clock menu more-arrow no yes

Filed under:

If Josh Allen succeeds, the Bills will have outsmarted basically all regular humans and the entirety of math itself

New, 39 comments

If this works, then stats really are for losers, I guess.

Wyoming v Nebraska Photo by Steven Branscombe/Getty Images

Josh Allen was hyped for a full year before the 2018 NFL Draft as a potential No. 1 pick, and he went No. 7 to the Bills, all to the bafflement of pretty much everyone not employed by the league.

There’s never been a bigger disconnect between what the NFL apparently thinks and what the armchair analysts think.

Personnel people see a prototype athlete whose flaws they can correct, while fans, statisticians, tape-grinders of varying expertise, and most media people see a flawed athlete with a prototypical shape.

He’s a project. Coaches, scouts, and the most prominent mock drafters disagree to extreme degree with eggheads, less prominent mock drafters, and the public on how many tuneups he needs, though.

In the end, somebody’s gonna be wrong.

If Allen succeeds, he’ll be a major statistical outlier, no matter how deeply into the numbers you go.

First, the raw numbers, some of which you’re sick of, but bear with me.

In two seasons as a starter, Allen ranked 32nd (2016) and 73rd (2017) in passer rating. Here’s how his final college season compares to his new peers:

Where Josh Allen’s peers ranked in college

QB Last college year Last college QB rating ranking
QB Last college year Last college QB rating ranking
Baker Mayfield 2017 1
Marcus Mariota 2014 1
Philip Rivers 2003 1
Russell Wilson 2011 1
Alex Smith 2004 2
Ben Roethlisberger 2003 2
Cam Newton 2010 2
Mason Rudolph 2017 3
Case Keenum 2011 4
Andy Dalton 2010 5
Carson Palmer 2002 5
Aaron Rodgers 2004 8
Blake Bortles 2013 9
Carson Wentz 2015 11 (FCS)
Mitchell Trubisky 2016 11
Brett Hundley 2014 13
Tyrod Taylor 2010 13
Derek Carr 2013 15
Matthew Stafford 2008 15
Tom Brady 1999 17
Eli Manning 2003 18
Joe Flacco 2007 18 (FCS)
Dak Prescott 2015 22
Non-Allen average - 23.4
Drew Brees 2000 25
Jameis Winston 2014 26
Sam Darnold 2017 26
Josh Rosen 2017 28
Kirk Cousins 2011 28
DeShone Kizer 2016 29
Lamar Jackson 2017 29
Jay Cutler 2005 60
Jacoby Brissett 2015 61
Matt Ryan 2007 61
Josh Allen 2017 73
C.J. Beathard 2016 78
Josh McCown 2000 90
Trevor Siemian 2014 104
NFL QBs with at least 225 attempts in 2017 and 2018’s top six draft QBs

Nothing is perfectly predictive, but as you can see, the only others to have below-average final college seasons were Beathard (lost his NFL starting job in 2017), McCown (didn’t become a steady NFL contributor until his 11th season), and Siemian (being replaced as an NFL starter). An attempt to find a true statistical comp for Allen usually ends up on Jake Locker, another big, agile QB.


A more advanced system by SB Nation’s Bill Connelly found that a QB’s ceiling in the pros is lower than his college stats, with dozens of previous examples. Well, based on that, Allen’s ceiling is the tall, strong Ryan Mallett. Not Allen’s projection. Allen’s ceiling.


Connelly’s opponent-adjusted Passing S&P+, which grades each unit in FBS on efficiency, explosiveness, and more, has Allen’s 2017 Wyoming passing attack ranked No. 119 out of 130 teams.

Allen didn’t have a lot of NFL talent in 2017, but Wyoming’s rushing offense, passing defense, rushing offense, and special teams all ranked better (or much better), so it’s not like the Cowboys had a hopeless roster.

In 2016, with two NFL draftees by his side (and thus mostly facing defenses with less pro talent), Allen’s passing game ranked No. 52, slightly above average.


So yeah, his best college season wasn’t exactly awesome. 2016’s All-Mountain West first-team QB was Boise State’s Brett Rypien.


Football Outsiders’ QBASE formula mixes conventional wisdom with stats, giving a boost to players who are popular among scouts and mockers. Its all-time college top 10 — Rivers, Palmer, McNabb, Mayfield, Wilson, Peyton Manning, Mariota, Byron Leftwich, Rodgers, and Roethlisberger — gives it credibility as an NFL predictor.

Well, Allen’s sub-zero results are “horrifying,” in FO’s words: “Since 1997, there have been 27 quarterbacks chosen in the top 100 with QBASE ratings below zero. The best was either Josh McCown or Brian Griese.”



Among Pro Football Focus’ grades, PFF finds “NFL throw” accuracy — when making difficult passes, basically — to be especially predictive. In versions of that metric, Allen ranks either fifth or sixth among 2018’s top six QB prospects.

PFF’s big board ranks Allen No. 6 among QBs and No. 35 among all 2018 prospects. PFF grades him below-average in six of 10 traits and above-average in just three: running, intermediate passes, and difficult/high-value passes.


From another system that adds eyeball context to raw numbers:

Sports Info Solutions’ [...] proprietary quarterback metric, IQR (Independent Quarterback Rating) accounts for events out of the passer’s control, including dropped passes and dropped interceptions. In 2017, Mayfield earned the highest IQR on 91 attempts under pressure by a mile—his 119.2 rating was better than Mason Rudolph by 12 points, and blew away Rosen (91.2), Darnold (79.8) and Allen (62.7).

In that stat, Allen also trails Luke Falk, Lamar Jackson, Virginia’s Kurt Benkert, and turnover-prone Nebraska QB Tanner Lee.


Those stumping for Allen as a top prospect might be vindicated, but they’ve still fallen into contortions along the way.

Elsewhere at ESPN, that gets somewhat debunked:

Supporters say he played in an offense that didn’t allow for many gimme completions. Some of that is myth. Allen threw a higher percentage of short passes than Darnold, Mayfield and Rosen, according to ESPN Stats & Information research. In fact, 30 percent of his attempts traveled zero to 5 yards in the air.

Also per ESPN, Allen was among the least accurate of his peers at each of three levels anyway:

ESPN

Also:


Another point made in Allen’s favor: his team was often out-talented. It’s true Wyoming has a middling MWC roster, but the defenses he faced were rarely anything special.

Which 2018 NFL draft QBs faced the toughest competition, over their full college careers?

QB Average opposing pass D ranking Top-10 pass defenses Top-30 pass defenses Top-60 pass defenses
QB Average opposing pass D ranking Top-10 pass defenses Top-30 pass defenses Top-60 pass defenses
Sam Darnold 51.5 3 9 14
Baker Mayfield 54.3 4 11 27
J.T. Barrett 54.4 7 15 27
Lamar Jackson 56.1 7 14 19
Mason Rudolph 60.4 1 8 21
Josh Rosen 66.3 1 5 14
Luke Falk 69.1 2 9 19
Kurt Benkert 69.3 3 6 10
Riley Ferguson 81.8 0 3 9
Josh Allen 82.8 0 3 9
Logan Woodside 83.3 0 4 10
Nick Stevens 85.3 2 5 13
Mike White 85.4 1 3 15
Chase Litton 89.9 0 0 10
Kyle Lauletta 152.6 0 0 2
FBS opponent-adjusted Defensive S&P+ rankings by SB Nation’s Bill Connelly. FCS team rankings by USA Today’s Jeff Sagarin.

And previous non-power first-rounders have put up better numbers against similar schedules:

How Josh Allen compares to previous 1st-round QBs from non-power conferences

Year Player College rating College yards/throw
Year Player College rating College yards/throw
2000 Chad Pennington (Marshall) 157.6 8.6
2002 Patrick Ramsey (Tulane) 126.0 6.8
2002 David Carr (Fresno State) 151.2 8.3
2003 Byron Leftwich (Marshall) 150.9 8.3
2004 J.P. Losman (Tulane) 129.8 6.8
2004 Ben Roethlisberger (Miami, OH) 151.3 8.3
2005 Alex Smith (Utah) 164.4 8.9
2008 Joe Flacco (Delaware) 137.8 7.5
2014 Blake Bortles (UCF) 153.8 8.5
2016 Paxton Lynch (Memphis) 137.0 7.4
2016 Carson Wentz (North Dakota State) 153.9 8.4
2018 Josh Allen? (Wyoming) 137.7 7.8
Sports-Reference

Another argument in Allen’s favor: he might have the strongest arm of any QB ever. But:

And:


Another defense of Allen: his receivers were bad.

Well, first, they were mostly playing against MWC cornerbacks, who are not the world’s greatest. Secondly, Allen’s WRs at least had good hands, even if they weren’t game-changing athletes:


Another: his offensive line was bad. However, he ended up with plenty of time in the pocket, even if he created a lot of it himself.

It’s fair to note the three QBs with the most time are also good at running around — nobody would argue Jackson’s offensive line was actually solid, for example.

Stats are for nerds (per Kiper), so let’s Go To The Tape.

I’m not gonna embed 1,000 video breakdowns, so here’s a mass summary:

Allen has the dream physique, makes throws that range from amazingly good to amazingly bad, has a whole lot to clean up, will be anything from Roger Staubach to Kyle Boller (old Boller scouting reports read a little eerily right now, actually), and might or might not be worth the risk.

And there the debate continues, exactly the same as it was back in February, December, October, and August.

Allen will have a chance to be an NFL quarterback. One side will be right.

The other side will ignore the result as a fluke and continue unfazed.