Make-or-break year for Josh Allen? A progress report on the Bills QB, and what's ahead in 2020

This week, I'm going to check in with progress reports for quarterbacks from the 2018 NFL draft. Five quarterbacks were drafted in the first round that year, and I'll take detailed looks at Lamar Jackson (Tuesday), Josh Allen (Wednesday), Baker Mayfield (Thursday) and Sam Darnold (Friday). Sorry, Josh Rosen; I'll get to you another time. 

Five months ago, Josh Allen delivered the most Josh Allen performance of his career in an overtime playoff loss to the Texans. In his first playoff start, he produced evidence supporting every single possible opinion you can hold about him after his second pro season. Allen made big plays with his arm and his legs and made incomprehensible mistakes, occasionally on the same drive. You could credibly argue that he nearly won the game for the Bills or that he was the biggest reason they lost in Houston. Allen optimists and pessimists alike were given more fuel for their respective fires.

The truth is that he is neither the quarterback Bills fans might want him to be or the quarterback who gets memed for missing wide-open receivers with errant passes. He's both of those guys, and there's no telling when Good Josh or Bad Josh is going to show up from possession to possession or even snap to snap. As a neutral observer, few quarterbacks in the league are more entertaining to watch.

There are two things I can say for sure about Allen. One is that he has improved since Buffalo drafted him No. 7 overall in 2018. The other is that the skills he was expected to have coming out of Wyoming bear little resemblance to even the good version of himself from a year ago. The Bills can win with him, as last season proved, and they're rightfully favorites to win the AFC East in 2020. If they want to win a playoff game and advance past the wild-card round for the first time since 1995, though, he needs to take another step forward.

Jump to a section:
How Allen improved last season
The ideal comp for Allen is ...
The big issue: Accuracy
Will he ever throw deep effectively?
How the Bills can help their QB
Why 2020 is a big year for Allen

Allen's 2019 in review

Let's start by taking a broader look at where Allen stands among starting quarterbacks. An uneven 2018 season led the Bills to invest in virtually an entire new offense around him: Twelve of the 15 players who lined up for more than 20% of the offensive snaps for the Bills were new to the roster, with Allen, left tackle Dion Dawkins and wideout Isaiah McKenzie the exceptions.

Buffalo finished last season 21st in offensive DVOA, up from 31st in 2018. Allen was able to stay healthy for the majority of the 2019 campaign after starting just 11 games as a rookie; he missed part of the second half against the Patriots in September after suffering a concussion and then sat out all but two drives of the Bills' meaningless Week 17 loss to the Jets. He faced the league's 12th-toughest slate of opponents.

The Bills went 10-6 but were 1-4 against playoff teams during the regular season, and that win came against the Titans with Marcus Mariota at quarterback. That game was one of five Buffalo wins that came against a quarterback who was either a backup or the lesser of his team's quarterback options during the season, including wins over the Giants (Eli Manning), Broncos (Brandon Allen), Steelers (Devlin Hodges) and Washington (Dwayne Haskins).

Allen authored a number of those victories, tying for the league lead with four fourth-quarter comebacks and five game-winning drives in the final quarter, according to Pro Football Reference. There's nothing wrong with leading those comebacks, of course, but it's tough to count on leading the league with those drives every season. When I looked at Derek Carr after his breakout 2016 season, I found that both of those statistics are relatively random from year to year.

Allen was more assured in his second season. While he still failed to hit the round numbers of a 60% conversion rate or 7.0 yards per attempt, he finished 24th in both passer rating and Total QBR. An optimist might point out that he was just behind former first overall pick Jared Goff in both categories; a pessimist could point out that the only starting quarterbacks with a worse passer rating than Allen who are assured of keeping their jobs in 2020 are Sam Darnold and Baker Mayfield. Allen and the rest of the Buffalo passing attack finished 22nd in the league in win probability added through the air.

The biggest stride he made, by a significant margin, was with his decision-making. He was far more careful with the ball as a passer, especially after a three-interception start against the Patriots in September. He nearly cut his interception rate in half, as it fell from 3.8% in 2018 to 2.0% in 2019. Football Outsiders tracks an adjusted interception rate statistic, which includes things like dropped interceptions and picks thrown on Hail Mary attempts. By that figure, Allen's adjusted interception rate was 2.6%, which ranked 15th in the league, ahead of Drew Brees, Matt Ryan and Dak Prescott.

That improvement wasn't a fluke on tape, either. While Allen wasn't always the most accurate quarterback, most of his misses were overthrows and in places where the ball wasn't likely to be intercepted. He made risky decisions in the playoff game against the Texans, but he routinely took safer throws and made decisions to protect the football, even when he might have been able to squeeze in a tougher throw downfield. He was Russell Wilson-esque when it came to scrambling away from pocket pressure, but while Wilson is able to turn many of those scrambles into big passing plays, Allen wasn't typically able to do the same.

When I wrote about Allen in December 2018, my concern was that he would find it tough to generate as much value from scrambling as he did as a rookie. He scrambled for 507 yards that season, the most scramble yards for any quarterback in a single season over the past five years. The only players to scramble for more yards than Allen in a single season over the past decade were Mike Vick and Colin Kaepernick.

Allen didn't scramble quite as much in 2019, but his 313 scramble yards still ranked fifth in the league. He continued to generate significant value as a runner by picking up more yardage on designed runs and moving the chains in key situations. He scored nine rushing touchdowns and racked up 42 first downs on 109 carries. He converted 16 times on 21 tries on third or fourth down with 2 yards to go or less, a figure topped only by Seahawks back Chris Carson.

By raw numbers, Allen was the second-most-productive rushing quarterback in the NFL behind Jackson, although the Ravens starter is in a league of his own. Allen was fourth in the league in expected points added as a runner (EPA) behind Jackson, Deshaun Watson and Daniel Jones, owing most likely to his fumble issues. According to Football Outsiders, Allen fumbled six times on 95 carries (excluding kneel-downs). The only quarterback who fumbled more frequently was Carr, who fumbled five times on just 19 attempts. Factor in six fumbles as a passer and Allen's 14 fumbles were third most in the NFL. It's something for him to work on in 2020.

The best comp for Allen is ...

At this point, Allen generates much of his value as a quarterback by running the football effectively and avoiding interceptions. For a guy who came into the league with comparisons to players like Ben Roethlisberger, Carson Wentz and even Patrick Mahomes, he has been a totally different sort of player. Let's throw out that 2018 season and the abbreviated Week 17 performance against the Jets and make a comparison between Allen and a familiar quarterback to Bills fans:

Those fans didn't love it when I made this comparison back in 2018, but with Allen cutting his interception rate and serving as a game manager at his best, his production is remarkably similar to that of Tyrod Taylor during Taylor's three-year run in Buffalo. They're about as productive in terms of yardage, with Taylor turning the ball over less frequently. Taylor attempted slightly more difficult passes, though they both threw deep at a rate well above league average.

There's nothing inherently wrong with Taylor, who I thought deserved more credit than he received by the end of his time in Buffalo. Allen can be a viable starting quarterback if he continues to approach Taylor's level of play, but he's still got a ways to go, as Allen's 48.1 QBR from the first 16 weeks of 2019 is still more than 13 points short of Taylor's 61.7 mark as the Bills' starter. It's just a little strange that Buffalo clearly grew frustrated with Taylor, traded up in the first round to draft his replacement, and then appear to have stumbled onto a less effective clone.

Does Allen have more upside than Taylor? Based on expectations before the draft, it's fair to say that's the case. It's clear that Allen has improved, and if he continues to improve at the pace he did over the past year, he'll eventually be better than Taylor was for Buffalo. Given that many of the improvements Allen made were about making safer decisions at the expense of the exact sort of big plays he was supposed to offer as part of that upside, though, it might be tough for him to make that sort of leap and then balance what he has learned as a pro with what the Bills drafted him to do.

Allen's accuracy issues

When you look closer at Allen as a passer, things aren't great. Let's use NFL Next Gen Stats. Allen completed 58.9% of his passes last season. When Next Gen Stats adjusts for the location and speed of his receivers and the defenders on the other side of the field to estimate what his completion percentage "should" have been, it wanted him to complete 62.6% of his throws. By that measure, Allen ranked 24th of the 26 passers, ahead of only Jacoby Brissett and Gardner Minshew.

As is the case for several other websites, Next Gen Stats also tracks its version of a success rate statistic, which is designed to measure how effective a quarterback is at keeping his team on schedule to move the chains. Their definition of a successful play comes when a quarterback picks up 40% of the yards needed for a conversion on first down, 50% on second down, or 100% on third or fourth down. (To cite another source, Football Outsiders doesn't publish a passing version of their success rate stat, but DVOA adjusts performance for down and distance, and Allen ranked 28th out of 34 qualifiers there.)

Allen finished last among the 26 passers in success rate, keeping the Bills on schedule a mere 42.4% of the time. Much of that comes as a result of what he did on third and fourth down. While he was a great runner in short yardage, his 27.2 QBR on third and fourth down ranked 24th out of those 26 qualifiers, ahead of only Darnold and Mitchell Trubisky. When you don't consider his runs, Allen converted just 32.8% of his third/fourth-down dropbacks into first downs as a passer, which was the second-worst rate in the league. He threw only two interceptions on 128 attempts in those situations, but the Bills were more concerned with having him avoid turnovers than they were with their second-year quarterback moving the chains.

While he made strides in this area, accuracy is still a major concern at all levels. Per Trumedia, 23% of Allen's passes were off target, which was the worst rate in the league among those 26 passers. As was the case with Tom Brady, who has been at or near the bottom of this category for years, I'm not concerned that this alone is proof Allen can't become accurate. After you remove the off-target passes from the equation, Brady completed 77.7% of his passes in 2019. Do the same thing for Allen and he completed 76.3% of his on-target throws. If he could just reduce his off-target rate from 23% to 20% or so, it would go a long way.

It's difficult for quarterbacks in the modern era to succeed with a completion percentage below 60, although it's not impossible. Cam Newton was able to pull it off for most of his Panthers career by serving as a valuable runner and throwing downfield effectively. Allen has the first part down. The second part is one of my biggest concerns about Allen and the step he absolutely needs to take to become even a league-average passer.

Can Allen throw deep?

Allen was drafted, in part, because of his incredible arm strength. Nearly one-third of his passes at Wyoming traveled 15 or more yards downfield, which was the highest rate among the five quarterbacks from the 2018 class by a comfortable margin. Last season, 23% of his passes traveled 15 or more yards downfield, which was the sixth-highest rate in football. Expand out to 20-plus yards and only Jameis Winston and Aaron Rodgers threw downfield more frequently than Allen.

The problem is that those throws weren't very productive. Allen ranked 21st in passer rating and 24th in QBR on deep passes (as defined by the NFL). He was 25th in completion percentage and yards per attempt. According to Next Gen Stats, he completed 33.7% of those deep throws when he actually should have hooked up on 39.5% of his targets, the fourth-worst mark in the league. Only Kyle Allen, Goff and Darnold were worse.

To be fair, the numbers also suggest that Allen didn't get much help from his receivers on those throws. Six of his 97 regular-season deep targets were dropped, which was the second-highest rate in the league behind Wentz. Much of that owed to an uneven season from rookie tight end Dawson Knox. I should also note that Allen didn't really have drop troubles elsewhere; his overall drop percentage in 2019 was 3.7%, which ranked 12th among those 26 passers and just above the league average of 3.4%.

This bore closer inspection, so in addition to watching a bunch of Allen starts from throughout the season, I watched every single deep pass he threw last season, including from the wild-card loss to the Texans. I can't argue that drops were a problem on deep passes; in addition to Knox, the otherwise sublime John Brown dropped a pass on what was the best rep of Allen football I've seen so far.

At the same time, it would be naive to pretend drops were what separated Allen from being a successful deep passer. By my estimation, more than 40% of his deep throws were uncatchable, either because they weren't in the wide receiver's range, carried out of bounds, or hit a defender on the way to the receiver. I counted at least six would-be touchdown passes he left on the field, including a pass against the Giants when Brown was 4 yards past the last defender and Allen overthrew him by 4 yards.

The second Patriots game is a good example. Allen hit two big passes in that game, including a near-touchdown to Knox to set up the first score and a beautiful throw under pressure to Brown for a 53-yard touchdown in the second half. At the same time, he flat-out missed two other possible touchdown passes to Knox off of the same smash concept, including one in the first quarter and then again on the final drive.

My biggest concern with Allen pops up in these deep throws but really shows up at all levels of the field: It's tough to find many plays where he throws with anticipation. It's something he worked on with Jordan Palmer during the offseason, but in so many cases, Allen has to see something actually come open before he gets rid of the football. This goes in the opposite direction, too, where he will see a passing window come open for a throw that isn't really available.

As an example, take one of the most famous plays from Allen's season, when he threw what appeared to be a bomb in double coverage to fullback Patrick DiMarco in overtime during the Texans game:

Allen escapes pressure here and works through his progressions to eventually get to DiMarco, who I'm guessing isn't high on the list of likely targets. When he spots the fullback, DiMarco is actually past the cornerback and open, although it's 27 yards downfield and all the way on the opposite sideline. It's an unrealistic throw; DiMarco is not fast enough to keep the cornerback and safety from converging on any possible throw, and even given his arm strength, there's no way Allen's going to be able to hit DiMarco in a hole between the two defenders without giving them time to catch up.

The routes Allen does throw with anticipation are the go, fade and back-shoulder routes, where the risk of throwing into traffic or a pass finding an unanticipated defender is at its lowest. The game-tying touchdown he threw to Brown in the comeback win over the Jets is a good example, although he isn't anywhere near as comfortable making those sorts of throws over the middle of the field. While he was best at completing dig and deep over routes, those completions really came where he could see a wide receiver clear out the coverage with a go route and open up the underneath throw. Even then, he usually waited for the dig to come clear before releasing the ball.

Can Allen learn to trust and anticipate those throws coming open and get his passes out before he actually sees them open on the field, or is that something quarterbacks simply have to have faith in before they get to the NFL? Allen took a few small strides in this direction versus what I saw from him as a rookie, but he needs to get much better at anticipating deeper routes coming open before he'll massively improve as a downfield passer. He also has to get better with his ball placement, as there are too many times in which he'll make even a completed pass hard for the receiver by forcing him to slow down, leap or dive for the ball, eliminating the after-catch opportunities that would have existed if the pass was thrown in stride.

The good thing is that he does usually find the open receiver when he has time, even if the throw doesn't end up where you might hope. He was more comfortable working in the pocket as a sophomore and didn't rely on his legs to bail him out when he didn't get the defensive look he was expecting. You can clearly see Allen going through his progressions and not panicking when his first or second read isn't there.

There are moments when you can see Allen developing almost in real time, like this rep against the Broncos, who have four defenders locking down the two-man route as he works toward the left side. Allen's footwork is choppy and he ends up almost horizontal to the line of scrimmage at times, but he tries to reset himself over and over again to maintain a passing angle, doesn't force a throw and eventually does something spectacular by making two Broncos miss before running to the opposite side of the field for a big gain.

How can the Bills help Allen develop?

This isn't all on the quarterback, obviously. The Bills can do their part to try to make things easier for Allen. They invested heavily in rebuilding the offense before last season, and general manager Brandon Beane followed that up by sending four draft picks to the Vikings for star wide receiver Stefon Diggs.

Everyone is expecting there to be a moment when Diggs gets angry for not getting an accurate throw from Allen, but as I wrote back in February, the Diggs trade is a risk worth taking for the Bills. Even in a draft with loads of wide receiver talent, the chances of landing someone as talented as Diggs were going to be pretty slim. Brown was a legitimate No. 1 receiver last season and would have had even bigger numbers with better quarterback play, but given his injury history, there's nothing wrong with the idea of having Diggs as the top wideout while shifting Brown into the second spot and using Cole Beasley as a third option.

Buffalo also drafted running back Zack Moss in the third round, presumably to replace Frank Gore in the lineup. Gore had basically ceded his playing time to Devin Singletary by the end of the season, and while Singletary was a much more effective player with the ball in his hands, the Bills had reservations about his pass-blocking. It almost cost them their season, as he missed a block on a crucial third-down snap that forced Allen into an intentional grounding penalty and knocked the Bills out of field goal range in the fourth quarter of the Texans game. Allen can scramble away from free rushers, but the Bills need to do a better job of protecting him.

The other rookie who struggled in pass protection last season was right tackle Cody Ford, who looked overmatched at times and committed eight penalties for a team-high 70 yards. Ford was rotating early in the year with veteran Ty Nsekhe before the 34-year-old suffered an ankle injury, which limited Nsekhe to just 35 snaps in the second half of the regular season. Buffalo undoubtedly wants to develop Ford into a starter, but if he can't protect Allen, the team might need to pick its priority and opt for a heavier dose of Nsekhe on the quarterback's frontside.

Offensive coordinator Brian Daboll did a good job of creating opportunities for Allen to use his legs, but there are some concepts worth stealing for the Bills over the offseason. Allen isn't Lamar Jackson, but the Bills have the depth at tight end and the quarterback to steal some of the concepts (and really the window dressing) the Ravens use with the 2019 MVP. We saw a hint of this on the opening play of the Texans game, when Buffalo used jet motion before the snap to tease the Texans into shifting their linebackers before running a pin/pull sweep the other way with Allen for a huge gain. They probably have Greg Roman's old playbook from the Taylor days laying around somewhere.

The Bills could also do more in terms of play-action, in part because play-action makes everyone better. The average passer rating for a pass attempt without play-action in 2019 was 88.8. The average play-action pass produced a passer rating of 105.5. That's essentially the difference between Philip Rivers' and Patrick Mahomes' passer rating last season. No disrespect to Rivers, but I'd rather do the thing that turns my quarterback into Mahomes as much as possible.

Allen was also much better when the Bills used play-action; his QBR jumped from 25th in the league among qualifying passers without a play-fake to the 13th-best mark in the league with play-action attached. Buffalo didn't really use a fake all that frequently, though, as his play-action rate was 21st in the NFL. As Allen gets more comfortable, more play-action concepts should increase his efficiency and effectiveness.

The 2020 outlook

This is a hugely important season for Allen. It's difficult to believe that the Bills would give up on him if he has a disappointing 2020 campaign, but look at the Bears with Trubisky. Like Allen, Trubisky had a difficult rookie campaign and followed it with a much-improved sophomore season after his organization surrounded him with talent. Like Allen, Trubisky took some of the credit while his defense propelled his team into the playoffs. And like Allen, Trubisky mixed ugly stretches of play in that playoff game with moments of brilliance, giving everyone something to back up their opinion heading into Year 3. Trubisky was a trendy MVP bet before the 2019 season, but by the offseason, the Bears were firing their offensive staff, trading for Nick Foles and declining Trubisky's fifth-year option.

It's easy to compare Allen to highly touted quarterbacks who made an impact as a runner while struggling with their accuracy, such as Trubisky, Blake Bortles or Jake Locker, whose era-adjusted passing stats come pretty close to Allen's. (Another name that pops up with remarkably similar second-year stats: Tyler Thigpen.) None of those guys worked out the way their organizations planned, and if Allen doesn't improve his accuracy and complete more deep passes, I suspect he'll go in the same direction.

At the same time, he has shown an interesting career path through two years. He came into the league expected to be a pocket passer who scrambled to make plays with his arm and struggled with his decision-making. Instead, he has evolved into something closer to Taylor, a run-happy quarterback who avoids mistakes, even at the expense of the big plays that were going to be his calling card.

If Allen can keep what he has learned as a pro and use the addition of Diggs to bring back some of the successful deep shots we saw from him at Wyoming, he can be a valuable quarterback in 2020. If he isn't able to evolve, though, we'll start hearing whispers about Allen's future as an unquestioned No. 1 quarterback. He has improved, but the Bills are too deep and talented to settle for what he was in 2019 as their long-term solution.