I have to confess, I was expecting something completely different when I sat down to do the analysis for this article. In a few weeks, Ryan Fitzpatrick will open the season for the WFT, the ninth NFL franchise he has played for in 16 seasons. FitzMagic comes to DC with a reputation as a gunslinger, but one whose game has matured as he approaches 40.
Even so, the general impression of Fitzpatrick around the NFL is that he is still a wildly uneven QB, capable of winning games and losing them from one week to the next. I was inspired to write this piece by a comment from Mike Tanier of Football Outsiders in an interview with Eboracum on Hogs Haven a few weeks ago.
Can Washington’s defense overcome a bad game by their QB? They can. Can they overcome the absolutely miserable game that Fitz sometimes has? Maybe the Steelers could do that, but you can’t necessarily count on that. When we do our DVOA and DYAR on QBs, Fitz comes out as a slightly below average starter over the last 10 years. But it’s like the old joke about the guy who’s got his hand in the oven and his hand in the ice cubes: he’s average! (laughs) On average his temperature is fine! That’s how Fitz is. So reasonably, you would say they didn’t get a rookie, they didn’t get Aaron Rodgers, but they got a competent veteran, which makes sense. I think the big difference with Fitz is most of the time, you’re expecting a competent veteran will just distribute the ball and take what the defense gives you. You don’t get that with Fitz, which is what makes it volatile and interesting for Washington. He’s the anti-Alex Smith.
Football Outsiders does some great work, and their out-of-the-box approach often reveals things that more conventional analytics approaches miss. So I thought it would be great fun to do some analysis of the variability of QB play, to see just how erratic Fitz is compared to the rest of the league. It was fun, but not for the reason I expected.
A key point about Fitz that has been drilled home by a few Hogs Haven writers who are more bullish on him than Tanier is that his play has improved in the last few years. Therefore, rather than looking over the last 10 years, which seems to have shaped Tanier’s impression, I decided to focus on his more recent play. I think the results might come as a surprise to anyone who hasn’t been a regular reader of Hogs Haven since Fitzpatrick’s signing. Which might mean they will surprise no one here, because this is a Hogs Haven article. Hmm.
Short and Sweet Methods Section
In order to examine the supposed volatility of Fitzpatrick’s play, and compare it to the rest of the league, I compiled ESPN’s total Quarterback Rating (QBR) values for each NFL team’s primary starting QB in 2020 over their last 16 complete regular season starts. Well, more or less. I made one exception by counting Dak Prescott as the Cowboy’s starter instead of injury replacement Andy Dalton for obvious reasons.
From the raw QBR values I calculated two simple measures:
Mean or Average QBR provides a measure of the overall performance over the last 16 games.
The Standard Deviation of the QBR values provides a measure of the game-to-game variability of each QB’s performance.
I discarded QBR values from partial games where the starter left with injury, such as Alex Smith’s Houston game in 2018, and games where the starter was benched after fewer than 15 pass attempts. When I had to skip games, I simply counted backward to the next start to get to 16 starts for each QB. Joe Burrow and Justin Herbert have only played 10 and 15 NFL games, respectively, so in their cases I simply calculated the mean and standard deviations of the available QBR values.
In Fitzpatrick’s case I had to go back to the November 18, 2019 Monday night game against Buffalo to get to 16 starts. In a few cases, like Alex Smith and Nick Mullins, I had to go back a few years to collect 16 starts.
How Erratic is FitzMagic?
The first figure shows the QBR values for Fitzpatrick and two other notable QBs, Patrick Mahomes and Alex Smith. Mahomes and Smith make useful comparisons for different reasons. Over his last 16 games, Mahomes has been the most consistently excellent QB in the league. Alex Smith was Fitzpatrick’s predecessor in DC, and readers should be interested to know how things might change with the new signal caller under center. He also had a reputation for steady, if unexciting, QB play –the opposite of what Tanier says we should expect from FitzMagic.
As I said, Mahomes (blue) consistently turns in performances with QBR greater than 70, which is a reasonable benchmark for good QB performance. Nine of his last 16 starts had QBR greater than or equal to 80 and one was a near perfect 98.5. Mahomes rarely turned in a sub-par performance, with only two games having QBR under 60.
In contrast to Mahomes, Smith was fairly consistently bad in his last 16 games, with 12 performances under 60 QBR. The high point in this span, game number one, was played with the stacked Kansas City Chiefs in 2017. While he did manage to turn in two decent performances by this measure in DC, he also notched up nine out of 16 games with a QBR less than 40, which is pretty terrible.
Fitzpatrick’s performance over his last 16 starts is clearly more variable from week to week and month to month than Mahomes. But a lot of that variability is pointed in a good direction. While Fitz did record three games with somewhat lower QBR than Mahomes’ worst performance, they weren’t that much worse. He also had three games with QBR over 90, on a par with Mahomes’ best four performances.
What might come as a surprise to most readers, given the common impressions of the two QBs, is that the variability of Fitzpatrick’s play over the last 16 starts, is slightly lower than that of Alex Smith. It is just expressed differently. While Fitzpatrick shows variability from week to week, Smith had more of a downward trend to his performance since moving to DC and getting injured. Both players’ performances over the same number of games spanned equivalent total ranges of QBR values. Smith just inhabited a lower range than Fitzpatrick.
What Does Really Erratic QB Performance Look Like?
Truth be told, as far as variability of QB play throughout a season goes, Ryan Fitzpatrick’s and Alex Smith’s recent performances are actually on the low side of the league average. The most variable performances in the NFL over the last 16 starts were turned in by Cam Newton and Baker Mayfield, as illustrated in the next figure.
Notice how both QBs use the whole space of the graph. Mayfield’s average performance over the last 16 starts is only slightly below Fitzpatrick’s, but he really explores the full range from three nearly perfect games to three games below 30 QBR and one below 10. Newton had three games below 10, and seven under 30, but still managed to put in five decent performances and one with QBR over 90.
I think it’s kind of unfair that Fitzpatrick gets labelled as the fire and ice quarterback when these two are playing in the same league.
Where Does Fitzpatrick Rank Among Starting QBs?
To answer the next question, I ranked the 32 starting QBs in two dimensions: Overall Performance (Average QBR) and game-to-game variability (QBR Standard Deviation). The two measures are plotted against each other in the next figure. In this plot, the top left corner represents consistent (low variability) excellent performance, and the bottom right represents highly variable poor overall performance. You want your starting QB to be at the top of the plot, and ideally as far to the left as possible.
Over his last 16 starts, Ryan Fitzpatrick showed well below average game-to-game variability and well above average overall performance. He had the 8th lowest QBR standard deviation and, coincidentally the 8th best average QBR across his most recent season’s worth of starts.
So much for the myth of Fitzpatrick being a QB who can win the game for you one week and lose one for you the next. While that may have been true earlier in his career with teams like Buffalo and the Jets, he seems to have really cleaned up his act in his most recent stint in Miami.
What About those Absolutely Miserable Games?
To get an idea of just how miserable the worst games of Fitzpatrick’s most recent starts were, I compared his worst games to those of the seven QBs ranked above him in overall performance. The lowest QBR value posted by Fitzpatrick in his last 16 starts was 33.8. Five of the seven better performing quarterbacks (Aaron Rodgers, Josh Allen, Drew Brees, Russell Wilson, Lamar Jackson) posted at least one game in their last 16 starts with a QBR lower than 33.8. Russell Wilson and Lamar Jackson each posted two worse performances.
The worst performance by any of the top eight QBs in this analysis was actually posted by the best overall performer, Aaron Rodgers. That was the 38-10 loss to Tampa on October 18, 2020 in which Rodgers went 16/35 for 160 yards, averaging 4.6 yards per completion with 0 TDs, 2 interceptions and 4 sacks. That is the kind of ugly QB performance that Tanier was talking about. The kind where Pittsburgh might still be able to pull off a win, but not the WFT, or apparently the Packers for that matter. Fitzpatrick didn’t have any games that bad as a starter in Miami.
How Often Does Fitzpatrick Have a Stinker?
To answer this question, I once again compared Fitzpatrick to the seven quarterbacks ranked above him in overall performance. In this case, though, I counted games with total QBR under 45. Fitzpatrick had three games with QBR 33.8, 38.5, and 39.6 in his last 16 starts in Miami.
All seven of the top performing QBs had at least one game with a QBR below 45, although Mahomes just squeaked in with a single game at 44.9. Aaron Rodgers and Drew Brees had one terrible game apiece with QBRs of 12.5 and 20.7, respectively. Josh Allen and Ryan Tannehill each had two games below 45 QBR. While Russell Wilson and Lamar Jackson each recorded three sub-45 games, like Fitzpatrick. Although one of Jackson’s games was the second lowest in the group with a miserable QBR of 13.6.
Recently, Fitzpatrick hasn’t had many more really bad games than some of the best QBs in the NFL. His worst games aren’t worse than some of the NFL’s best QBs.
What Can WFT Fans Expect from Ryan Fitzpatrick in DC?
Vastly improved QB play compared to what we have got used to in the last three seasons. Terry McLaurin, Curtis Samuel and Dyami Brown should get plenty of opportunities to showcase their deep ball tracking skills.
Contrary to widespread misconceptions, in recent years, Fitzpatrick’s play has been no more inconsistent from week to week than many of the best QBs in the league, including names like Drew Brees, Russell Wilson, Aaron Rodgers, Lamar Jackson, and Tom Brady.
Fitzpatrick does not play mistake-free football, and he has a bad game from time to time. However, his worst performances are not appreciably worse than some of the best quarterbacks in the NFL, and he doesn’t have really bad games any more frequently than the best quarterbacks either.
Forget what you thought you knew about Fitzpatrick and enjoy the ride.
Thanks to James Dorsett for his usual expert editorial assistance
The Football Team’s offense ranked 30th in in net yards per game in 2020, and Fitzpatrick’s Dolphins ranked 22nd. Where will the Football Team rank in 2021?
This poll is closed
Crappy for Corral!
Who will start for the WFT in Week 17?
This poll is closed
FitzMagic, unless they rest him for the playoff run
Not currently on the roster