Writers of Pro Football Prospectus 2008

Most Recent FO Features

RamseyJal16.jpg

» Week 6 DVOA Ratings

The No. 1 Chiefs lost their first game. The No. 30 team is 3-2. The Jaguars have the worst DVOA ever for the top-ranked defense after Week 6. These are among the signs of how strong parity has been so far during the 2017 NFL season.

21 Mar 2017

NFL Draft: Round-by-Round QB Study (1994-2016)

by Scott Kacsmar

"You can get a quarterback later in the draft."

Three seasons have passed since the last time we did this study, and that little phrase is not going to go away after Dak Prescott, a late fourth-round pick, won the AP Offensive Rookie of the Year for the 2016 season. While some teams traded a mint to move to the head of the queue to grab Jared Goff and Carson Wentz, the Cowboys simply used their fourth-round compensatory pick to take Prescott. The performance they got from Prescott was arguably the greatest rookie quarterback season in NFL history. Prescott deserves high praise, but FO readers surely know that these are not the normal expectations for a fourth-round pick. Even at the top of the draft, a performance this good is tough to find at the game's most crucial position.

Should he continue his success over a long career, Prescott will stand as a great outlier among late-round quarterbacks. His Dallas predecessor, Tony Romo, already is a great outlier among undrafted quarterbacks, but neither fans nor NFL executives view the large, undrafted talent pool as a reasonable method of finding a franchise quarterback. The data supports that stance as well, but it also paints a gloomy picture of what the late rounds produce.

Still, we can count on a decent quarterback prospect to slip in the 2017 draft, a class that does not have an obvious Andrew Luck-like No. 1 pick, but does have a lot of "developmental quarterbacks" not expected to contribute right away. Some team will champion the thought that they got a real steal. Some writer (on another site) will put together a slideshow to rank which late-round quarterbacks are most likely to duplicate Prescott's historic 2016 results.

The truth is that teams still have to consider the top 40 picks as the most probable spot to land a quarterback of Prescott's caliber. His success keeps the dream alive that a quarterback can be found later, but it's just not the reality of the NFL. Every NFL draft now comes with a reminder that Tom Brady was a sixth-round pick, yet among the 92 quarterbacks drafted in the fifth round or later since 2001, only three have thrown at least 1,000 passes: Derek Anderson, Matt Cassel, and Ryan Fitzpatrick. That trio's "success" is driven mostly by their hard-to-explain continued employment rather than the actual success of their passes.

Teams are desperate to find a quarterback, but will Prescott's instant success really change anything in that process? The last few years suggest it won't if we use Russell Wilson as a direct comparison. Not only do they have a comparable playing style, but Wilson and Prescott are the only quarterbacks not drafted in the top 40 since 1978 to start all 16 games as rookies.

Height brought the 5-foot-11 Wilson down in the draft despite some stellar projections for pro success. A total of 47 players have been drafted since 2013 (post-Wilson) to play quarterback in the NFL, and only three were shorter than 6-foot-2. That includes 5-foot-11 B.J. Daniels, a 2013 seventh-round pick by the 49ers who went to Seattle where he was eventually converted to wide receiver -- the only player among the group of 47 switched to another position. The other two shorties were both chosen by the Browns: the troubled Johnny Manziel (6-foot-0) and 2016 rookie Cody Kessler (6-foot-1). Similar things can be said of the Buffalo Bills, but they have been starting Tyrod Taylor (6-foot-1) the last two seasons, and he is very much in the Wilson mold of being a magician at quarterback. He has also provided the team with some of its best passing numbers since Hall of Famer Jim Kelly retired, yet the Bills still waffled this offseason on bringing Taylor back for 2017. (However, if Taylor starts for the entire 2017 season, he will join the Anderson-Cassel-Fitzpatrick trio as a late-round pick with 1,000 career pass attempts.)

None of the 12 highest-graded quarterbacks at NFL.com in this year's draft are shorter than 6-foot-2. A lack of other short quarterbacks post-Wilson would make perfect sense if there were no viable prospects out there at that height. However, that has not been the case. Former Oregon quarterback Vernon Adams earned some supporters during last year's draft process, but no one took a chance on the 5-foot-11 quarterback in the draft. Seattle was the first team to give him a tryout before he headed to the CFL. Seattle did add TCU's Trevone Boykin, another Wilson-esque player who only stands at 6-foot-0, as an undrafted free agent. Bowling Green's Matt Johnson (6-foot-0) struck out despite some high praise, only earning a rookie mini-camp tryout with the Bengals. So it's basically been Seattle or bust for short quarterbacks trying to make it in the NFL.

Meanwhile, tall human beings Brock Osweiler (6-foot-7) and Mike Glennon (6-foot-6) were drafted higher into the NFL than Wilson and Prescott were. Osweiler and Glennon were also both given contracts in excess of $15 million per year despite very few career achievements to that point. Having ESPN's Jon Gruden fawn over how tall you are is not an achievement. Even if they look like Gumby and Big Bird having a "who can throw it the farthest?" contest, Glennon and Osweiler still look the part of what many have been engrained to view as the traditional NFL quarterback: tall, strong arm, and if we're being completely honest here, white. That's not to accuse any teams of being prejudiced in their team building. However, even if it's only subconsciously done, there still may be some teams who view a Glennon as a more "coachable" player, and avoid taking a chance on an athlete who may go off script more often simply because he has the physical tools to do so. If a DVOA dynasty like Seattle is one of the only teams willing to be forward thinking enough to gain a competitive advantage by trying out different athletes, then more power to them.

Not to toot our own horn, but our college projection systems have always loved Wilson ("The Asterisk"), and last year's QBASE outlook had Prescott with the second-best DYAR projection among quarterbacks in his draft class. (The caveat there is that QBASE is supposed to be used only on the top 100 picks, and Prescott ended up falling lower than that.) Osweiler has the lowest QBASE of any top-100 drafted quarterback since 1997, and Glennon is fairly close to the bottom. The early signs suggest that Christian Hackenberg (fifth-worst projection since 1996) going 84 spots ahead of Prescott in the draft is another case of perception failing. The only eye test a quarterback should need to pass is whether or not he can effectively move the offense on a consistent basis. If he can do that, then the attributes that the player has no control over should be irrelevant.

So when your team drafts a quarterback in the fourth round next month, please do not expect him to be Prescott. If anything, he is more likely to be Cardale Jones, the Ohio State quarterback who five of our staff members (including myself) predicted back in September 2015 would be the No. 1 pick in the 2016 draft. He actually went 139th, four spots after Prescott, to Buffalo.

No one has ever said that finding a quarterback was supposed to be easy.

The Updated Quarterback Draft Study (1994-2016)

I collected data for all 272 quarterbacks* drafted in the NFL from 1994 to 2016. The 1994 draft is a pivotal one since it's the first of the salary cap era and features the current seven-round format. This still presents the common problem in draft studies of including active players with incomplete career data, but I think that's suitable for this. In theory, the NFL quarterback climate should always consist of some old veterans with huge career numbers, some guys in their prime with the best rate stats, mediocre compilers, unproven youngsters, and a few guys destined for the clipboard (or Cleveland). Here we also have plenty of averages and rate stats to consider instead of just focusing on raw totals.

(*Note: In my draft database, Pat White and Joe Webb are quarterbacks. Webb in particular has bounced around from wide receiver back to quarterback to kick returner in Carolina, so none of his statistics since 2013 are included here. Brad Smith, Reggie McNeal, Isaiah Stanback, and B.J. Daniels are considered wide receivers. Steve Bellisari (2002) was converted to safety.)

NFL Draft: Quarterback Games (GP) and Games Started (GS) for 1994-2016
Round #QB 1+GS 0 GP GP GS Record Pct. PO GS PO Record Pct.
1 57 57 0 4539 4192 2150-2033-7 0.514 204 106-98 0.520
2 25 23 1 1368 1045 534-509-2 0.512 39 16-23 0.410
3 28 23 2 807 567 249-316-2 0.441 16 9-7 0.563
4 34 20 3 753 464 200-262-2 0.433 9 2-7 0.222
5 32 13 14 222 98 37-61-0 0.378 3 1-2 0.333
6 48 18 22 898 679 373-306-0 0.549 50 31-19 0.620
7 48 19 20 726 386 155-229-2 0.404 3 0-3 0.000

Our first table looks at game appearances, which are obviously skewed towards quarterbacks in the first three rounds getting many more opportunities to play than the later-round quarterbacks. Out of 110 quarterbacks drafted in the first three rounds, only Giovanni Carmazzi, David Greene, and second-rounder Christian Hackenberg failed to play in a regular-season game. However, Hackenberg was just a rookie last season and he should get a chance eventually. Pat White, the athletic running quarterback from West Virginia, is still the only second-round quarterback in this study who retired without starting a game. He was really more of a fool's gold pick for the Dolphins after they had some success with the Wildcat in 2008.

The first round has produced 953 more starts than the other six rounds combined. Even the winning percentage dips after the second round, but of course we see a huge upswing in the sixth round thanks to Tom Brady and the Patriots. In fact, Brady has as many playoff starts (34) as the other 214 quarterbacks drafted in rounds two to seven have playoff wins (34) since 1994. We'll get to the sixth round later, but it should be noted that 43.8 percent of the quarterbacks drafted in rounds six or seven failed to play in a game.

It is one thing to play in a game, but how about the performance in those games? Here are the passing stats for the average quarterback by draft round:

Average Quarterback's Passing Stats by Round (1994-2016)
Round #QB Att. Cmp Pct. Yds YPA TD TD% INT INT% PR Sack%
1 57 2,418 1,464 60.6% 16,975 7.02 103 4.27% 68 2.83% 84.2 6.31%
2 25 1,392 836 60.1% 9,579 6.88 56 4.03% 42 3.02% 81.7 6.33%
3 28 685 415 60.7% 4,703 6.87 26 3.76% 20 2.92% 81.6 7.56%
4 34 467 277 59.2% 3,155 6.75 18 3.77% 14 3.08% 79.3 7.07%
5 32 113 60 53.0% 686 6.05 3 3.06% 4 3.86% 65.6 7.62%
6 48 496 300 60.5% 3,419 6.89 21 4.25% 13 2.71% 84.1 6.08%
7 48 279 160 57.3% 1,834 6.58 11 3.80% 9 3.38% 75.8 6.46%

For a lot of these numbers, we see a general pattern where they are best in the first round and then taper off as you advance a round, with the fifth round being the worst (more on that later). But then the numbers go back up in the sixth round (again, strong Brady factor) and down for the final round.

I collected rushing data too, but it was not worthy of a table. The first-round quarterbacks again have the best production, shocking no one. Great athletes are always going high in the draft and players such as Cam Newton, Donovan McNabb, Michael Vick, and Robert Griffin III have certainly boosted those rushing stats.

Speaking of rushing, I was able to collect total DYAR (rushing and passing) for every quarterback's career. I also calculated a weighted average for passing DVOA. Citing the average DYAR or especially average DVOA of a sixth-round quarterback can be more misleading than other metrics. It's not so problematic for the earlier rounds, since most of those quarterbacks play. Without playing, a quarterback really has no DYAR/DVOA, which of course can have a negative value for poor performance. So crediting quarterbacks who don't play with zero DYAR or DVOA and using that in the average would result in an inaccurate measure of the quality of the group, when you consider that 16 ranked quarterbacks had negative DVOA in 2016 alone. Even if Ryan Tannehill and Cam Newton were not statistical marvels in 2016, they still had more value than Jeff Driskel and Brandon Doughty (late-round rookies who did not play).

Technically, a quarterback who never throws a pass does not have a completion percentage either, but we can still calculate his round's completion percentage with him omitted. We have to omit him from DVOA too, however. We can't just assume he would be replacement level (roughly -13.3%) since we're weighing things by number of plays -- something these quarterbacks have none of. On a small sample size, which is expected of a poor, late-round quarterback, the DVOA can be frighteningly bad.

For example, Joe Hamilton was a seventh-round pick in 2000. He had one career dropback, on which he was strip-sacked on fourth-and-8 in the red zone. Granted, it was at the two-minute warning with Tampa Bay leading 27-7, but his career passing DVOA is minus-2,167.2%. This does get weighed based on the number of plays (one in this case), but players like that are why we need to take a different approach for the following table.

The solution was to weigh all players in the round equally. This can raise eyebrows for the early rounds where the first pick has much different expectations than the 32nd pick, but first-round quarterbacks were not the problem since almost everyone plays. In the later rounds, it's really just a marginal difference at best in value, so that should not be a problem. To get the average passing DVOA by round, I weighed each round's quarterbacks with at least one pass play equally to get the average DVOA.

For example, the sixth round has 48 quarterbacks, but only 26 had a dropback and calculated DVOA. I divided each quarterback's DVOA by 26, and the sum of those figures was the sixth round's DVOA (-29.8%). That is more accurate then weighing by number of attempts, which would be skewed so heavily by Brady. I also collected the Weighted Career Approximate Value (AV) from Pro Football Reference for every draft pick.

Total DYAR and Passing DVOA by Round
Round #QB AVG CarAV AVG Total DYAR Total DYAR/Game AVG PASS DVOA
1 57 46 2,142 26.9 -11.1%
2 25 27 1,154 20.2 -19.6%
3 28 13 275 8.5 -22.2%
4 34 9 228 9.1 -26.9%
5 32 2 -214 -16.4 -43.2%
6 48 9 926 26.8 -29.8%
7 48 5 24 0.9 -105.4%

This makes the seventh round look really terrible, but it's hard to disagree with that conclusion. Also, Hamilton still brings the DVOA down dramatically since the round would be -28.0% with him excluded. The closest thing to a notable seventh-round quarterback with a positive DVOA was Steve Matthews, who only had 44 pass plays in his career.

While the average first-round quarterback's DVOA (minus-11.1%) would rank 25th in 2016, there's no denying where most of the best quarterbacks of the last two-plus decades come from. Here's a weighted career passing DVOA leaderboard of the top 20 quarterbacks drafted in the NFL since 1994 (also included Total DYAR):

Leaders in Weighted Passing DVOA (1994-2016)
Rk Quarterback Round Total DYAR DVOA
1 Peyton Manning 1 26,356 29.9%
2 Tom Brady 6 21,278 26.1%
3 Aaron Rodgers 1 11,214 20.5%
4 Drew Brees 2 18,814 19.8%
5 Matt Ryan 1 9,848 17.0%
6 Ben Roethlisberger 1 11,504 16.3%
7 Philip Rivers 1 11,075 16.2%
8 Chad Pennington 1 4,581 14.8%
9 Russell Wilson 3 4,577 13.4%
10 Kirk Cousins 4 2,377 10.9%
11 Steve McNair 1 7,764 10.5%
12 Carson Palmer 1 8,907 10.2%
13 Matt Schaub 3 4,565 9.0%
14 Daunte Culpepper 1 4,919 7.0%
15 Andy Dalton 2 3,613 5.2%
16 David Garrard 4 2,859 4.0%
17 Matthew Stafford 1 4,564 4.0%
18 Byron Leftwich 1 1,608 3.4%
19 Donovan McNabb 1 6,229 3.1%
20 Jameis Winston 1 1,050 2.9%
Only includes QBs drafted since 1994 (min. 1,000 passes)

No real surprises, right? The cream really rose to the top here, and 13 of the top 20 quarterbacks were first-round picks. We'll spend the second half of our draft study by going over some of the drafting dynamics in each round.

First Round

Since a simple "bust/success rate" can be so largely subjective, once again we will use a tier format of how I view these quarterbacks as of March 2017.

Tiers In My Eyes: 57 First-Round Quarterbacks
The Good The Undecided The Mediocre The Ugly The Winning Ugly
Peyton Manning Blake Bortles Byron Leftiwch Akili Smith Mark Sanchez
Aaron Rodgers Carson Wentz Jason Campbell Blaine Gabbert Rex Grossman
Ben Roethlisberger Jameis Winston Jay Cutler Brady Quinn Tim Tebow
Philip Rivers Jared Goff Kerry Collins Brandon Weeden Trent Dilfer
Matt Ryan Marcus Mariota Ryan Tannehill Cade McNown Vince Young
Andrew Luck Paxton Lynch Sam Bradford Christian Ponder
Eli Manning Teddy Bridgewater David Carr
Steve McNair The ALEX EJ Manuel
Chad Pennington Alex Smith Heath Shuler
Matthew Stafford J.P. Losman
Donovan McNabb Jake Locker
Carson Palmer JaMarcus Russell
Daunte Culpepper Jim Druckenmiller
Cam Newton Joey Harrington
Joe Flacco Johnny Manziel
Michael Vick Josh Freeman
Kyle Boller
Matt Leinart
Patrick Ramsey
Ryan Leaf
Tim Couch
Robert Griffin

I placed 16 quarterbacks in The Good heap here, and that number could certainly rise with so many recent picks still needing to prove their worth. That may not sound good out of 57 players (28.1 percent), but we're only giving 19 quarterbacks a "Good" designation in the final six rounds combined, for a success rate of 8.8 percent. Matt Ryan is the fifth first-round quarterback listed to win an MVP award. The other six rounds have produced one MVP winner in this period (Brady).

When it comes to The Undecided, Blake Bortles is really one more lousy season away in Jacksonville from sliding over to The Ugly pile. Marcus Mariota looks to be trending in the right direction, and hopefully he will recover well from his broken leg. Unfortunately, Teddy Bridgewater's career could be over after two seasons after his significant injury last summer.

Alex Smith could arguably fit into any category here, so I gave him his own suitable tier. He's certainly had his share of The Winning Ugly days where he leaned on the rest of his team to carry the load. Those quarterbacks all had winning records in the NFL and will be hard to forget for observers of this era. That's much different from The Ugly collection of 22 quarterbacks, including a lot of major busts. Robert Griffin III and Josh Freeman could probably deserve one-year wonder tags, but the failure to pick their careers back up makes it feel right that they belong where they are.

Sam Bradford and Ryan Tannehill are hard to place since we're still hearing annual excuses for why next year will be the breakout year that never comes. They absolutely belong in The Mediocre, as does Jay Cutler after 11 seasons.

Second Round

The second round has been the least utilized round for getting a quarterback, with only 25 selections since 1994.

Tiers In My Eyes: 25 Second-Round Quarterbacks
The Good The Average The Failed Replacements The Ugly The Drowned Dolphins
Drew Brees Charlie Batch Brian Brohm Brock Osweiler John Beck
Colin Kaepernick Kevin Kolb Drew Stanton Chad Henne
Andy Dalton The Future? Quincy Carter Geno Smith Pat White
Jake Plummer Jimmy Garoppolo Tarvaris Jackson Jimmy Clausen
Kordell Stewart Todd Collins Kellen Clemens
Derek Carr Him? Marques Tuiasosopo
Christian Hackenberg Shaun King
Tony Banks

Most of the successes here were drafted in the range of 32 to 42, including Derek Carr (36th) in 2014. If Carr can sustain his success for a couple more years, he'll be the second-best quarterback behind Brees here. Charlie Batch (1,199) actually had more total DYAR than Kordell Stewart (700) in their careers, but Stewart will always be remembered more fondly for his 1997 and 2001 seasons. Well, at least up until he imploded at home in two AFC Championship Game losses, but he was better at his peak than Batch, who was a really solid backup for the Steelers for a long time after escaping Matt Millen's mess in Detroit.

Still, the messiest thing here might be Miami failing on three straight second-round picks with John Beck, Chad Henne, and Pat White from 2007 to 2009. The Jets may not look so hot either if Christian Hackenberg doesn't work out so soon after Geno Smith, who is now a backup for the other New York team.

The Patriots will have some interesting decisions to make with Jimmy Garoppolo, but he could actually just be Tom Brady's successor. What's interesting here is how many of these players were supposed to be successors to notable quarterbacks, but that did not pan out. Todd Collins wasn't a Jim Kelly in Buffalo, nor was Quincy Carter able to make people forget about Troy Aikman. Kevin Kolb was out of Philadelphia not long after Donovan McNabb was traded. Brian Brohm was a rare misstep by the Packers at quarterback in a 2008 bridge year between Brett Favre and Aaron Rodgers.

Brock Osweiler could go under a few categories here. He was insurance for Peyton Manning's neck in Denver in 2012, but he was not the true successor after taking the money from Houston last offseason. The Texans already made a salary-cap dump trade to the Browns this offseason to unload Osweiler after one miserable season.

Third Round

Warning: a really short and a really tall quarterback are about to be discussed again.

Tiers In My Eyes: 28 Third-Round Quarterbacks
The Good The Tall One The Ugly The Patriots
Brian Griese Mike Glennon Andrew Walter Kevin O'Connell
Matt Schaub Bobby Hoying Ryan Mallett
Nick Foles The Cockroach Brock Huard Jacoby Brissett
Russell Wilson Josh McCown Brodie Croyle
Chris Simms The Browns
The Decent Backup Dave Ragone Eric Zeier
Chris Redman David Greene Charlie Frye
Giovanni Carmazzi Colt McCoy
Clipboard Jesus Jonathan Quinn Cody Kessler
Charlie Whitehurst Stoney Case
Captain Checkdown The Future? Meh
Trent Edwards Garrett Grayson
Sean Mannion

With the third round we get to the point of the draft where teams are not necessarily looking for a starter, but just a future plan or a competent backup. Garrett Grayson does not seem to be on track to be Drew Brees' successor in New Orleans, which makes that look like a wasted pick in 2015 when defenders such as Jordan Hicks (Eagles) and Danielle Hunter (Vikings) were still on the board. Likewise, with Jared Goff in Los Angeles, Sean Mannion is not likely to do anything of value for the Rams in his career. Of course, no quarterback has done anything of value for the Rams in the last decade.

The far right has a group of players drafted by the Patriots and Browns. Ironically, Eric Zeier was a third-round pick in 1995 when Bill Belichick was calling the shots in the swansong for Cleveland 1.0. In New England, Kevin O'Connell and Ryan Mallett were not the wisest choices when Tom Brady was still obviously in his prime. Now last year's selection of Jacoby Brissett made more sense with Brady inching towards 40, and Brissett even started two games while Brady was suspended.

The Browns 2.0 have used their picks to try finding cheap starters without any real success so far. Charlie Whitehurst and Josh McCown also played for Cleveland last year, but which teams haven't they played for yet? The world could go to nuclear war and Whitehurst would still resurrect his playing career somewhere while McCown would survive with the cockroaches. We couldn't even finish this study without McCown picking up a deal for $6 million with the Jets.

The last time we did this study, the 2014 Buccaneers were choosing between McCown and Mike Glennon as their starting quarterback. McCown got the nod and led a bad offense as was predicted, and the Buccaneers ended up getting Jameis Winston with the No. 1 pick in 2015. Glennon showed enough in his first two seasons to deserve a shot at more playing time, but it was really a huge surprise to see Chicago so amped up to dump Jay Cutler to sign Glennon to a three-year deal for $45 million this offseason. Even if it is really a one-year audition, the thought of Glennon throwing to the likes of Eddie Royal, Kevin White, Cameron Meredith, and Markus Wheaton does not sound like an imposing offense. Glennon is also not going to a notable offensive system in the way that some other third-round successes did.

We threw some nicknames out for fun for these third-round picks, but the fact is that only Russell Wilson has really been a consistently strong starter for a good period of time. Brian Griese and Matt Schaub had their moments in that Mike Shanahan-Gary Kubiak style of offense, and Nick Foles really took advantage of Chip Kelly's system in 2013, but none of those players have ever had a five-year stretch of success like Wilson has started his career with. He truly is The Asterisk.

Fourth Round

Dak Prescott is not even the first NFC East quarterback from the fourth round to shine in recent years. Kirk Cousins may have been insurance to Robert Griffin III back when Washington drafted both in 2012, but he has proven to be the superior quarterback in the NFL. Cousins is set to play on the franchise tag for the second year in a row and has been on a pretty strong run of play over the last season and a half. Can he be frustrating at times? Sure, but given what Washington is used to at the position, Cousins has been an excellent fourth-round pick.

Tiers In My Eyes: 34 Fourth-Round Quarterbacks
The Good The Wait and See The Ugly The Fun The Bachelor
Aaron Brooks Bryce Petty Chad May Landry Jones Jesse Palmer
Dak Prescott Cardale Jones Chris Weinke Luke McCown
David Garrard Connor Cook Danny Kanell Matt Barkley
Kirk Cousins Tom Savage Danny Wuerffel Rob Johnson
Kyle Orton Dave Barr Sage Rosenfels
Doug Nussmeier Seneca Wallace
Jeff Lewis
Joe Germaine
Logan Thomas
Mike Kafka
Pat Barnes
Perry Klein
Rohan Davey
Ryan Nassib
Stefan LeFors
Stephen McGee
Steve Stenstrom
Tyler Wilson

While Prescott's lack of mistakes was one of the hallmarks of his rookie campaign, many of the others in this group loved to have fun on the field. Who can forget this backwards pass from Aaron Brooks? Rob Johnson never turned down a sack from any defender. Seneca Wallace was like an early prototype for Russell Wilson, right down to being manufactured in Seattle. Slingin' Sage Rosenfels threw 30 touchdowns to 29 interceptions in his career, and took one memorable flight on the Rosencopter. Josh McCown wasn't the only McCown brother to have improbable longevity in this league. And who can forget Jesse "The Bachelor" Palmer?

The 2013 draft was an interesting one for the four fourth-round choices. We once thought Matt Barkley was going No. 1 overall instead of 98th to Philadelphia. In 2016, Barkley had an intriguing season for the Bears with a high ALEX and the third-lowest failed completion rate since 1989. Ryan Nassib was once thought to go as high as No. 8 to Buffalo in the 2013 draft. He went 110th to the Giants and led the greatest irrelevant comeback in NFL history, helping to erase a 26-0 deficit in the fourth quarter of a 2014 preseason game against the Colts. Tyler Wilson had some good tape, but never caught on in the NFL. Then there is the case of Landry Jones, whom general manager Bill Polian thought would be the best quarterback in this draft class. Jones' first snap in his rookie preseason involved running into his own running back in the end zone for a safety. Last preseason, Jones managed to throw four interceptions in one half against the Eagles. He still might be more serviceable than Bryce Petty or Tom Savage, but time will tell.

Fifth Round

The fifth round has been so shockingly barren that it makes sense to just post all of the quarterbacks with their stats together again.

Tiers In My Eyes: 32 Fifth-Round Quarterbacks Since 1994
Year Quarterback Att. Cmp Pct. Yards YPA TD INT PR CarAV TOT DYAR DVOA
2002 Craig Nall 48 30 62.50 402 8.38 5 0 123.8 1 206 43.7%
2014 AJ McCarron 119 79 66.39 854 7.18 6 2 97.1 3 153 6.9%
2005 Dan Orlovsky 512 298 58.20 3132 6.12 15 13 75.3 7 231 -4.4%
2007 Troy Smith 234 121 51.71 1734 7.41 8 5 78.5 6 8 -12.9%
2001 A.J. Feeley 762 425 55.77 4618 6.06 28 31 69.1 7 -353 -17.9%
2011 T.J. Yates 227 132 58.15 1534 6.76 6 8 72.8 4 -244 -27.3%
2001 Mike McMahon 515 229 44.47 2867 5.57 15 21 55.1 6 -718 -32.4%
2010 John Skelton 602 320 53.16 3707 6.16 15 25 63.0 4 -955 -35.4%
2008 Dennis Dixon 59 35 59.32 402 6.81 1 2 71.4 2 -77 -36.6%
2008 Josh Johnson 177 96 54.24 1042 5.89 5 10 57.7 4 -337 -42.7%
2002 Kurt Kittner 114 44 38.60 391 3.43 2 6 32.5 1 -302 -52.7%
2002 Randy Fasani 44 15 34.09 171 3.89 0 4 8.8 1 -182 -74.2%
2000 Tee Martin 16 6 37.50 69 4.31 0 1 25.3 0 -75 -78.4%
2004 Craig Krenzel 127 59 46.46 718 5.65 3 6 52.5 1 -684 -85.4%
2015 Brett Hundley 10 2 20.00 17 1.70 0 1 0.0 0 -86 -86.4%
2003 Brian St. Pierre 33 15 45.45 185 5.61 2 3 45.6 0 -150 -88.8%
2016 Kevin Hogan 26 14 53.85 104 4.00 0 2 31.6 1 -80 -108.9%
Year Quarterback Att. Cmp Pct. Yards YPA TD INT PR CarAV TOT DYAR DVOA
1995 Jay Barker - - - - - - - - 0 - -
1999 Kevin Daft - - - - - - - - 0 - -
2002 Brandon Doman - - - - - - - - 0 - -
2005 Adrian McPherson - - - - - - - - 0 - -
2006 Ingle Martin - - - - - - - - 0 - -
2006 Omar Jacobs - - - - - - - - 0 - -
2007 Jeff Rowe - - - - - - - - 0 - -
2008 Erik Ainge - - - - - - - - 0 - -
2008 John David Booty - - - - - - - - 0 - -
2009 Nate Davis - - - - - - - - 0 - -
2009 Rhett Bomar - - - - - - - - 0 - -
2010 Jonathan Crompton - - - - - - - - 0 - -
2011 Nathan Enderle - - - - - - - - 0 - -
2011 Ricky Stanzi - - - - - - - - 0 - -
2014 Aaron Murray - - - - - - - - 0 - -

This is a round where even the "notable" quarterbacks are remembered for the wrong reasons. Dan Orlovsky is famous for his unintentional safety. A.J. Feeley is probably best remembered for some epic games in prime time against the Patriots in 2004 and 2007. T.J. Yates is the only fifth-round quarterback to win a playoff game in the salary-cap era. Mike McMahon might be remembered for somehow being worse than Joey Harrington. Tee Martin is still used in irrational arguments against Peyton Manning for his time at the University of Tennessee.

Kevin Hogan saw the field some for Cleveland last year, and may have had the most improbable 100-yard rushing game in NFL history when he rushed for 104 yards off the bench against the Bengals in his debut game. That will likely be his career highlight. Brett Hundley is also a current backup project for Green Bay, but he hasn't taken his garbage-time opportunities as seriously as Craig Nall once did.

The last hope here is really Cincinnati's AJ McCarron. He started three games in 2015 when Andy Dalton was injured, and also nearly led a 15-point fourth-quarter comeback win over the Steelers in the wild-card round. When backups are floated around as trade bait, McCarron's name will be high on that list, or he could just take over in Cincinnati should something drastic happen with Dalton's career path. In all likelihood, McCarron, a typical game manager who never threw the ball more than 35 times in a game for loaded Alabama, is not going to be some franchise's long-term answer at quarterback. However, he at least still has a chance at relevancy, which puts him ahead of the pack in this round.

Sixth Round

Among the nine quarterbacks drafted in the sixth round since 2014, only Zach Mettenberger and David Fales have appeared in a game so far. I came down pretty hard on Mettenberger after his rookie season, and the Titans rightly drafted Marcus Mariota with the No. 2 pick in 2015. Still, Mettenberger remains one of the few notable quarterbacks in this list, and was last seen as a backup in Pittsburgh, which is exactly what Bruce Gradkowski has been lately, and what Jim Miller was in the 1990s.

Tiers In My Eyes: 48 Sixth-Round Quarterbacks
The Good The Member Berries The 2016 Gang The Forgotten
Tom Brady Derek Anderson Brandon Allen Andre Woodson Jordan Palmer
Marc Bulger Jim Miller Jeff Driskel Andy Hall Josh Booty
Matt Hasselbeck Bruce Gradkowski Jake Rudock Brooks Bollinger Josh Harris
Tyrod Taylor Zach Mettenberger Nate Sudfeld Chuck Clements Josh Heupel
Colt Brennan Keith Null
Craig Whelihan Keith Wenning
Curtis Painter Kliff Kingsbury
Dan LeFevour Mike Cawley
David Fales Mike Cherry
Drew Henson Mike Teel
Garrett Gilbert Rusty Smith
J.T. O'Sullivan Ryan Lindley
JaJuan Seider Spence Fischer
Jeff Smoker Spergon Wynn
Jerry Colquitt Tajh Boyd
Jim Sorgi Todd Husak
Joe Webb Tom Brandstater
John Dutton Tony Pike

While Brady continues to widen the gap between his success and the rest of his round's production, we have seen one other competent quarterback step up here. Tyrod Taylor was Joe Flacco's backup in Baltimore, but he was largely relegated to preseason action since Flacco never missed any time in his first seven seasons. Taylor went to the Bills in 2015 and has thrown 37 touchdowns to 12 interceptions so far. He's a .500 starter with only one 300-yard passing game in 28 starts, but he has been good at protecting the ball and running a run-heavy offense. He also hasn't had much luck with the health of his skill players or adequate defensive help over the last two years. Taylor is not quite up to the level of a Marc Bulger or Matt Hasselbeck when it comes to being a consistent passer, but for what we're used to from the sixth round, he looks like a gem. Some fans in Baltimore probably wish the team would have just paid him cheaply these past two years given what the Ravens are getting from the expensive Flacco. Taylor is a rare case of a quarterback riding the bench for four years before finally showing he can handle a starting job.

Seventh Round

Only two quarterbacks have gone in the seventh round of the last three drafts, but one has at least made a bit of a name for himself.

Tiers In My Eyes: 48 Seventh-Round Quarterbacks
The Notable The Who?
Gus Frerotte Alex Brink James Kilian Moses Moreno
Matt Cassel B.J. Coleman Jarious Jackson Ronnie McAda
Matt Flynn B.J. Symons Jay Walker Scott Covington
Ryan Fitzpatrick Brad Sorensen Jeff Kelly Sean Canfield
Tim Rattay Bradlee Van Pelt Joe Hamilton Sean Renfree
Trevor Siemian Brandon Doughty John Navarre Seth Burford
Casey Bramlet John Walsh Steve Matthews
Chandler Harnish Jon Stark Tony Corbin
Chris Greisen Ken Dorsey Tony Graziani
Cody Pickett Koy Detmer Tyler Thigpen
D.J. Shockley Kyle Wachholtz Wally Richardson
Gibran Hamdan Levi Brown Wes Pate
Glenn Foley Matt Mauck Zac Dysert
Greg McElroy Michael Bishop Zac Robinson

Who ultimately replaced Peyton Manning in Denver? It wasn't Brock Osweiler like what was once planned, nor was it the insufferable Mark Sanchez. Instead the job went to Trevor Siemian, an underdog story out of Northwestern. Siemian started 14 games last season, and while the team drafted Paxton Lynch in the first round and has flirted with Tony Romo this offseason, Siemian was at least adequate as far as seventh-round picks go. He is likely not the long-term solution in Denver, but if Josh McCown and Ryan Fitzpatrick can find more than a decade of work in this league, then Siemian should be fine too. You can already make the argument that Siemian is a top-five quarterback in this group of seventh-round quarterbacks, but it's not like we have a "good" category here.

Of course, there's one other place to get a quarterback: free agents left over after the draft. In a second article later this week, we'll look at the search for the next Kurt Warner or Tony Romo.

Posted by: Scott Kacsmar on 21 Mar 2017

37 comments, Last at 30 Mar 2017, 6:38am by jefeweiss

Comments

1
by JoeyHarringtonsPiano :: Tue, 03/21/2017 - 4:32pm

Fascinating stuff. I have just one minor quibble. I would argue that putting Drew Stanton in the "ugly" category might be a bit harsh. True, he was a disaster (even by Millen-era Lions standards) every time he took the field early on in his career. But from 2010 onwards, I think he's proven himself a serviceable backup.

2
by Eleutheria :: Tue, 03/21/2017 - 4:58pm

I love that Alex is in a category all by himself. HAHA!

3
by Theo :: Wed, 03/22/2017 - 9:53am

And the Undrafted?

4
by Raiderjoe :: Wed, 03/22/2017 - 10:54am

is poster drunk? article is abotuy drafted quaretrbacks. guy would have to do separate article about nopn-drafted ones

5
by whateverdude :: Wed, 03/22/2017 - 11:23am

Clearly someone didn't read until the end. Busted!

6
by Theo :: Wed, 03/22/2017 - 11:28am

Whoops. Thanks and my appologies.

7
by Raiderjoe :: Wed, 03/22/2017 - 12:59pm

NO prob'

8
by justanothersteve :: Wed, 03/22/2017 - 1:14pm

I think the lesson here is never, ever draft a QB in the fifth round.

9
by MilkmanDanimal :: Wed, 03/22/2017 - 1:45pm

I feel like Josh Freeman needs a "One Day Someone Will Do A Documentary On His Collapse And It Will Be Fascinating But Until Then WTF Happened" category.

Josh McCown as "the Cockroach" could not be more perfect.

10
by Mike B. In Va :: Wed, 03/22/2017 - 1:48pm

I still think it's sad that John David Booty washed out. That was a name made to be called on TV..

11
by BJR :: Wed, 03/22/2017 - 1:50pm

Same goes for Jeff Smoker

15
by Raiderjoe :: Wed, 03/22/2017 - 3:27pm

there is a josh smoker curuenlty pitching for n.y. mets

12
by MilkmanDanimal :: Wed, 03/22/2017 - 2:05pm

Real shame he didn't last long enough to throw passes to Jake Butt.

16
by Raiderjoe :: Wed, 03/22/2017 - 3:28pm

there was john booty who palyed for jets and eags

22
by PatsFan :: Thu, 03/23/2017 - 1:15pm

Dunno that we want Lectroids playing in the NFL.

23
by Vincent Verhei :: Thu, 03/23/2017 - 3:21pm

* Golf clap. *

35
by Mike B. In Va :: Mon, 03/27/2017 - 10:03am

Indeed.

25
by Raiderjoe :: Thu, 03/23/2017 - 5:16pm

Asusmign Lectroid js type of Transforms.r did not see Transformers movies so do not understand point you are trying to make.

32
by justanothersteve :: Fri, 03/24/2017 - 3:03pm

Lectroids are an alien species who live among us in Buckaroo Banzai, a fun cult SF movie from the early 80's.

13
by lofkri :: Wed, 03/22/2017 - 2:27pm

It would be interesting to see how this kind of a study would change if you look at the draft pick with the current setup of 32 teams. Drew Brees was a 2nd round pick but he was the 32nd player drafted in 2001, the last year of 31 picks in each non-compensatory pick round. That would definitely make the difference between the 1st and 2nd round greater.

That being said, the distribution of compensatory picks is not always the same year to year by rounds as far as I know. However, now that they can be traded, that will make them more "fluid" and more like traditional draft picks.

Since the number of compensatory picks is always 32, in theory you end up having 8 rounds in the draft. The study could be done using 8 rounds and where the players got drafted within that setup. It surely won't change the main takeaway of the study but would still be interesting.

29
by SandyRiver :: Fri, 03/24/2017 - 12:43pm

Might make sense to use a set number of picks for the tiers, 1 thru XY, then XY+1 thru 2*XY, etc. Comp picks (and/or # of teams) may not remain as they are now. Except for the day 1,2,3 thing, there's nothing magic about draft round; the pick number is more relevant.

14
by Mr Shush :: Wed, 03/22/2017 - 2:58pm

I'm not sure there enough categories in some of these groups. The gap between Peyton or Rodgers and Flacco in terms of how valuable it is to draft one is probably greater than between Flacco and Ryan Leaf.

Also, I'm not sure I buy the idea that Vick was a better quarterback or draft pick than Cutler. That probably means moving Vick down a tier, not Cutler up one.

19
by RickD :: Wed, 03/22/2017 - 11:36pm

Peak Vick was definitely a better QB than Cutler. He was nearly the league MVP in 2010. And in a league where it seems like every half-competent makes the Pro Bowl, Cutler has only done that once.

Certainly Vick had huge accuracy issues. As a passer over the course of a career, Cutler is slightly better and certainly more accurate, simply because of Vick's inconsistency. But then we have to account for Vick rushing for 6109 yards over his career.

As a GM I would prefer to have Vick. Certainly the more dangerous of the two, even though his completion percentage was crap.

17
by Led :: Wed, 03/22/2017 - 8:12pm

I'm still sad that Pennington could not stay healthy. When his shoulder was intact, he was a pleasure to watch. And the future...sigh. Tomorrow, and tomorrow, and tomorrow. Creeps in Bryce Petty pace from day to day until the last syllable of recorded time. And all our Hackenbergs have lighted Bowles the way to dusty death....

21
by James-London :: Thu, 03/23/2017 - 6:25am

+ 1 on Pennington. The only other QB to win the AFCE during Brady's time as a starter. 2002 with the Jets and 2008 with Miami, albeit with Brady on IR. That year with Pennington, the Brown/Williams wildcat and Parcells actually running the Dolphins was a (short) period of real hope.
Then Parcells started mailing it in, Pennington's shoulder imploded and Jeff Ireland drafted Pat freakin' White...

Phil Simms is a Cretin.

24
by MilkmanDanimal :: Thu, 03/23/2017 - 3:36pm

"How would you characterize the Jets' draft strategy for the last 20 years?"

"'Tis a tale told by an idiot, full of sound and fury, signifying nothing."

33
by mehllageman56 :: Fri, 03/24/2017 - 8:47pm

They drafted Darelle Revis, Nick Mangold and Leonard Williams within the last ten years. They're not the worst drafting team in their division, much less the NFL. As far as quarterbacks go, they've missed on the last three(possibly 4 if you include Hackenberg), but spent only one first round pick on any of them. They also drafted two franchise level qbs; unfortunately, they broke both of them, Ken O'Brien and Chad Pennington.

18
by Theo :: Wed, 03/22/2017 - 9:44pm

Why is "Wee" Brooks Bollinger in the 'The Forgotten' column??
Did you forget the Jason Beattie comics about him???
http://www.jasonabeattie.com/fo092805.html

20
by James-London :: Thu, 03/23/2017 - 6:19am

"Drowned Dolphins". This makes me sad, and goes a long way to explaining why Miami have broadly sucked for 15 years

Phil Simms is a Cretin.

26
by rj1 :: Fri, 03/24/2017 - 8:29am

Nice study and all, but comparing the performance of QBs to other positions, is it that out of the norm? How do 6th-round runningbacks do compared to 1st-rounders?

27
by Theo :: Fri, 03/24/2017 - 10:08am

Wow.
Josh McCown has been in the league for 14 seasons and has been on 7 teams.
He has thrown passes in 78 games, worth of almost 5 seasons or 1/3 of his career.

28
by Hoodie_Sleeves :: Fri, 03/24/2017 - 10:58am

I can't help but think that there are so many confounding factors with this sort of thing that it's almost a waste of time to do.

On the subject of "You can get a quarterback late in the draft," the real question isn't "How many late round QBs pan out?" but "How many late round QBs could pan out?"

Beyond the obvious fact that late round QBs are generally inferior prospects as compared to the higher round guys (because of poorer skills, size, experience, whatever), they're also given nowhere near the opportunity, not just to start, but to practice.

Your average 1st round QB pick, no matter how terrible he is, is getting the lion's share of practice snaps. Because of this, it may be that the average first round pick has a 28% chance of being a viable starter when he is drafted,and still has a roughly 28% chance a year later. A 5th round pick may walk in with a 15% chance, and a year later have a 5% chance, and a year after that have almost no chance, largely because he's had about 10% of the practice snaps for most of the year. Film session, and drills are useful, but they can't teach you everything.

There's also the problem that higher picks are essentially playing with a bigger bank - they're able to ride out downswings, whereas late picks get cut if they have a bad camp, or get hurt, or are a little lippy. The whole thing is a bit of an apples and oranges discussion.

30
by Steve in WI :: Fri, 03/24/2017 - 1:58pm

That is an excellent point. I hate to bring up Tom Brady again since he is the outlier of all outliers, but there is probably a plausible version of history where he never starts an NFL game. There are almost certainly other late-round picks who might have had the potential to become serviceable starters but never got the opportunity.

34
by Jerry :: Sun, 03/26/2017 - 4:44am

All we can do is evaluate what actually happened. And, since this particular study looks at a couple of decades of data, the aggregate results probably account for a lot of wouldas and couldas and shouldas in getting to the conclusions.

36
by ChrisS :: Mon, 03/27/2017 - 1:11pm

A big contributor to that effect is that the front office and perhaps HC have an investment in seeing their early draft picks succeed as that success then redounds back to them.

37
by jefeweiss :: Thu, 03/30/2017 - 6:38am

Every time I see this article I lobby for an analysis of quarterbacks by games started or even by snaps. So every QB who started a game would have their first game compared to every other QBs first game and so on. This would be an attempt to control for opportunity, although higher round QBs would still probably have more practice snaps early in their careers.

I think there would be a lot smaller difference between QBs by round in that analysis than one might imagine. Just comparing the tables in this article about games started by round and the success of QBs by round, the second round looks less successful than the third round. Also AV tends to overvalue the contribution of below average players, if they start. This means that a bad first round QB who gets more starts has a higher AV than a lower round QB with fewer starts, which would skew the analysis towards first round QBs.

If NFL teams were very good at evaluating QB talent, the number of games started would be a smooth curve with the 1st rounders having the most, followed by the 2nd round and on down to the 7th. When you are trying to figure out if NFL teams are good evaluators of QB talent, using a stat that assumes that they are good evaluators of talent is problematic. The best stats to use would probably be something that is independent of the number of trials, which is one reason that I would suggest comparing 1st games started with 1st games started, or even 1st snaps of a QB with 1st snaps and so on.

31
by Steve in WI :: Fri, 03/24/2017 - 2:29pm

This doesn't take anything away from the analysis regarding the history of QBs, but I am a little weary of hearing fans justify this as "see, that's why we have to draft a QB in the first round, because most of the ones afterward don't work out."

That view ignores the opportunity cost of a given pick. The fact that, say, a 4th round QB is far more likely to fail than a 1st round QB is mitigating by the fact that a 4th round pick is way less valuable. I would much rather see my team draft several mid-round QBs before finding one that pans out (even on a middling level) than spend 1 first round pick on a total bust. (This is why I'm firmly on the don't-spend-the-#3-pick-on-any-of-the-QBs train this year for the Bears).