Writers of Pro Football Prospectus 2008

Most Recent FO Features


» 2017 Defensive Personnel Analysis

Defenses have taken a wide variety of responses to the rise of 11 personnel. Is any one system better than another? And how has the rise of the "moneybacker" changed defensive philosophy?

05 Dec 2016

OFI: Committee Got It Right

by Chad Peltier

We have made it to bowl season. Your College Football Playoff final four are Alabama, Clemson, Ohio State, and Washington, which is a pretty similar lineup to what we expected entering the season. Penn State made a surprising, dominant nine-game run, and Wisconsin and Ohio State outperformed expectations (while Florida State didn't meet them), but overall we have a more or less blue blood group to play for the title this year.

Despite cries from Penn State fans and the Big 12, the playoff picks were good. The committee's stated goal is to find the "best" four teams for the playoff -- not the most deserving, however you want to define that. All four teams were among the top five in the F/+ rankings (with Michigan third and Washington fifth), so the committee did its job if you go by the advanced statistics to determine the best four teams. Notably, in the F/+ actual scores, there's a large gap between Alabama (75.9 percent) and Ohio State (66.3 percent), and then between the Buckeyes and both Clemson and Washington (55.5 percent and 54.2 percent). Based on the matchups and early Vegas lines, that suggests you'll get a rematch of the 2014 Sugar Bowl in the championship game this year.


  • We expected the Pac-12 Championship to be close between Washington and Colorado, but it ended up being the game with the largest success rate differential (Washington plus-32 percent) on the weekend. One in five of Colorado's offensive snaps was successful, compared to half of Washington's! Washington managed four times as many scoring opportunities as Colorado and was plus-3 in turnover margin. This was just an absolute beatdown and likely was just enough to keep Penn State at fifth in the final College Football Playoff rankings.

    There were two problems for Colorado. First is that the Buffaloes offense is heavily reliant on quarterback Sefo Liufau's play. So if he is injured or has an off day, then the entire offense is in trouble. Both things happened against Washington -- backup quarterback Steven Montez was forced into action for part of the game, and Liufau threw three interceptions (or, one interception for each pass he completed) and averaged just 2.2 yards rushing on the day. This of course all speaks to Washington's excellent pass defense -- something Alabama will certainly be reviewing ahead of their playoff matchup. Washington's defense is eighth in passing S&P+, fueled by an incredible pass rush (fifth in adjusted sack rate) and aggressive secondary (eighth in defensive back havoc rate). In case you haven't been following much Pac-12 action this season, the Huskies defense is built on the strength of their back seven as opposed to the defensive line -- the linebackers provide an excellent pass rush while the defensive backs load up on interceptions and passes defensed. The defensive line, while solid, doesn't create much pressure on its own.

    Notable for the Alabama matchup, this was not quarterback Jake Browning's finest day. He went 9-of-24 with 4.9 yards per attempt, as the Huskies were both inconsistent and unable to generate explosive plays through the air. Instead, the Huskies relied on their duo of Myles Gaskin and Lavon Coleman to be steady (if not flashy) on the ground to consistently move the ball. In the end, with a 51 percent offensive success rate, the Huskies didn't need to be anything more than steady on the ground -- an incredible defensive effort and solid day running the ball were more than enough to capture the Pac-12.

  • The Crimson Tide could also make a claim to the most dominant championship game performance. Alabama raced out to a 33-16 halftime lead and then completely stonewalled the Gators in the second half. Alabama's win was similar to Washington's in that the Tide can claim the largest yards per play margin of the weekend (plus-2.8, even though Washington had the highest success rate margin). Besides the high margin of victory and statistical dominance, their respective wins were similar in that the quarterbacks didn't exactly have prolific days through the air (Jalen Hurts was 11-of-20 for 138 yards), and their dominant wins were instead fueled by incredible defense, plus-3 turnover margins, and efficient rushing performances. Alabama's trio of running backs -- Bo Scarbrough, Damien Harris, and Joshua Jacobs -- combined for 25 carries for 212 yards, an 8.5 yards per carry average.

    On the other side of the ball, the Crimson Tide were dominant against the run -- holding leading rusher Jordan Scarlett to 1.5 yards per carry and 17 total yards -- and created enough turnovers to make up for Florida quarterback Austin Appleby's second-most prolific passing performance of the year. Early on it seemed like the Gators might have something with a pass-heavy game plan, throwing for all but two of their 64 yards on their opening drive and converting three third downs of 6 or more yards. But that was essentially it for the Gators. Austin Appleby threw interceptions that resulted in ten Alabama points on his next two drives, then outside of two scoring opportunity drives on either end of halftime, the Gators' drive chart looks like this: punt (returned for a touchdown!), three-and-out, downs, three-and-out, interception, three-and-out, and punt. For the last quarter and a half, the Gators wouldn't cross midfield.

    That's how Alabama stealthily dominates this season -- turnovers, field position, efficient rushing, and overpowering defense. Little about the Tide's performance was necessarily flashy or explosive -- the passing game was essentially just three explosive passes and little else -- but the Tide nevertheless won by nearly 40 points.

  • Compared to the contests in the Pac-12 and SEC, the Big Ten and ACC championship games were close, one-score games. Clemson went up 21-7 midway through the second quarter and looked like it was on its way to blowout win over Virginia Tech, but the Hokies' strong fourth quarter put things close at the very end. In fact, with roughly one and a half minutes left in the game, Virginia Tech was driving and hit the Clemson red zone. But a fourth down interception ultimately preserved the win for the Tigers, sending them to their second-straight playoff appearance. All in all, the Hokies and Tigers looked fairly evenly matched for most of the game. The Hokies actually created one more scoring opportunity despite having a 20 percent lower offensive success rate, but the Tigers were able to get touchdowns on every one of their scoring opportunities and the Hokies averaged only five points per opportunity. That final interception -- in the red zone on fourth down -- ended up being just about the only advanced statistical difference between the two teams. While it was a disappointing ending for Virginia Tech, Justin Fuente's first season in Blacksburg was an unqualified success at 9-4.

    Notable for the Tigers' playoff run, neither they nor the Hokies were effective running the ball. Deshaun Watson was the leading rusher and averaged 5 yards per carry, and he had just a single explosive run on the night. Similarly, Virginia Tech quarterback Jerod Evans had three times as many carries as running back Travon McMillian, but still averaged just 2.2 yards per carry. The Hokies and Tigers were mostly only able to move the ball on one another through the air (though both were effective, with 40 percent and 50 percent offensive success rates!), as Jerod Evans had a 47 percent passing success rate but also managed a 15 percent passing explosiveness rate. While Ohio State's J.T. Barrett has had efficient days throwing this year, he has been widely variable, and the Buckeyes haven't been very explosive -- so it might be difficult to create a similar level of big plays against Clemson in the first playoff round.

  • Probably the best conference championship game this year was in the Big Ten, where Penn State recovered from a 28-7 deficit in the second quarter to down Wisconsin. The Badgers were incredible in the first half, forcing two three-and-outs, returning a fumble for a touchdown, stopping a fourth-down attempt on the Wisconsin side of the field, and recovering a second Penn State fumble. All in all, the Badgers scored touchdowns on three of five first-half drives, created three turnovers (if you include the turnover on downs), and had two three-and-out stops on defense. It looked like the Badgers were on their way to a blowout win, similar to Clemson. But then Penn State launched an explosive two-minute drive to close the first half -- which saw four plays of 12-plus yards -- and then Penn State went off from there.

    Beginning with that drive, the Nittany Lions would score on every single possession of the rest of the game (outside of the final, clock-killing possession with a minute left), with four touchdowns and a single field goal. The defense stepped up as well, allowing just three points on three second-half scoring opportunities (one point per opportunity) after allowing a touchdown on every opportunity in the first half. In the end it came down to a final, potentially game-tying drive with five minutes left for the Badgers, but Wisconsin was unable to pick up a fourth-and-1 on the Penn State 24-yard line with a minute left.

    In the Nittany Lions' nine-game winning streak, they have beaten two top-ten teams in Ohio State and Wisconsin and averaged a 40.1-to-19.2 margin per game. That dominance wasn't enough for the Playoff Selection Committee, however, which slotted Penn State at fifth behind Washington, despite Penn State winning the Big Ten and beating the team from their own division who did get in -- Ohio State. The advanced stats rankings here support that call -- the Nittany Lions are 12th, sixth, and ninth in the S&P+, FEI, and F/+, while Ohio State is third, second, and second in the same metrics.

  • Although the de facto Big 12 championship was far less consequential compared to the others -- since it was extremely improbable that either the Sooners or Cowboys would get in to the playoff race -- it was nevertheless a tight, interesting game for the first half and one that impacted Ohio State's strength of schedule.

    The Oklahoma offense didn't heat up until its last drive of the first half. Before, it managed three three-and-outs, but two explosive plays fueled a 76-yard drive that would take the Sooners into halftime tied at 17.

    The second half was all Oklahoma: while holding the Cowboys to just three points, the Sooners scored touchdowns on three of four second-half drives, and ended the game with an incredible 13-play, eight-plus-minute drive to ice the game. As incredible as Oklahoma's second-half drive efficiency is the fact that the Sooners were able to chew up enough clock in the second half to hold the Cowboys to just four drives. The Sooners, true to form, managed ten explosive plays, including nearly a third of Baker Mayfield's completions (Mayfield had three passes of 40-plus yards). After a dismal September where Bob Stoops' Oklahoma lost to both Houston and Ohio State, the Sooners managed another ten-win season. Plus, Baker Mayfield decided to return for his senior season in 2017.


  • Samaje Perine, RB, Oklahoma. The Sooners offense is full of star talent, but Samaje Perine was trusted to shoulder the load with 37 carries for 239 yards against Oklahoma State. Everyone on the Sooners offense ripped off a 40-plus-yard play (literally five different Sooners had a play longer than 42 yards), but only Perine could handle that many touches.
  • Deshaun Watson, QB, Clemson. You could make an argument that no player is as critical to his team as Deshaun Watson. Watson threw for 288 yards with 8.5 per attempt, and led the team in rushing with 17 carries for 85 yards. He may be more turnover prone this season (with an interception against the Hokies), but Watson was incredibly efficient overall and recorded five explosive passes against the Hokies.


  • Reuben Foster, LB, Alabama. The Butkus Award-winning linebacker left a heck of a final impression for voters, with two sacks and 2.5 tackles for loss along with a team-leading 11 total tackles against Florida. This prototypical linebacker was all over the field against the Florida offense despite the high percentage of passes from Austin Appleby. Reuben Foster is a surefire first-round pick in the draft next year.
  • Dexter Lawrence, DL, Clemson. Dexter Lawrence is a true freshman, but he was the most impactful player for either defense in the ACC Championship Game. Lawrence nearly equaled Reuben Foster's production with two sacks and tackles for loss on the defensive line.

Posted by: Chad Peltier on 05 Dec 2016

24 comments, Last at 07 Dec 2016, 1:31pm by Tomlin_Is_Infallible


by Aaron Brooks Go... :: Mon, 12/05/2016 - 7:07pm

Nothing about Temple dismantling Navy's historically-efficient offense in the AAC title game?

by Will Allen :: Mon, 12/05/2016 - 11:55pm

I think it would be fun if, this Saturday, we had Wisconsin at Alabama, Oklahoma at Clemson, Michigan with a rematch at Ohio State, and Penn State at Washington. If you wanted to really reward conference champs, and not get too heavy in one conference, thus limiting any conference to 3 participants, we'd have USC at Alabama in a rematch, Oklahoma at Clemson, Miichigan at Washington, and a Ohio State at Penn State rematch. Start the games at 11 eastern tiime, then 230 eastern, 6 eastern, and 930 eastern for the West Coast game. I'd watch that.

by dbostedo :: Tue, 12/06/2016 - 12:24am

I'd watch that too.

But I'd be bitter over how the expanded playoffs made the earlier OSU/Michigan game much less meaningful, as well as Penn State's loss to Pitt, and the early season Wisconsin and USC losses.

I love, love, love the feeling that a single loss can (potentially) end a college football teams hopes for the title. Then again, I liked the BCS (but hated how they kept tweaking it to try to match the polls, when it was put in place in part to balance out and adjust what the polls said).

by Will Allen :: Tue, 12/06/2016 - 1:05am

If you made the 5-8 selections highly dependent on wins and close losses to high quality out of conference opponents, you'd really get rid of the worst aspect of college football, the deliberate scheduling of cupcakes. I'll take that trade.

by Cloister :: Tue, 12/06/2016 - 1:13am

You need to keep in mind these OOC games get scheduled years in advance. Rutgers was a solid team at the time Washington scheduled them, but gets denigrated now for having such a weak team in their schedule. Meanwhile, Michigan gets bonus points for playing Colorado, who was terrible and an expected cupcake back when the game was scheduled. You just can't really know how things are going to work out.

by Will Allen :: Tue, 12/06/2016 - 10:56am

Sometimes you get lucky or unlucly in the recruiting/coaching cycles. Sometimes you get exactly what you were looking for, like when Washington put Idaho and Portland State on the schedule, or when Ohio State decided to go play Oklahoma in Norman. In my preferred setup, Washington would be in, given they won their power conference, maybe even given a playoff game at home ( I really would prefer to reward conference champs), but short of a conference championship, I'd look for every opportunity to leave somebody home for putting Idaho and Portland State on their schedule, especially if they didn't play a powerhouse out of conference.

Frankly, I think if they ever get to five 16 team power conferences, which only play each other, that would be better for the sport, and for academics.

by Sleet :: Tue, 12/06/2016 - 4:27am

Guess I'm old school. If you lose head-to-head and don't win you conference championship, I think it takes more than a lucky victory at home against Michigan to get the nod over Penn State, whose last loss was back in September. If you are not #1 in your own conference and lost head-to-head to the conference champion, you should not be playing for #1 and depriving the team that beat you, the champion, a spot. If it was about the best team getting in, USC would be playing. Heck, why don't we give the nod to Michigan, who sure looked like the better team but for a referree blocking the vision of the QB on the second pick (and the one-sided calls). Justify giving OSU the nod all you want, they lost to Penn State, the Big 10 champ, who earned the right to play for the National title. OSU should be playing SC in the Rose Bowl.

by Aaron Brooks Go... :: Tue, 12/06/2016 - 11:11am

"Guess I'm old school. If you lose head-to-head and don't win you conference championship, I think it takes more than a lucky victory at home against Michigan to get the nod over Penn State, whose last loss was back in September."

Like road victories over Oklahoma and Wisconsin?

by Will Allen :: Tue, 12/06/2016 - 12:28pm

I love conference championsips too, but when you don't have a round robin conference schedule, you really aren't determining the conference championship in a sound manner. I really would prefer an 8 team tournamnent, with 4 games this weekend, featuring 5 conference champs and 3 at large, no more than 3 teams from any conference. I can't decide if I'd rather give 4 home games to the highest 4 seeded conference champs, or to the 4 teams that had the most impressive non-conference performance, which means wins and close losses against strong competition. The latter would really incentivize teams to seek out tough nonconference games, which would be great for the sport.

by Aaron Brooks Go... :: Tue, 12/06/2016 - 2:01pm
by LyleNM :: Tue, 12/06/2016 - 2:09pm

Still not a round robin, though.

by ramirez :: Tue, 12/06/2016 - 11:22am

It'll be interesting to see how much Alabama is favored by the time the game against the Huskies is played. What do we really know about them? They've beaten up on a dumpster fire of a conference in the SEC this year, and their big win over USC was ages ago, and against a very different team from the one that beat Washington. I think Washington has a good chance to win, though the game being played in Atlanta give the tide a big advantage.

I support an 8 team playoff, because leaving out 1-2 deserving teams every year for a set of reasons that changes season-to-season is dumb. But how could Western Michigan be left out of an 8 team playoff when they're 13-0? You have to give those teams a chance.

by Aaron Brooks Go... :: Tue, 12/06/2016 - 12:25pm

Alabama is basically a better Colorado.

If I'm U-W, I'm terrified that I couldn't run at all against USC, and Browning crapped the bed against two of the three real defense he played all season (USC, Colorado) and was middling against the third (Utah).

If Gaskin can't run (and no one could against Alabama. Ole Miss and Texas A&M broke 100 yards, but at about three yards per rush), U-W has no chance. Tennessee, LSU, Auburn, and Kentucky all run well, but couldn't get anything going against Alabama.

That wounded duck Browning threw across his body for a TD against Colorado is a pick-6 against Alabama. A really good passing team can cause Alabama problems. Ole Miss the last few years, Arkansas this year, Clemson last year. You can pass on Alabama if you're good. But good Browning hasn't appeared against the non-cupcake portion of the schedule.

Penn State might have been able to beat 'Bama. I think Washington loses by as much as Saban wants to beat them by.

by Will Allen :: Tue, 12/06/2016 - 12:34pm

Yeah, the big, athletic Penn State receivers may have been able to make a few big plays, and make the game competitive, although the way Wisconsin moved the ball tells me that Alabama would have sugnificant offensive success against Penn State. I don't think the Huskies have a chance, either, but this may be a historically great Alabama team that just rolls everybody. I don't think Ohio State can move the ball against them. Clemson? I dunno, I think the Sabanites will have an easy time with them as well.

by Aaron Brooks Go... :: Tue, 12/06/2016 - 2:03pm

I think OSU can stop their offense, though. They can win a rock fight against Alabama, ala their wins over Wisconsin and Michigan.

'Bama's offense is good, but it's situationally good. It can be turnover prone, and OSU is really good at turning teams over.

Clemson's defense is good enough to keep their offense in the game.

by ramirez :: Tue, 12/06/2016 - 1:25pm

You guys are buying into the SEC hype way too much. Obviously Alabama will be favored, but this idea that they will just steamroll the Huskies, what is it based upon? Alabama starts a freshman QB, and he's not going to be facing the sucky pass defenses of the SEC west. I also remember hearing about how invincible Alabama was before they faced teams like Ohio State, Utah, and Oklahoma in the postseason. And how did those games turn out for Saban and the Tide?

I'm concerned about Browning, too, but if Alabama doesn't get any TDs from defense and special teams, can Hurts carry the load? You shouldn't assume the Tide is a lock because a very differnet Alabama team crushed Michigan State last year. Again, who has Alabama beaten?

by Will Allen :: Tue, 12/06/2016 - 2:07pm

It is an error to speak of Alabama's past teams, when evaluating the current team.

by Aaron Brooks Go... :: Tue, 12/06/2016 - 2:13pm

gain, who has Alabama beaten?

They beat USC by 46.

When it comes to Washington, it's more a question of the matchup. I have no confidence U-W can run their way to victory over Alabama like they did to Stanford and Colorado, and to a lesser extent Utah. Frankly, I think Utah might be a batter matchup against Alabama than U-W is. If you're really good, you're looking at 100 rushing yards. But I am confident that they can holding Browning to something like a 12-24-200-1-1. I don't think 300 total yards beats Alabama.

It's not impossible to win a rock fight with Alabama, but it's hard:

But it works better if you have a running QB and can pass it over Alabama.

I just don't think Washington's strengths work well against Alabama, but I think Alabama will be able to run on Washington, and that means they don't have to try to let Hurd win the game. I think OSU and Clemson can force him to do that, though.

by ramirez :: Tue, 12/06/2016 - 2:35pm

"They beat USC by 46."

I addressed this in my previous posts, but you're talking about a USC team that since Alabama played them, has changed their coach and quarterback, and massively improved in other areas as well. So they're not the same team as in September, not even close. I live on the west coast, so perhaps I see more Pac-12 games than you do, but the Pac-12 has been a better league than the SEC this year, and it's not close. That's why I'm sekptical that Alabama will manhandle Washington. The Huskies have looked impressive against Wash State and Colorado. Alabama has looked impressive against Florida and Auburn. The problem is that I saw the games where Auburn was beaten by Georgia, and Florida was manhandled by the Seminoles, a team that didn't even win their division in the ACC.

Once you accept the fact that the SEC as a whole is massively overrated, and Alabama has padded their stats beating up on a bunch of weak opponents, it's easy to see why I'm not buying the hype.

by Aaron Brooks Go... :: Tue, 12/06/2016 - 3:34pm

Yes, it's a closer game if Darnold plays. But that doesn't make it close. Alabama didn't even play particularly well against USC. USC led after 1, and won the TO battle.

That you dismiss the USC game doesn't make it disappear, or become irrelevant. And yes, the SEC sucks this year. But the ACC sucked in 2013, but that didn't mean FSU wasn't the best team in the nation.

Don't get me wrong; stranger things have happened. I watched Penn State lose to Pitt in a game where they gave up 340 yards rushing (a game a lot like the Wisconsin game, in retrospect), squeak past Temple (who was terrible in the beginning of the year) and then get annihilated by Michigan. I never expected they could beat Ohio State.

Penn State, USC, and Temple might be the most improved teams in the country. I'm not entirely convinced USC's improvement isn't a scheduling quirk, though. Washington is the only decent team on the second half of their schedule.

Long story short, Alabama may not really be 46 points better than USC, and USC may not be really 13 points better than Washington. But Alabama not really being 59 points better than Washington doesn't mean Washington is going to beat them.

If I had to rank teams this year, it's probably something like this:

1. Clemson when they care.
2. Good Louisville.
3. Alabama
4. Good U-W
5. OSU when Meyer isn't running the turtle offense.
6. Michigan with a healthy QB
7. Passing PSU
8. Ole Miss in the first half.

by Will Allen :: Tue, 12/06/2016 - 3:57pm

Yeah, USC just was physically overmatched in their game against Alabama. It wouldn't be the biggest upset in the histry of college football, and I hate giving 14 points in any kind of playoff game, but I sure wouldn't feel good about laying on the dog, either, in this case. Not an attractive play in any fashion, in my view.

by ramirez :: Tue, 12/06/2016 - 4:11pm

The problem with your analysis and rankings is that they're larely based on hypotheticals. For the most part, these teams haven't played each other, so we don't really know how they compare head-to-head at all. My point is that Alabama hasn't played a very good schedule this season, so the fact that they have blown out most of their opponents doesn't surprise me. You're focusing on the results against USC since that is their only common opponent. It's a major point in Alabama's favor, and I've already agreed that the Tide should be favored over Washington.

What I'm not buying is your argument that Alabama will be able to toy with Washington. I don't see any reason to believe that. I think your claim that Saban can run up the score at will, or whatever your exact quote was, is based on impressive results achieved by other Saban teams in previous seasons. Your overlooking the fact that Alabama doesn't have any impressive wins since the USC game, and that for reasons I've already explained, the USC win wasn't nearly as impressive as it looked at the time. You're also overlooking that Washington has a few impressive wins of their own, that they came recently, and that the Pac-12 was a better league than the SEC.

Alabama has big wins over teams like Tennessee (lost to Vandy, barely beat App St.), Miss State (lost to South Alabama at home) and Texas A&M (lost 3 of their last 4, haven't beaten a power 5 team since October 8). The claim that SEC teams are simply more physical than schools from other leagues has been debunked again and again. If that were really the case, shouldn't the SEC win more than half their games against power 5 nonconference opponents?

by Cuenca Guy :: Wed, 12/07/2016 - 12:02pm

I find the problem with most of this discussion to be that we're looking at just one common opponent, and one that looks like a far, far different team today. Let's play the game with Stanford. Stanford beat USC, UW destroyed Stanford, so the Huskies should handle USC in Seattle. Except it didn't happen that way. Or look at the way USC barely beat Colorado at home but won at UW. So the Buffs should have the advantage over the Huskies.

Let's look at matchups, Browning's struggles in big games, key injuries on both sides, and the like, but I think it's silly to try to make predictions based on the result of one game played four months before the semifinal. Even the result at Husky Stadium had a team playing their worst game and seeming to suffer some emotional effects from the loss of their leading tackler.

Should Alabama be favored? Absolutely. However, I wouldn't see it as a major upset for Washington to win. Even ESPN's FPI has Alabama favored only about 2:1. Favoring the Tide by 5-7 points makes sense to me. 14-15 is crazy. Alabama might cover, but that has to be one of the most favorable point spreads for betting the underdog that I've seen in a while.

As to the Huskies not getting pressure up front, Washington very rarely sent more than four the first half of the year and got tremendous pressure on the QB. After the loss of Joe Mathis, the team has had to rely on sending five much more frequently.

by Tomlin_Is_Infallible :: Wed, 12/07/2016 - 1:31pm

so long as "every week matters" is proven to be an outright lie, the committee gets it/got it wrong

The standard is the standard!