Tuesday, May 21, 2013
Ranking Quarterbacks By Records
I've written this before: judging individual players by the results of their teams is stupid. If the Packers replaced everyone except Aaron Rodgers with the roster from say, Georgia Tech, they'd lose every game. If Justin Verlander was backed up by your old high school team, he'd never win. Take the best goalie in the world, but if his teammates suck badly enough, they're never going to score, and eventually he's going to let something by. That's half the appeal of team sports: you can't win by yourself.
So judging individual players by the results of their teams is stupid. But that hasn't stopped me before, and it's not going to now. Other than maybe MLB pitchers, no player in sports is more often evaluated by the results of the team than an NFL quarterback. Below, I've looked at active QBs to see how their teams did when they played, and when they didn't. The qualifications for inclusion:
* Started at least 20 games for a single team (not necessarily the current team)
* Missed at least 10 games with that same team
* Was on an NFL roster in 2012
Some players have started 20+ games for multiple teams, so they're eligible twice. This study does include players missing most or all of a season, like Tom Brady in 2008 and Peyton Manning in 2011, as well as the reverse: backups who went a whole year without starting. The most notable example of the latter is Aaron Rodgers.
For cases like Brady and Rodgers, though, I only went two years out. Do we really want to compare the Colts' record in 1999 (with Peyton Manning) to 2011 (without him)? Instead, I looked at the Colts from 2009-11. For players who have repeatedly missed games, like Ben Roethlisberger or Tony Romo, we're looking at their whole careers. Twenty-six players qualified for this study, but duplicates mean we're looking at 31 sets of data.
Of course, most players listed here did better than their replacements. Backups are usually backups for a reason — and when they're not, they often stop being backups. Players are listed in ascending order of success.
- .177: Jay Cutler, Denver Broncos, 2006-08
Broncos with Cutler: 17-20 (.459)
Broncos without Cutler: 7-4 (.636)
We start with one that might be a little unfair. The 2005 Broncos were very good. They went 13-3, made it to the AFC Championship Game. In '06, the defense was sensational and the offense a disaster. The team started 7-4, but Jake Plummer was so ineffective that the coaches replaced him with rookie Cutler, who finished the season 2-3. Denver was essentially a .500 team the next couple of years, and that's probably more about the decline of the rest of the team than it is the switch from Plummer to Cutler.
He hasn't missed enough games with Chicago to qualify for this exercise, but since he joined them, the Bears are 34-22 with Cutler and 2-6 without him. More disappointing for Cutler is that by this measure he rates basically even with Kyle Orton. When Denver traded Cutler to Chicago, for Orton and two 1st-round draft picks, the Broncos stayed 8-8 and the Bears dropped from 9-7 to 7-9.
- .146: Kyle Orton, Chicago Bears, 2005-08, Denver Broncos, 2009-11
Bears with Orton: 21-12 (.636)
Bears without Orton: 19-12 (.613)
Broncos with Orton: 12-21 (.364)
Broncos without Orton: 8-7 (.533)
With the Bears, he was a little better than other quarterbacks. With the Broncos, he couldn't hold a candle to Tim Tebow.
I struggled with how to weigh players who qualify for the study with more than one team. I decided on straight addition, so Orton's rating is computed by adding his score with Chicago (+ .023) to his score with Denver (- .169). I realize that's not perfect, but you can't compare a .600 Bears team to the .400 Broncos, and he missed a lot more games with the Bears. Also, just so we're clear, I know that for most the games Orton "missed" he was actually a deliberate backup.
- .115: Jason Campbell, Washington Redskins, 2006-09
Washington with Campbell: 20-32 (.385)
Washington without Campbell: 6-6 (.500)
Like Cutler, Campbell suffers from the decline of his team. In his first two seasons, Washington went 3-6 with Mark Brunell, 8-12 with Campbell, and 3-0 with Todd Collins. Then Joe Gibbs retired and Jim Zorn became head coach. In '08, Campbell had his best season and the team finished 8-8. The next year, they plummeted to 4-12, and that's heavy on Campbell's record. He didn't play enough games with the Raiders to draw many conclusions, but his 11-7 record with Oakland was far better than 5-11 with Bruce Gradkowski and Carson Palmer.
- .033: Tarvaris Jackson, Minnesota Vikings, 2006-10
Vikings with Jackson: 10-10 (.500)
Vikings without Jackson: 32-28 (.533)
In 2007, Minnesota was 8-4 in Jackson's starts, compared to 0-4 without him. What mostly hurts his record is the '09 season, when Brett Favre started every game and the team went 12-4. That team had a good defense, an easy schedule, and Adrian Peterson. Jackson almost certainly would have won double-digit games, too, if he'd been the starter all season.
- .013: Matt Schaub, Houston Texans, 2007-12
Texans with Schaub: 44-36 (.550)
Texans without Schaub: 9-7 (.563)
This is surprising, isn't it? Schaub started every game in 2009, 2010, and 2012. Houston went 27-21 in those seasons, and Schaub made two Pro Bowls. In 2011, the Texans started 7-3, then Schaub got hurt and they went 3-3 with T.J. Yates. They were also better with Schaub (6-5) than without him (2-3) in '08. But in 2007, Schaub posted a 4-7 record, then he got hurt and the mighty Sage Rosenfels went 4-1 in his absence. The improvement was much more about Mario Williams coming into his own than it was about Rosenfels, but that's why Schaub rates where he does. I warned you at the beginning that this whole idea was kind of stupid.
- .010: Colt McCoy, Cleveland Browns, 2010-12
Browns with McCoy: 6-15 (.286)
Browns without McCoy: 8-19 (.296)
Ever since they returned to the NFL in 1999, it hasn't really mattered who plays quarterback for the Browns.
+ .006: Carson Palmer, Cincinnati Bengals, 2003-10
Bengals with Palmer: 46-51 (.474)
Bengals without Palmer: 14-16-1 (.468)
Palmer started every game for the Bengals from 2005-07 and 2009-10. Jon Kitna went 10-9 in Palmer's first two seasons, and Ryan Fitzpatrick managed a 4-7-1 mark in '08. Palmer himself posted a 0-4 record that season.
+ .007: Chad Henne, Miami Dolphins, 2009-11
Dolphins with Henne: 13-18 (.419)
Dolphins without Henne: 7-10 (.412)
I remember Chad Henne's second game. For one week, he was the next Dan Marino. The 2009 Dolphins started 0-3 with Chad Pennington, but rebounded to win 7 games with Henne. He sunk back to basically even with the team in 2011, going 0-4 before he was injured and replaced by Matt Moore, who went 6-6.
+ .020: Aaron Rodgers, Green Bay Packers, 2006-12
Packers with Rodgers: 52-26 (.667)
Packers without Rodgers: 22-12 (.647)
I know Rodgers was on the Packers in '05, but I don't see what we gain by comparing the Packers now to the Packers three years before Rodgers became starter. The non-Rodgers record is 21-11 Brett Favre, 1-1 Matt Flynn.
+ .042: David Garrard, Jacksonville Jaguars, 2002-10
Jaguars with Garrard: 39-37 (.513)
Jaguars without Garrard: 32-36 (.471)
You wouldn't guess it from his so-so rating in this exercise, but I believe Garrard is the most underrated QB of his generation. He came into his own around 2007, passing for 18 TDs and only 3 INTs, with a 102.2 passer rating. He went 9-3 as starter and led Jacksonville to a postseason win before they were eliminated by the 16-0 Patriots.
The next two seasons, Garrard played every game, and the Jags didn't do particularly well (12-20). What a lot of people didn't realize was that the team was falling apart, with Garrard and Maurice Jones-Drew the main pieces holding it together. In 2010, Garrard went 8-6 as starter, posted a 90.8 passer rating, and rushed for 5 TDs. In 2011, the Jaguars cut him at the end of preseason. They fell from 8-8 (8-6 with Garrard) to 5-11, then 2-14. Those numbers don't count toward the "Jaguars without Garrard" figures, but I think it's obvious at this point that the Jags cut their best QB and crippled their own offense. The table below shows Jacksonville's ranks in offensive yardage, scoring, and team passer rating.
Garrard's 2010 rating, 90.8, would have tied for 10th. In just two games, his backups dropped that to 83.4. In the team's first year without Garrard, its collective passer rating checked in at 62.2. The offensive drop-off is similar (though less dramatic) to the Colts when Peyton Manning got hurt. Such a mild positive rating for Garrard, + .042, doesn't reflect his true value.
+ .045: Byron Leftwich, Jacksonville Jaguars, 2003-06
Jaguars with Leftwich: 24-20 (.545)
Jaguars without Leftwich: 10-10 (.500)
I did not rig the numbers to get Leftwich and Garrard back-to-back. Actually, a math error caused me to initially rate Garrard at + .048, one spot ahead, and I was tempted to leave him there to reinforce my point. Oh well.
Leftwich's advantage is all from 2003, when he went 5-8 in place of Mark Brunell (0-3). Compared to David Garrard, he's basically even, 19-12 (.613) against Garrard's 10-7 (.588). That puts Leftwich up by a game or two, but Simpson's Paradox is in play. Look at their records year by year:
Each season, they're basically equal. Leftwich has a better overall winning percentage because he started more games in 2005, when the team as a whole was at its best.
+ .079: Alex Smith, San Francisco 49ers, 2005-12
49ers with Smith: 38-36-1 (.513)
49ers without Smith: 23-30 (.434)
From 2005-10, Smith was 19-31 (.380) and other Niner QBs were 18-28 (.391). When Jim Harbaugh became coach and the team suddenly got good, Smith's record (19-5-1) is basically the same as Colin Kaepernick's (5-2). This is Simpson's Paradox again: Smith's backups played a lot when the team was bad, but Smith has handled 80% of the games since they got good.
+ .105: Rex Grossman, Chicago Bears, 2003-08
Bears with Grossman: 19-12 (.613)
Bears without Grossman: 33-32 (.508)
Grossman comes out roughly equal to Kyle Orton, .640 to .636. But Grossman played little his first two years, going 3-3 while his fellow QBs compiled a record of 9-17. I actually think Orton is a better QB than Grossman or Tim Tebow, but I'm following my own rules.
+ .115: Matthew Stafford, Detroit Lions, 2009-12
Lions with Stafford: 17-28 (.378)
Lions without Stafford: 5-14 (.263)
Stafford has started every game the last two seasons. He missed time mostly when the Lions were coming out of their 0-16 hole, so both he and his backups have poor records for that time. The difference is that his backups didn't play in the surprising 2011 season, when Stafford passed for 5,038 yards and Detroit went 10-6.
+ .117: Tony Romo, Dallas Cowboys, 2006-12
Cowboys with Romo: 55-38 (.591)
Cowboys without Romo: 9-10 (.474)
Romo actually was the first quarterback who got me thinking about this project. Few players are more often criticized, so it was no surprise when Romo's offseason contract garnered negative attention. But other than in 2010 — when the Cowboys started 1-7 with Wade Phillips and (mostly) Romo, compared to 5-3 with Jason Garrett and (mostly) Jon Kitna — the team has consistently done better with Romo than with other QBs, including established players like Drew Bledsoe and Kitna.
Romo has made some high-profile mistakes, but in seven years as starter, he's never had a passer rating below 90, and he's thrown almost twice as many TDs (177) as INTs (91). Romo is not Aaron Rodgers or Tom Brady, but he's not easily replaceable. I get that a lot of people don't like him, but he's a very good player.
+ .125: Trent Edwards, Buffalo Bills, 2007-10
Bills with Edwards: 14-18 (.438)
Bills without Edwards: 10-22 (.313)
Yes, the Bills were better with Captain Checkdown than they were without him. At first, anyway. From 2007-08, Buffalo went 12-11 when Edwards started, compared to 2-7 without him. In the 2009-10 seasons, Edwards was 2-7 (.222) and the other starters 8-15 (.348). His confidence and decision-making eventually became a serious problem, but many Buffalo fans are too hard on the QB who brought them out of the J.P. Losman era.
This study only includes active players, and Edwards is still in the NFL. Last season, he threw 2 passes for the Philadelphia Eagles, both completions.
+ .134: Ben Roethlisberger, Pittsburgh Steelers, 2004-12
Steelers with Ben: 87-39 (.690)
Steelers without Ben: 10-8 (.556)
In a nine-year career, Roethlisberger has started every game only once, but he's never missed more than four starts in any season. The Steelers are good enough that they've done well even without him, and the team has done well in holding on to capable backups like Byron Leftwich and Charlie Batch.
+ .135: Charlie Batch, Detroit Lions, 1998-2001
Lions with Batch: 19-27 (.413)
Lions without Batch: 5-13 (.278)
The Lions were a decent team in the '90s, with good players like Barry Sanders and Chris Spielman. In 1998, though, they began 0-4 and were so desperate they turned to a rookie from Eastern Michigan. Batch had one of the finest rookie seasons ever and the team won five of its last 12 games. Batch played well again the following season and Detroit made the playoffs. After a decent year in 2000, Matt Millen came on as GM and the Lions became the Lions. Batch went 0-9 in 2001 and spent the rest of his career as a reliable backup for the Steelers.
+ .166: Ryan Fitzpatrick, Buffalo Bills, 2009-12
Bills with Fitzpatrick: 19-31 (.380)
Bills without Fitzpatrick: 3-11 (.214)
We spoke about Trent Edwards like three entries ago. Edwards eventually became so risk-averse that Bills fan dubbed him Captain Checkdown for his refusal to throw downfield. In his last two seasons, Edwards was 2-7, Fitzpatrick 8-13, and Brian Brohm 0-2.
+ .169: Matt Cassel, Kansas City Chiefs, 2009-12
Chiefs with Cassel: 19-28 (.404)
Chiefs without Cassel: 4-13 (.235)
This doesn't include his year with the Patriots, because he only made 15 starts. Last season, both Cassel and Brady Quinn went 1-7 as starter. But in 2010, Cassel played well and the Chiefs made a surprise playoff run, going 10-5 with Cassel and 10-6 overall.
+ .204: Matt Hasselbeck, Seattle Seahawks, 2001-10, Tennessee Titans, 2011-12
Seahawks with Hasselbeck: 69-62 (.527)
Seahawks without Hasselbeck: 14-15 (.483)
Titans with Hasselbeck: 11-10 (.524)
Titans without Hasselbeck: 4-7 (.364)
The Titans have better results with Hasselbeck than with Jake Locker, and almost all of his score is derived from that, a young QB who's only started a dozen games. More surprising is that with Seattle, Hasselbeck did not noticeably outperform his backups.
From 2003-07, when the Seahawks made the playoffs every year, Hasselbeck went 47-27, and Seattle was 4-2 when he didn't play. But in his first two seasons, Hasselbeck was less effective (10-12) than other QBs (6-4), and after '07, he was basically equal (12-23) to Seneca Wallace (4-9).
+ .238: Derek Anderson, Cleveland Browns, 2006-09
Browns with Anderson: 16-18 (.471)
Browns without Anderson: 7-23 (.233)
I don't think anyone quite understands what happened to Anderson and the Browns in 2007, when they went 10-6. Over the rest of his career in Cleveland, Anderson was only marginally better (.316) than other QBs (.241).
+ .294: Michael Vick, Atlanta Falcons, 2001-06, Philadelphia Eagles, 2009-12
Falcons with Vick: 38-28-1 (.575)
Falcons without Vick: 9-20 (.310)
Eagles with Vick: 18-16 (.529)
Eagles without Vick: 15-15 (.500)
This is basically about one season: 2003. The Falcons went 2-10 while Vick recovered from injury, far below the team's standard from 2001-06. Atlanta didn't bottom out again until ... 2007, when Vick went to jail.
The "Eagles without Vick" record includes an 11-5 mark in 2009, the last year with Donovan McNabb. It's easy to forget that Philadelphia's most successful season since Vick joined the team was the year he didn't play. It still blows me away that so many people in the sports media promoted Vick as an MVP candidate in 2010, when he led the Eagles to a 10-6 record that was worse than their performance the year before.
+ .325: Jon Kitna, Seattle Seahawks, 1998-2000, Cincinnati Bengals, 2001-05, Detroit Lions, 2006-08
Seahawks with Kitna: 17-15 (.531)
Seahawks without Kitna: 6-10 (.375)
Bengals with Kitna: 18-28 (.391)
Bengals without Kitna: 17-17 (.500)
Lions with Kitna: 10-26 (.278)
Lions without Kitna: 0-12 (.000)
Jon Kitna is one of those players who makes me wonder, one of those "what if things had happened differently" guys. Every time he's gotten a chance to play, it's been with a below-average team. That's understandable, because Kitna's never played well enough that a good team would want him, and he's never quite been good enough to change a bad team into a good one.
But if he had ended up on a team that got good — quality offensive line, solid defense, things like that — I think he was good enough that he would have won a lot of games, and we'd view him a lot differently. Even in Cincinnati, he didn't play badly. The 2001-02 Bengals were amazingly bad. Kitna went 8-19 with those teams, which is awful (.296), but the other QBs went 0-5. His record is lower than the team's because he played a lot when the Bengals were bad, and by the time they got good he'd given way to Carson Palmer. Palmer started every game in the playoff year (2005), and that counts against Kitna's record. On the other hand, Kitna had the good fortune to miss most of the Lions' winless 2008, and that's why he rates so highly here.
+ .329: Tom Brady, New England Patriots, 2000-12
Patriots with Brady: 87-24 (.784)
Patriots without Brady: 15-18 (.455)
The records here only include 2000-03 and 2006-10; Brady's overall regular record with New England is 110-32 (.775). From 2000-01, Drew Bledsoe's record as starter was 5-13. In Week 1 of 2008, Brady suffered a season-ending injury, and Matt Cassel went 10-5 in his stead. Comparing Brady only to Cassel (.667), he'd rate much lower.
+ .625: Peyton Manning, Indianapolis Colts, 1998-2012
Colts with Manning: 24-8 (.750)
Colts without Manning: 2-14 (.125)
The Colts' collapse without Manning in 2011 is one of the most remarkable in the history of sports, and a powerful testament to the importance of an extraordinary player.
I only evaluated Manning's record in the two seasons before his injury. His complete record with the Colts was 141-67 (.678). That includes some rocky seasons early — Indianapolis compiled a record of 32-32 in his first four seasons — followed by nine years of consistent excellence (.757), during which Manning won a record four NFL MVP Awards.
The full list of quarterbacks:
- .177: Jay Cutler
- .146: Kyle Orton
- .115: Jason Campbell
- .033: Tarvaris Jackson
- .013: Matt Schaub
- .010: Colt McCoy
+ .006: Carson Palmer
+ .007: Chad Henne
+ .020: Aaron Rodgers
+ .042: David Garrard
+ .045: Byron Leftwich
+ .079: Alex Smith
+ .105: Rex Grossman
+ .115: Matthew Stafford
+ .117: Tony Romo
+ .125: Trent Edwards
+ .134: Ben Roethlisberger
+ .135: Charlie Batch
+ .166: Ryan Fitzpatrick
+ .169: Matt Cassel
+ .204: Matt Hasselbeck
+ .238: Derek Anderson
+ .294: Michael Vick
+ .325: Jon Kitna
+ .329: Tom Brady
+ .625: Peyton Manning
It's a strange ranking, with undistinguished players like Cassel and Kitna rating far ahead of stars like Rodgers and Romo. I warned you at the beginning that judging individuals by team results was ridiculous. These numbers are subject to dozens of important variables, most notably the quality of the other QBs. Calculating these numbers for Joe Montana would rank him about the same as Alex Smith.
The takeaway here is probably a second look at players we may have dismissed. Guys like Trent Edwards, Derek Anderson, and Jon Kitna all rate higher than we'd probably expect. Maybe that's just the nature of this silly project, but maybe some of these players have gotten more criticism than they deserve.