QBASE: Pickett, Corral Lead Lukewarm Quarterback Class
NFL Draft - Guest column by Alex Olbrecht and Jeremy Rosen with Aaron Schatz
According to Football Outsiders' QBASE 2.0 model, there is no statistically significant difference between the top five quarterbacks in this year's NFL draft. Moreover, the model predicts that this year's class is more likely to disappear into obscurity than walk into the Hall of Fame. After all, for every class of 1983, there are years that don't yield any starters (see 2007 and 2013).
Likewise, Scouts Inc. doesn't project any of this year's quarterbacks highly, which is perhaps why there has been so much action this offseason via trades. The Washington Commanders have traded for Carson Wentz, the Denver Broncos for Russell Wilson, the Cleveland Browns for Deshaun Watson, and the Indianapolis Colts for Matt Ryan. Jimmy Garoppolo and Baker Mayfield may be traded too. Yet some teams could still enter the draft desperate for a quarterback. Given the extensive discussion in Brian Billick's book The Q-Factor, we know that drafting a quarterback out of need can lead to red flags being ignored and, say, EJ Manuel becoming a franchise's first-round pick. This year's model should serve as a warning for any general manager thinking of drafting a first-round quarterback out of desperation.
In summary, QBASE 2.0 combines Andrew Healy's original QBASE model (2015) with Jeremy Rosen and Alex Olbrecht's functional mobility model (2018) by factoring in a quarterback's rushing ability while also using his adjusted college passing statistics and adjusted years started. The adjustments consider the quality of both the quarterback's teammates and opponents, and while they reward quarterbacks who have steadily improved over time, they penalize one-year wonders.
As always, Football Outsiders' 2022 NFL draft coverage is presented by Underdog Fantasy!
Projections for the 2022 Class
Prospects are listed by order of their overall Scouts Inc. ranking in this draft class.
Malik Willis, Liberty
Scouts Inc. Rank: 21
|Mean Projection||-0.26 TDYAR/A|
|Bust (< 0 TDYAR/A)||59.9%|
|Adequate Starter (0-0.75 TDYAR/A)||23.0%|
|Upper Tier (0.75-1.5 TDYAR/A)||12.2%|
|Elite (> 1.5 TDYAR/A)||4.9%|
Willis originally played for Auburn but lost the starting job as an incoming junior to true freshman Bo Nix. However, after transferring to Liberty, he displayed excellent athleticism and vaulted himself up draft boards. Yet despite displaying the arm strength and mobility required to play at the next level, he has been inconsistent at anticipating receivers coming open and at following progressions, which can result in him breaking the pocket and scrambling too quickly.
Many view him as having the highest upside of this year's class because of his arm talent. But his QBASE 2.0 projection shows the inherent risk that he will not be able to function in the NFL as a pass-first quarterback (though if he plays early on, the offense can be adjusted to highlight his mobility and big-play ability). In the end, he's in line with the other quarterbacks we evaluate: most likely unimpressive but with the potential to exceed expectations. That said, part of his relatively low projection is due to his Scouts Inc. ranking; if the rumors of him going second overall to the Detroit Lions come true, his projection would jump to 0.28 TDYAR/A.
Kenny Pickett, Pittsburgh
Scouts Inc. Rank: 22
|Mean Projection||0.00 TDYAR/A|
|Bust (< 0 TDYAR/A)||49.7%|
|Adequate Starter (0-0.75 TDYAR/A)||26.2%|
|Upper Tier (0.75-1.5 TDYAR/A)||16.0%|
|Elite (> 1.5 TDYAR/A)||8.1%|
Pickett showed significant improvement last season, displaying better pocket movement and above-average arm strength. In addition, while he lacks Willis' athleticism, he has pro-style experience and makes good decisions in general (though many of his throws came off quick reads). Ultimately, he is more of your typical pocket passer, albeit with some rushing ability, and this better balance is reflected in his lower "bust" rate than Willis. Notwithstanding, the model projects him to be a rather risky pick for a team looking for a franchise quarterback.
Matt Corral, Ole Miss
Scouts Inc. Rank: 34
|Mean Projection||-0.03 TDYAR/A|
|Bust (< 0 TDYAR/A)||51.0%|
|Adequate Starter (0-0.75 TDYAR/A)||25.7%|
|Upper Tier (0.75-1.5 TDYAR/A)||15.6%|
|Elite (> 1.5 TDYAR/A)||7.7%|
Scouts view Corral as a relatively athletic but slender quarterback who has exhibited the ability to make good throws while anticipating his receivers coming open. In addition, he threw for over 8,200 yards in college, won the Conerly Trophy in 2021, and was a two-time Manning Award finalist and two-time Davey Award semifinalist. However, Ole Miss' run-pass option offense didn't often require him to make more than quick, relatively simple reads.
As a result, the big question mark is how he will adjust to the significantly more complex NFL game. Our projections are aligned with this hesitancy, but they also give him 23% chance of developing into an upper tier or better starter. His career trajectory is likely to be determined by the type of situation he is drafted into: will he fall into the arms of a spread-friendly team?
Desmond Ridder, Cincinnati
Scouts Inc. Rank: 36
|Mean Projection||-0.22 TDYAR/A|
|Bust (< 0 TDYAR/A)||58.9%|
|Adequate Starter (0-0.75 TDYAR/A)||23.1%|
|Upper Tier (0.75-1.5 TDYAR/A)||12.7%|
|Elite (> 1.5 TDYAR/A)||5.4%|
Ridder was a four-year starter at Cincinnati with dual-threat ability, two factors which are both positively correlated with NFL success. However, his completion percentage and rushing yards per attempt declined last year, which hurts his projection. In addition, scouts have concerns about his accuracy and ball placement. Seen by some scouts as a weaker version of Marcus Mariota or Alex Smith, Ridder enters this draft with a middling projection.
Sam Howell, North Carolina
Scouts Inc. Rank: 50
|Mean Projection||-0.27 TDYAR/A|
|Bust (< 0 TDYAR/A)||59.9%|
|Adequate Starter (0-0.75 TDYAR/A)||23.2%|
|Upper Tier (0.75-1.5 TDYAR/A)||12.0%|
|Elite (> 1.5 TDYAR/A)||5.0%|
A three-year starter at North Carolina, Howell helped his QBASE 2.0 projection by improving his rushing game significantly last year (his rushing yards per attempt increased from 1.6 to 4.5). However, he also hurt his projection by regressing as a passer (his completion rate decreased from 68.1% to 62.5%). This is why our model still projects him as a relatively risky pick in comparison to the rest of our sample. An interesting question is whether NFL teams will be nervous about North Carolina quarterbacks after what happened with Mitch Trubisky.
Carson Strong, Nevada
Scouts Inc. Rank: 93
|Mean Projection||-1.67 TDYAR/A|
|Bust (< 0 TDYAR/A)||93.9%|
|Adequate Starter (0-0.75 TDYAR/A)||4.8%|
|Upper Tier (0.75-1.5 TDYAR/A)||1.1%|
|Elite (> 1.5 TDYAR/A)||0.2%|
Since the functional mobility model first debuted in 2018, NFL teams have increasingly used a college quarterback's rushing ability as a key factor in their evaluations. To be clear, pure rushing ability doesn't dictate success by itself, since each quarterback must also have the requisite accuracy, arm strength, and intangibles to succeed in the NFL.
However, Strong is a clear demonstration of how QBASE 2.0 will grade a quarterback who put up -4.1 rushing yards per attempt in his last college season. This lack of mobility means that he has a very high chance of being a bust. Is it possible that he could be successful as a classic pocket passer in the right situation? Sure, but it's not likely.
Will a quarterback be drafted highly this year? Most likely yes, given that need and desperation have historically clouded teams' judgement. And hitting on a quarterback is so important that teams are understandably willing to take bigger risks than with other positions. But the warning signs this year are strong, with our model projecting that none of these quarterbacks will perform significantly above replacement value. In turn, these low projections mean that our model is less predictive than usual about the order in which these quarterbacks will be drafted, especially considering a team could fall in love with one of them as the New York Giants did with Daniel Jones. Ultimately, the story of this year's QBASE 2.0 is that buyers should beware.
That being said, each of the first five quarterbacks listed above has a 15% to 25% chance of becoming a high-quality starter. Which means that collectively, there is a very good chance one of these quarterbacks outperforms our projections. Specifically, there's a 28% chance that at least one of them becomes an elite starter and a 68% chance that at least one becomes an upper-tier starter. We just aren't willing to go bet on any of them to be the one.
Our methodology this year is consistent with that from last year. This year, our sample begins in 2005, and our data have been updated in line with last year's NFL season. A big challenge in 2020 was the effect of the pandemic on opponent abilities (for example, Zach Wilson) and time lost (in the case of Trey Lance, a whole season). And there are still concerns this year due to players missing games due to COVID precautions. We hope these issues dissipate over time.
Our main dependent variable is total DYAR per attempt (TDYAR/A), which we defined last year. When compared to the more traditional variable DYAR/A, our variable gives a small boost to mobile quarterbacks. Interpreting each quarterback's projection is straightforward. A value of 0 is replacement level, whereas any value over 1.5 is indicative of a Hall of Fame career. We run 50,000 simulations to provide a distribution that each quarterback falls within a particular range. Since 2005, few quarterbacks have put up elite numbers in the NFL: Patrick Mahomes, Aaron Rodgers, Matt Ryan, Justin Herbert, Deshaun Watson, and, with rounding, Russell Wilson. Anything is possible, but no one in this year's class is projected to come close.
This article originally appeared on ESPN+.
Jeremy Rosen is a doctoral student of economics at Georgetown University. Alexandre Olbrecht is a professor of economics at Ramapo College of New Jersey and the Executive Director of the Eastern Economic Association. The views in this column are expressly our own and do not represent the views of Georgetown University, Ramapo College, the State of New Jersey, or the Eastern Economic Association.