On Sept. 28, just seven weeks ago, Clemson found itself on the brink. The Tigers were 4-0, having begun their national title defense with three wins of 5+ touchdowns and a comfortable defeat of Texas A&M. But thanks to a short North Carolina touchdown, they were facing the prospect of falling behind the Heels with 1:17 left in regulation.
Trailing 21-20, UNC went for the go-ahead two-point conversion, but Clemson stuffed it and survived. Still, there was a pall hovering over what many expected to be a wire-to-wire national championship team. All-world quarterback Trevor Lawrence had already thrown more interceptions than in all of 2018. Receiver Justyn Ross was catching only half his targets. After a dynamite first game against Georgia Tech, the run game was stuck in containment.
The defense was perhaps overachieving, considering how much pro talent had left the school in the previous offseason. But this team was not cruising in fifth gear, having fallen to eighth in SP+ and, a few weeks later, fourth in the AP poll and fifth in the initial College Football Playoff rankings. This team was taking its sweet time living up to its potential.
This was nothing new.
Any coach in any sport will tell you that he or she wants their team to peak late in the season, when it counts the most, but this is easier said than done. You don't know what injuries you're going to have to account for. You don't necessarily know how opponents will counter what you show them early in the year, and whether you have the right counters for the counters. And you only have so much control over when the biggest challenges on your schedule come -- do they come early, forcing you to show more of your cards early? Do they come late, allowing you to keep cards close to your vest (but risking late-year burnout)?
That this act is so difficult makes it increasingly remarkable that Clemson pulls it off basically every year. Dabo Swinney's team is like clockwork in the way it reveals its best self over the latter third or half of the season.
In 2016, the Tigers eked by Troy in September and NC State (among others) in October, then got upset by Pitt in mid-November. But they won their final two regular-season games by a combined 71 points, kept Virginia Tech at arm's length in the ACC title game, beat Ohio State by a historic 31-0 margin in the Fiesta Bowl, and beat Alabama to win the national title.
In 2017, they lost to Syracuse and struggled to find fifth gear offensively, but won their last four regular-season games by an average of 34 points and walloped Miami 38-3 in the ACC title game. The offense couldn't do enough to get past Alabama in the CFP semis, but the Tigers still peaked late in the year.
In 2018, they found fifth gear more quickly. They narrowly escaped Texas A&M in Week 2, and following a combination QB change and QB injury, they needed a big comeback to get past Syracuse in Week 5. From that point forward, however, they won their final 10 games -- including the CFP semis against Notre Dame and finals against Alabama -- by an average of 36.
Whatever Clemson's best self is, you're not going to see it until a trial or two. It's like the Tigers can't play their best ball until someone makes them bleed. "Clemsoning" used to mean something bad. Now it means this.
I'll give you one guess what has happened since Clemson's survival in Chapel Hill.
• Average scoring margin: Clemson +26 in the first five games, +43 in the next six
• Yards per game: Clemson +231 in the first five, +367 in the next six
• Yards per play: Clemson +3.2 in the first five, +5.0 in the next six
• Performance vs. spread: -3.9 points per game in the first five, +10.3 in the last six
• Performance vs. SP+ projection: -3.6 points per game in the first five, +13.3 in the last six
Because the ACC is so incredibly dreadful this season, the 11-0 Tigers haven't had a marquee game of late to show off how far they've come, but rest assured: they have righted the ship. Both Vegas and SP+ -- which obviously take opponent quality into account (so go ahead and refrain from yelling "AIN'T PLAYED NOBODY") -- haven't been able to catch up to their progress in recent weeks. Clemson has covered in five straight games, mostly by large margins.
What has changed?
Not the defense, that's for sure. This has been maybe the best coaching performance of defensive coordinator Brent Venables' absurdly successful run. Despite five freshmen and sophomores among the six leading tacklers on the line, Venables' unit has dominated from basically the first snap. Like many top teams, Clemson deploys large rotations on D and it has served to build incredible depth.
Projected to fall to ninth in defensive SP+ (oh, the indignity), the Tigers leaped to fourth two weeks in and currently rank third. They are perhaps more vulnerable to the run than they have been in the past -- 12th in rushing success rate allowed, 27th in percentage of carries gaining at least four yards -- but they're still pretty good in that regard, and they make up for any speck of deficiency by fielding the best pass defense in college football.
Opponents are completing only 47% of their passes, 41% on passes thrown beyond the line of scrimmage. The Tigers are first in passing success rate allowed, first in passing marginal explosiveness allowed, second in adjusted net yards per attempt (ANY/A, a yardage measure that takes touchdowns, interceptions, and sack yardage into account), and third in sack rate.
Although they aren't getting as much of a contribution from their front four, the linebackers and safeties have more than made up the difference. Isaiah Simmons, the country's prototype nickel linebacker, has a combined 12 tackles for loss, six sacks and 13.5 run stuffs with five passes defensed. He is ridiculous, and he has lots of help.
Fewer hero throws
So does Trevor Lawrence. And he's leaning on it more.
Five games into the season, Lawrence's numbers obviously weren't outright bad. He was completing 62% of his passes at 13.5 yards per completion and he had taken only three sacks. His ANY/A average of 7.4 yards was perfectly solid.
But this was Trevor Lawrence, Alabama conqueror and future No. 1 pick. He was supposed to be more than solid. He seemed to be making a lot of decisions through a "what would a No. 1 pick do?" prism, forgoing easier, more open passes to throw challenging passes into tight windows. It wasn't going very well. He was completing just 67% of his passes between 0-10 yards downfield (with three interceptions) and just 40% of his passes between 10 and 20. And although receiver Tee Higgins had started the year on fire, 2018 CFP star Ross had not (18 catches, 218 yards, 50% catch rate).
The problems were both frustrating and fixable.
According to ESPN Stats & Information research, Lawrence is having less success on intermediate throws between 6-14 yards this year compared to last year. Perhaps that has something to do with Hunter Renfrow, the ultimate possession receiver, being gone. The Tigers are also throwing the ball deeper with less success, a definite test of that patience. ... There were passes he missed to wide-open receivers that he knows he should have made, and there were catches his receivers should have also made.
The Tigers got a bye week after UNC, and Lawrence has looked primarily like the Lawrence we thought we'd see ever since. He again got baited into a couple of intermediate mistakes early against Louisville, but his numbers over the past six games are brilliant: 74% completion rate at 13.6 yards per completion, an 89.8 QBR, an 80% completion rate on balls thrown 0-10 yards downfield, and a 78% completion rate on balls 10-20 yards downfield.
What's he doing differently? He's taking what he's given. If that means forgoing deeper passes and checking down to his amazing pair of running backs -- Travis Etienne and Lyn-J Dixon have caught 21 of 25 passes for 254 yards and two scores in the past six games -- so be it. In those first five games, he was averaging six passes per game 20+ yards downfield; since, only 2.7. But he's also averaging more yards per pass (11.1 ANY/A). That's a pretty good trade.
Your résumé should ask both who and how you've played
So this is pretty wild. Based on my post-game win expectancy figure, we should have seen a *crazy* string of upsets yesterday.
— Bill Connelly (@ESPN_BillC) November 17, 2019
Cincinnati > USF (22% postgame win exp)
Penn State > Indiana (27%)
Oklahoma > Baylor (29%)
UL-Laf > S Alabama (33%)
Georgia > Auburn (49%)
OK, you've kept it bottled up this long, and I appreciate it, but I'll give you a moment. Go ahead, you know you want to.
YEAH, BUT WHO HAVE THEY PLAYED??????
Feel better now?
Indeed, the ACC is only slightly better than the AAC this year, if at all. Per SP+, the second-best team in the conference is 6-4 Miami, and the Coastal champion will likely be either No. 39 Virginia or No. 44 Virginia Tech.
The Tigers have played only one team currently ranked in the SP+ top 40 -- compare that to Georgia (5), LSU (4), Ohio State (4), Alabama (3), Penn State (4), Oregon (4), Oklahoma (3), Utah (3). Still, you can only play who's on your schedule, and we can measure your performance pretty accurately all the same.
First of all, SP+ is itself an opponent-adjusted predictive measure and it ranks Clemson fourth after dropping the Tigers to eighth early in the year. (If not for a dreadful special teams unit, they would probably be third.) Beyond that, though, I have also come up with not only the predictive SP+ but another measure intended to gauge the strength of your résumé. I call it Résumé SP+. Wildly creative, I know.
Résumé SP+ compares your actual output to what would be expected of the average top-five team (per SP+). It says, "the average top-five team would enjoy a win percentage of X against your schedule, and an average scoring margin of Y," then compares your averages (with a 50-point cap on blowouts) to expectation.
Clemson's schedule has indeed been absurdly weak. The average top-five team -- in this case, a team with the average ratings of Ohio State, Alabama, LSU, Clemson and Georgia -- would, over a long period of time, enjoy a 0.952 win percentage against the teams Clemson has played. That ranks 93rd in SP+ SOS, barely ahead of schedules such as Memphis' (98th) or UCF's (103rd).
However, Clemson's performance against this schedule has been exceptional, even with the early-season glitches. An average top-five team would have a scoring margin of plus-31.9 against Clemson's 11 opponents. Again capping blowouts at 50 points (meaning, anything over 50 points is deemed a 50-point win), Clemson's scoring margin is plus-35, 3.1 points higher. The Tigers are one of only three teams with a positive differential.
Top 12 teams, per Résumé SP+
1. Ohio State (+10.4 points per game above expected scoring margin, No. 59 SOS)
2. Clemson (+3.1, No. 93 SOS)
3. Alabama (+1.2, No. 39 SOS)
4. LSU (-2.1, No. 15 SOS)
5. Georgia (-4.3, No. 24 SOS)
6. Wisconsin (-5.6, No. 17 SOS)
7. Penn State (-5.7, No. 30 SOS)
8. Florida (-5.9, No. 9 SOS)
9. Utah (-6.0, No. 65 SOS)
10. Auburn (-7.4, No. 2 SOS)
11. Oregon (-7.9, No. 45 SOS)
12. Oklahoma (-8.6, No. 70 SOS)
Ohio State is ahead of the pack in both SP+ and Résumé SP+; the Buckeyes have beaten four top-40 opponents by an average of 34.5 points, which is crazy. But Clemson has done more damage against its given opponents than anybody else, including LSU or Alabama. And Clemson has been in fifth gear for only about half the year. That, plus last season's exploits, gives us all the evidence we need that the Tigers are once again serious contenders.
Schedule strength is overrated, by the way
We laud teams that play big, bold schedules and we have all heard the words "scheduling intent" come out of the mouths of CFP committee members. But you know what tends to happen when you play a bunch of really good teams? You lose games.
Top 10 hardest schedules, per SP+ SOS ratings
1. 4-6 Michigan State (the average top-five team would have a 0.807 win percentage, or about an 8-2 record, against MSU's schedule)
2. 7-3 Auburn (0.812)
3. 3-7 Maryland (0.813)
4. 4-6 Mississippi State (0.815)
5. 4-7 Ole Miss (0.826)
6. 4-7 South Carolina (0.827)
7. 2-8 Rutgers (0.832)
8. 2-8 Vanderbilt (0.845)
9. 9-2 Florida (0.847)
10. 5-5 Tennessee (0.847)
Only two teams with top-10 schedules, and seven in the top 20, have winning records.
Auburn is 7-3 with wins over Oregon (15th in SP+) and Texas A&M (17th) and losses to the Nos. 3, 5, and 9 teams. Florida has beaten Auburn and three other top-40 teams and lost to Nos. 3 and 5.
Michigan State and South Carolina are 40th and 41st, respectively, in SP+ at the moment. The five teams above them have gone a combined 35-15; the five teams below, 34-16. But the Spartans and Gamecocks have earned all the scheduling intent cred in the world. They're also a combined 8-13.
Playing nobody is a pretty good way to go through life if you ask me.
Week 12: one of near upsets
Here's one final note for you.
Top teams like Clemson, Ohio State and Alabama all cruised in Week 12 (though with a cost, in Alabama's case), and all in all, we didn't see that many upsets. But if the stats had their way, we could have.
Postgame win expectancy is my stab at looking at how randomness affected the outcome of a given game. It takes all the key predictive measures from a game (all the things that end up in the SP+ formula), tosses them into the air, and says, "With these stats, you could have expected to win this game X% of the time." I wrote about this in a Chalk piece recently.
A couple of weeks ago, only one team with a postgame win expectancy under 50% actually prevailed. Week 12, however, was defined by some top teams getting a bit lucky. Penn State and Oklahoma were both under 30% against Indiana and Baylor, respectively, and Georgia's win over Auburn came with just a 49% win expectancy.
In a more statistically just universe, then, Georgia, Oklahoma and Penn State would all have two losses now, Baylor would be unbeaten and ranked around eighth, and the primary one-loss teams we'd be arguing about would be Alabama, Utah and Oregon. Instead, Baylor is unbeaten no more and the pool of one-loss contenders remains enormous.