Tua Tagovailoa's Tale of the Tape

Miami Dolphins QB Tua Tagovailoa
Miami Dolphins QB Tua Tagovailoa
Photo: USA Today Sports Images

NFL Offseason - Tua Tagovailoa Truthers raise some very valid points.

Take this finny fella. He's clearly onto something. Or at least on something:

Perhaps those arguments are a teensy bit pleading. But Tagovailoa finished 18th in both DVOA and DYAR last season: not bad at all for a second-year quarterback. The Miami Dolphins passing game earned a DVOA rating over 20.0% in four of the seven games after Tua returned from his midseason kinda-sorta-quasi-injury-benching, a sign that Tua may be turning the developmental corner.

Sure, Tua appeared to be propped up by an RPO-intensive, dink-and-dunk-heavy system in 2021. He also had a knack for crushing interceptions. But his extenuating circumstances were really extenuating. The Dolphins offensive line finished 30th in adjusted line yards and needed all of that quick passing to rank 19th in adjusted sack rate. The Dolphins offense ranked 29th in rushing DVOA, leaving them playing behind the sticks while forcing their quick game to supplement their rushing game. And while the details of Brian Flores' allegations against the Dolphins and the NFL are beyond the scope of this article (and therefore, preferably, won't dominate the comment threads), his dueling offensive coordinators and quick hooks hardly made him a quarterback whisperer.

Flores has been replaced by The Adorkable Mike McDaniel, a Kyle Shanahan disciple with a quarterback-friendly personality and scheme. Tyreek Hill and Terron Armstead headline a productive Dolphins offseason, particularly on offense. So are the Tua Truthers right? Or are they just engaging in typical local-fan wish-casting about a prospect whose career is listing sideways?

Let's see if there's an answer lurking somewhere in the numbers.

Tua Tagovailoa: The RPO Man

Any discussion of Tua Tagovailoa's performance or the 2021 Dolphins offense in general must begin with the run-pass option. The Dolphins used an RPO concept on 13.0% of their passing plays last season, easily the highest rate in the league. (All data in this article is from Sports Info Solutions unless otherwise noted.) Tagovailoa himself attempted 77 RPO passes, the third-highest total in the NFL, completing 53 for 559 yards.

As Walkthrough discussed in our Matt Corral feature a few weeks ago, the RPO is often used as a crutch for a quarterback who has little idea what he is doing. But it can also be a tactic for mitigating a horrible offensive line or as a jab for an offense that's perfectly capable of throwing some haymakers. As a result, the top five teams in RPOs as a percentage of pass plays in 2021 was a diverse bunch. We'll use adjusted net yards/attempt to get a sense of how successful each team was at executing the RPO:

Top Five Teams in RPO Usage, 2021
Team RPO% ANY/A
MIA 13.0 7.7
PIT 11.4 4.6
GB 10.5 6.5
KC 10.4 6.2
BUF 8.3 8.7

The Steelers used the RPO to conceal the fact that their offensive line was embarrassing, and that Ben Roethlisberger was held together with duct tape and rubber bands. As the ANY/A figure above illustrates, no one was fooled.

The Packers, Chiefs, and Bills, on the other hand, had pretty good quarterbacks and offenses. They used the RPO as a counterpunch when teams played a lot of two-deep shells against them (all three ranked in the top 10 of rushing attempts from an RPO concept, a sign they were facing reduced boxes) and as a red zone tactic.

So what were the Dolphins doing? Protecting Tagovailoa from his awful offensive line, no doubt. Covering for his relative inexperience at times. But they were also enjoying relative success with the tactic, with a higher ANY/A rate than the Packers, Chiefs, and many other teams that used the RPO frequently, such as the Chargers, Eagles, and Colts.

The numbers suggest that we should not write Tagovailoa off simply because he threw a few more RPO passes than Aaron Rodgers and Patrick Mahomes. Being skilled at making easy reads and delivering accurate quick passes is, after all, a good thing. But those cannot be the only things a young quarterback is good at.

What we need to do is take a closer look at what Tua was doing when he wasn't just faking a handoff and flicking a short slant to Jaylen Waddle or Mike Gesicki.

Tua Tagovailoa: Beyond the Valley of the RPOs

Let's filter all those RPOs out of Tagovailoa's passing stats and take a look at where he ranked among quarterbacks with at least 100 pass attempts in 2021:

Tua Tagovailoa, Non-RPO Passes
Comp % 68.5% 14th
Y/A 6.6 31st
ANY/A 5.4 30th
Average Throw Depth 7.1 40th

Ew. This is red meat for the Tagovailoa skeptics. Tua tied Roethlisberger and Daniel Jones with that 7.1-yard average throw depth on non-RPOs. He's wedged between Taylor Heinicke and Mike White in ANY/A, with Taysom Hill just behind him. This is some bad company.

The numbers above suggest that even when the Dolphins weren't executing RPOs, their passing game was loaded with RPO-like substances: short, quick, non-nourishing tosses. There's also lots of evidence to support the fact that Tagovailoa wasn't getting much support from his receivers. Per Pro Football Reference, for example, the Dolphins averaged just 4.3 yards after catch per reception, the third-lowest figure in the NFL, ahead of only the Ravens and Bills. The Dolphins' dropped pass rate of 5.7%, per Pro Football Reference, ranked fourth in the NFL.

In other words, Tua may have been forced to throw short passes to protect his offensive line, only for his receivers to either drop those passes (lowering his completion rate) or get tackled immediately after the catch (lowering his yards per attempt).

If that's indeed what happened, Tagovailoa could indeed enjoy a breakout year now that Tyreek Hill and some running backs who can generate YAC have joined his arsenal and Terron Armstead is anchoring his protection.

But is that really what happened? Let's torture the numbers some more!

Tua Tagovailoa: Pressure Principles and Stick Figures

Tua Tagovailoa only attempted 96 non-RPO passes when pressured in 2021. On the one hand, that's a surprisingly low number (28th in the NFL) for a young quarterback behind an objectively woeful offensive line. On the other hand, it shouldn't be surprising, because the Dolphins offense twisted itself in knots to avoid pressure, as most of the splits we have cited so far illustrate.

Tagovailoa was horrible under pressure. His ANY/A of 1.0 ranked 32nd in the NFL. He threw seven interceptions under pressure, tied for second in the NFL. All quarterbacks are worse under pressure than from a clean pocket, but Tua's issues were pronounced in a way that's not surprising for a young quarterback, but not encouraging either.

Take the pressure AND the crutch of the RPO away and we start to see a slightly different quarterback, though still not a particularly effective one:

Tua Tagovailoa, Not Pressured, Non-RPO
Comp% 75.0% 3rd
Yards/Att 7.2 26th
ANY/A 7.8 17th
Average Throw Depth 6.8 32nd

Tagovailoa threw 220 passes in the sample above, so we have not driven off the sample size cliff. Tua still looks like Dinky McDunkenstein, but he's at least an efficient Dinky McDunkenstein who completes three-fourths of his passes and avoids turnovers. If Myles Gaskin eluded a few more tackles, if three defenders didn't immediately convene on Jaylen Waddle after every quick screen, if holding onto the ball for more than a nanosecond was an option … you get the idea.

Let's get all of those dinks and dunks out of the data by focusing on passes beyond the sticks. Let's also focus on passes where Tagovailoa is not pressured. Finally, let's mix the RPOs back in so we don't granulate the data too finely: throwing past the sticks naturally filters out some of the RPO flings to the flat anyway. Here's what we get:

Tua Tagovailoa, Not Pressured, Past the Sticks
Comp% 54.6% 27th
Yards/Att 8.7 27th
ANY/A 9.4 20th
Average Throw Depth 13.0 31st

Sorry that this is not better news, Tagovailoa truthers. That 20th-overall finish in ANY/A isn't terrible: it's just below Josh Allen and ahead of Mac Jones and Jalen Hurts, though it's also below Carson Wentz, Daniel Jones, and Davis Mills. Only Jimmy Garoppolo ranked below Tua in average throw depth among quarterbacks with 50-plus attempts in this split. Also, Tua's high completion rate evaporates when he's forcing the ball a little further downfield.

Overall, the figures above suggest that Tagovailoa was probably a fourth-quartile NFL starter in situations where the Dolphins offensive line was doing its job and his success wasn't dependent on a teammate doing something after the catch.

On the bright side, the predictive value of non-pressured, past-the-sticks passes is sketchy at best: we should probably not trust any split that rates Carson Wentz ahead of Josh Allen. Also, no matter how we sift through the data, Tagovailoa always ranks near the bottom of the NFL in throw depth. Even if Tua had a noodle arm and his offensive line were historically bad, the Dolphins' reluctance to let 'er rip once in a while would be a little extreme.

Tagovailoa only threw 14 passes of 25-plus air yards last season, a reasonable minimum benchmark for what we would consider a "bomb." Garoppolo threw 18 bombs. Drew Lock threw 19 in six appearances. Tyler Huntley threw 14 in four appearances. Justin Fields, a rookie quarterback behind a bad offensive line, threw 23.

Tagovailoa completed eight of those 14 passes for 328 yards, one touchdown, and one interception. Those are excellent rate stats for this tiny sample. We shouldn't hang our hats on 14 pass attempts, but it looks like he could have been a semi-capable deep passer if given a few more opportunities.

With Tyreek Hill around, he'll get those opportunities.

The Garoppolo-ization of Tua Tagovailoa

To recap:

  • The 2021 Dolphins offense relied heavily on dink-and-dunk tactics, even when they weren't using RPOs at a league-high frequency;
  • Their reliance on short passing was abnormally high, even for a team with a bad offensive line and inexperienced quarterback;
  • An inability to generate YAC limited the effectiveness of the Dolphins short passing game and made Tua Tagovailoa's stats and performance look worse than they really were;
  • Whether despite of or because of all of these factors, Tua produced a nearly league-average DVOA;
  • But when we try to isolate Tua from his environmental factors, we get a mixed message about his capabilities as a downfield passer.

So we're stuck in the sort of circular argument that breeds fan theories and encourages sports talk debate, informed and otherwise.

What's certain is that the circumstances around Tagovailoa have improved in every way: the line is better, the playmaking corps better and deeper, and the coaching staff better pedigreed from an offensive standpoint. Tua should get more mileage out of short passes to Tyreek Hill, Chase Edmonds, and others with Mike McDaniel designing 49ers-flavored game plans. He should have more time to throw deep to Hill, Jaylen Waddle, and others with Terron Armstead anchoring an upgraded offensive line. At the very least, Tua should benefit from the advantages Jimmy Garoppolo has enjoyed when he and his supporting cast have been healthy over the past three seasons. Garoppolo may be sitting by the curb with a "Take Me" sign on his back right now, but the Dolphins would not mind a ball-control offense capable of some deep playoff runs led by a still-young-and-affordable quarterback with room to grow.

I was a Tagovailoa skeptic in 2021, and I remain one, because a "skeptic" is someone who challenges the evidence, not some hater or naysayer. Despite some of the discouraging splits cited in this article, I'm sanguine about his future, in part because he doesn't fit the profile of a fading prospect with a stat profile full of fluff.

If Tua ranked third in YAC instead of third from last, it would be a sign that his league-average DVOA was the result of his supporting cast. If his RPO data looked like the Steelers RPO data, it might be a sign that the Dolphins coaches were desperately throwing tactics at the wall. But the Dolphins were better at the RPO than the Chiefs or Packers, despite the fact that defenses had no reason to fear their deep game or their running game. The numbers suggest that Tua could be very efficient in a timing-based offense with better game plans and better playmakers. Both of which he now has.

So Tua Truthers have a point, as do Tua skeptics, as it almost always the case when we slog through the statistical splits. And the Dolphins have done everything imaginable to help Tagovailoa succeed in 2022. If he fails, it will be Teddy Bridgewater time. And if he does something in between success and failure, the Dolphins possess two first-round picks in 2023. Even a hardcore truther will be able to figure out what happens next.

Meanwhile, in Ashburn, Virginia…

VLAD THE EVIL ACCOUNTANT: Greetings! You must be the new accounting intern. Follow me.

CARSON WENTZ: There must be some mistake. I'm not an accounting intern.

VLAD: Of course you are. You dress like you've never seen a mirror and walk around with a vaguely lost expression on your face. Now sit at this workstation and let me coach you on our system.

CARSON WENTZ: Well, I'm out to prove how receptive I am to coaching, so let's give this a whirl. (Reading screen.) Hmm, it seems that some of our NFL revenues are being funneled into accounts labeled "soccer revenues," "concert revenues," and "monster truck rally revenues." That cannot be right.

VLAD: Don't worry about that! It's called The Juice. That's a macroeconomics term.

CARSON WENTZ: OK, well what about this item here: money that's earmarked for "orphan charities" is being spent on something labeled "rare vintage wines for D.S. to bathe in."

VLAD: Ignore that too. That's just … The Zest.

CARSON WENTZ: Wow. There's sure a lot about accounting that I don't understand. What about this part of the spreadsheet: cutting down Amazon rainforest for wood paneling for the wet bar on a superyacht, displacing the indigenous residents to serve as sweatshop labor in a knockoff NFL jersey factory, then re-labelling the jersey revenues as something called "capital gains losses from regrettable quarterback decisions?"

VLAD: That's called The Nom-Nom Yummy-Yums. Ignore it.

CARSON WENTZ: Gee, this all sounds slightly unethical.

VLAD: Listen kid, this is an easy job. All you have to do to hide those files is check the boxes, all the way down.

CARSON WENTZ: Check … down?

VLAD: Yes! This information is very, um, private! We cannot let anyone intercept it.

CARSON WENTZ: Check down to avoid an interception?

VLAD: Easy, right? Just work within the system. The accounting system, that is.

CARSON WENTZ: (Sweating profusely.) Oh no, the pressure is getting to me. Must … attempt … something heroic! (Pressing keys frantically.) SEND! SEND! SEND TO WASHINGTON POST. SEND TO THE ATHLETIC. SEND TO THE FBI. SEND TO THE JACKSONVILLE JAGUARS DEFENSE. SENNNNNNNDDDDDDDD!

(Chasm opens up in earth, swallows the Washington Commanders organization without a trace. Don't worry though, folks: Ron Rivera, Terry McLaurin, and the other likeable employees were out getting bagels.)

CHRIS BALLARD: And that's how I toppled the most evil owner in the NFL. And you folks thought I was a bad general manager!

Comments

66 comments, Last at 15 Apr 2022, 8:17pm

5 It's refreshing to see an…

It's refreshing to see an analytics analysis that admits the data is confounding and that they don't know.

Because a lot of analytics arguments are just Stephen A. Smith wearing glasses.

16 In my experience with…

In my experience with football analytics, they're pretty good at telling you when something seemingly good is probably a mirage and/or not actually that beneficial, decent at pointing to overlooked players, and not especially strong at solving gordian knot type problems like "is Tua good actually?" Compared to something like baseball, where these days you can dive into stuff like spin rate, pitch sequencing, exit velocity, and launch angle, football analytics are just much noisier and more entangled with other players, positions, and schemes. This analysis did a great job of being more descriptive in terms of what the numbers can tell us about how Tua has played so far rather than trying to manufacture a conclusion that just isn't really supported by the data.

3 On the bright side, the…

On the bright side, the predictive value of non-pressured, past-the-sticks passes is sketchy at best: we should probably not trust any split that rates Carson Wentz ahead of Josh Allen. 

Don't let Carson Wentz: the Meme get in the way of Carson Wentz: Starting QB Pursued by Three Franchises. Non-pressured Carson Wentz is really good. In the aggregate, he's been better than Allen, and has been better than Allen in two of their four common seasons; Allen has only been markedly superior to Wentz in one of their common years. 

Given your cut was non-pressured and deep, it's not crazy that the best version of Carson Wentz comes out as better than up-and-down-year Josh Allen.

14 Agreed. Wentz is far from…

Agreed. Wentz is far from the only quarterback who looks good when non-pressured while falling substantially under pressure; that even fits the narrative around him that Mike literally used in the skit at the end of the article!

4 At the very least, Tua…

At the very least, Tua should benefit from the advantages Jimmy Garoppolo has enjoyed when he and his supporting cast have been healthy over the past three seasons. Garoppolo may be sitting by the curb with a "Take Me" sign on his back right now, but the Dolphins would not mind a ball-control offense capable of some deep playoff runs led by a still-young-and-affordable quarterback with room to grow.

I suppose the gamble the Dolphins are taking is that Tua is Good Trey Lance and not Jimmy Garoppolo?

I suppose, even then, if you can get Good Alex Smith out of him, that's a pretty effective QB.

10 Yeah, but his tool set is…

Yeah, but his tool set is not like any of those QBs. With Tua, the upside you dream of is Drew Brees. As far as the number crunching, it's interesting, but the offense the Dolphins were running was so odd the only thing I know about Tua is he's very accurate, has great pocket awareness, and he needs to cut down on dumb interceptions. I don't even trust the fact that he's an ALEX king because the offense design didn't even include deep routes most of the time. I actually suspect Tua is way less conservative than he showed last year.

6 CHRIS BALLARD: And that's…

CHRIS BALLARD: And that's how I toppled the most evil owner in the NFL.

The really sad part is that NFL owners are probably only the third-most disgusting set of owners in American professional sports. They aren't nearly as craven and openly corrupt and greedy as MLB owners and appear to at least be capable of the concept of embarrassment.

Nor do they actually own honest-to-god, no-foolin' slaves and have shackled themselves to authoritarian foreign regimes like NBA owners have.

So, by virtue of not openly conducting a genocide and being proud of it, they slot in behind NHL owners at the top of the ethics list. NHL owners are generally smart enough to stay out of sight.

12 Nor do they actually own…

Nor do they actually own honest-to-god, no-foolin' slaves and have shackled themselves to authoritarian foreign regimes like NBA owners have.

 

I don't understand what you mean by this. It's arguable that NBA players have more power than players of any other league in the US

18 The NBA almost assuredly has…

The NBA almost assuredly has the most powerful union. It's the youngest, weakest league; it involves the fewest players; fans are physically closest to the players; and players enjoy long careers and solid health in a generally non-collision, safe game.

But I'm talking about the owners. The Nets, for instance, were sold from one foreign oligarch busy conducting a genocide to a completely different foreign oligarch busy conducting a genocide. A few other slightly less open owners are not far behind. The league almost shut down due to a tweet from a GM, and basically only didn't because the majority of the federal government pointed out that was a national security problem.

So while MLB owners like to treat people like they were slaves and dream of crushing their competition to death, some NBA owners do this literally rather than figuratively. 

15 Reading about the latest…

Reading about the latest Snyder mess I initially thought that maybe DC fans will finally be rid of this dysfunctional ethically challenged owner, but then I realized the pool of potential buyers is probably composed of similar people.

43 If you wanted a nearby,…

If you wanted a nearby, poorly-run franchise with a beleaguered but loyal fanbase, but one not actually under federal investigation, wouldn't it be cheaper to buy the Orioles?

7 "If he fails, it will be Teddy Bridgewater time."

That's depressing. There's no point in failing him midseason for a journeyman. Just give him the full year and don't do 2020 redux with the wishy washy pulling back and forther. Gather as much data as you can on him. We have plenty on Teddy. 

8 I can't tell if the league…

I can't tell if the league has changed or it's always been this way, but QBs are now effectively given two years to prove they are capable of really good seasons or they are earmarked for the dust bin.

Lest we forget, Josh Allen was essentially written off as a bust after year 2.

Is that fair? I am not sure. Who are the qbs that teams truly would have felt buyer's remorse for letting go? In point of fact, I am asking which QBs became tier 1 and tier 2 by showing so little the first two years?

Brees and Allen come to mind with lesser versions being Eli Manning, Ryan Tanny, and Alex Smith.

I guess that suggests Tuas chances at being tier 1 or tier 2 are fairly small.

13 Lest we forget, Josh Allen…

Lest we forget, Josh Allen was essentially written off as a bust after year 2.

I really don't think this was true at all. There was certainly questions about whether he could be a franchise quarterback, but at worst his rushing ability gave him a fairly high floor. Obviously his 2020 exceeded essentially everybody's expectations, but he was always a project and so his improvement wasn't totally out of nowhere. They made the playoffs in 2019, after all, and he even won an AFC offensive player of the week award. I have no idea who was calling him a bust, and if somebody did they were laughably premature and likely holding onto some pre-draft bias.

21 FO took the piss out of Josh…

FO took the piss out of Josh Allen through the end of his 2019 season (https://www.footballoutsiders.com/quick-reads/2020/quick-reads-decade-review-qb-totals) and up until week 2 of 2020.

https://www.footballoutsiders.com/quick-reads/2020/week-2-quick-reads

 

22 Honestly, QB have till they…

Honestly, QB have till they reach FA, esp as a top 5 pick(yes, allen wasn't). You have to be epicly bad to lose a starting job before that. Fans pretty much want every QB to be a pro bowler/HOF and anything else is a bust. 

I think Allen was expected to be a low end starter before his breakout. Even in 2019 Allens rushing was good enough to make him not totally terrible. 

 

24 Even that first one though…

Even that first one though notes his rushing ability, which is why it's hard for me to believe anybody would truly call him a bust after those first two years. Doubts about him being a franchise quarterback, sure, and the potential for him busting was still probably there, but I don't think anybody was writing him off already.

25 Here's the thing, his…

Here's the thing, his passing numbers were so bad that it's sort of renders this all moot. No amount of Tim Tebow rushing was going to overcome the fact that he was a god-awful passer. And the same could be said for a slew of other Rush first passing optional quarterback types.

The fact that Josh Allen was given three years to develop says more about the fact that he was a first round pick and a high first round pick than anything about what he showed his first two years. I don't quite understand why more people don't just accept the most standard explanation, is breakout season came out of nowhere and sometimes people Buck trends. If anything, it makes Josh Allen look even more impressive that he was able to continue to believe in himself and see that come to fruition.

27 his passing numbers were so…

his passing numbers were so bad that it's sort of renders this all moot. 

Josh Allen was a replacement-level passer his second year. You're making this out like he was some sort of giant mega-disaster.

He's not a good example of someone who turned it around. He had a bad rookie year, a mediocre second year, and stellar after that. FO was super-down on him primarily because of his college numbers, not his NFL ones. No one's going to outdo Alex Smith for that award. Alex was over 1500 DYAR in the hole by the time he started turning things around, and he still ended up leaving San Francisco negative. 

Allen wasn't even the biggest disaster his rookie year. Not even close!

28 Sometimes I read responses…

Sometimes I read responses that leave me utterly slackjawed.

" He had a bad rookie year"

Notice the proliferation of the anodyne language being used.  He was bad his rookie year. Well gee...when you finish second to dead last in the NFL in terms of DVOA and DYAR, that's just garden variety, could have happened by accident bad. 

Oh and this line is even a bigger douzey. 

"Allen wasn't even the biggest disaster his rookie year. Not even close!"

Its a good thing there are people like Josh Rosen lurking around  to make sure we aren't damned with praise way too faint. Take a bow, he wasn't as a bad as one of the all time worst rookies. 

And then we get to this part. 

"a mediocre second year"

Ah, the word mediocre. Its like the word interesting...meant to mask how we really feel lest we stick our neck out too far and get labeled an extremist. Gotta love the word mediocre...it encompases everyone to the left of Patrick Mahomes' pinky toe so we can't be accused of polemics. 

He finished 28th in DVOA and 28th in DYAR. Yes, below fellow mediocrity Mitch Trubisky and the perennially mediocre Andy Dalton, 10 year + veteran and now a journeyman for a reason. I guess if those guys are mediocre, Fitzpatrick is a damn near hall of famer.

If second to last is bad, and 28th is mediocre, I guess we can reserve god awful for wide receiver, forced into QB duty, Kendall Hilton. Oh wait though, even he...Kendall Hilton, had a better DVOA than Josh Rosen. Oops.  

 

41 Allen is an outlier. The…

Allen is an outlier. The only pass I'll give him is that he wasn't supposed to play his rookie season AT ALL, and McDermott really screwed the pooch on that by thinking Peterman was good enough to be a placeholder. He might have showed better in year two, or he might not have.

I was cautiously optimistic about year three - you could see that he was getting better at understanding what he was supposed to do, even if the results weren't there yet - I just thought it would be more like year 4. He blew through the ceiling in a way that was hard to predict.

Allen is a bad measuring stick. People need to stop using him as such. Tua is his own QB with his own set of skills and issues. I don't think he's the answer in Miami, but I don't see any reason for them not to try and find out.

47 I really don't understand…

I really don't understand viewing Allen's rookie season as an outlier. I mean, if Allen's is an outlier, so is Stafford's. And McNabb's. And Jared Goff's (regardless of your long term opinion on Goff).  I mean, OK, those are outliers in the same sense that flipping three heads in a row is an outlier, but "outlier" is usually something I reserve for "once in a generation," like Tom Brady. Not something I can come up with three other examples for off the top of my head.

Sure, you can say "Allen's first three years are hard to find a comparison to" but that's not super-surprising. Add in the 4th and smooth it out and there are a lot more comparisons (although not all that favorable to Allen).

45 Well gee...when you finish…

Well gee...when you finish second to dead last in the NFL in terms of DVOA and DYAR, that's just garden variety, could have happened by accident bad. 

Yes, exactly! Glad we agree. I mean, the starting quarterback for the current Super Bowl champions certainly would've done better than that, right? And, of course, add in Allen's rushing ability and around -300 DYAR total is completely normal for a rookie year. Donovan McNabb is an even worse example.

 

He finished 28th in DVOA and 28th in DYAR. Yes, below fellow mediocrity Mitch Trubisky and the perennially mediocre Andy Dalton, 10 year + veteran and now a journeyman for a reason. I guess if those guys are mediocre, Fitzpatrick is a damn near hall of famer.

Elite, great, good, average, mediocre, bad, awful. Lot of room between mediocre and elite.

To be more technical, I can just say "he was a replacement level passer his second year," since for me "replacement level" is "mediocre..." because there are a lot of quarterbacks who don't show up on that list with a minimum number of passes. 

Quarterbacks who get to replacement level that quickly can't (shouldn't?) be viewed as busts. Certainly it's totally fair to question whether they're going to work out long term at that point, but considering Allen was a strong runner, replacement-level passer by year 2 and strong runner doesn't equal "bust."

I mean, Allen certainly had tons more potential than Sam Darnold or Daniel Jones by year 2. 

Keep in mind, I'm not saying that Allen was good or anything. Or that he'd turn into where he is now. I'm just saying that the opinion of him (mainly here) was too low, because of the college data. I mean, jeez, I'm basically saying "nah, he was basically Blake Bortles at that point." The kind of guy who can frustrate the hell out of a franchise because he's just not obviously bad enough to abandon yet.

49           Well gee...when…

          Well gee...when you finish second to dead last in the NFL in terms of DVOA and DYAR, that's just garden variety, could have                  happened by accident bad. 

Yes, exactly! Glad we agree. I mean, the starting quarterback for the current Super Bowl champions certainly would've done better than that, right? And, of course, add in Allen's rushing ability and around -300 DYAR total is completely normal for a rookie year. Donovan McNabb is an even worse example.

The situation Stafford and Sam Bradford walked into in 2009 was a universe removed from what Allen walked into in 2018. Allen walked into a 6-7 win team that had lucked its way to 9 wins but also underperformed, in part because McDermott was an idiot and played The Human Turnover for three games. It was roughly a -10% roster. 

Bradford and Stafford were drafted by wretched, execrable, incompetence messes that were historically awful (in 2009, too). Their combined pythagorean wins were 1.5 games worse than Buffalo in 2017. (St. Louis and Detroit combined for 5 wins in four team-seasons -- 2017 Buffalo had that many by game 7.) At -45 to -48% DVOA, Buffalo was actually closer to #1 NE (+27%) than they were to 2008 Detroit or St. Louis. Allen's situation was closer to Mahomes' than to Stafford or Bradford. The situations are in no way comparable. 

51 Allen walked into a 6-7 win…

Allen walked into a 6-7 win team that had lucked its way to 9 wins but also underperformed, in part because McDermott was an idiot and played The Human Turnover for three games. It was roughly a -10% roster. 

I think that undersells Taylor a bit, who was really an absolutely solid QB whose body failed him early. And even with him, the offense was still 26th in the league. Sure, it's fun to rag on Nathan Peterman but Derek Anderson started in 2018 as well and was worse than Allen. OK, Matt Barkley was great in one game, sure, but it's the Jets. 

Plus McCoy fell off a cliff from 2017 to 2018 and he was like the entirety of their offense. And yes, I know he was basically replacement-level in 2017: like I said, 26th ranked offense. The other alternatives were worst-in-the-league Zay Jones and Charles Clay. McCoy had the highest number of targets (and receptions) on the team in 2017! 

Buffalo's success (such as it was) in 2017 was on defense, not on offense. 

Allen's situation was closer to Mahomes' than to Stafford or Bradford.

I... don't agree. The '18 Bills had nothing on offense. A shell of a running back and a second round WR bust.

Matthew Stafford walked into the Lions with Calvin Johnson.

52 2017 Buffalo was the worst…

2017 Buffalo was the worst receiving team I have ever seen. No one could get open, and even if they could, they couldn't catch the ball. That Buffalo-Jacksonville game was excruciating the watch, because no one on either team could catch a pass.

But whatever the question was, Peterman was the wrong answer. He's something like the second-most turnover-prone QB in NFL history. You or I could probably go under center and turn the ball over less than Peterman did.

The 2009 Lions were sort of like the 2020 Bengals. They had a rookie QB, one functioning receiver, and nothing else. (They also had a black hole on defense, but we'll ignore that for now) The 2008 Lions were actually more talented on offense, as they had Roy Williams for part of the season, too.

But the defense matters. Buffalo could keep games close even if their offense did nothing. They could also run the ball. Detroit could not. Stafford had to throw, because his defense hemorrhaged points and his running game was completely useless. (Bradford had less in his receiving room, but at least had Stephen Jackson) Buffalo could get to the playoffs by playing conservatively (which, again, makes the entire Peterman decision so frustrating). Detroit basically had to throw bombs to Johnson, because that was the only functional part of their offense. Which was trouble, because they couldn't pass block, either. (Which is why Stafford was hurt so often those first years)

54 I'm just missing your logic…

I'm just missing your logic. The 2017 Bills had (by your own statement) the worst receiving team you've ever seen. They gained no one from 2017 to 2018. And while the 2017 Bills could run the ball, the 2018 Bills could not. McCoy completely collapsed, and the top two RBs averaged 3.2 and 3.3 yards, respectively. Josh Allen had the most rushing yards on the team!

So the 2018 Bills had terrible receivers and no running game. Stafford's Lions at least had "one functioning receiver" (which is the strangest way to describe Calvin Johnson). So how in the world is Josh Allen's rookie season "a far better situation"?

I mean, I get the whole "better defense means it puts less pressure on the offense" but Josh Allen was literally the only functioning thing on the offense at all. I think "playing conservative" is literally what he did - drop back, look around, say "oh hell no, not these guys" and take off.

56 My point is that Buffalo…

My point is that Buffalo could survive with a zero at QB, so long as you weren't an enormous liability (like Peterman). Taylor actually generated +200 DYAR despite dinking and dunking to a bunch of Agholors (Peterman somehow gave all of that back in two games). Buffalo could win a 14-13 game.

Buffalo, in 2017 and 2018, as inefficient as they were, actually outrushed their opponents at about the same average, and ran about as much as they passed. They could do so both because they could kind of run effectively, and because they could keep a game close enough long enough that they didn't have to go to desperate catchup passing.

The 2008-2009 Lions couldn't run when they wanted to, and didn't want to because they weren't close enough behind that anyone believed they were going to run. So they threw early, often, and because they had to. They threw all the time. (They even did so when they had Kitna's corpse and Orlovsky's rare moments of not stepping out of the back of the end zone -- so it wasn't because they enjoyed passing) The Lions basically always had a deficit (they were a full TD worse per game than Buffalo -- they turned the average opponent into the Saints' offense), so the entire offense was Stafford trying to survive a collapsing pocket long-enough to throw a go-route to Johnson. And while Johnson was really good, he wasn't quite HOFer Calvin Johnson yet. At the time of the Roy Williams trade, Williams was still the better player, if not the more talented one.

So basically, Buffalo just needed Allen to not be a disaster, but he sort of was. (see 2019 Buffalo, which was basically the same team as 2017 Buffalo) 2009 Detroit needed Stafford and Johnson to be 1984 Marino throwing to 1987 Jerry Rice, and it would be nice if they were 1943 Sammy Baugh and Don Hutson instead, because they also needed someone to kick and play defense, too. Stafford ended up having to be inefficient at high-volume because the team's situation was always dire, but because they were so limited, where he was going to go was obvious to everyone in the building. Allen needed to not trip over his dick; Stafford needed to stay upright after he was thrown off a cliff.

58     Buffalo, in 2017 and…

 

 

Buffalo, in 2017 and 2018, as inefficient as they were, actually outrushed their opponents at about the same average,

Yeah. Because Josh Allen in 2018 was running at 7.1 yards per carry. Take away his 89 attempts and 631 yards and as a team they went 379 carries for 1353 yards, or 3.56 yards/carry.

They couldn't run effectively. At all. The only guy who was running effectively was the guy you said was a disaster.

Again, I get that having a solid defense puts less pressure on the offense, but I don't care how little pressure you have, at least when Stafford threw the ball there was someone to catch it.

 

46 Here's the thing, his…

Here's the thing, his passing numbers were so bad that it's sort of renders this all moot. No amount of Tim Tebow rushing was going to overcome the fact that he was a god-awful passer.

I'm going to assume you didn't actually look up any numbers for this, because this statement is laughable.

Josh Allen's passing DYAR/DVOA/attempts & rushing DYAR/DVOA/attempts:
(2018) -534 / -35.9% / 350  &  192 / 33.3% / 81
(2019) -21 / -11.8% / 497  &  100 / 6.5% / 95

Tim Tebow:
(2011) -222 / -22.8% / 301  &  -36 / -18.3% / 117

So not only was Josh Allen's 2019 significantly better passing than Tebow, he actually contributed positive value rushing, unlike Tebow's literally worst in the league rushing value. You're also neglecting that he jumped from truly horrendous as a rookie to merely bad in year 2 (as a comparison, Tannehill and Daniel Jones were around -11% DVOA last year). That's obviously not what you want from your starting quarterback, but given he'd made quite a jump in a year I think most people viewed 2020 as his make it or break it year. Either he would continue improving to be somewhere around average passing while still contributing rushing value and thus be a viable starting quarterback, or he would plateau and force some tough decisions for the Bills going forward. But crucially to this conversation, both possibilities were still reasonable outcomes, meaning it's silly to say that people had already written him off as a bust. Again, I'm not saying people saw his 2020 coming, nor should they use that specific outcome as a prediction for other players (cough Malik Willis), only that giving Josh Allen a third year of development isn't some crazy longshot decision even based purely on his numbers over his first two years.

63 There are times when people…

There are times when people are so married to stats they refuse to believe what their eyes tell them. By the later part of 19 it was clear Tebow didn’t  belong in the same discussion as Allen. Watching the pitt game on snf followed by the Ne showdown it was clear Allen had become a viable franchise quarterback. That’s different than elite of course, that conversation didn’t end until the 2020 playoffs.  But by the end of year 2 it was clear that Allen was the right choice. 
 

tua didnt make a similar year 2 leap. His rookie year ended getting benched twice for fitzy and an embarrassment against the playing for nothing bills. This year he played no meaningful games after October, and he didn’t show he could make all the throws needed to be a franchise quarterback.  

29 I agree. From watching the…

I agree. From watching the Bills in 2018 and 2019 it was clear Josh Allen was far better his second season and was poised to improve further in 2020. Even his so called terrible 2019 stats were as good as any Bills QB since Bledsoe in 2002.

31 Tyrod Taylor's best bills…

Tyrod Taylor's best bills season:2015: 3035 yards, 20 pass td, 6 int, 64%/8.0ypa/99.9rate, 4 rush td, 24 total td,

Fitzpatrick:2012:3400yds, 61%, 6.7ypa, 83 rate, 24 pass td,16 int,1 rush td,25 total td,

Josh Allen 2019:3089 yards, 59%, 6.7ypa, 85 rate, 20 pass td, 9 int, 9 rush td, 29 total td

Based on the stats I guess Taylor's passing stats are better but with the catch that he threw almost 20% fewer passes, and his overall stats are slightly better but not by much. Allen's 2019 was as good or better than Fitz in 2012(or 2011).

So then Allen's 2019 stats were as good as any Bills QB season in the 2000s before 2019 except two seasons.  And his 2019 was still a significant improvement over 2018, and gave plenty of reason to be optimistic for further improvement.

Tyrod Taylor's 2015 season definitely gave promise for his future in Buffalo, but unlike Allen, Taylor never came through for Buffalo after that, and even though Buffalo would have been justified dumping him after 2016 his 2017 performance further cemented moving on from him.

23 Rookie and Young QB's

I do think that the league was changed by the rookie salary ceiling that was established during 2011. If you have a young QB that is not living up to expectations, it is now much less costly to pursue a possible improvement.

 

Additionally, over the past fifteen seasons, we've seen rookie or young QB's like Matt Ryan, Joe Flacco, Russell Wilson, Andrew Luck, Dak Prescott, Justin Herbert, Patrick Mahomes, DeShaun Watson, and Joe Burrow immediately find success, and, in some cases, revitalize their respective franchises. I think that this has also led some organizations to become much less patient in a young QB's development. It seems the NFL has reached a point where if you are not showing the goods right away, then the organization will be tempted to seek someone who will.

50 Gonna respond to everyone…

Gonna respond to everyone with one example.

Gardner Minshew.

Gardner Minshew's first two seasons were both better than Josh Allen's. His rookie year was way better and his second year was also better.

And yet no one here nor anywhere else was talking themselves into Gardner Minshew, either as a long term starter or even vaguely capable of reaching superstardom. 

The difference as far as I can see is Josh Allen is a first round pick and Minshew was a 6th rounder. Which goes back to my original point.

 

53 Oh, first round picks get a…

Oh, first round picks get a much longer leash, because of the Sunk Cost fallacy, even if it was someone else's cost (see Trubitsky, Mitchell). Having said that, Minshew declined in his second season (not unexpectedly, given the team around him and injuries), but yes, Jacksonville is a shitty place to try and succeed. Allen's situation was better.

60 Oh, first round picks get a…

Oh, first round picks get a much longer leash, because of the Sunk Cost fallacy,

It's not a sunk cost fallacy. It's the strength of the initial evaluation. Even if you see 2 years from someone in the NFL, that's still not enough to toss the predraft evaluation away completely.

59 and his second year was also…

and his second year was also better.

Huh? Minshew's second year was virtually indistinguishable from Allen's as a passer and adding in Allen's rushing, I'd be seriously hard pressed to say it was better. Only justification there is because, well, Jacksonville.

The difference as far as I can see is Josh Allen is a first round pick and Minshew was a 6th rounder.

Yeah? Which is the entire point? You drafted Allen high because you saw a ton of potential there. You never saw any potential from Minshew. Of course you're going to give Allen more chance to figure it out than Minshew, because you've already seen years of tape from him and didn't see any value there.

I mean, from a Bayesian perspective you had a low prior on Minshew, the first two years pulled it up a little but not enough to discard the prior. You had a high prior on Allen, and the first two years pulled it down but again, not enough to discard the prior.

I don't see how this is anything other than completely logical behavior.

61 Self-fulfilling prophesies…

Self-fulfilling prophesies are self-fulfilling.

It's amusing to note that of the top-10 or so QBs in DVOA/DYAR in 2001/2002, when Brady -- famously a 6th-round QB -- debuted, there were more QBs who were from the 6th round or later than from before the 6th round. Brady was actually about median for the group. 

The list included Fiedler, Garcia, and Warner (undrafted), Brad Johnson (9th), Trent Green (8th), and Matt Hasselbeck (6th). Gannon was a 4th rounder on his 4th team and 15th season. There was also Kordell Stewart, who was a 2nd rounder, but had been drafted to play a position other than QB.

The 1st rounders were Collins, Manning (P), Pennington (pre-injury), and McNair.

So the pursuit of sunk-costs and prior-assumptions hasn't always been so ingrained.

62   So the pursuit of sunk…

 

So the pursuit of sunk-costs and prior-assumptions hasn't always been so ingrained.

It's because in 2000, free agency was less than a decade old and quarterbacks took up probably a quarter of the cap hit they take now. Giving cheap QBs a chance for years made perfect sense. Nowadays there's no such thing as a cheap QB.

Note that for all of the QBs you listed, the number of top 5 seasons they had was actually a pretty low fraction of their career. Even for Warner. Probably the best one there is Matt Hasselbeck, who gave Seattle like 7-8 years of high end play? Even then you're still talking about 50% of their career.

It's the same issue as Minnesota has with Kirk Cousins. With contracts as high as they are now you want slam dunk obvious choices, not "hey, he might have a 2-3 year stretch for us where he's great."

65 Top 10 picks are generally…

Top 10 picks are generally still given a chance to start into their 3rd season even if they struggle in their first 2 seasons. Josh Rosen is the only exception I can think of in recent years. Unless I'm missing someone, last one before that was also in Arizona (and the 10th pick), Matt Leinart. Plenty of guys who weren't so good in their first 2 seasons still got to start at the beginning of year 3: Allen, D Jones, Darnold, Trubisky, Bortles, etc.

Outside the top 10 it's more common to not get a 3rd year: Dwayne Haskins, Paxton Lynch, Johnny Manziel, EJ Manuel, Brandon Weeden.

9 Miami draft

They should still shore up OL and WR to make sure it's good just in case there's some new team jitters from the free agents.

20 The RPO stuff is very…

The RPO stuff is very incomplete: you're only looking at a selected portion of them - the ones that developed into passes. Tua's ANY/A on RPO passes is nice... but if, for instance, they're only a small fraction of the total number of RPOs and the ones that developed into runs were disasters, then it's a selection effect. You're only looking at the times when a defense was dumb enough to let Tua pass to an open spot. It'd be like looking at the success rate on "defense jumps offsides" plays. You'd be like, damn, he's awesome in these plays, they should run "idiot gives them a free play" all the time.

You can't split RPOs into "runs" and "passes" like that. It's all one play. The offense doesn't choose what it turns into.

48 I do agree with this when…

I do agree with this when looking at RPOs in the context of an offense as a whole, but in this case the goal is to examine Tua's performance specifically. I do think you can tell a few things from looking at just RPO pass numbers, especially relative to other offenses - first, they were a high percentage of the offense, which is probably a factor of poor pass blocking, lack of faith in Tua's ability to fully go through progressions, or both. Second, the Dolphins did quite well on those plays, even relative to other good offenses, which seems to indicate that Tua can make good quick decisions off the snap. Caveat here of course being that perhaps he only threw on the most obvious of passing situations (your "defense jumps offsides" analogy above). I think the comparison to other good offenses that heavily used RPOs, as well as the sheer number of plays, indicates this probably isn't the only factor here, but ideally you'd have the RPO rush splits as well. Of course, I'm not sure that's even possible, since the best way to tell a play is an RPO is when the QB throws the ball while everybody else is playing it as a run.

 

Tl;dr I think using those numbers alone as proof of anything is problematic for the reason you mentioned, but I think using them in combination with the other stats to make some inferences about Tua is perfectly valid. Just, larger error bars.

55 Of course, I'm not sure that…

Of course, I'm not sure that's even possible, since the best way to tell a play is an RPO is when the QB throws the ball while everybody else is playing it as a run.

A "typical" RPO is easy to recognize because when the ball is snapped, the quarterback's eyes go downfield rather than watching the handoff.

35 The Dolphins offensive line…

The Dolphins offensive line finished 30th in adjusted line yards and needed all of that quick passing to rank 19th in adjusted sack rate.

Is there really much evidence that adjusted line yards is related to pass blocking?

We routinely accept that a defensive lineman or entire line can specialize in run-stopping or pass pressure. But somehow run blocking stats are frequently used as evidence about pass protection.

66 Stutter

Wow… I’ve been here longer than most… and this was my first experience of the dreaded double post.  I guess I had to make up for lost time.

Jeesh. No idea how that happened.