Wednesday, December 26, 2018

On the 2018-19 College Football Postseason

On Clemson vs. Notre Dame

  • The prevailing consensus for this game seems to be that Notre Dame will be lucky if the game is still winnable at halftime. I understand the skepticism of Notre Dame. I am not as high on them as I was at the midseason - since that time, they have had to eke out close wins against Pitt and USC, and their big wins over Stanford, Virginia Tech, and Michigan look significantly less impressive. Their schedule looked daunting in the preseason, but it ended up being a really bad schedule when Stanford, VaTech, Florida State, and USC all collapsed in the same season. It is fair to say they they haven't been tested against a great or even very good opponent since week 1. And I think to some extent, Alabama's beatdown of Notre Dame in the 2013 Orange Bowl still colors the way many perceive the program. But I caution viewers against underrating this iteration of the Irish. First of all, Notre Dame mostly dominated their schedule - since Ian Book became the starter, they haven't had a single game with a post-game win expectancy below 95%. Second, this team has recruited at a higher overall level than they had in 2012, and they have high draft picks at all levels of the defense (by contrast, I don't think anyone from the 2012 team's secondary was drafted higher than round 6). Since that Alabama game, Notre Dame has beaten LSU twice, and gone blow for blow with three playoff teams (2014 Florida State, 2015 Clemson, and 2017 Georgia). Third, this is a very balanced team that can beat you in many different ways. Their offense isn't elite in any one area, which is why they don't always look like world beaters against mediocre teams. However, with Book under center, it's extremely difficult to stop them completely, and I think their offense will scale better than the spread-to-run offense they were running with Brandon Wimbush.
  • The main reason why I am simply not convinced that game will be a blowout is that I think people are overrating just how good this Clemson team really is. Don't get me wrong, Clemson is elite. They have the best freshman quarterback since Peyton Manning, along with usual stars at wide receiver and defensive line. Travis Etienne may well be the best running back in the country. Their rushing defense is historically good. However, I have major concerns about their passing defense. Their statistics look pretty good because the ACC was pitiful this year - in most games, Clemson's defensive line was able to overwhelm the past protection so badly that the opposing quarterback didn't stand a chance. I also think their games against two triple option teams (Georgia Southern and Georgia Tech) because those teams are disproportionately dependent on being efficient in the run game to be competent in the passing game, and Clemson is hard-wired to shut down power running games. However, their two toughest opponents, Texas A&M and South Carolina, were able to keep their quarterbacks upright and pick apart Clemson's shaky secondary. Had Clemson played a tougher schedule I question whether this game would be viewed as a warm-up for the national championship.
  • The South Carolina game is especially concerning to me. Like Jake Bentley, Ian Book makes fast decisions and has a quick release. Also, Notre Dame's big, physical receivers that can punish Clemson's safeties for their inability to play the ball when it's in the air. Book is not as mobile as Kellen Mond, but I think he can escape pressure and make plays outside the pocket better than most of the quarterbacks Clemson faced in ACC play. It is certainly possible that Clemson's defensive line will wreck the Irish offense on their own - they are that good, and Notre Dame's offensive line is not as elite as it was last year. But if they can even slow down Clemson's pass rush, there will be chunk plays to be had.
  • Notre Dame's run defense is decent, but not elite. I don't think they will be able to shut down Travis Etienne, but I do think their front seven is good enough that Clemson won't just run it down Notre Dame's throat. The key is to hold Clemson to around 5 yards per carry and bank on their ability to win the chess match on passing downs. If Clemson can get around 6 yards per carry, they'll probably wear Notre Dame down by the 4th quarter.
  • I said all year that Clemson needed to play Trevor Lawrence early and often, because a weaponized Lawrence gave them a better chance to win in the playoffs than Kelly Bryant would have. Clemson gave Lawrence the reps and his inexperience didn't cost them in the regular season, and now we get to see if he raises their playoff ceiling. Unlike Clemson (and unlike the 2012 Fighting Irish), this year Notre Dame's passing defense is elite both by the numbers and the eye test. For this particular matchup, I really like that they excel at preventing big plays. I am as high as anyone on Lawrence, but he is a true freshman; if Notre Dame can contain Tee Higgins (and I believe Julian Love can do that) and test Lawrence's ability to stay patient on 80 yard drives, I like their chances of keeping the game close. I wouldn't be shocked at all if Lawrence does indeed nickel and dime Notre Dame into submission, but if that happens you tip your cap knowing you made the game as hard as possible against a once in a generation passer.
  • The biggest worry for Notre Dame is that they aren't a great rushing team to begin with, and they are playing one of the best rushing defenses in recent memory. If they base their gameplan around establishing the run on standard downs, that will only lead to 3-and-outs. I think it is crucial for Notre Dame to have success throwing the ball on first down. If they can stay ahead of the chains and set up second and short, I think their RPO game can really take advantage of Clemson's poor safety play. If they get into 3rd and long and let Clemson's best pass rushers pin their ears back, they won't be able to keep up with Clemson's offense.
  • Dexter Williams is much more explosive than he is efficient as a running back - and I think that actually helps Notre Dame in this matchup. It is almost impossible for any team to be efficient against Clemson anyways, and I would rather have someone that is able to strike gold on the few holes that do open up against this front four.
  • Prediction: Clemson 31, Notre Dame 21
On Alabama vs. Oklahoma
  • I thought Alabama's secondary would struggle this year, but their front seven would protect them. Instead, Alabama's front seven wasn't up to their usual standards, but their phenomenal secondary made up the difference. Both units give up a few more explosive plays than I think Nick Saban is comfortable with though, but I think they've cleaned that up as the season went along. Their front seven in particular has improved drastically. Coming into the year I posited that any marginal improvements or declines in the defense wouldn't really matter if Tua Tagavailoa made good on the superstar potential he flashed in the national championship game against Georgia. I was conservative in predicting how Tagavailoa would transition into being the full-time starter, but he smashed even the most optimistic predictions. When Tagavailoa was healthy, this Alabama team was the most dominant of Saban's tenure. He overcame Alabama's tendency to give up the occasional big play by unlocking Alabama's long-suppressed wideout talent and producing his own big plays at an astronomical rate. Modest quarterback play was the one thing that leveled the playing field against Alabama, and when that advantage dissipated, the results were downright scary.
  • I was correct in predicting that Oklahoma's defense would continue to be a huge problem. Where I erred was in assuming that Kyler Murray wouldn't be quite as prolific as Baker Mayfield. I'm still not convinced that Murray is actually the better player - I would still like to see him slice up a defense as good as Auburn, Ohio State, or Georgia - but Murray was arguably more productive. Oklahoma's offense was better this year on a per-play basis, but they faced a weaker slate of defenses, and their offense was significantly more dependent on the big play. I have doubts about how this offense will translate against a team with elite athletes.
  • Most Big 12 teams played soft zone against Oklahoma, because they were so scared of getting burned for the big play by Oklahoma's speedy wideouts or by Kyler Murray's legs. The two teams that had the most defensive success against Oklahoma were Texas and Iowa State. From what I can tell, these two teams did the best job of containing Kyler Murray's running ability. I think this is because these two teams had the best linebacker play in the conference, and were able to play man coverage while still keeping Murray in the pocket. Alabama's linebacker play has been suspect at times this year, but they're still probably better than any big 12 team. Furthermore, Alabama has the athletes to leave their corners on an island and not get annihilated (especially if Marquise Brown is injured). Oklahoma will score points, but I think they are going to have to work a lot harder for those points than they're used to.
  • Oklahoma's offensive line versus Alabama's defensive line is a marvelous matchup. I think Oklahoma has the best offensive line in college football - all five of these guys are going to be drafted - but they will not have seen lineman like Quinnen Williams and Raekwon Davis in conference play.
  • Oklahoma's pass defense is pitiful. There is no way to sugarcoat it. In particular, they do a terrible job of preventing completions on standard downs. I don't see any way they don't struggle to stop Alabama from moving the ball. 
  • If Tagavailoa were 100% healthy I would take Alabama over the playoff field. But his ankle clearly became a bigger and bigger issue as the season progressed. He was sailing throws against LSU and Mississippi State, and he was flat out ineffective against Georgia. For as many advantages as Alabama has over Oklahoma, if Tagavailoa can't perform at a functional level, Alabama could certainly fall into a whole out of which they can't crawl.
  • If the Alabama ends up turning to Jalen Hurts, I still like their chances. Statistically, Oklahoma's run defense is better than their pass defense, which in theory would give them more confidence in this scenario. However, from what I've seen, Hurts has actually looked markedly improved as a passer this year - not just against Georgia, but in garbage time during the regular season. More importantly, I think in this scenario Saban would be perfectly content to run QB power until Oklahoma cries uncle, the way Texas did with Sam Ehlinger. Oklahoma's shoddy tackling would become a bigger problem here.
  • To me, the bottom line is that Alabama has the upper hand in a slugfest, and they probably have the upper hand in a shootout. Alabama's offense was just as good as Oklahoma's despite playing tougher defenses, and Alabama actually plays defense of their own. To win comfortably, all Alabama needs to do is focus on keeping Murray in the pocket, and hope that their secondary doesn't get completely torched. I think that's more than doable.
  • Prediction: Alabama 65, Oklahoma 38
On Michigan vs. Florida
  • Let's be clear: as poorly as the Ohio State game went, Michigan has a very, very good team, and they should be proud of the season they had. They are a legit top-10 team who would, and in my opinion they were the best team not in consideration for the playoffs. But Jim Harbaugh's goal is surely to compete for a national championship, and to that end the program still has steps to take. Michigan has a good defense, and their defensive statistics were phenomenal for most of the year. In retrospect, this is because they were built specifically to overwhelm ground and pound Big 10 offenses, without being very versatile (in statistics and machine learning, this called overfitting). The Ohio State game exposed how vulnerable their secondary is if you can protect the quarterback from 3 wideout sets. Furthermore, I was disappointed by how vanilla their offensive schemes were against a defense that had shown a complete inability to deal with misdirection. To truly compete for a championship, Harbaugh and his coaching staff are going to need to seriously re-evaluate their identity on both sides of the ball.
  • I thought Florida would struggle this year because Jim McElwain left the team at such a talent deficit. But Dan Mullen's staff truly squeezed every ounce he could of this roster (and they were pretty lucky with turnovers). Their bread and butter has been out-executing their opponents on standard downs, so that the offense is always in third and short and the defense is always in 3rd and long. Interestingly, Florida's offense isn't all that great (relatively) on third and short, and their offense isn't great (relatively) on third and long - but the bet is that even the worst 3rd and short team will have an easier time than the best 3rd and long team. To me, being significantly better on standard downs than on passing downs means your play callers are getting the players to overachieve.
  • If Michigan were playing a spread passing game in the Peach Bowl I would be very worried. Instead they drew their ideal matchup. Like Florida, Michigan also banks on their ability to be more efficient than you on standard downs and force you into more passing downs. Where Michigan could get into trouble is against a high-caliber passing team that can take advantage of their lack of corner depth to generate big plays on standard downs. Florida is not that team. Florida will try to beat Michigan at their own game but with worse players. Even if Florida tries to air it out, I do not think Michigan has any reason to fear Felipe Franks.
  • Prediction: Michigan 21, Florida 7
On Washington vs. Ohio State
  • I think recency bias has impacted how this game is perceived. Ohio State looked like world beaters against Michigan, and thus they should destroy Washington - right? The problem is, the Michigan game is not at all representative of how they played against the rest of their schedule. Ohio State's defensive line was pretty good at stuffing runs and creating pressure (though not as good as I thought they would be). However, if you managed to get the ball to the second level, their linebackers and secondary were atrocious at limiting the damage on big plays. This was fine when Michigan was content to simply plough forward 4 yards at a time, but Ohio State got gouged against almost every other team on their schedule - even lowly Oregon State was able to gash them! What saved them this season was their decision to transition from the smashmouth spread to the air raid. Their new offense, and Dwayne Haskins's ability to stretch the field, allowed them to maximize the ability of their speedy wideouts. But don't assume their leaky defense is fixed because of one particularly favorable matchup.
  • I was all in on Washington in the preseason. I thought their team had no holes, and that Jake Browning and Trey Adams would get back to the high level of play they showed in 2016. Instead, Adams was hurt most of the year, the rest of the offensive line stagnated, Jake Browning continued to regress, and the team was was one of the worst in the country at finishing drives. Luckily, their wideouts took a monumental step forward to prevent the offense from cratering - a lot of times their best offense was Jake Browning throwing jump balls and counting on his receivers to snatch the ball over helpless cornerbacks. Their defense was phenomenal on paper, but I thought their secondary had some bad games at the wrong times against Auburn and Oregon. Washington is good, but disappointing.
  • Schematically, Washington is much more prepared to take on Ohio State than Michigan (or almost any Big 10 team). Washington has high end talent and depth at cornerback, and their linebackers are comfortable in space. If Ohio State tries to force Washington's linebackers into man coverage, Ben Burr-Kirven can deal with a parade of crossing routes better than Michigan's linebackers did.
  • Washington doesn't run a ton of RPOs, but I wonder if they'll install some packages to take advantages of Ohio State's lack of discipline on defense. I also expect Washington to use a lot of shifts and motions to confuse Ohio State, much like what Chris Petersen did to a superior Oklahoma defense in the 2007 Fiesta Bowl.
  • By far, Washington's biggest weakness on offense was their inability to convert touchdowns in the red zone. If they can get it together for one game, they could outstrip their usual offensive output.
  • If Washington can give Browning enough time to throw, I think their wide receivers and tight ends can tear apart Ohio State's coverage. Unfortunately, I simply can't see their offensive line holding up. Ohio State's defensive line hasn't lived up to their billing, but they're still very good. Washington's offensive line is mediocre, and they struggle mightily to protect Browning if you can force them into obvious passing situations. One area that could favor Washington is that they are pretty good at running on passing downs, and Ohio State struggles to stop the run - Washington might be able to steal some first downs by running in 3rd and long.
  • When Ohio State did not make the playoff, I thought it was possible they would not take the Rose Bowl seriously, and thought Washington had a chance to shell-shock them early. That probably won't happen now that the team is trying to send off Urban Meyer with a win. And if Ryan Day gets to take over a bigger share of the game plan, I think that will help Ohio State - at this point I would say Day is the better X's and O's coach than Meyer.
  • Prediction: Ohio State 27, Washington 14
On Georgia vs. Texas
  • I was really high on Georgia going into the year - I even picked them to win the SEC. Their run defense ended up suffering more from their losses at linebacker than I anticipated. Luckily, Jake Fromm took the step forward that I thought he would, they were able to win even when their defense was mortal. Jake Fromm was absolutely unconscious against Alabama, and if he can somehow play at that level consistently next year he will be a top five pick in the draft. Georgia isn't without holes, but they are nothing short of elite at the things they do well.
  • Coming into the year, I was higher on Texas than most people - I picked them to win the Big 12. They came very close to proving me right, and yet I feel somewhat disappointed by their season. I thought they would field the best defense in the Big 12 while pounding teams with an SEC style power running game. Their defense was merely decent after being excellent in 2017, and while their rushing offense was efficient, it wasn't explosive at all. If anything, the passing game is what saved the team. Sam Ehlinger is a better passer than given him credit for, and Colin Johnson and Lil'Jordan Humphrey were excellent at keeping Texas ahead of the chains. I thought Texas hit a home run when they hired Tom Herman, but I think he goes into next season with a lot of pressure to prove that he is an elite, rather than good, coach.
  • It seems unlikely that Texas will be able to beat Georgia with a pass-first approach, so the question becomes how well Texas can run the ball. Texas has an efficient rushing offense, and Georgia has an inefficient rushing defense - thus, I think Texas can generate a decent yards per carry average. The problem is that Texas generates very few big plays, and Georgia gives up few big plays, and thus Texas will have a very small margin for error if they want to convert yardage into points.
  • It's also possible that Texas's running efficiency is inflated simply because most Big 12 teams defenses built to stop a power running game - whereas Georgia sees those kinds of offenses regularly in conference play.
  • Georgia's offense may not be pretty, but don't be fooled - it is very effective. Per S&P+, Georgia is a top ten team both at running and passing the ball. Texas is outside the top 40 in both areas. Georgia's offense is more efficient than Texas's defense; Georgia's offense is significantly better at creating big plays than Texas is at stopping them. This holds for both rushing and passing. It's hard to see Georgia's offense not having its way with Texas.
  • Prediction: Georgia 35, Texas 17
On LSU vs. UCF
  • I thought LSU would underachieve this year, mostly because I did not think Ed Orgeron was a good coach, and I thought Steve Ensminger was a clear downgrade from Matt Canada. The offense did indeed take a step back, but their defense took a corresponding step forward. I suspect their defense should have been even better given the talent they have, but it's hard to know how much better they would be with a more efficient offense. Overall I don't think that LSU is as good as Auburn was last year, but I think they are better than the version of Auburn that showed up at the Peach Bowl with no Kerryon Johnson.
  • I really do feel bad for UCF. At a basic level, we play sports to win, and UCF accomplished the goal that every team has going into the season. I understand why UCF fans would feel unfair that their team won every game on their schedule and that wasn't sufficient to compete for a title. But if we are objectively evaluating their merits relative to the other top teams, there simply is very little evidence that UCF is either one of the four best teams, or that they had one of the four most accomplished seasons. UCF did not play the same quality of week to week opponent as a power 5 conference champion. This would be fine if they dominated the way a power five champion would have. Instead, they had multiple close games against teams that were destroyed by power 5 opponents (and in Temple's case, lost to an FCS team). UCF was impressive against Pitt, but Pitt went 7-6 in the what was the worst power five conference - I don't think this proves that UCF is on par with teams like Ohio State, Oklahoma, and Georgia. They are 8th in S&P+, but I think this is illustrative of the limits of this metric - it can be fooled by teams that are exponentially more dominant against bad teams but mortal against good teams. I would prefer to see them dominate the conference uniformly, along with wins against high level Power 5 teams. I think UCF is better than most power 5 teams, and they probably would have won the Pac-12 south, the Big 10 west, and the ACC Atlantic. But the fact that they couldn't dominate the AAC to a larger degree leads me to believe they are not a playoff caliber team.
  • It seems like every LSU game involved the announcers explaining how Joe Burrow transformed LSU's offense. I don't see it. Burrow was a 58% passer this year, and his 7.2 yards per attempt was decent but not spectacular. Their passing offense went from 16th to 74th in S&P+. Not all of that falls on Burrow - perhaps Burrow would be dynamite with a better coordinator. But don't excuse competency with excellence.
  • Per S&P+, LSU had a below average running game, and UCF had an above average run defense. If UCF can actually hold up at the point of attack on standard downs, I actually do think they can pull the upset - LSU is atrocious on passing downs. But LSU's numbers were based on playing in a conference that recruits the best, and more to the point, the biggest defensive linemen. UCF's defense is fast but undersized. Until I see it, I am skeptical that their defensive line won't get pushed around by LSU's offensive line.
  • UCF likes to run it more frequently than the average team, both on standard and passing downs. I was shocked to find that LSU's opponents actually run with extremely low frequency, despite LSU's excellent secondary.  Something's gotta give. On one hand UCF is down their starting quarterback, but LSU is down two starting cornerbacks. I have no idea who will be the one to break tendency. LSU's defensive line is not up to their usual standards, and I could see UCF being able to run the ball efficiently. I think what's more likely is that they try to throw on standard downs so that they can pick on LSU's backup corners and not force Darriel Mack to make plays on 3rd and long.
  • This game is important for LSU - it's important for recruiting, it's important for the players trying to get to the NFL, and it's important for building your program credibility for future playoff arguments. UCF is a known quantity at this point - they are not sneaking up on anybody. I do not want to hear any excuses if LSU loses - just give UCF the credit and move on.
  • This is a big if, but: if UCF beats LSU, beats Stanford by an amount commensurate with the best Pac-12 team, and they run roughshod over the AAC, there will be no excuse for keeping them out of the playoff. If all of those things happen and UCF still is left out, that will be a tacit admission that there is basically nothing that a Group of 5 team can do to finish in the top 4.
  • Prediction: LSU 35, UCF 24
On other college football storylines
  • Urban Meyer is one of the all time great college football coaches - he might even be the second best of all time. But he also repeatedly turned a blind eye and gave second chances to players and coaches who did horrible things. Meyer's legacy should be his willingness to win no matter what the human cost. Any team that hires him knowing his history should be criticized heavily.
  • Recently there has been talk of expanding the college football playoff to 8 teams with 6 auto-bids. I have mixed feelings about auto-bids. On one hand, I believe every team deserves to enter the season with an unambiguous path to the title. It just doesn't feel right whenever a team wins every game on its schedule and still doesn't sniff the championship. On the other hand, if you can auto-bid your way in with a conference championship, teams would have little incentive to schedule tough out of conference games to prove your worth. I am guessing most teams would rather preserve themselves for conference play than try to play for the the two at large spots. This is a bad unintended consequence. If we don't do autobids, I think we should have a 6 team playoff. This gives enough room for every Power 5 champ that is truly playoff worthy, and it gives room for a truly elite G5 team (the best mid-major teams from the past 15 years typically finished in the 5-8 range), but the lack of autobids would prevent someone like Washington from making the playoffs over Georgia. I think 8 with no autobids is too much, because we almost never have 8 teams that are seriously capable of competing for the title. If we are committing to autobids, I think we should go to 16 teams with autobids for every conference. It's arbitrary and unfair to limit the autobids to the Power 5. Furthermore, with 5 at large spots available, we will have enough room for elite teams stuck in tough conferences. With that, along with the huge benefit going to the high seeds who play the worst Group of 5 teams in the first round, teams would still be incentivized to schedule tough games.
  • Regardless of what playoff format we use, every non-championship game should be played on college campuses. The magic of college football is the mythic venues, the thrill and fear of carrying your region's flag into distant lands, and the pageantry of an entire and campus/town being consumed by game day. It is simply awful that these schools have been suckered into letting the revenue generated by the postseason be poured into the the pockets of bowl executives rather than the local economy.
  • The caveat to the above is that it is hypothetical - we should think long and hard before actually forcing players to play an extra 2-4 games for no extra pay. If you oppose any playoff expansion based on player safety, or the morality of the economics of it, I won't push back.
  • I like that USC went through with much needed staff overhauls, and I really like that they clearly did make it a priority to hire coaches with past ties to USC. Kliff Kingsbury with USC's quarterback and receiver talent is going to shred Pac-12 defenses. However, the fact that they kept the same strength and conditioning coach tells me that USC wasn't as serious as they should have been about changing the culture of the program. I will probably pick USC to win the conference next year, but I don't think they are a national title contender, and I worry that once the new assistants leave for promotions elsewhere, the program will spiral out of control.
  • Justin Herbert was the only quarterback that looked like a franchise NFL quarterback to me. With the news that he is staying for his senior season, I imagine guys on the fence, such as Will Grier, Dwayne Haskins, and Ryan Finley, will be more likely to turn pro. However, none of those players really excite me if I am picking in the first round - I would probably prefer to take one of the stud defensive players and try to get Tagavailoa, Herbert, or Fromm in 2020. Side note: the fact that both Herbert and Marcus Mariota chose not to leave Oregon early speaks well to the culture and stability of the program.
  • Finally, will someone please explain to me why the playoff games can't all be on New Year's day? What is wrong with having the two playoff games sandwich the Rose Bowl? Playing the playoff games on December 29th makes New Year's day feel so anticlimactic when it should be the best day of the season.