Philadelphia Eagles QB Carson Wentz

Adjusted Interceptions 2020

Officially, the NFL's interception leaders last year played for Denver and Philadelphia, but one of those quarterbacks was largely a victim of bad luck. Other unlucky quarterbacks played in Seattle and Minnesota, while the football gods smiled on quarterbacks in Pittsburgh, Arizona, and Miami.

Today we are going to discuss adjusted interceptions. Unlike the NFL's raw interception totals, these numbers account for plays when a defender drops a pass that he should have caught, or when a wide receiver makes a big play to turn what should have been a turnover into an incompletion instead. On the other hand, sometimes quarterbacks are charged with interceptions that aren't really their fault—passes that bounce off a receiver's hands and straight to a defender—or interceptions that don't matter, like Hail Mary passes.


How it Works

After each season, we go back and account for these discrepancies and tabulate each quarterback's adjusted interceptions. Here's the process:

  • We start with each player's actual interception total. Denver's Drew Lock and Philadelphia's Carson Wentz tied for the league lead in this category with 15 apiece in 2020.
  • We then add plays where the quarterback threw a ball that could have or should have been intercepted but was not, either because the defender outright dropped the ball (which we have been tracking in game charting since 2007), or he had it knocked out of his hands by an offensive receiver (a "defensed interception," which we have been tracking since 2012). These are listed as "Drop/Def INT" in the table at the end of this page. Ben Roethlisberger had 10 such plays last season, most in the league; Kyler Murray, Sam Darnold, and Nick Foles were tied for second with nine each.
  • Next, we subtract those interceptions that were tipped by receivers into the hands of defenders (as established in 2017, these plays can be thought of as Matt Ryan specials). Derek Carr, Daniel Jones, and Russell Wilson tied for the league lead in this department with two each. We also subtract passes that are tipped by receivers but then dropped by defenders to make sure they are not double-counted. There were only four of these in 2020; no player threw more than one.
  • We subtract Hail Mary interceptions, as well as interceptions thrown in desperation on fourth down in the final two minutes of a game. We're flexible on these definitions, but this year there was only one play that was in the gray area: a Wentz interception in Week 5 against Pittsburgh, thrown on fourth down while trailing by nine points with 2:18 to go, which we counted as a Hail Mary interception. Kirk Cousins led all players with three Hail Mary interceptions last season.
  • We subtract dropped interceptions that occur in Hail Mary situations, since those plays wouldn't count as adjusted interceptions even if they had been caught. Cousins, Darnold, and Lock each had one of those throws.

Long-Term Trends

When we started running these numbers, we had to get the data from our own in-house volunteer game charters. For the last six seasons, we've had access to data from Sports Info Solutions. Determining whether or not a defender should be charged with a dropped interception will always be subjective on some plays, but you can rest assured that all the obvious calls have been counted here.

Officially, only 2.2% of all passes thrown in the NFL in 2020 were intercepted, the lowest rate for any season in the record books. That number has been trending downwards for years—decades, really—but one reason it bottomed out was that defenders all over the league developed a severe case of butterfingers. There were 236 dropped interceptions last season, most of any year in our records going back to 2007. Dropped interceptions have actually been climbing for several years now, even as official interception numbers have plummeted. The leaguewide adjusted interception rate was 3.4%, the highest in any season since 2012.

NFL Year-by-Year Adjusted Interceptions, 2011-2020
Year INT HM/
End Q4
Drop/
Def INT
Tip INT Tip AND
Drop INT
Drop INT
also
HM/End Q4
Adj INT Pass Att
(no DPI/
spikes)
INT
Rate
Adj
Rate
2011 477 14 158 33 588 16681 2.9% 3.5%
2012 463 14 192 27 4 610 17665 2.6% 3.5%
2013 502 16 151 27 7 603 18036 2.8% 3.3%
2014 448 18 159 22 4 563 17759 2.5% 3.2%
2015 435 9 152 19 6 553 19372 2.2% 2.9%
2016 415 20 175 23 0 547 18064 2.3% 3.0%
2017 429 14 192 41 3 563 17242 2.5% 3.3%
2018 419 15 192 25 1 570 17470 2.4% 3.3%
2019 410 15 205 33 8 5 554 17684 2.3% 3.1%
2020 395 9 236 20 4 3 595 17761 2.2% 3.4%

Note that we didn't start counting "tipped and dropped interceptions" until 2012 and "dropped interceptions that occur in Hail Mary situations" until 2019, which is why no numbers are listed in those categories in earlier seasons.

It looks like quarterbacks haven't gotten any better at avoiding interceptions, but defenders have gotten worse at reeling them in. Given the critical value of turnovers in the game of football, this seems like something defenses should be working on more in practice. Anyway, as a result of these trends, the average quarterback threw about 50% more adjusted interceptions than actual interceptions, up from the usual rate of 30% or so.


Individual Leaders

Adjusted Interceptions, 2020
Name Team INT HM/
End Q4
Drop/
Def INT
Tip
INT
Tip AND
Drop INT
Drop INT
also
HM/End Q4
Adj INT Pass Att
(no DPI/
spikes)
INT
Rate
Adj
Rate
Rate
Dif
11-C.Wentz PHI 15 1 7 0 0 0 21 436 3.4% 4.8% +1.4%
1-K.Murray ARI 12 0 9 1 0 0 20 550 2.2% 3.6% +1.5%
7-B.Roethlisberger PIT 10 0 10 0 0 0 20 608 1.6% 3.3% +1.6%
14-S.Darnold NYJ 11 0 9 0 0 1 19 361 3.0% 5.3% +2.2%
16-J.Goff LAR 13 0 5 0 0 0 18 550 2.4% 3.3% +0.9%
5-T.Bridgewater CAR 11 0 7 0 0 0 18 490 2.2% 3.7% +1.4%
2-M.Ryan ATL 11 1 8 0 0 0 18 625 1.8% 2.9% +1.1%
9-N.Foles CHI 8 0 9 0 0 0 17 311 2.6% 5.5% +2.9%
3-D.Lock DEN 15 2 5 1 0 1 16 441 3.4% 3.6% +0.2%
17-P.Rivers IND 11 0 6 0 1 0 16 540 2.0% 3.0% +0.9%
8-L.Jackson BAL 9 0 8 1 0 0 16 373 2.4% 4.3% +1.9%
17-J.Allen BUF 10 0 6 1 0 0 15 572 1.7% 2.6% +0.9%
3-R.Wilson SEA 13 0 3 2 0 0 14 557 2.3% 2.5% +0.2%
4-N.Mullens SF 12 0 3 1 0 0 14 326 3.7% 4.3% +0.6%
12-T.Brady TB 12 0 3 1 0 0 14 606 2.0% 2.3% +0.3%
9-M.Stafford DET 10 0 4 0 0 0 14 521 1.9% 2.7% +0.8%
10-J.Herbert LAC 10 0 5 0 1 0 14 591 1.7% 2.4% +0.7%
10-M.Trubisky CHI 8 0 6 0 0 0 14 297 2.7% 4.7% +2.0%
14-R.Fitzpatrick MIA 8 0 5 0 0 0 13 266 3.0% 4.9% +1.9%
9-D.Brees NO 6 0 8 1 0 0 13 388 1.5% 3.4% +1.8%
15-P.Mahomes KC 6 0 7 0 0 0 13 587 1.0% 2.2% +1.2%
1-T.Tagovailoa MIA 5 0 8 0 0 0 13 287 1.7% 4.5% +2.8%
8-K.Cousins MIN 13 3 4 1 0 1 12 513 2.5% 2.3% -0.2%
8-D.Jones NYG 10 0 4 2 0 0 12 446 2.2% 2.7% +0.4%
14-A.Dalton DAL 8 0 5 1 0 0 12 333 2.4% 3.6% +1.2%
6-B.Mayfield CLE 8 0 4 0 0 0 12 481 1.7% 2.5% +0.8%
17-R.Tannehill TEN 7 0 7 1 1 0 12 481 1.5% 2.5% +1.0%
4-D.Watson HOU 7 0 5 0 0 0 12 542 1.3% 2.2% +0.9%
1-C.Newton NE 10 0 1 0 0 0 11 366 2.7% 3.0% +0.3%
12-A.Rodgers GB 5 0 6 0 0 0 11 523 1.0% 2.1% +1.1%
4-D.Carr LV 9 0 3 2 0 0 10 513 1.8% 1.9% +0.2%
9-J.Burrow CIN 5 0 5 0 0 0 10 401 1.2% 2.5% +1.2%
11-A.Smith WAS 8 0 1 0 0 0 9 252 3.2% 3.6% +0.4%
4-D.Prescott DAL 4 0 5 0 0 0 9 221 1.8% 4.1% +2.3%
7-D.Haskins WAS 7 1 2 0 0 0 8 241 2.9% 3.3% +0.4%
15-G.Minshew JAX 5 0 2 0 0 0 7 326 1.5% 2.1% +0.6%

When we run these numbers for Lock and Wentz, the interception co-leaders of 2020, we get very different results. Lock had 15 interceptions, but when we add five that were dropped, and subtract four more (two on Hail Marys, one that was tipped by a receiver, and one dropped on a Hail Mary), we get 16 adjusted interceptions—still in the top 10, but not at the top of the pile. The math for Wentz is simpler: we add seven drops and subtract one turnover on a Hail Mary and get 21 adjusted interceptions, most in the league. That's a remarkable feat for a quarterback who was benched for the last four games of the year. Only one passer has ever led the league in adjusted interceptions on fewer throws, and that comes with an asterisk: in 2018, Sam Darnold threw 21 adjusted interceptions, tied for the league lead with Patrick Mahomes, on only 413 passes.

Two players tied for second behind Wentz with 20 adjusted interceptions: Pittsburgh's Ben Roethlisberger and Arizona's Kyler Murray. As noted earlier, Roethlisberger led all players with 10 dropped interceptions, and Murray was just one behind him. Add those 10 picks to Roethlisberger's 10 interceptions that weren't dropped and you get 20. Murray threw 12 interceptions with nine drops, but we also subtract one that was tipped by a receiver to get his total of 20.

Though Wentz led the league in adjusted interceptions, it was his old teammate Nick Foles who had the highest adjusted interception rate (minimum 200 throws). Foles threw eight interceptions with nine more dropped in only 311 throws, a rate of 5.5%. Foles is followed by Darnold (5.3%), Ryan Fitzpatrick (4.9%), Wentz (4.8%), and another Chicago quarterback, Mitchell Trubisky (4.7%). The Bears actually led all teams with 31 adjusted interceptions, which largely explains why they traded up in the draft for Justin Fields. (The Eagles were second with 29—21 by Wentz, seven by Jalen Hurts, and one by Nate Sudfeld.)

Derek Carr was the league's best passer at avoiding interceptions, with an adjusted rate of 1.9%. The next five names behind Carr include four quarterbacks with Hall of Fame-caliber performances (Aaron Rodgers, Deshaun Watson, Patrick Mahomes, and Tom Brady) and also Gardner Minshew, who excels at avoiding turnovers and is below average at pretty much everything else.


Interception Luck

When we chart adjusted interceptions against actual interceptions, we see which quarterbacks were unlucky when it came to interceptions (those above the line) and those who benefited from good fortune (below the line). And it's very clear who the most unlucky quarterback of the year was (click to open a larger image in a new window).

Adjusted Interceptions 2020

Minnesota's Kirk Cousins threw 13 interceptions, tied for third-most in the league, but had only four dropped, which left him outside the top 25 in that category. Jacksonville's Jake Luton had more interceptions dropped than Cousins (five) despite throwing over 400 fewer passes. And then there's Cousins' league-high three interceptions on Hail Marys. Cousins finishes with only 12 adjusted interceptions, making him the only full-time quarterback whose interceptions went down after adjustments. (It also happened to New England's Jarrett Stidham in limited action—in 44 throws, he had three interceptions but only two adjusted interceptions.)

Other unlucky quarterbacks include Lock (who threw 15 interceptions with only five dropped) and Russell Wilson (who had two interceptions tipped by his teammates, nearly as many as the three passes that were dropped by defenders).

Three players also stick out from their peers on the other side of the line, meaning they finished with fewer interceptions than expected based on their adjusted interceptions would suggest. Two of them (Ben Roethlisberger and Nick Foles) we have already discussed in detail. The third was Miami rookie Tua Tagovailoa, who threw five interceptions with eight dropped for a total of 13 adjusted interceptions in only 287 passes. Tagovailoa's adjusted interception rate (4.5%) was 2.8% higher than his raw interception rate (1.7%); only Foles had a higher gap among qualified quarterbacks.

Comments

55 comments, Last at 24 May 2021, 11:40pm

1 what really matters; pretty much only thing, really ...

So how well do Adjusted Interceptions predict future interceptions? How well did they do that last season?

I recall Bill Walsh saying that when the quarterback puts the ball right where it's supposed to go, receivers don't drop the ball. Just how 'sticky' are tipped interceptions? To whatever extent they are, that would make them a quarterback (un)skill rather than mere bad luck.

7 Good news then! All the…

In reply to by BigRichie

Good news then! All the prior editions of this article are readily available. Just google “football outsiders adjusted interceptions” and the year. 

11 Nostradamus ain't got nothin' on you!

So I can go to a year(s)-old article to see how well actual 2020 interceptions lined up with Adjusted Interceptions?? Wow! So they have Biff from Back to the Future II on FO staff now and they're keeping it quiet! (actually, I would too)

I've no problem that Vince and the other FO staffers have more important uses for their time. The idea that I'm supposed to go back to see just how it works, then tabulate the 2020 results and next run what I decide are the appropriate fit tests and see the results ... umm, that's a pretty funny one.

They could've run the tests to see what actual 2020 results Adjusted Interceptions projected. They decided it wasn't a good use of their professional time. Okie-dokie. As is pointing out that hasn't been done.

16 We have usually found that…

We have usually found that adjusted interceptions predict future interceptions better than actual interceptions do, but that has not been the case the last two years. Among qualifying QBs who did not change uniform numbers (because our tables are generated straight from the play-by-play, and that's how they are identified) from year to year:

2015 correlation between interception rate and next year's interception rate: 0.415. Correlation between adjusted interception rate and next year's interception rate: 0.468.

For 2016, the numbers are 0.111 and 0.285.

For 2017, the numbers are 0.200 and 0.458.

For 2018, the numbers are 0.610 and 0.479.

For 2019, the numbers are 0.251 and 0.187.

The last year there can be written off as 2020 being a weird year, but I don't know why adjusted interceptions would have been the less accurate predictor in 2018. It could be that as interceptions become more rare, they are also becoming more inherently random, and thus more difficult to predict.

2 is it interception luck?

or do some quarterbacks just throw more catchable interceptions? (e.g. throw velocity perhaps)....

46 Can't imagine it's pure luck

Otherwise what are we evaluating? More so called bad decision making (hello Jameis) that leads to "what did you think would happen to that throw besides be intercepted"

3 not all bad for Tua

Given Fitzmagic was at 4.9% you wonder if the lack of quality at wide out had something to do with both QBs high adjusted INT rate. The good news was, apparently no matter who passed the ball in Miami, the defenses refuse to catch it :)

25 He has 7 in the table. Which…

He has 7 in the table. Which is not low. I saw 6 of them myself even though I watched only 4 regular season games of Chiefs. One of them was San Diego game , which if I remember correctly he had either 2 or 3 dropped interceptions. I don’t know if it means FO has a different standard than me for what is considered dropped or they just did not count right or Mahomes had most of those in the games I watched. 

9 rate difference

Hmm...we'd expect rate difference to be roughly proportional to 1/sqrt(n), if it were sampling noise. That's not an exact conceptual match, but it would be interesting to see rate difference plotted versus passing attempts.

10 Apparently you are not…

Apparently you are not including Hail Marys at the end of first halves? Roethlisberger had one against the Titans last year.

13 Hail Mary

A Hail Mary on the end of the first half needs to be the final play of the half. Roethlisberger's interception came with 14 seconds left and represented a choice to throw deep rather than trying to get closer for a field goal attempt. The point of a Hail Mary is that there's no downside to the interception, but with this play, there was a downside, because the Steelers lost the chance to try for 3 points.

Roethlisberger actually had two such plays, as he also threw a deep pick in the end zone with 9 seconds left in the first half against Jacksonville.

14 Ok. But that choice was made…

In reply to by Aaron Schatz

Ok. But that choice was made by the offensive coordinator, not the QB. It was a bad play call, and I agree the opportunity was there to run another play to get into FG range. But the play call absolutely was an all vert Hail Mary. I have no problem crediting the team or the offense with an INT in this situation. But if you're going to exclude Hail Marys from a QBs stats for the purpose of adjusting INTs for more context/accuracy, then that is one play that should be excluded. Roethlisberger executed the play that was called. And that play was a Hail Mary.

I have no problem, btw, with the pick from the Jags game counting. IIRC, they were at/near FG range, with a TO and shorter underneath options. 

17 Fair point, but did the…

Fair point, but did the Steelers really send everyone deep and the defense keep everyone back when the Steelers when in FG range? Sounds pretty odd to me. Sending everyone deep isn't a Hail Mary if the offense is hoping to catch the defense off-guard when they are still defending the underneath zones.

19 Sending everyone deep late…

Sending everyone deep late in the half worked out great for Tampa Bay against Green Bay in the NFC Championship Game.

https://twitter.com/tometrics/status/1354066568896892929

Here's the Roethlisberger INT we're discussing here. This is not a Hail Mary; this is a lob to one receiver who is double-covered in the corner of the end zone. It's true that all of Pittsburgh's receivers were running verticals, which is odd, but Roethlisberger should have thrown this away and tried something else on second down.

https://twitter.com/titansfilmroom/status/1320831182666264719

21 And then do what? Run the…

And then do what? Run the same play again and then it would count as a Hail Mary?

DCs make dumb calls too. The Steelers benefitted from one the year before vs Miami iirc.

But here the Titans shows the same defense and coverage and the Steelers false started backing them out of FG range. Titans kept the same defense and look for the INT play. The Titans had this play locked down long before the snap. They gave Fichtner a free look and he still called this dumb doomed play. And still had time to change it because the radio helmet doesnt cut out til 15 seconds on play clock.

But, hey, suit yourself. I get that you've got to establish criteria and stick to it. Nobody wants a committee to vet every case and vote on it. I'll get over it.

34 It does look like the…

It does look like the offensive play call left Ben out of options, but I guess you have to draw the line somewhere. And Ben might have changed the play at the line if he thought it was that bad.

37 Personally, I think that…

Personally, I think that throw was meant as one of those "my guy or out the back of the end zone" attempts. I'm basing this on where Diontae Johnson is at the time of the INT, and the fact that Cruickshank jumped pretty high to catch it in his chest. Get that ball a little higher and it reaches the corner pylon, which puts Johnson in much better position. If it goes out the back, you've still got time for a long FG attempt.

The problem is that Ben just didn't have the strength and/or accuracy to get it there, which was unfortunately pretty common for him last season.

20 They were not in FG range vs…

They were not in FG range vs Tennessee. Tho needed maybe five more for Boswells range. They were vs Jax. 

Vs Tenn, two DL standing on the LOS. Only pass rushers. Everyone well off the line, and I think they were all DBs, nine. Two deep safeties 27 yards deep presnap and bailing. Two more 'intermediate safeties' at 12 deep and bailing. Three corners all 10 off with outside leverage. And two more for the TE and back. 

If they thought they might catch them sitting on field goal range, it was entirely evident presnap that that wouldnt be the case. That's why most coordinators call two plays in those situations, and kill one depending on look. Or pay attention when the DC sends nine DBs into the game, and call a play accordingly.

Idk. Seems a play like this warrants more grace than that Wentz pick.

27 Is your position that…

Is your position that Rothlesberger had no discretion to audible out of the look?  I don't follow Pittsburgh, but it seems reasonable to assume a long-tenured veteran QB has the discretion - and perhaps even affirmative responsibility placed on him by the coaching staff - to change the play there.

12 When Wentz was drafted I…

When Wentz was drafted I said taking quy that high, when he had faced such a small number of college defenses with high level athleticism, was an unwarranted gamble. Then I thought the gamble had paid off for the Eagles. Now it seems certain that he just won't be able to make the jump. Shanahan's a lot better suited than most coaches to maximize the chance that such a bet will pay out, and the Niners roster's better able to withstand the wager busting out, compared to the typical team that selects in the 3 spot, but I really disagree with the idea of trading that much draft value to move up to that spot, to get a terrific athlete whose toughest competition was against James Madison University, and who really hasn't competed much against anyone.

15 "He won't be able to make…

"He won't be able to make the jump" is a strange thing to claim about a guy who has put up three seasons north of 0% DVOA (although only just in '19), including a season where he arguably could've been MVP if he hadn't gotten hurt. Sure, he fell off a cliff last year, but that doesn't erase the fact that in his prior three years (and maybe arguably even his rookie year at -4.4% DVOA), he was pretty clearly an NFL starting-caliber quarterback. I don't think it's at all arguable that he made the jump. He clearly did. If anything, I think Wentz shows that it's possible for an FCS QB to succeed quickly.

22 I trust the Eagles judgement…

I trust the Eagles judgement most, since they know him best. He was in a setting where a lot of qbs could have excelled. When the setting deteriorated, he lost his job. Maybe he'll excel again, but that's not the way to bet.

23 I'd normally agree, except…

I'd normally agree, except the Eagles situation was weird. They torched themselves due to a bunch of bad roster decisions which really put them in a bad cap situation, and then 2020 just detonated them. I mean, I'm not sure I'd say that the Pederson fallout wasn't due to that as well - I can easily see Pederson being incredibly frustrated with Roseman and then just being like "yeah, eff this, no one can make this work." Not saying that's what happened, of course, but it's a possibility.

So I'm not completely convinced the Eagles didn't let him go just because they couldn't afford the gamble - 2020 + the bad decisions left them in horrible shape - so bad that even if Wentz did rebound to like, average, they'd still be a bad team for several more years. To be clear, the reason I think that is that that's what I was pushing for. I don't see how Philly becomes competitive again any sooner than 2024.

However the better argument against Wentz is that they had zero trade partners for him - meaning the rest of the league sees him as a very unlikely bet. The Colts have so many mitigating factors that it makes total sense for them to risk it, but for everyone else, there are better options.

Goff, for instance, is different - I'd actually have bet the Rams would've had several trade partners for Goff on their own.

29 I'm not taking a stance on…

I'm not taking a stance on the decision to trade him, I'm just addressing the idea of drafting FCS quarterbacks. You stated that Wentz is a cautionary tale about making the jump from FCS-level competition to NFL-level competition, seemingly on the basis of his abysmal fifth year. My counter-argument is that his first four years actually show quite the opposite - that it's possible (not likely, but possible) for an FCS QB to come in and fairly quickly have success at the NFL level. He absolutely made the leap.

Why a guy would fall off a cliff in his fifth year almost certainly has nothing to do with the level of competition he faced in college. He's thrown more passes in the NFL than he ever had in college.

30 By "made the jump" I mean …

By "made the jump" I mean "demonstrated that he was worth a number 2 selection in the draft". The trade, and what he commanded in value, would seem to indicate that he was not. I was not making a statement with regard to the spees with which an FCS player could start or contribute in the NFL. I was making a statement about whether an FCS career, a brief one at that, could provide enough useful data to generate confidence that a top 5 draft pick would be a great player. Unless it's just a horrible year for the draft, if you're taking somebody that high, you ought to have pronounced confidence, relative to lower picks, that the player will be a starter for many years, and top 5-10 at that position as well. I'm doubtful that such a college career can produce an adequate data set for that purpose, especially at qb.

33 Ehhh. I mean... it's tough…

Ehhh.

I mean... it's tough to argue that from a single data point. Wentz's trade value after ~4 years is likely average for a #2 pick, at worst - so a "push". And expanding it out to the other two high FCS picks (Flacco and Garoppolo) they were "push" and "boom" picks. Even with 3 data points it's obviously nothing solid, but still.

"you ought to have pronounced confidence, relative to lower picks, that the player will be a starter for many years, and top 5-10 at that position as well."

Yeah, no, that's impossible. If that's your standard, teams would, on average, be failing, because obviously you can't have a top 5-10 player at their position every year in the top of the draft. There's a difference between what you hope for and what you expect

I mean... arguing "it's not worth the added uncertainty..." I'm just not sure about that. Obviously Wentz at #2 and Flacco at #18 were below what you would hope for, although they certainly did about/above what you'd expect. Garoppolo is certainly well above what you would hope for.

So, I mean... if you look at FCS QBs it's hard to say they're "underperforming" when literally the 3 highest-drafted ones in the past years all got starter-level contracts. Obviously you can't say the same thing about the FBS ones.

Yeah, you could argue "but none of them have become top 5-10 QBs!" and that's definitely true. But if you give me 3 random FBS QBs drafted in the first round, a large part of the time I'm gonna end up with zero top 5-10 QBs, too. I mean, going from 2014-2016, you've got 9 FBS QBs drafted in the first 2 rounds and at most a single top 10 QB (Carr) - so you grab 3 from that set and 2/3 of the time you're not ending up with anyone, either.

Just not positive that competition level matters in terms of scouting evaluation. In fact I tend to be more wary of the QBs from the highest end schools. 

36 We want to look at FBS qbs…

We want to look at FBS qbs drafted in picks 1-5. Yes, there is also a risk of over-weighting college performance of qbs at the perennial top 5 schools. I just don't think you learn much more in watching how somebody did against markedly inferior athletes, than you do in pro day workout.

Yes, you are going to be wrong more than you're right, but if you don't have sound reason to be confident that a guy picked #3 will be a top 5-10 player at the position, compared to a guy picked #35, then GMs and Directors of College Scouting are being overpaid.

38 "We want to look at FBS qbs…

"We want to look at FBS qbs drafted in picks 1-5"

Bwaha, you're making things worse. FCS set reduces to Wentz, and even if I go from say 2013-2018, you've now got 7 FBS QBs drafted in picks 1-5, and the best one out of the set is either Goff or Mayfield, and the "average" of those 7 is worse than Wentz (Mayfield, Darnold, Trubisky, Goff, Winston, Mariota, Bortles - Mayfield's above, let's call Goff above, Darnold's a push, Trubisky/Winston/Mariota/Bortles are below based on contract valuation).

I mean, I get your concern, but I don't think Wentz going from #2 pick to 2nd/conditional 1st after 5 years is confirmation of your concern. Like I said, that's likely the average result of a top-5 pick.

 

39 Well, I've been saying for…

Well, I've been saying for 20 years that sample sizes with regard to nearly all draft analysis are too small to make it useful. Names like Trubisky, Goff, Bortles, etc. lead to my other suspicion, that college QBs are overvalued in terms of drafting priority. 

41 QB free-agent salaries have…

QB free-agent salaries have been growing faster than cap for the entirety of the salary cap era, and they're by far the most expensive position - which means it's basically impossible for college QBs to be overvalued in drafting priority. Giving Goff a second contract was a mistake, but Goff and Trubisky provided significant excess value to their teams relative to the typical top 5 pick.

Do you want a list for the non-QB positions? It's gonna be just as full of mediocre players. The actual value of a draft pick is significantly lower than what it's made out to be by draftniks. I mean, for DE, you've got Ziggy Ansah, Dion Jordan, Clowney, Bosa, and Thomas. Yay, you've got 1 top 10 player in that list, a few guys who had maybe a good season or two and then bounced around the league like a pinball. And if it were me, in that list, I'd call Bosa a boom (likely a 'ka-boom'), Clowney and Ansah are hits, and Jordan and Thomas are busts.

But in that list, for instance, Bosa can't ever be the standard. Clowney was the first overall pick, and if I'm the Texans I wouldn't be going and looking at the draft pick and saying "how did we go wrong." I'd look at that pick and say yeah, that's fine. If teams are going into the draft expecting guys to blow up the league, I mean, they'd be jumping off buildings with depression.

43 "The only issue with Clowney…

"The only issue with Clowney's performance has been injury, a random event."

Uh... why in the world do you think that player injury rates are random? Of course they're not. If they were, there wouldn't be buckets of studies in sports medicine detailing correlation rates. Clowney missed multiple games in college his final year due to injuries. He's had a recurring foot injury since high school. How in the world can a "recurring injury" be random?

I'm not saying that NFL execs could've predicted "oh, he'll be limited his entire career by injuries." I'm saying that him being limited by injuries was always a possibility. There's never a player you can point to in the draft and say "no doubt, he'll be a Hall of Fame player." For most college players they've just barely finished physically growing!

Draftnik expectations for top picks are just hype. They're just there to get people excited. They're not real. Clowney being drafted by the Texans didn't get anyone fired for that evaluation. Same thing with Wentz. I'm sure Wentz's successes and struggles didn't surprise anyone in the Eagles. I'm also sure they hoped he'd surpass his limitations, but I'd bet they knew they could manage them for a few years. I mean - Foles in 2017 wasn't a cheap pickup! They grabbed him because they knew that Wentz might have durability issues.

44 Er, I neither stated or…

Er, I neither stated or implied that any top draft pick could be predicted to be a Hall of Fame player, so it strange that you would write as if such a contention was made. Yes, people injured a lot in college are more likely to be injured in the NFL, but Clowney's been an obviously good player in multiple years, and there was good reason to have high confidence that injury would be the only thing that prevented that from happening. Fundamentally, this is a debate about market efficiency, and I just don't think there are nearly enough transactions to be confident that this is on the whole an efficient market. That same small number of transactions, of course, means you really can't establish inefficiency, either. My inclination would be to avoid drafting players at the top of the 1st round if they never competed against top level athletes in college, but as I said elsewhere, if the team is not typical of those that pick at the top of the draft, meaning it doesn't have glaring deficiencies at multiple positions, it makes more sense to take that risk.

 

45 A frequent top 5 player is a…

A frequent top 5 player is a Hall of Fame player, so I don't think there's much difference between the statements.

Clowney's been through 3 teams. He's clearly never been valued at a top 5 DE, which is more like 20M+. But he's likely easily average (or close) for top overall picks in terms of his valuation. It's what you expect: a guy who has the physical tools, but you don't know if he'll handle the mental/physical toll.

*Everyone* has asterisks coming out of college. If you scout right, I don't see FCS QBs being intrinsically riskier than any other QB. It's fine to say "I just think FCS QBs are too risky and the sample size is too small" but you can't say "you've gotta be confident this guy will turn out to be a top 5-10 player," because no one's *ever* confident of that.

52 That's why I said there's …

That's why I said there's *not much* difference, and I included the "not much" for a reason, too. Someone who's between 5-10 is a borderline Hall of Fame player. You can't even have any confidence of that, either.

48 Wentz was a top 5-6 QB by…

Wentz was a top 5-6 QB by advanced metrics until the Eagles offense, and organization, imploded around him. Where they drafted a backup QB 25-50 picks too high in R2 instead of shoring up the OL, or providing him with weapons other than injured Alshon, injured DeSean and rookie WRs/UDFAs to throw to. 

49 5-6?

That's generous. Was 17th/32 in EPA+CPOE Composite from 16-19. Dropped to 23rd when 2020 is included (or 24th if I drop the min plays to 1400 to include Kyler).

53 Ehh. The Eagles offense was…

In reply to by ImNewAroundThe…

Ehh. The Eagles offense was already detonated in '19 (by the end of that year it was "anyone wanna play WR?") and '18 was a post-ACL recovery year, so Wentz actually putting up the numbers he did was fine. Problem with that is that you're basing everything on less than a year, and there've been plenty of flash-in-the-pan QBs.

That being said, if we go back to the original argument and ask "did it make sense to draft Wentz" I think the answer there pretty clearly is "yes" - Wentz pretty clearly demonstrated that the skills he showed in college translated to the NFL. Now, obviously, those skills don't seem to last, but that's not something you can really predict in the draft, and it's also a viable question as to whether that's Wentz's problem or the coaching staff.

Really hard for me to understand why people don't seem to get the argument that a player can be a good draft pick and an NFL long-term failure at the same time.

55 So

it's all based on the 2017 season. A season that's already been shown to be unstable (was insane on 3rd down, and and not the other 3 downs). And knowing that, hard to ever actually put him that high at the end of the day/season. Especially what Foles did shortly after, and we knew he wasn't ever that high (outside another shortened part of a season early). 

15th/32 from 16-18 (lower than Dak even with the insane 17 3rd downs). Sure the pick was fine but long term or not, he wasn't ever that high. No advanced metrics were putting him that high before the "collapse" of the team. Unless you're looking a very small sliver of his profile (in which everything was perfect around him). There was a reason Dak won ROTY (yet Dak v Wentz (v Goff) was still a debate a year ago...not so much anymore).

47 He successfully made the jump from FCS

and it's patently absurd to say that he didn't, MVP-type season in only his 2nd year, top 6 in QBR from 2017-2019 and top 12 each year, before the Iggles absurd OL injuries AND awful lack of WR weapons [even Reagor got hurt] led him to having a terrible year. 

And this nonsense about no other teams wanted him is due to his big salary cap number [why the Iggles moved him] not that 31 other teams don't think he's a competent starter anymore. Due to injuries on the OL/lack of talent at skill positions, injuries to starting WRs the past two seasons, Wentz singlehandedly carried them to the Division Title in 2019 relying on Boston Scott [lol] and Greg Ward [lmao]. There were games last year where both Ertz and Goedert didn't play with injuries. Miles Sanders got hurt again.

Wentz had bad throws partially due to no time to throw, and no WRs who could get open or catch 50-50 balls. TY Hilton will be the best receiver he's played with in 4 seasons, if not ever. 

W.Allen doesn't like the Iggles nor Wentz, we get it. But his is a world-class terrible take. 

24 Effect on DVOA

We can see the spread of INTs. If all quarterbacks were equal on the successful throws, how big a difference would the spread of INTs have on Offensive DVOA?

26 This would be a very…

In reply to by ChicagoRaider

This would be a very complicated question to answer, because you'd have go back and not just count how many interceptions a quarterback "should" have thrown, but which ones should count and which ones should not. There is a huge spread in the DVOA spread between the best and worst interceptions. An interception on a screen pass on first down at your own 20 is much worse than an interception 40 yards downfield on third-and-forever.

32 Interceptions account for…

In reply to by ChicagoRaider

Interceptions account for something like 1/6 of the spread in ANY/A, so their effect on passing DVOA is probably in the same ballpark.

31 based on AI rate....

Just like we'll be seeing in the record books for years to come, Rodgers and Mahomes are right near each other at the bottom of the rate stack, separated only by Gardner Minshew.
Wait, what?

Herbert and Burrow had fantastic rates for rookies!