THE BEST AND WORST QBs (AT SPECIFIC TASKS)
This from FootballOutsiders.com:
As young, athletic quarterbacks have started to replace the previous generation of pocket passers, it has become clear that traditional passing statistics cannot on their own explain which players do the most and least for their teams. But that realization doesn’t make quarterback evaluation any easier. Quarterback success depends in part on the success of the players on offense around them, the success of their team’s defense and special teams, and the success of the strategies their coaches’ implement. It is difficult to disentangle.
ESPN Stats & Information makes that effort a little bit easier with their expected points added metric. It works by comparing the points a team is expected to score at the start and at the end of their plays based on factors like down and distance, field position and time remaining. It does not split the credit between a quarterback and his receiver for a long touchdown or split the blame between a quarterback and his linemen for a blindside sack. But it does frame every play on the same scale, making it possible to compare a team’s success on plays that involve its quarterback to plays without him on offense — typically runs but also plays with a backup under center — and on defense and special teams. It isn’t perfect, but that approach can help identify real trends that traditional statistics overlook.
By comparing the total EPA of each quarterback’s pass and run attempts to that of his team’s running game, defense and special teams, we have identified the top 10 and bottom 10 quarterbacks in comparative value. Essentially, this is how much value quarterbacks brought in comparison to the rest of their teammates. It is not a raw measurement of “most valuable,” because that measurement would end up with Lamar Jackson No. 1 in 2019. But Jackson shared the spotlight with a good running game, a good defense and good special teams. Someone like Ryan Fitzpatrick emphatically did not.
For our rankings, we included passers who made only 10 or more starts in 2019, which excluded four teams in the Broncos, Lions, Steelers, and Washington Football Team. We’ll start with the players who brought the most comparative value compared to their teammates.
Most comparative value
1. Derek Carr, Las Vegas Raiders
QB EPA: +70.7 (eighth)
Team EPA: -177.1 (27th)
QB comparative value: +247.9
Since Jon Gruden returned to the Raiders, Carr has heard rumors that he would be replaced as the team’s starting quarterback. Apparently, he’s tired of it. Even before his comments, Carr made that point clear with a tremendously efficient 2019 season. He finished top 10 among regular starters with both his 62.2 QBR and 18.2% passing DVOA (DVOA measures efficiency adjusted for situation and opponent, explained further here.). His 6.7-yard average depth of target (aDOT) was bottom five and does back up the scouts who lament that Carr is sometimes unwilling to push the ball to open receivers downfield. But while conservative, Carr’s approach is one that can lead to tremendous success — just look at the Saints, and Drew Brees’ even lower 6.6-yard aDOT. If the Raiders want to make changes to improve their playoff chances in future seasons, they should aim to fix their bottom-10 defense (14.8% DVOA, 31st) and special teams (minus-3.2%, 25th) before they worry about their quarterback.
2. Ryan Fitzpatrick, Miami Dolphins
QB EPA: +25.2 (19th)
Team EPA: -213.4 (28th)
QB comparative value: +238.5
3. Deshaun Watson, Houston Texans
QB EPA: +103.8 (fifth)
Team EPA: -98.3 (23rd)
QB comparative value: +202.1
4. Dak Prescott, Dallas Cowboys
QB EPA: +138.9 (third)
Team EPA: -26.2 (12th)
QB comparative value: +165.1
5. Patrick Mahomes, Kansas City Chiefs
QB EPA: +140.1 (second)
Team EPA: +2.6 (seventh)
QB comparative value: +137.6
Mahomes is unquestionably the best quarterback in football. Despite a lower volume because of the two games he missed because of a knee injury, Mahomes again finished top three at the position with his 30.0% passing DVOA and 76.3 QBR. But the Chiefs’ success with Mahomes sidelined showed just how talented their roster is.
6. Lamar Jackson, Baltimore Ravens
QB EPA: +190.6 (first)
Team EPA: +57.4 (third)
QB comparative value: +133.2
Jackson seems more essential to the Ravens than even Mahomes is to the Chiefs. But if GM Eric DeCosta and head coach John Harbaugh and can tailor their roster and strategies to fit their MVP quarterback, then I suspect they could do the same for another player under center. Meanwhile, many of the team’s players who are perfect fits for Jackson’s skill set are also simply great players who would be assets to any team.
T-7. Andy Dalton, Cincinnati Bengals
QB EPA: -5.6 (25th)
Team EPA: -135.4 (26th)
QB comparative value: +129.8
On the opposite end of the spectrum from the talented Mahomes and Jackson on their talent-rich teams, Dalton did not play his best in 2019. But the Bengals’ 14-loss season was also clearly not his fault.
T-7. Matt Ryan, Atlanta Falcons
QB EPA: +56.8 (13th)
Team EPA: -72.9 (19th)
QB comparative value: +129.8
However you want to measure quarterback effectiveness, Ryan has been unable to duplicate the success of his 2016 MVP season. But the steady decline of his efficiency from 39.1% DVOA and 79.6 QBR that season to 7.0% and 57.6 last season and the team’s decline from 11 to seven wins says more about the talent around Ryan than Ryan himself.
9. Russell Wilson, Seattle Seahawks
QB EPA: +62.6 (11th)
Team EPA: -57.3 (17th)
QB comparative value: +119.9
10. Philip Rivers, Los Angeles Chargers
QB EPA: +53.5 (14th)
Team EPA: -61.1 (18th)
QB comparative value: +114.6
Least comparative value
1. Tom Brady, New England Patriots
QB EPA: +28.2 (18th)
Team EPA: +166.8 (first)
QB comparative value: -138.6
The quality of Brady’s play has slipped in recent seasons. He might be able to blame his decade-low 60.8% completion percentage and 2.4% DVOA on a lack of receiving talent in New England last season. But his yearly decline from a 73.9% accuracy percentage in 2015 to 72.7% in 2016, 72.1% in 2017, 70.4% in 2018, and 66.9% in 2019 probably captures the skill decline one would expect of a quarterback who has entered his 40s. That said, Brady was far from a bad passer last season. His placement at the bottom of this list reflects the Patriots’ otherworldly defensive talent (minus-25.5% DVOA, first) that would make almost any quarterback seem worse by comparison. That talent distribution is what makes the Patriots’ transition to a new quarterback starter so fascinating for 2020. Unfortunately, the team’s free-agency losses of defensive starters Kyle Van Noy and Jamie Collins were exacerbated by the opt-out losses of Dont’a Hightower and Patrick Chung. They still have their elite secondary, but the 2020 Patriots’ defense could be dramatically worse than the 2019 edition.
2. Jimmy Garoppolo, San Francisco 49ers
QB EPA: +70.5 (ninth)
Team EPA: +98.3 (second)
QB comparative value: -27.8
GM John Lynch and head coach Kyle Shanahan have engineered a team in San Francisco that probably could succeed with any competent quarterback.
3. Josh Allen, Buffalo Bills
QB EPA: +19.2 (20th)
Team EPA: +36.2 (fourth)
QB comparative value: -17.0
Allen made major strides in his sophomore season, but even his marked improvement from a minus-35.9% to a minus-11.8% passing DVOA couldn’t pull him out of the bottom 10 quarterbacks in passing efficiency.
4. Mitchell Trubisky, Chicago Bears
QB EPA: -14.2 (27th)
Team EPA: -3.6 (10th)
QB comparative value: -10.8
5. Carson Wentz, Philadelphia Eagles
QB EPA: +38.8 (16th)
Team EPA: -8.4 (11th)
QB comparative value: +47.2
6. Baker Mayfield, Cleveland Browns
QB EPA: -1.2 (23rd)
Team EPA: -56.5 (16th)
QB comparative value: +55.4
7. Kirk Cousins, Minnesota Vikings
QB EPA: +81.1 (sixth)
Team EPA: +22.7 (fifth)
QB comparative value: +58.4
8. Aaron Rodgers, Green Bay Packers
QB EPA: +63.0 (10th)
Team EPA: -0.6 (eighth)
QB comparative value: +63.6
9. Sam Darnold, New York Jets
QB EPA: -5.2 (24th)
Team EPA: -78.3 (20th)
QB comparative value: +73.1
10. Ryan Tannehill, Tennessee Titans
QB EPA: +73.8 (seventh)
Team EPA: -2.9 (ninth)
QB comparative value: +76.6
Most accurate
In 2019, Drew Brees became the second quarterback in NFL history to complete 74% of his passes in a single season, joining only … Drew Brees from 2018. In fact, Brees has five of the six highest completion percentages on record, as he has turned the short, accurate passing game into an art form in its own right.
He’s not alone, however; NFL teams are completing passes at rates higher than ever before. When Brees entered the league in 2001, teams completed just 59% of their passes. Last season, teams completed 64% of their passes, as the short, lower-risk passing game becomes more and more popular. This leaves us with a couple of problems when using raw completion percentage to compare players. An astonishing number a generation ago can seem downright pedestrian by modern standards; passers are simply expected to complete more passes nowadays than they were a generation ago. Degree of difficulty also starts muddying the waters. Was Brees’ 74% completion rate, on a steady diet of short, high-percentage plays, really better than Ryan Tannehill’s 70% completion rate on passes thrown more than three yards further downfield on average?
To help answer these questions, Football Outsiders has been using game-charting data since 2006 to create a stat called passing plus-minus, which estimates how many passes a quarterback completed compared to what an average quarterback would have completed based on each throw’s distance, the yards needed for a first down, and to which side of the field the pass was thrown.
This is turned into a rate stat: completion percentage over expected, or CPOE, that shows just how much more accurate a passer is than league average, adjusted for the situations the passer found himself in. Note that our CPOE might be different from other models found elsewhere on the internet because we’re removing passes thrown away on purpose, batted down at the line or thrown when the quarterback was hit in motion.
Here are the 10 most accurate QBs from the 2019 season by CPOE:
1. Drew Brees, New Orleans Saints
CPOE: +8.5%
Take out throwaways and tipped passes, and Brees completed 79.8% of his passes in 2019; an average quarterback would have been expected to complete just 71.3%. For most quarterbacks, that would be a career-best number, but for Brees, it’s just another year on a Hall of Fame résumé. He had a plus-6.1% CPOE for the entire 2010s; no one else topped plus-3.6%. Our database goes back to 2006, and Brees is the top performer there, too, with his plus-5.6% CPOE beating Kurt Warner’s plus-5.1% to take the top spot. It’s not just racking up completions on easy, quick routes, either; last year, Brees had a CPOE of plus-8.6% on passes traveling at least 10 yards in the air, second best in the league. Even as he enters his 40s, Brees remains the most accurate passer of the 21st century.
2. Ryan Tannehill, Tennessee Titans
CPOE: +7.7%
Tannehill’s career revival in Tennessee is both a career outlier and not entirely out of the blue.
3. Kirk Cousins, Minnesota Vikings
CPOE: +6.2%
4. Derek Carr, Las Vegas Raiders
CPOE: +6.1%
5. Russell Wilson, Seattle Seahawks
CPOE: +4.9%
6. Matt Ryan, Atlanta Falcons
CPOE: +3.2%
7. Philip Rivers, Los Angeles Chargers
CPOE: +2.9%
8. Patrick Mahomes, Kansas City Chiefs
CPOE: +2.7%
Mahomes makes the impossible look easy. On plays with an expected completion percentage of 50.0% or less, Mahomes completed 51.7% of his passes, good for a plus-16.5% CPOE. That was second highest in the league, just behind Jimmy Garoppolo, but Mahomes had 58 such targets and Garoppolo just 35. No quarterback created more value out of low-percentage throws than Mahomes did in 2019. The flip side is that his CPOE on throws with an expected completion percentage above 50% was just plus-0.7%, 16th in the league, and he had a negative CPOE on passes that traveled between 0 and 9 yards through the air. If he can ever recalibrate his howitzer to hit the short stuff as well as he hits the deep stuff, he might just make it in the NFL.
9. Deshaun Watson, Houston Texans
CPOE: +2.6%
10. Lamar Jackson, Baltimore Ravens
CPOE: +2.4%
Least accurate
The least accurate quarterbacks of 2019 form a motley list, filled with rookies struggling to find their footing, backups forced into action by necessity, and players who lost their starting job for 2020. Oh, and one of the greatest quarterbacks of all time, too.
1. Dwayne Haskins, Washington
CPOE: -6.7%
We have two pieces of optimism for Haskins going forward. Yes, Haskins’ CPOE was worst in the league, but minus-6.7% is a pretty decent number for “worst in the league.” The last time the league’s worst CPOE was better than that was Brandon Weeden’s minus-5.7% in 2013. Haskins is last because no one was extraordinarily terrible last season. In addition, Haskins’ CPOE improved as the season went along, going from minus-10.9% in his first 100 attempts to minus-1.9% in his last 90. Haskins needs to improve dramatically in Year 2, but at least we saw progress toward the end of a tumultuous rookie season.
2. Gardner Minshew II, Jacksonville Jaguars
CPOE: -4.0%
Proven: Exceptional mustaches do not bring with them tremendous accuracy. Minshew’s minus-6.4% CPOE on short passes was the worst in the league, though he did manage a plus-8.4% CPOE when asked to throw deep. For the record, Nick Foles provided the opposite skill set, with a plus-6.2% CPOE short and a minus-12.1% CPOE deep, for a total of plus-2.4% CPOE. If only the Jaguars could have used them to their respective strengths. Does anyone know how Foles would look with a false mustache?
3. Tom Brady, New England Patriots
CPOE: -3.8%
One of these things is not like the others …
While Brady’s accuracy was slightly better than that of Haskins or Minshew, he threw 130 more passes than either one, meaning he ended up dead last in passing plus-minus, the counting version of this stat. Brady was simply not on the same page as his receivers; his minus-5.1% CPOE to wideouts put him ahead of Andy Dalton. He struggled with what should be his easiest throws: On passes thrown five yards downfield or less, he had a minus-4.7% CPOE…This is not a death sentence; Brett Favre had the worst plus-minus in the league in 2006 and went on to make the Pro Bowl in his next three seasons. But Brady’s accuracy and rapport with his receivers need to improve if the Buccaneers are going to meet their high expectations for 2020.
4. Mitchell Trubisky, Chicago Bears
CPOE: -3.7%
After popping to just above average in 2018, Trubisky matched his rookie mark in CPOE last season. Matt Nagy and the Bears coaches made things as easy as possible for Trubisky, giving him the easiest throws of his career; his expected completion percentage was 70.3%. Trubisky still could not deliver. His worst numbers came in the most crucial situations: a minus-4.5% CPOE on third down, a minus-8.2% CPOE in the fourth quarter, a minus-6.2% CPOE on deep passes, all at or near the worst marks in the league. In his career, Trubisky has been most accurate throwing to his tight ends, which might explain why the Bears added roughly one thousand of them in the offseason. It’s one last attempt to see Trubisky realize his potential.
5. Andy Dalton, Cincinnati Bengals
CPOE: -3.7%
6. Baker Mayfield, Cleveland Browns
CPOE: -2.8%
7. Jacoby Brissett, Indianapolis Colts
CPOE: -2.6%
T-8. Daniel Jones, New York Giants
CPOE: -2.4%
Jones ended up with the second-best CPOE among rookies, though there was a significant gap between him and Kyler Murray at plus-0.5%. Jones had a confidence that his arm couldn’t always match; he had a minus-5.6% CPOE on throws 10 or more yards downfield, with below-expected accuracy to every intended receiver with 10 or more targets. We appreciate the urge to gamble, but Jones was far better getting the ball out on quick drops than when he embraced his inner gunslinger.
T-8. Josh Allen, Buffalo Bills
CPOE: -2.4%
10. Case Keenum, Washington
CPOE: -1.5%
Most aggressive
Football Outsiders created a metric called ALEX to measure quarterbacks’ aggressiveness on third downs. (Yes, of course, Alex Smith had something to do with it.) ALEX stands for “air yards less expected,” and measures the average difference between a quarterback’s pass depth and the yards he needs to gain on third down to create a new first down. The league’s ALEX had actually been on a downward trend the past few seasons before rising to plus-1.9 in 2019, up from plus-1.1 in 2018.
The most aggressive quarterback in 2018, Patrick Mahomes, actually tumbled off the top 10 list in 2019. Some of that undoubtedly has to do with the ankle (and later knee) injuries that befell the Super Bowl MVP, but another factor is Mahomes’ average intended air yards falling from 9.2 in his MVP year to 8.6 last season. That doesn’t sound like a huge drop, but 9.2 would have put him in the top 10 in 2019; 8.6 put him in a four-way tie for 12th.
So instead, our most aggressive quarterback on third down is a man who is no stranger to the top of these leaderboards:
1. Aaron Rodgers, Green Bay Packers: +4.1 ALEX
Rodgers has finished in the top five in ALEX each of the past five seasons, and even in an extremely down 2019 season on the whole, he still managed to be ultra aggressive on third down. However, his efficiency was way, way down overall on those third-down throws, as he found just 5.6% DVOA on them. (DVOA measures efficiency adjusted for situation and opponent, explained further here.) He completed just 47 first downs or touchdowns on 140 total attempts that didn’t involve Davante Adams. The Packers looked at this, and then decided to do nothing about it, so tough luck, Aaron. Jordan Love will listen to your anger on the bench.
2. Ryan Fitzpatrick, Miami Dolphins: +4.0 ALEX
3. Russell Wilson, Seattle Seahawks: +3.7 ALEX.
4. Josh Allen, Buffalo Bills: +3.4 ALEX
5. Matthew Stafford, Detroit Lions: +3.0 ALEX
6. Ryan Tannehill, Tennessee Titans: +2.9 ALEX
7. Carson Wentz, Philadelphia Eagles: +2.7 ALEX
8. Sam Darnold, New York Jets: +2.5 ALEX
9. Matt Ryan, Atlanta Falcons: +2.5 ALEX
Traditionally not a big deep passer, Ryan actually finished with negative ALEX in his MVP season in 2016. He hadn’t had an ALEX above 2.0 since 2014 — which, not so coincidentally, was also the last time Dirk Koetter coordinated the Atlanta offense before being hired by Tampa Bay
10. Dak Prescott, Dallas Cowboys: +2.3 ALEX
Least aggressive
1. Teddy Bridgewater, New Orleans Saints: -1.4 ALEX
Bridgewater, of course, hasn’t been a qualified starter (minimum 200 passes) for a long time since his devastating preseason knee injury in 2016. But when he was a starter in 2014 and 2015, this was part of the normal game plan for him. Bridgewater carried a minus-0.7 ALEX in 2015, and a minus-0.8 ALEX in 2014, finishing ahead of only Alex Smith and Blaine Gabbert among qualified starters in both seasons. But part of last year’s number was a general predilection to throwing short on the part of the entire Saints offense; nobody had a lower average depth of target than Bridgewater last season, at 6.2. Bridgewater’s average actually increased to 6.9 intended air yards on third-down targets. The 0.9% DVOA, however, wasn’t great. The Panthers will want to make sure Joe Brady’s offense offers a lot of Christian McCaffrey underneath targets on third down in 2020.
2. Case Keenum, Washington: -1.3 ALEX
3. Jacoby Brissett, Indianapolis Colts: -0.3 ALEX
4. Joe Flacco, Denver Broncos: -0.3 ALEX
5. Mason Rudolph, Pittsburgh Steelers: -0.2 ALEX
6. Jimmy Garoppolo, San Francisco 49ers: -0.1 ALEX
7. Dwayne Haskins Jr., Washington: +0.2 ALEX
8. Daniel Jones, New York Giants: +0.7 ALEX
9. Drew Brees, New Orleans Saints: +0.8 ALEX
10. Tom Brady, New England Patriots: +1.0 ALEX
|