Like plenty of others in the football nerd/analyst community, I was shocked by the rise of Wyoming’s Josh Allen up the boards heading into the 2018 NFL draft.
Whereas Oklahoma’s Baker Mayfield had been one of college football’s most productive and efficient quarterbacks, and Louisville’s Lamar Jackson had produced 7,000+ passing yards and 3,000+ rushing yards in two years, and Sam Darnold had supercharged a USC offense that was growing stale (and would again be stale after he left), Allen’s stats were merely OK. In 2016-17, he had completed 56% of his passes with a Total QBR ranking of 45th, and Wyoming’s offense averaged a No. 88 offensive SP+ ranking.
To see Allen in the same light as Mayfield or Darnold, or in a greater light than Jackson, was to see theoretical tools — namely, an enormous arm and good athleticism — over actual production and knowledge of the QB position. It was the Kyle Boller fiasco all over again only worse because Boller had gone 19th in the draft — Allen would go in the top 10!
I was quite right, until the moment when I became quite wrong.
This year, there are five quarterbacks who appear to be locks for the first half of the first round: Trevor Lawrence, Justin Fields, Zach Wilson, Trey Lance and Mac Jones. History says all five won’t succeed, but which ones do may be more correlated with where they land than anything else. Which teams will put the best infrastructure around their new quarterbacks and help mitigate the flaws that each signal-caller comes into the league with? (More on that later.)
And that brings us back to Allen.
Buffalo took Allen seventh overall, and for two years the skepticism proved warranted. Nathan Peterman beat Allen for the starting job but flamed out so quickly and famously that Allen was starting by Week 2 as a rookie. He averaged only 4.4 adjusted net yards per pass attempt* (ANY/A) that year and improved only to 5.7, still below average, in 2019. His Total QBR over the two years ranked 27th. He was using his athleticism to decent effect — 1,141 rushing yards, second among QBs behind Jackson — but outside the pocket, where this athleticism should shine, his QBR was an abysmal 12.1.
* ANY/A combines total passing and sack yards, applying a 20-yard bonus for touchdowns and a 45-yard penalty for interceptions, and dividing by total passes and sacks.
Buffalo made the playoffs in 2019, but that was primarily the product of a top-10 defense and three wins over the lowly Jets and Dolphins. The Bills were succeeding despite their top 2018 pick, not because of him. Mayfield had performed only slightly better, and both Darnold and fellow first-rounder Josh Rosen had been abject disasters, but that didn’t justify Buffalo’s decision, especially when Jackson, taken 32nd, was the reigning MVP.
Instead of panicking, the Bills remained dedicated both to their young QB and to finding ways to surround him with talent.
In 2019-20, they signed sturdy linemen Mitch Morse, Daryl Williams and Brian Winters, added an efficient possession man (Cole Beasley) and a burner (John Brown) to the receiving corps, and scored huge wins in the draft’s middle rounds, landing players like RB Devin Singletary, WR Gabriel Davis and TE Dawson Knox. They made a huge trade in 2020, too, sending a first-round pick and an assortment of others to Minnesota for receiver Stefon Diggs.
Heading into his third season, Allen had an excellent line, a top receiver with an array of complementary weapons and what would turn out to be some of the most progressive play-calling in the league from coordinator Brian Daboll. The defense was still good. If Allen had it in him to take a big step forward, this was the time to do so.
He took a very, very big step forward.
“Third in Total QBR” big. Allen’s ANY/A average rose to 7.8, and he cut his sack rate in half. First-down passes are life hacks for avoiding blitzes and pass pressure, and Daboll called the most first-down dropbacks in the NFL. Allen ranked sixth in first-down yards per dropback (8.2) and third in QBR on play action (90.9). Almost 25% of his throws were to receivers deemed “wide open” by Sports Info Solutions (sixth in the league), and only 13% were into tight windows (27th).
Daboll utilized Allen’s arm strength not by having him throw 60-yard bombs or the “off-platform” mixtape passes increasingly common at prospects’ pro days. Instead, Allen threw long sideline-to-sideline passes — his average pass traveled 21.4 yards of true air distance, most in the league. Allen’s sturdy line and the lack of blitzing gave him time to throw, and he even improved to 13th in QBR out of the pocket.
Teams will adjust, of course, just as they did against Jackson following his MVP season. To remain one of the best quarterbacks in the NFL is to constantly adapt and evolve, and not everyone pulls that off. Still, scheme, personnel and personal development converged to produce a third-year leap unlike just about any we’ve seen.
Allen’s success to date has altered my perspective on the importance of nature vs. nurture in the creation of a successful quarterback.
We spend months each winter and spring arguing about top quarterback prospects. We talk about hand size and pocket presence and arm strength and whatever minute details we can find to occupy debate time through April, but it’s possible that we’re looking at things too much in one direction (the player) and not enough in another (the team).
To rank the success of 2018’s first-round draft class to date is to also rank the level of organizational quality and commitment provided to each — the guys drafted by the Ravens and Bills (Jackson and Allen) have fared the best, followed by the Browns (Mayfield) and, far further down the list, the Jets (Darnold) and the two organizations in two years who helped assure that Rosen had no chance.
Individual quality matters, of course — there’s nothing saying Rosen would have had the same level of success if he had gone to Buffalo instead of Allen — but where said individual lands, and what his organization does for him, might matter just as much.
What does this mean for the 2021 draft class?
There’s really only so much you can do with a player’s college stats. Football Outsiders’ QBASE projections components like experience, basic averages, opponent adjustments and projected draft position to project performance. It does a solid job, but the margin for error remains high.
My primary contribution to the projections conversation has been more proof-of-a-negative than projection of anything positive: Unless your name is Mahomes, you probably aren’t going to top your college stats.
Below is a chart for every quarterback who 1) was drafted between 2011-20, 2) played at the FBS level and 3) threw at least 300 passes in his first four pro seasons. (Why four years? Because that’s typically the length of a rookie contract, aka how long you’ve got to figure out what you have in a given QB.)
Of the 52 quarterbacks in this sample, only two have produced an ANY/A average higher in the pros than in college: Patrick Mahomes and Trevor Siemian. A third, Lamar Jackson, has almost perfectly matched his numbers: 7.20 in college, 7.19 in the pros.
Mahomes and Siemian have different reasons for their outlier status: Siemian’s average at Northwestern (5.07 ANY/A) was so bad that he exceeded the numbers even while playing at an extremely average professional level. Mahomes, meanwhile, is Patrick Damn Mahomes, an outlier in almost every possible way (and a player who landed in an extremely QB-friendly situation in Kansas City).
Jackson could add his name to this outliers list with a standout 2021 passing performance, though it bears mentioning that he’s topped 6.7 ANY/A only once in three seasons. Allen, meanwhile, would likely have to lead the league in ANY/A to overcome his terrible first two seasons and drag his four-year average above his college average (6.7).
There’s only the slightest correlation between strong college stats and strong pro stats. Strength-of-schedule differences, the variety of schemes at the college level and, yes, the organizational effects of where a given QB lands in the pros all cloud the picture. But we can say with near certainty that your college averages are your pro ceiling, especially during your first contract.
This is good news for 2021 prospects because, unlike in previous seasons, all of the top guys have great stat lines.
Each of the five quarterbacks likely to go in the first round averaged at least 8.7 ANY/A in their respective careers, which means that their respective “ceilings” are all high. There are no screaming red flags — a Daniel Jones (5.4 college ANY/A) or Jake Locker (same) — in the bunch, even when you add in the second-tier prospects.
Good stats or not, odds are minimal that all five top prospects will become successful pros. But instead of trying to predict which players will succeed or fail, let’s talk about one thing that each prospect could use from his professional home the most to assure early success. Anyone would benefit from a comprehensive Allen-in-Buffalo situation — unwavering commitment, great personnel moves, progressive and finely tuned playcalling — but let’s discuss what each player could use the most.
Trevor Lawrence needs help in man coverage
An all-world recruiting prospect, Lawrence stepped into the lineup early in his first season at Clemson, surged late in the year and played an almost perfect national title game against Alabama. He didn’t lose his first game until the national title game the next season, and it almost felt unjust that his career ended with only one title and zero Heismans.
Lawrence can make every throw from every angle, and he’s proved to be a natural-born team leader as well. He’s everything you could want in a No. 1 pick, but his statistical résumé does have a pretty big hole in it when it came to facing man coverage.
Top QB prospects (sans Lance) vs. man coverage, 2018-20:
Completion rate: Mac Jones 66%, Zach Wilson 61%, Justin Fields 58%, Trevor Lawrence 53%
Yards per dropback: Jones 11.4, Lawrence 8.4, Fields 7.9, Wilson 7.9
Raw QBR: Jones 95.8, Fields 91.5, Lawrence 83.1, Wilson 69.0
After completing 58% of his passes, with a raw QBR of 90.2, as a freshman against what Sports Info Solutions deemed man coverage, these numbers fell to 48% and 74.1, respectively, in 2019-20. To be sure, Lawrence’s receiving corps weren’t as good as Jones’ or Fields’ in 2020, with Tee Higgins off to the pros (he had 908 yards and six TDs as a rookie for the Bengals) and Justyn Ross missing the season with injury. But these numbers had begun to fall sharply in 2019, when both Higgins and Ross were available.
Looking through film — especially in the 2019 College Football Playoff, when he went a combined 7-for-25 for 105 yards, two sacks and no TDs — there was no real pattern to the problems he had. You’d see a poor throw, a drop, a good defensive play, a batted ball at the line, etc. It did seem like he put a lot of trust in his receivers to come down with 1-on-1 balls.
In college, the ratio of zone coverage to man coverage is in the neighborhood of 1.5-to-1 zone. In the NFL, it’s closer to 50-50. If this is an actual Lawrence problem, it would behoove whoever drafts him (OK, it’s going to be Jacksonville) to have a couple of receivers capable of consistently standing out against man coverage.
The Jaguars’ 2020 receivers were below average in this regard. Against man coverage, NFL wide receivers and tight ends tend to average about a 62% catch rate and 12.1 yards per catch, with about 25% of these passes thrown into what is defined as a tight window (separation between receiver and nearest defender is less than 1 yard at the ball’s arrival); Jacksonville’s top five (Laviska Shenault, DJ Chark, Keelan Cole, Chris Conley and tight end Tyler Eifert) averaged 56%, 11.5 and 30%, respectively. New acquisition Phillip Dorsett, who missed 2020 with injury, averaged 63%, 11.5 and 22% in 2018-19. To maximize its investment in Lawrence, Jacksonville needs an obvious upgrade out wide.
Justin Fields needs a playcaller who uses the entire field
After transferring from Georgia in 2019, Justin Fields succeeded Dwayne Haskins at Ohio State and finished his career with extremely similar stats.
Fields’ career stats (2018-20): 68% completion rate, 9.3 ANY/A, 91.7 Total QBR
Haskins (2017-18): 70% completion rate, 9.7 ANY/A, 85.2 Total QBR
This was great for the Buckeyes — after Haskins went 13-1, Fields proceeded to go 20-2 with two CFP appearances — but it doesn’t ring the right bells when it comes to projecting him as a pro: Haskins has thus far been a first-round bust.
With extra context, however, we see that while Fields’ raw stats were similar, he was a very different quarterback.
Pct. of passes thrown behind the line of scrimmage: Haskins 24%, Fields 17%
Pct. of passes thrown at least 10 yards downfield: Fields 38%, Haskins 27%
Pct. of passes thrown between the numbers and the sideline: Fields 53%, Haskins 37%
Fields allowed Ohio State to space the field almost perfectly, throwing at every level both vertically and horizontally. Fields was far more mobile than Haskins, as well.
Fields’ confidence in his own playmaking ability occasionally got him into trouble. Indiana and Northwestern baited him into five interceptions and sacked him eight times in 2020, and of this year’s top QB prospects, Fields’ performance against pressure this past fall was easily the worst — his QBR vs. pressure was a paltry 14.6.
Of course, you don’t want any rookie QB facing pressure — no one’s going to be particularly good at handling it early on. And despite his struggles against IU and Northwestern, he still threw for 300 yards against the Hoosiers, still used his legs to solid effect against the Wildcats, and oh yeah, threw for 579 yards and seven touchdowns, with 126 non-sack rushing yards, in two CFP games. Pretty good.
If Fields is to succeed early on at the pro level, though, it will likely come in a system that, like Ohio State’s, uses the entire field. He did that better than any of the other top prospects.
Percentage of career passes thrown between the numbers and the sidelines:
The NFL average for these throws varied significantly, from 57% for Aaron Rodgers and 55% for Josh Allen to 37% for Nick Mullens and 38% for Cam Newton. It’s a strong-armed quarterback’s throw, in other words, and Fields’ arm is both strong and wonderfully accurate.
Simply throwing passes to the sidelines isn’t a marker of quality, obviously. Rodgers and Allen were both great at it and played in great offenses, but other quarterbacks throwing a high percentage of passes outside the numbers — Drew Lock (54%), Nick Foles (50%), Sam Darnold (49%) — didn’t fare as well. Still, whatever Fields’ early potential may be, it will be maximized if the offense that deploys him looks to spread the defense in every way, especially considering what that could do for running lanes on the twice per quarter that he looks to run.
Zach Wilson needs a strong pair of tackles
Zach Wilson’s arm is obnoxiously strong, his mixtape second to none. But for his career, Wilson’s stats are so context-dependent that it’s hard to pull much from them.
In 2019, he produced mixed results against strong early opposition, then missed four games with injury. In his last two games of the year, against San Diego State and Hawaii, he completed only 59% of his passes with four interceptions and no TDs.
In 2020, he was healthy, the skill corps around him was experienced, and he played with a point to prove. He also barely played anyone good — the Cougars’ schedule fell apart in August due to coronavirus-related cancellations and altered schedules. BYU had to piece together a schedule of willing and available teams, at one point playing Coastal Carolina on barely three days’ notice, and only three of 12 opponents finished better than 59th in SP+. Six were 86th or worse.
BYU still ranked seventh in offensive SP+ — an opponent-adjusted figure — and against the three best teams they played (Coastal, Boise State and UCF) Wilson averaged 342 yards per game with a 73% completion rate, seven touchdowns and one pick.
A weak schedule wasn’t the sole reason for his otherworldly 2020 performance, in other words. But the time he had to throw sure helped.
Pct. of passes thrown after at least three seconds in 2020: Wilson 35%, Fields 34%, Jones 34%, Lawrence 23%
Pct. of passes thrown after at least four seconds: Wilson 14%, Fields 8%, Jones 6%, Lawrence 6%
Wilson averaged a 69% completion rate at 20.3 yards per completion on passes thrown after three seconds, and he averaged 62% and 22.7, respectively, after four. He faced so little pressure in 2020 that it altered his internal clock; against the three defenses he faced with a top-30 sack rate (Houston, BSU, Coastal), he went just 10-for-22 under pressure with an ANY/A of 3.5.
To his credit, he quickly showed growth in this regard late in 2020: Against pressure-happy SDSU and UCF defenses following the loss to Coastal Carolina, he went 14-for-21 for 248 yards, two scores and three sacks under pressure. ANY/A: 12.5.
This isn’t a definitive weakness so much as an area where Wilson will have to further adjust, but landing on a team with pass blocking that can last an extra beat might help him most of any top prospect.
As you would probably expect, most of the teams with the top picks in the draft didn’t block very well last year — it’s why they’ve got top picks. Of teams with the top 11 picks, 10 finished 20th or worse in Pass Rush Win Rate (PBWR). The Falcons, with the fourth pick, finished 16th, but most accounts suggest he’ll likely either land at No. 2 (with the Jets) or No. 3 (49ers). The Niners do at least have the freshly re-signed Trent Williams (fourth overall in PBWR as a tackle), at least.
Trey Lance needs a redshirt
One of the factors Football Outsiders’ QBASE takes into account is starting experience. A lack of quality reps is a good predictor of potential NFL struggles, and according to ESPN’s Stats & Information, only three first-rounders since 2002 had fewer career starts than Trey Lance’s 17: Mitch Trubisky, Haskins and Mark Sanchez. This is basically the only box Lance doesn’t check, but it’s a potentially important one, it might be rendered even more important considering all 17 of his starts came against FCS defenses.
He did everything he could to make up for that lack of competition. He averaged 10.3 ANY/A, easily the second-best average of this year’s top QB prospects. And while he had some inconsistent moments against strong James Madison and Illinois State defenses in the playoffs, he had excellent moments, too: against Northern Iowa and Illinois State in the regular season, he was 22-for-33 for 334 yards, six TDs and only one sack. He dominated.
Lance was good inside the pocket (10.0 yards per pass attempt, per ESPN Stats & Info) and outside of it (8.8), he was good under center (10.9) and in the shotgun (9.2), and he didn’t throw his first career interception until his last game, a rusty, only-game-of-the-fall showcase game against Central Arkansas last October 3. You had to blitz him to have any chance (he averaged 10.4 yards per pass attempt against four or fewer pass rushers and 8.3 against 5+), but if you blitzed and he escaped, he could either throw or run a very long way. In the smallest of samples, he showed virtually every skill you could possibly want.
Still, he’s taking an even bigger leap in competition than his 2021 draft counterparts, and he’s doing it with the least experience. The best possible situation for him — and a luxury most top QB picks are not granted — would be one in which a decent veteran is still employed to take most of the snaps this fall.
Plenty of teams seem to have the right intentions in this regard. The Chargers drafted Justin Herbert sixth last year but had veteran Tyrod Taylor signed for until Herbert was ready. (The Dolphins also had Ryan Fitzpatrick for breaking in No. 5 pick Tua Tagovailoa, too, though Tagovailoa also had injury concerns.) In 2019, the Giants still had one more year of Eli Manning to ease in No. 6 pick Daniel Jones.
Granted, Manning struggled enough that the pressure to play Jones grew quickly, and Taylor suffered a freak injury that necessitated tossing Herbert into the deep end. (Turns out, he was ready.) Sometimes it’s hard to keep your top picks off the field, and if Lance ends up on the field early, then as with Herbert, it might work out fine. But if he were to land in a place like San Francisco (which still has Jimmy Garoppolo) or, better yet, Atlanta (Matt Ryan), he could learn and ease in.
Mac Jones needs a star receiver between the numbers
It has proved impossible to separate Mac Jones’ quality from that of his supporting cast. Heisman voters sure couldn’t.
Compare Jones’ production in 2020 to that of Joe Burrow in 2019.
Completion rate: Jones 77.4%, Burrow 76.3%
Yards per completion: Jones 14.5, Burrow 14.1
Total QBR: Jones 96.1, Burrow 94.9
Heisman votes: Burrow 2,608 (first), Jones 1,130 (third)
Burrow’s supporting cast was brilliant — Ja’Marr Chase, Justin Jefferson, Clyde Edwards-Helaire, Terrace Marshall Jr. — but he received due credit for an all-time great season in 2019. Jones topped Burrow in nearly every way, and against an all-SEC and CFP schedule, but finished second on his own team in Heisman voting. Even this writer gave Burrow more credit, ranking him the fifth-best QB of the 2000s and placing Jones in the 30s.
Did Jones have just about the greatest supporting cast imaginable? Of course. He had Heisman winner DeVonta Smith, plus Najee Harris, Jaylen Waddle, John Metchie, etc., around him and an almost infallible offensive line in front of him. He was operating from Steve Sarkisian’s perfectly modernized, motion- and RPO-heavy offense. But he still did his job perfectly. His anticipation and footwork were impeccable, which meant his accuracy was, too. When something didn’t come open immediately, he either stepped up properly in the pocket or wafted out of it to buy time. His QBR under pressure was 78.3, and no one else in FBS was above 51.3.
Accuracy, anticipation, brains, footwork … this seems to be something an NFL-caliber play-caller could work with. But said coordinator won’t have a booming, sideline-to-sideline arm to utilize. See those “passes outside the numbers” percentages above again; Jones wasn’t just lower than the rest, he was much lower. Having a true go-to receiver out of the gates — especially one that can do damage between the numbers — would help him immensely, even if his supporting cast lacks in other areas.
Of the nine teams most likely to draft Jones, here are the leading receivers currently on the roster for each on passes thrown between the numbers:
49ers (No. 3 pick): Brandon Aiyuk (542 yards between the numbers, 7.9 per target), George Kittle (352, 8.8)
Falcons (4): Calvin Ridley (550, 7.7 per target), Russell Gage (472, 7.9)
Lions (7): T.J. Hockenson (462, 7.3), Breshad Perriman (303, 12.1)
Panthers (8): D.J. Moore (622, 9.6), Robby Anderson (527, 7.2)
Broncos (9): Jerry Jeudy (468, 9.4), Tim Patrick (10.2)
Patriots (15): Jakobi Meyers (480, 9.1), Hunter Henry (477, 7.6)
Raiders (17): Darren Waller (550, 7.9), Hunter Renfrow (423, 9.0)
Washington (19): Terry McLaurin (628, 8.1), Curtis Samuel (623, 9.7)
Bears (20): Allen Robinson (653, 8.4), Darnell Mooney (293, 8.6)
Jones worked with Jeudy at Alabama in 2019, who stood out as a rookie in this area of the field. And if Jones falls into the teens, most of the teams in that neighborhood have exciting weapons in that regard. Wherever he lands, though, these could be the players he leans on, and his success will depend on theirs.