Which Teams’ Quarterbacks Do Better Elsewhere?
Which Teams’ Quarterbacks Do Better Elsewhere?
Bryce Young isn't the first Panthers QB to struggle — but do they really succeed when they leave?
In an early stretch of the 2024 NFL season defined by quarterback struggles, the saddest story might belong to Carolina’s Bryce Young — the former No. 1 pick in the 2023 draft who was benched this week in favor of 36-year-old journeyman backup Andy Dalton.
It’s unclear what the move means for Young’s future with the Panthers, who infamously traded WR D.J. Moore and a ton of draft picks for the right to slide up and take Young one spot above future star QB C.J. Stroud. But let’s just say it’s certainly not a good development. According to ESPN: “Young, 23, becomes the first quarterback selected No. 1 in the common draft era (since 1967) to be benched for noninjury reasons in his second season.”
I’ve often argued that the process of drafting QBs is fundamentally broken in the NFL — that the brightest college talents are, by definition, thrown to the wolves with the worst supporting casts and coaching systems and left to fend for themselves, which might explain why so many of them get eaten alive by NFL defenses. Some are able to rise above it, but most aren’t. (Even Trevor Lawrence, the most hyped QB prospect since Andrew Luck — if not Peyton Manning — has had trouble consistently producing or winning ballgames with the Jaguars so far.)
That line of thinking is bolstered by the notion that certain teams simply seem cursed to see their highly drafted QBs go bust, then perform better with other teams. “Baker Mayfield and Sam Darnold Show That Bryce Young Isn't the Problem—The Panthers Are,” one headline at Barstool Sports read after the Young benching.
With all of this in mind, I set out to find the teams whose QBs play better — or worse — when they are with other teams in recent years.
My methodology looks at QB Expected Points Added (EPA) via passing and rushing per dropback1 since 2019, breaking things out by each quarterback and the teams they played for. Then, for each team, I aggregated the stats for their QBs while playing for both that franchise and all other franchises, and compared EPA per dropback rates in each situation — weighting each rate (and the differences between them) by the harmonic mean of dropbacks under each category.
When we do this, here is a plot of each team’s QBs’ EPA per dropback with — and without — that team’s uniform on:
(If you’re curious why the averages are below zero in each situation, it’s a consequence of weighting by QBs with a greater sample across multiple teams — those QBs don’t tend to be as good, or else the team would have just kept them.)
Some teams are outliers with small samples. Because of Blaine Gabbert, the Chiefs have the biggest negative gap of any team, followed by the Bills (thanks to Mitchell Trubisky and Case Keenum) and Titans (because of Marcus Mariota, Josh Dobbs and Malik Willis). And on the positive side, the Packers get a boost from Aaron Rodgers and Malik Willis being better in Green Bay, and the Ravens benefit from a similar effect with Josh Johnson and Trace McSorley.
But among teams with real samples for comparison, the Vikings are No. 1 — buoyed by Nick Mullens, Darnold and Kirk Cousins performing better in Minnesota than elsewhere. They’re followed by the Seahawks (Russell Wilson!), Bucs (Baker Mayfield and Tom Brady),2 Jaguars (Gardner Minshew II, Nick Foles, Mike Glennon), Cardinals (Dobbs, Colt McCoy, David Blough) and Texans (Deshaun Watson, ugh).
At the other end of things, we have the Dolphins (with whom Jacoby Brissett and Teddy Bridgewater were much worse), Broncos (Wilson, Bridgewater, Joe Flacco, Brandon Allen, Jeff Driskel — damn, that is a lot of QBs), Giants (Tyrod Taylor has been better with the G-Men, but Glennon and McCoy were worse), Jets (Darnold, Flacco, Trevor Siemian, Rodgers, Tim Boyle), Chargers (late-career Philip Rivers was better in Indy, as was Taylor with both N.Y. teams) and Browns (Mayfield, Watson, P.J. Walker).3
Where do the Panthers rank? Perhaps surprisingly, they’re only 15th-worst overall, and 12th-worst among teams with a sample weight of at least 500 harmonic-mean dropbacks playing for both that club and all others. Three of their five highest-weighted QBs — Mayfield, Cam Newton and Kyle Allen — have indeed performed worse in Carolina than elsewhere since 2019. But their two highest-weighted QBs — Darnold and Bridgewater — were better as Panthers, as were both Walker and Dalton.
Overall, that leaves the Panthers’ passers only slightly worse in Carolina than anywhere else they played, on average. So more than ruining otherwise good QBs, the Panthers have just had bad ones in recent years, no matter who they’ve played for.
Right now, Young is one of those bad QBs. Sooner or later, it seems like we’ll get a chance to see how he does with another team — and with it, whether he helps or hurts Carolina’s ledger in this stat.
Filed under: NFL
With a modification that counts QB rush attempts as 0.56 of a pass attempt or sack.
Remember, this is just since 2019, so Brady’s Bucs comparison is with his relatively mediocre swan song in New England.
As we can see from some of these recurring names, one team’s positive differential is driven by the same QBs that create negative differentials for other teams.
Love this idea and data, but ... "But among teams with real samples for comparison, the Vikings are No. 1 — buoyed by Nick Mullens, Darnold and KIRK COUSINS performing better in Minnesota than elsewhere." This had to be filed before the two-minute warning in Philly Monday night.
Three quick reactions. I find it difficult to believe that Cam Newton was better in New England than in Carolina...no?
Second, instead of teams, would not looking at head coaches be a better potential high level initial correlative variable than jerseys given the churn in staff over time? Playing QB for Kevin O'Connell feels a lot different than Mike Zimmer. You can then potentially segment the "types" of coaches that are more likely to be successful at transfer e.g. offense v. defense minded, McVay tree etc.
Finally, it would be interesting to examine those that did better and what exogenous factors likely contributed to that improvement. Factors that feel relevant are Pressure% (OL); Team Defense (lower opponent scores); Approximate Value receiving corp; age/experience etc.
Take Darnold:
McVay Tree offensive HC - check
Team Defense - check
WR/TE AV - check
Pressure % - check
Age - check
Darnold feels like a great gamble. That sort of deeper dive would help front offices decide whether to pursue a suboptimal QB to wager on future improvement. Thanks very much.