.

Ironically, the Washington defense in 2012 came to play. Their EPA/pass allowed was not elite but very good. And the defense was as good at forcing turnovers and defending in the red zone as the offense was bad.

-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Toledo #56

The average FBS team scores 4.8 points per possession in the red zone. Toledo in 2012 scored 4. Because Toledo reached the red zone in 33% of possessions, poor performance in the red zone (versus average) cost Toledo (4.8-4.0)/3=.27 points per possession. This difference would have boosted the Rockets from 63rd nationally in points per possession to 34th, just head of Tennessee.

Defensively, Toledo compensated for generally poor play with 95.6 points off turnovers and by allowing only 4.1 points per possession in the red zone.

____________________

The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Iowa State #57

Iowa State had a top 40 defense. They were 38th in points per games allowed and 33rd in points per possession allowed. And they did this despite being among the nation's worst in sacks per pass, TFL per run, and in the bottom half nationally in yards per carry and yards per pass allowed. Fairly high yards per play and low negative play rate meant that teams averaged 5.6 plays per possession against Iowa State. How were they able to turn all these negative fundamentals into a successful defense?

First, opponents started with poor average starting field position (29th best nationally). Second, they were fairly good at preventing  explosive plays (7th fewest nationally in explosive plays per rush). As a result, while the yards per play were high, the EPA per play was lower (nationally average in nominal terms, top 40 when adjusted for strength of schedule). Finally, while Iowa State didn't turn teams over at an exceptionally high rate they made the most of the turnovers they did force, which resulted in a high EPA per turnover.

On offense, Iowa State again outperformed its fundamentals. They were not good on 3rd down, gained 25 or more yards on only 3% of plays, and were below average in EPA per pass and per carry, and yet the offense was slightly above average in adjusted effective points per possession (EP3+). They didn't score a lot of points on those possessions, but they tended to turn bad field position into good field position for the defense. It's a silver lining.

-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Missouri #58

Oh, what a difference a conference makes. Missouri in 2012 had the toughest schedule in the country both in terms of bowl eligibility and going undefeated. And it wasn't the non-conference schedule. Missouri got Alabama and Texas A&M out of the West, who were each, arguably, the best team in the country when Missouri played them. They played at Florida and at South Carolina. The only given on their conference schedule was Kentucky at home, and they won that game handily.

Oh, what a bigger difference a half-dozen injuries can make. In 2011, with a little help from the opposite sideline, Missouri beat Texas A&M. In 2012, A&M literally could have called students from the stands to play the second half and still won. That has nothing to do with playing in the SEC. Pinkel got .045 points per pass from his quarterbacks in 2012, half as many as Rutgers.

The defense was underrated, top 20 by most metrics. It allowed 1.9 points per possession, but 142 of those points came against Georgia, Alabama and Texas A&M, three of the nation's four most efficient offenses. Performances against Tennessee, Syracuse and Texas A&M to close out the season were uninspired, but Missouri's D was quite good through the rest of the season when playing against mortals. They could have been much better if they hadn't allowed more than 5 points per possession in the red zone.

-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

What Vegas thinks of your college football team

Before the 2012 NFL season I used Vegas win totals to reverse engineer power rankings. Now, Kegs 'n Eggs is listing Golden Nugget's very early spreads for 248 games in 2013. I bet you can guess where this is going.

It's a pretty simple process to slip these values in a model and reverse engineer power ratings. And it turns out the Golden Nugget spreads are pretty consistent with the power ratings I've estimated. The chart below compares the reported point spreads with the points spreads generated by my model using the reverse engineered power ratings. The average difference between the Nugget's spreads and the model's is just under a point.

Shockingly, according to the reversed engineered power rankings, the best team in college football in 2013 is . . . .  Alabama. Gasp. What is actually surprising is that Oregon is just a field goal below Alabama and Texas A&M just a touchdown from the Tide (though the 6.5 points they're getting in Kyle Field suggests a bigger gap). The Aggies are followed by Ohio State, Georgia, a run of Big 12 teams with LSU thrown in the middle to finish out the top 10.  

Ole Miss at 14 is notable. There are four SEC West teams in the top 14, and then there's a huge jump before we get to Mississippi State, Auburn and Arkansas. 

By my estimates, Notre Dame would be around 11th/12th if Golson were still on campus. 

Oregon is almost ten points better than Stanford. That seems fair.

Staying in the Pac, the Golden Nugget doesn't seem to think much of Sonny Dykes. Only 13 teams have lower power ratings than Cal.

Texas at 6 sets alarms off in my head. Then again, there's no reason they can't be there (outside of the terrible, terrible coaching staff). 

Michigan and TCU are two teams that I would put higher, and both for the same reason: very good quarterback play.

Then there's Louisville. I've said it before and I'll say it again. Louisville is getting hyped because they beat a flawed Florida team by 10 points a month after any other games. In the three games before that they lost to Syracuse by 19, UConn by 3 and beat Rutgers by 3. Teddy Bridgewater is good, and may have a great NFL future ahead of him, but he was the 16th most efficient passer last year - behind, among others, Johnny Manziel, Marcus Mariota, Connor Shaw, Tajh Boyd, and Collin Klein, guys that were asked do to a lot more than sit in the pocket and sling while playing against much better competition. The only thing Bridgewater did better than Michigan's Devin Gardner and Oklahoma State's JW Walsh is take more snaps. The Golden Nugget spreads place Louisville 24th among the 83 teams with data. I think anything above 30th is optimistic.

There are some notable exclusions: Utah State finished last year's power rankings at 14 and San Jose State is one more team with a better college quarterback than Louisville.

Statistical Review: Virginia Tech #59

Oh, Logan Thomas. In 454 pass attempts he added 4.3 points. That's less than Jordan Wynn, who retired mid-season, less than 98 skill-position players averaged per game over at least 10 games, less than Johnny Manziel in 11 different games, less than Manziel's backup Jameill Showers, and even less than Manziel on this one play.

He completed 51% of his passes and averaged less than 7 yards per pass. He was twice as likely to gain no yards as he was to gain 10 or more yards. He threw interceptions on 3.7% of passes, twice as often as Everett Golson and four times more often than AJ McCarron and Colby Cameron. Most troubling is that Thomas was only 1.4 points per game better in Virginia Tech's wins than in losses. By way of comparison, Manziel was 30 points per game better in wins than in losses. At quarterback, Thomas wasn't a difference maker.

What was the difference maker? Tech allowed 27 or more points in every loss and 23 or fewer points in every win. The defense was solid, not spectacular, across the board, except they were mediocre at preventing explosive plays and didn't force enough turnovers. They forced turnovers on only one in nine possessions. They were in the bottom quarter nationally in yards per kick and punt return allowed. Consequently, they were below average in starting field position on offense and defense. Not exactly Beamer ball.

-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Georgia Tech #60

Georgia Tech allowed 40+ points in six of 12 regular season games. That list includes a win against UNC despite allowing 50 points (one of nine games in 2012 in which the winning team allowed 50+ points). If not for offensive lightweights Maryland, Boston College, Presbyterian and USC sans Barkley, Georgia Tech would have allowed more than 35 points per game. Despite that, the stats say they were really an eight win team living in a seven win team's season.

 Tech ran the ball 81% of the time, so that merits the majority of our attention, but they also averaged 9.37 yards per pass. Only Georgia was better. And they led the nation with 17.8 yards per completion, best in the nation. In fact, in the Paul Johnson era, Georgia Tech has set up camp near the top of the yards/pass stats. If you watch Aaron Murray and then watch Washington or Lee (the quarterbacks, Washington and Lee is an institution of higher learning), you'll see that they aren't in the same class (and by that I mean one is good at throwing a football and two are much less good), but their yards per pass are similar. How is that?

The easy explanation is that Tech runs the ball 80% of the time so defenses are willing to give up something in pass defense to focus on defending the option. If that's the full explanation, defensive coordinators are morons. Georgia Tech runs the ball 90% of the time in most situations but throws the ball more than half the time on 3rd and 6+ and 85% of the time on 3rd and 11+. It doesn't take a PhD in statistics to figure out what they're going to try on the next play. Only Army is more predictable.

The more important explanation is that Georgia Tech only employs the forward lateral when it makes sense to employ a high risk/high return strategy; that is to say, they don't always throw the ball, but when they do, they prefer to throw it deep . The result is a lot of incomplete passes (47.4% incompletions), a lot of interceptions (4.12% of passes), and a lot of long completions. Every team in college football (except Georgia Tech) could increase their yards per pass by employing this strategy. But if you are doing it always on all downs you won't have a lot of success moving the ball down the field. If you're not convinced, watch a noob play Madden. That is why Georgia Tech has one of the nation's most efficient pass offenses in terms of yards per pass, but only a moderately effective passing game in terms of EPA per pass. Because they are often throwing on 3rd and long, defenses are giving them the first 6 or 8 yards and are focused on taking away anything further down field. They don't always succeed.

This Tech team was like all others of recent history in that they ran the ball a lot and depended heavily on explosive plays. This team was unique because it was fairly good at preventing explosive plays (only 3.4% of plays). Unfortunately, they were brought down by poor run defense (proof that teams aren't just good at what they practice against) and on 3rd downs.

-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Ohio #61

By my metrics, Ohio had the absolutely easiest path to bowl eligibility and the 6th best path to an undefeated season. To get six wins they needed to beat Norfolk State, Massachusetts (#123), New Mexico State (#122), Akron (#118), Eastern Michigan (#115) and Miami (OH) (#107). They lost to Miami (OH) but were already playing with house money after beating Penn State to start the season. If a couple calls went against them in games against Marshall, UMass and Akron, Ohio could have fell from BCS bowl aspirations to no bowl at all.

Ohio was more than a little bit lucky. They benefited most in those aspects of the game that are least predictive. By predictive I mean that some that some stats are very good at predicting future performance, like yards per carry. Other stats are less predictive, e.g., opponent field goal percentage (see Penn State v. Virginia). Ohio lost only five fumbles (but dropped the ball 13 times) and threw only five picks all season. They turned the ball over on 6.7% of possessions, third best in the country. Turnovers cost them less than 3 points per game. Holding on to the ball is a skill, but turnovers are as much or more about luck as skill, and Ohio had a lot of it.

Besides avoiding turnovers, Ohio benefited from good average starting field position (own 32) and good play on 3rd down, which led to longer drives (5.7 plays per possession). That's not luck. That's good football. They attempted a field goal on 17.5% of possessions and converted on 13.3%, 8th most nationally. In other words, the offense did a good job of squeezing points out of possessions, despite averaging only 3.7 points in the red zone, and of not putting the defense in bad situations.

On defense, aspirations for mediocrity were undermined by poor performance in the red zone (5 points per trip) and on 3rd downs (44% conversions allowed on 3rd and 4th downs, 25th worst nationally).
-------------------- The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Tennessee #62

With Tennessee at #62 we officially reach the halfway point in the 2012 statistical reviews. Also, Tennessee makes the third SEC team; Bob would be proud to hear we've only listed one Big 12 team so far.

Tennessee really was fire-your-coach bad in 2012. They won five games, and they reached that number by beating Georgia State, Akron, Troy (with little room to spare), Kentucky and North Carolina State. When your marquee win came against a 7-6 team, you have only one other win against a top 100 team, and you can fit a moderate size city in your stadium, it's time to consider other coaching options. 

Tennessee had talent on offense. We know about the receivers. Bray could have been better, especially with the receiver talent, but he did have an EPA per play better than EJ Manuel among others. And Tennessee had the second lowest sack rate in the country, which might help explain why they were such a tough out for Clowney and the Gamecocks. They were fairly explosive, e.g., 13.3 yards per completion, but this should have been a top five explosive offense instead of top 25. Added explosiveness would have lifted the 2.5 points per possession closer to 3 and made Tennessee a nationally formidable offense. (You know how they could resolve this problem? Go get Cincinnati's coach. Oh, right, they did that, just a season too late.)

The offense wasn't as explosive as it needed to be, but the defense allowed too many explosive plays. Shoddy pass defense in general really held the Volunteers back.

The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

xxxx

Statistical Review: North Carolina State #63

The per play statistics for the NC State defense are not overwhelming: 4.1 yards per rush, 7.1 yards per pass. But they made up ground by being elite on 3rd downs, especially against the pass. The Wolfpack did not sack the quarterback at an exceptionally high rate, but the data suggests they brought effective pressure on 3rd downs - the benefit of long, athletic linebackers.

Now to Mike Glennon. The NC State run game was abysmal (3.1 yards per carry), so a lot of pressure fell on Glennon to keep the ball moving. In this he was not overly successful. He completed less than 60% of his passes. His .208 EPA per pass was worse than Bo Wallace, Braxton Miller, and Matt McGloin to name a few. Against Virginia he scored an Expected Points Added of -19.8, one of the ten worst performances by a college football player in 2012; to put that in perspective, David Piland's five interception, 47% completion game against UCLA added up to an EPA of -22. Rubbing salt in the wound, Glennon's three very good performances against FBS opponents were all in losing efforts - Miami, North Carolina and Clemson.

The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Louisiana-Monroe #64

For some stupid reason we thought it was an upset when Louisiana-Monroe beat Arkansas. After Tyler Wilson was knocked out of the game, which more or less neutralized Hamilton and Cragg, Louisiana-Monroe was hands down the better team. Kolton Browning and crew racked up 550 yards of offense and held the ball for 40 minutes!

In week one we should have learned that Arkansas wasn't very good. In week two we should have learned that Auburn was even worse. What we did learn is that Louisiana-Monroe is much more effective when they go for it on 4th down than when they kick field goals. Few teams were more effective in the red zone (5.3 points per trip) but the Warhawks made only five of nine field goal attempts.

The offense was best characterized by the 5.9 plays per possession. It lacked explosiveness (e.g., 11 yards per completion) and Browning was the leading rusher despite a negative EPA on running plays. But Browning completed 64% of his pass attempts and the Warhawks were very good on 3rd down.

The defense was sub-par, average against the run and less than average on passing plays.

The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

The Top College Fantasy Football Quarterbacks In 2013 From Non-BCS Schools

Today we have a guest blog from the fine folks at FanDuel:

While college fantasy football might not be as popular as the NFL, more and more people are playing it. With significantly more teams to pick players from, a person must do more research. Here are the top five quarterbacks from the schools you might see on television that often.

Jordan Lynch

Northern Illinois might lose some players from their Orange Bowl team, but star quarterback Jordan Lynch returns. Perhaps no other player in the country is more important to their team than Lynch, who will look to navigate through a tough MAC once again. Lynch is a true dual-threat quarterback, throwing for 3138 yards and picking up another 1815 on the ground. In total, he was responsible for 44 touchdowns for the Huskies, and the entire offense is built around his talent, it will be hard to ignore him in fantasy football leagues.

Brett Smith

The Wyoming Cowboys rarely play on national television, but Brett Smith will be looking to help them win more than just four games like they did in 2012. Smith has good size and is a very capable pocket passer. He completed over 62% of his passes in 2012, throwing for 27 touchdowns. As long as he controls the offense and doesn’t become reckless, he will be a top 10 performer.

Kolton Browning

Stuck in the heart of SEC country, Louisiana-Monroe often gets overlooked. However, more than likely, only Johnny Manziel will put up better numbers than Browning in the SEC in 2013. ULM is a passing team, but Browning can also create offense on the ground as well. He led the team in rushing, although he only notched 488 yards. His 10 picks in 2012 are a bit disturbing, but when you throw the ball 428 times in one year, you are bound to have a few bad passes.

Derek Carr

Not only does Carr have a strong college arm, but he is by far the best pro prospect out of this bunch as well. He eclipsed 4000 yards in 2012, and he rarely forced throws into coverage. Fresno State could do some damage with their schedule this season, which makes Carr a highly-touted fantasy option.

Rakeem Cato

Marshall has fallen on hard times since moving from the MAC to C-USA, but Rakeem Cato could give them a shot at winning the league in 2013. He plays high-percentage football with his passes, connecting on just under 70%. He threw for 37 touchdowns in 2012, and that number could have been higher if not for a few off games.

Statistical Review: Bowling Green #65

When Florida struggled against Bowling Green to kickoff the season I interpreted it as a product of first game rust. By no means did I believe that the Florida offense was a dynamo in disguise, so I wasn't anticipating a 70-14 thumping, but this game was tied at 14 late in the 3rd quarter. Now with the a little perspective, the only surprising result from that game is that Bowling Green scored at all. Let me explain.

Relative to their level of competition, the Bowling Green defense was elite. They were one of six teams nationally to allow fewer than 300 yards per game despite being league average in possessions per game (in other words, opponents had their chances). They allowed only 3.4 yards per rush and 10.4 yards per reception, touchdowns on less than 15% of possessions and 1.2 points per possession. These numbers are deflated by weak competition - Idaho, UMass, Eastern Michigan, Akron to name a few. But the best offense on their schedule, San Jose State, scored only 29 - only BYU, Utah State, Stanford and Navy (?) held Fales and Co. under 29. The surprise isn't that Florida scored only 27 against Bowling Green. If anything, it is surprising the Gators scored than many points..

Relative to their level of competition the Bowling Green offense was poor. Two numbers - 54% completions, 11.5 yards per completion - best reflect the general offensive malaise. Florida should have shut Bowling Green down completely. For the season, Bowling Green averaged 365 yards per game but they racked up 327 yards, including more than 100 yards rushing, against one of the nation's elite defenses.

Finally, Bowling Green missed two field goals against Florida. Those six points would have made the game much more interesting in the 4th quarter. But Bowling Green made only 53% of their field goals, good for 120th nationally, and their field goal team was .74 points worse per attempt than the national average. Missing field goals was just part of the M. O.

The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Statistical Review: Middle Tennessee State #66

MTSU beat Georgia Tech by 21 points. Georgia Tech played for the ACC championship and was one play from beating Florida State and claiming the ACC title. MTSU lost to McNeese State and lost to Arkansas State by 45. They won eight games, six more than in 2011, but they didn't play in a bowl game. An odd season to say the least.

The Georgia Tech win really says more about Georgia Tech than Middle Tennessee. Paul Johnson's Tech teams tend to take quarters, halves, even games off on defense, and they definitely did that in this case. Also, Georgia Tech just wasn't that great all season; they were just outside of the top 1/3rd nationally. That they played for the ACC title is more about the ACC and the cheating cheaters that cheat in the ACC than anything else. Another important factor, though, is that Georgia Tech is not built to exploit MTSU's biggest weakness: a porous pass defense and 67.5% completions against. To their credit, the Blue Raiders were tough on defense in that game.

On offense, MTSU was effective at the line of scrimmage in 2012. They had among the nation's lowest sack and TFL rates. They ran 5.7 plays per possession and had good average starting field position (own 33), but performed poorly in the red zone (3.4 points per opportunity), and so they kicked field goals on 30% of their scoring drives. In 2013, they must depend more on the passing game with Kilgore back and Terry Pettis set to make an impact at receiver; in 2012, they ran the ball 58% of the time despite it being the relatively less effective part of their offense.

The biggest strength of the defense was turnovers. They forced turnovers on just under one possession in every five - which is a big reason for the good average field position for the offense. They were also good at preventing explosive plays. But they failed to bring much pressure; MTSU sacked the opponent's quarterback on only 3.7% of pass plays.

The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Stanford, yards per carry and run/pass ratio

@kennylong wanted to know how Stanford's yards per carry compares with other teams with a similar run/pass ratio. My original point was that Stanford's yards per carry was unimpressive for a team getting title buzz and known for physical play at the line of scrimmage. Was I out of line? Would a little more context explain Stanford's lackluster yards per carry performance?

While it might seem counterintuitive to some, teams that run the ball more often also average more yards per carry. For all 124 teams, the correlation between yards per carry and %run is greater than .5. There are two reasons for this. First, run-heavy teams run the ball in all situations at all points on the field. Pass-heavy teams reserve run plays for specific, short yardage situations. The yards per carry are lower because 1) defenses stack the line and 2) the goal is different - the play is oriented towards a conversion instead of getting as many yards as possible. Second, run-heavy teams average more yards per carry because teams that excel at running the ball run the ball more often. No one is satisfied with three yards and a cloud of dust these days. Effect becomes cause. 

The best approach for accounting for context-specific success (for example, two yards on 4th and goal at the 2 is a huge success, while two yards on 3rd and 4 at midfield is a failure) is to use the EPA per carry. EPA measures success in points - how many points can a team expect to score after a play versus before the play. Converting on 3rd and 10 , for example, is more valuable than ten yards on 1st and 10. Stanford ranks 43rd in EPA per carry, slightly better than in yards per carry, but not much. Stanford finished 24th in fewest TFLs per carry, and Taylor finished 35th out of 240 qualifying running backs in that statistic. Good, but not elite. Taylor finished 160th among those same 240 in conversions per carry. Poor, but not terrible.

Now, to actually answer the question I've prepared a chart below which locates 118 teams by yards per carry (ypr) and % run (ratio); I've excluded the most run and pass heavy teams - e.g., Army, Washington State. I've overlaid the regression the line - the line that tracks the relationship between ypr and ratio for these 118 teams - and then added the names for teams in the vicinity of Stanford in terms of run/pass ratio. Stanford falls right on the regression line, at the same level as Arizona State and UCLA, behind Cal and Michigan, and well behind the elite explosive teams - Florida State, Baylor, Cincinnati, etc. On the other hand, Stanford is ahead of LSU, Clemson, and South Carolina.

Statistical Review: Pittsburgh #67

Pitt threw only three interceptions all season. Twenty five teams threw more than three interceptions in individual games, and only 48 teams avoided a 3+ interception game. Army had fewer interceptions in 2012, with two, but Army threw 115 passes all season. Pitt was second nationally in fewest turnovers per possession: 6.4%.

Pitt's Sunseri did more right than not throw the ball to the other team. He also completed 66% of his attempts for 13 yards per completion. All together, Sunseri had an EPA per pass of .395, better than Geno Smith, Seth Doege, Landry Jones and EJ Manuel among many, many others. Some of that was offset, though, by 38 sacks, so that Sunseri had an EPA per play (run + pass) of only .176, worse that Case McCoy.

The Pitt defense was solid but not spectacular. The biggest weakness was a high play per possession from relatively poor 3rd down defense.

-

The Statistical Review breaks down teams along a number of performance categories, everything from red zone scoring to field goal percentage, and compares that performance against the rest of the FBS. All 124 teams will be reviewed from 124 to 1 by the hybrid rankings. You can find short descriptions of the stats used in the table below.

Measuring the Golson Effect

With Everett Golson out at Notre Dame, what impact will his poor academic judgement have on the Domers in 2013? That's a very reasonable question, but it's not the question I'm going to answer. There are too many variables and too little information. Instead, I'm going to try to answer a different question: What would Notre Dame 2012 have looked like without Golson? Then, I will use that answer to speculate on 2013.

To answer my question, I make a few assumptions/simplifications. First, I assume that if Golson were not playing Tommy Rees would have played in his place. In the case of the Navy game that is obviously not true. I don't care. Notre Dame would have won that game with me at quarterback, so it doesn't really matter.

Second, I measure passing performance using EPA: expected points added. EPA is a statistic that converts a player's performance into points by estimating each play's impact on the expected points a team will score on that drive. It has a long track record of predicting Heisman winners, and I find it to be the best single metric for measuring performance. Using the EPA I find that Rees was .053 points worse per pass attempt than Golson (.204-.152 plus rounding error). I assume that we can estimate Rees' performance in place of Golson by subtracting .053 per play from Golson's performance in that particular game. (An EPA/per pass of .204 is not good, but above average; the elite quarterbacks hover around .4.)

Third, I assume that Rees would run at his own rate at his own average level of success, not at the same rate and at the same level of success as Golson. I assume that the rest of the running plays would be picked up by the running backs, and that their performance on those plays would have been average (EPA=0).

Findings

For the season as a whole, I find that Golson made Notre Dame 20 points better than they would have been with Rees at quarterback. Most of that difference comes from the Pitt game (12.6 points). Against Purdue, Michigan and Stanford, the model predicts that Notre Dame would have been better off with Rees in at quaterback for the entire game, and because Rees played the entire game against BYU, those results are unaffected by the substitution. In four games, the difference is less than a field goal, and in seven games the difference between Rees and Golson at quarterback is less than five points. Only against Pitt would the hypothetical performance gap between Rees and Golson been enough to change the outcome of the game.

What does this mean for 2013? First, Golson is clearly better than Rees. Higher upside for Golson means the gap between the two would have grown. This difference will probably be more important in 2013 as Notre Dame will depend more on the quarterback as a play maker. But the quarterback is not the only player on the field. I'd put the conservative estimate of the cost of losing Golson in 2013 at 2 points per game, and the unconservative estimate at a field goal per game. That figure balloons if Rees goes down early with injury. But in the end, if Notre Dame can keep six opponents under seven points and all but two under 20, they'll still have a puncher's chance at 10 wins.