2020 Passing Plus-Minus: The Search for Drew Brees' Successor
The King is dead. Long live the King.
This week, we're continuing our annual review of the passing game with our look at 2020 plus-minus stats. And for the first time in a while, we're searching for a new gold standard in the field of accuracy.
Passing plus-minus is a stat we annually track to help provide context to completion percentage. Given the location of a quarterback's passes, it compares his completion percentage in each area to historical baselines. This stat does not consider passes listed as "Thrown Away," "Tipped at Line," "Miscommunication," or "Quarterback Hit in Motion" by Sports Info Solutions charting. Metrics are based on how often a pass is completed based on the pass distance, the distance required for a first down, and whether the ball was thrown to the left, middle, or right side of the field. This is a counting stat, so more attempts are obviously a great thing for the purposes of what we're talking about here. Our Completion Percentage Over Expectation (CPOE) numbers may differ from other models around the Internet.
Throughout the 2010s, all this was preamble for our annual statement that Drew Brees was very good. Brees owned the stat in the 2010s, in what I suspect will be a historically dominant performance. Brees was the single-season champion eight times in the decade. His total of +342.8 completions over expected was over 50% higher than any other passer in the last 10 years, and his 6.1% CPOE was by far the best in the league. He was the only quarterback with at least 30 pass attempts to top a 5.0% mark—and again, he did that over 5,621 pass attempts. Even in the years he didn't win (losing in 2012 to Peyton Manning and 2015 to Russell Wilson), he was always in the top 10. No one had passed him; they were just keeping his seat warm.
But now Brees has retired, after his worst season ever in this metric. That leaves a void at the top of these tables, a throne waiting for the taking. Someone is going to lead the league this decade by completing more passes than they should, and it's a wide-open field.
With Brees' noodle-armed reputation over the last few years in his career, you may think that our search for a true successor to Brees should start with the best in the short game. In the 2010s, +100.7 of Brees' plus-minus came on passes with an expected completion percentage of 75% or more, those short slants and screens that came to define the Saints' offense over the decade. That's more plus-minus than all but four other passers put up in the entire decade. Never screwing up the easy stuff is a solid foundation for a great career. But then again, Brees also led the league in plus-minus on passes 10 yards or more downfield (+156.1 to Philip Rivers' +90.3) and on passes with less than a 50% chance of being completed (+73.5 to Russell Wilson's +51.3). You don't dominate this stat just by feasting on empty calories.
So, who will take Brees' crown for the 2020s? Let's take a look at where last year's passers stood, ranking all 36 passers who qualified for our leaderboards.
|2020 Passing Plus-Minus|
An Empty Throne
Deshaun Watson's +27.9 is the lowest total to lead the league since Peyton Manning's +27.7 in 2006, the first year we ran these numbers. His 5.5% CPOE is the lowest mark to lead the league that we've ever recorded. Don't get us wrong; +27.9 is still a very good result. In a Brees-less world, +27.9 would have led the league in 2017 and has always been at least a top-five mark. We just may have to get used to a world where the leader in this category is in the high-20s, not the mid-30s.
But let's stop holding Watson up to a historic outlier and instead look at what he actually managed to do himself. Watson put up these league-leading numbers despite having an ALEX of +0.2; the next six guys on the list all had negative ALEXes. To find someone who beats Watson's 8.9 aDOT, you have to scroll down to Tom Brady. Watson was putting up fantastic completion numbers despite his offense's lack of the checkdown. When viewed in that light, Watson's season seems far less run-of-the-mill. The last player to top Watson's +27.9 while having a positive ALEX was Tony Romo in 2014; it has been quite a while since we saw someone push the ball downfield as frequently as Watson had to in 2020 while still maintaining high success rates.
The Texans were terrible in 2020. They were the third-worst defense in the league, and had the worst rushing game in football by DVOA. That means lots of huge deficits, and lots of call for deep passes just to keep the game remotely competitive. And Watson thrived in that environment in a way we haven't seen in a long time. Yes, some of that came from Will Fuller having an amazing season; Watson had a +6.6 plus-minus targeting Fuller at least 10 yards down the field. But this wasn't a one-receiver show; he also had at least a +1.0 plus-minus on deep targets to Brandin Cooks, Randall Cobb, Chad Hansen, Pharaoh Brown, and Keke Coutee. And Watson's CPOE actually went up by 2.9% in 2020 despite the loss of DeAndre Hopkins, so losing yet another top receiver might not have as big an effect as you might think on his numbers.
Watson has been in positive-double digits for three seasons straight now, tied for the second-longest active streak with Derek Carr and Kirk Cousins and only trailing Russell Wilson's eight-year streak. It's not a one-year wonder result for Watson, even if 2020 was fairly clearly his best season to date. Yes sir, it looks like nothing is going to derail the Deshaun Watson train entering 2021. Now, if you'll excuse me, I just need to catch up on my reading of this past offseason and…
… oh, right. All that news that happened. It's always awkward writing articles like these about a player with an active legal situation, and the odds of Watson playing a full season in 2021 are about as likely as the odds of this writer leading the league in CPOE next year. You can see why teams like the Broncos would still be campaigning to bring Watson in for their starter despite the investigation into sexual misconduct allegations, but football remains firmly on the backburner for Watson at this point, and so we shall move on.
But speaking of passers the Broncos hope are under center in 2021, we have Aaron Rodgers sitting in second place. There were worries last offseason that Rodgers' career might be winding to a close; all his stats were dropping and plus-minus was no exception. Rodgers had a -0.1 combined plus-minus in 2018 and 2019 with a CPOE of 0.0%, a poor finish to the decade that saw Rodgers get passed by Wilson and fall into fifth place for the 2010s. Well, Rodgers roared back with his best season since 2012, a +26.2 mark for the MVP. His CPOE jumped up 5.2%, making him one of three players to see a five-point jump from 2019. You always hear the stories that the LaFleur/Shanahan offense takes a year or two for a passer to grasp, and maybe that's what we saw here: Rodgers and Matt LaFleur finally getting on the same page offensively. Rodgers is unlikely to lead the league in the 2020s in this stat, as he would be 46 in 2029, but Packers fans can relax knowing that Rodgers and LaFleur are working smoothly together now.
And while we're on the subject of aging passers showing improvement in 2020, Tom Brady was dead last in this stat in 2019. We bemoaned the fact that Brady and his receivers were never on the same page, and pointed out that Brady's on-target and catchable pass rates were also in the basement. We said that it "wasn't a death sentence," and that there had been instances of players bouncing back before to have quality seasons. I would consider a Super Bowl MVP a bounceback season; Brady joins Rodgers as one of the three players to see his CPOE jump by at least five percentage points. Is there more to it than just "having good receivers helps you look good?" Well, Brady went from fifth worst to seventh best in on-target rate, per Sports Info Solutions' charting, so there was a real jump in the number of Brady's passes that were even catchable last season. Sorting out how much of that is better rapport and more trust in his receiving corps versus adjusting for his slowly declining physical abilities is a difficult process. It seems unlikely that Brady will lead the league in the 2020s in this stat, as he would be 52 in 2029, but that's assuming human aging curves ever apply to Brady. Similarly, Kirk Cousins, Russell Wilson, and Matt Ryan are all on the wrong side of 30 if we're looking for a long-term successor to Brees.
We turn instead to the highest player on this list who is both still in his 20s and looks set to play a full season next year: Josh Allen. Allen saw his CPOE jump 7.0% from 2019, which was in itself already a 5.9% jump from his league-worst mark in 2018. In last year's article, we noted that if Allen could even duplicate half of his improvement from 2018, he would end up on the positive side of the ledger. Well, he shot well past that target, placing him in some fairly rarified air. Just like we've always said: accuracy is a real strength of Josh Allen, and part of what makes him such a dangerous player. Whether or not this is sustainable remains an open question; we go into great lengths about it in Football Outsiders Almanac 2021 (cheap plug!)
Meanwhile, in the Basement…
We come now to the bottom of the table. Allen's successor as worst CPOE in the league was Dwayne Haskins, who has now become the first passer to have back-to-back league-worst CPOEs since 2011-2012 Blaine Gabbert. Haskins now joins the Pittsburgh passing room, which features Haskins (-6.8% CPOE in 2020), Ben Roethlisberger (-2.3%), and Mason Rudolph (-6.0%) atop their depth chart. Good luck.
The race to the very bottom in plus-minus was close. For the record, it is not a tie. Carson Wentz takes the crown over Drew Lock at -25.82 to -25.81, but it if you flip one pass for either player, you get a new champion. Wentz's CPOE fell -6.0% from 2019; while he's never been a plus-minus machine with a career total of -5.9 coming into 2020, last year was an entirely new level of terrible. Frank Reich has his work cut out for him restoring Wentz to adequate NFL starter levels.
As for Lock, his -1.9 and -1.3% marks as a rookie were passable in a small sample size, but things really took a swing in the wrong direction. He's now battling Teddy Bridgewater for the starting job in Denver. It's not hard to see why Broncos fans are pining for a Watson or a Rodgers to make the trip over, but it is hard to see why the Broncos opted to pass on Justin Fields with the ninth pick in the draft.
Two other players joined Wentz in seeing their CPOE drop at least five percentage points between 2019 and 2020. One was Drew Brees, whose last season with a noodle arm apparently incapable of even challenging defenses downfield still managed a +4.5 plus/minus and a 1.2% CPOE. He retires with 15 straight years of positive seasons in both of these stats, and the only reason that stretch isn't longer is that we don't have the data before 2006. Long live the King.
The other player to see a drop-off is more relevant and interesting—Ryan Tannehill. In 2019, Tannehill was second in the league behind Brees with a 7.7% CPOE and fifth with a +20.7 plus-minus. He fell back to earth last season. His +5.1 and +1.1% marks aren't bad, mind you, but they're more in line with the player Tannehill was in Miami than with 2019's Pro Bowl season. With the Dolphins, Tannehill averaged a +3.9 plus/minus and had a career CPOE of +0.9%, so his numbers were still up last season. But you can put this in the pile with 2020's YAC+ numbers as another stat that shows Tannehill playing closer to his historic averages than his breakout season two years ago. And yet, Tannehill never came close to sniffing double-digit-DVOA in Miami. These small stat regressions saw Tannehill tumbled all the way from fifth in DVOA in 2019 to, uh, sixth in 2020. Tannehill's burst of accuracy in 2019 may have been somewhat of a small sample size illusion, but his understanding and running of Tennessee's offense is light years beyond what he managed in Miami. What that says about Adam Gase and/or Joe Philbin is left as an exercise for the reader.