≡ Menu

Guest Post: Quarterback Tiers Based On Age and Talent

Today we have a guest post from James Deyerle, a longtime reader of this blog and the PFR blog, and who lives in Richmond, VA. As always, we thank our guests for contributing to the site.

If there’s one thing NFL fans can’t resist, it’s ranking quarterbacks, but while those conversations are often framed with stats, playoff success, awards, and more, it misses a big part of how fans and front offices treat quarterbacks. For example, despite similar stat lines in 2014 Vikings fans are justified in feeling very differently about Teddy Bridgewater than Bears fans feel about Jay Cutler.

One of the biggest reasons is age: Cutler was 31 and in his 9th season last year, while Bridgewater was only a 22 year old rookie, making him one of the youngest rookies in the past 15 years to see significant playing time. A collection of studies on quarterback aging from Chase, Neil Paine, and Brian Burke show that as a group, quarterbacks rapidly improve into their late 20s, peak for a few years, and then begin an accelerating decline throughout their 30s. This expected improvement lends promise to Bridgewater’s young career while the projected decline condemns Cutler’s, which informs our opinions and feelings on these quarterbacks. As such, I decided to create a system that more accurately reflects a team’s quarterback situation.

Still a work in progress, I started simple while I worked out the concepts. Based on the aging studies referenced above, I created four age buckets: Young, Prime, Aging, and Old. The theory behind each bucket is straightforward – Young quarterbacks (25 and under) are expected to significantly improve in the upcoming seasons, Prime quarterbacks (26-30) are at or near their peak, Aging quarterbacks (31-35) aren’t likely to get any better and may decline slightly, and Old quarterbacks (36+) are in danger of falling off a cliff (or even retiring!) in the very near future. This makes Old quarterbacks a high risk proposition – the upside is excellent QB play as only the best QBs last long enough to still be starting caliber into their late 30s, but the end can come suddenly and without much warning.

The next step is creating present-day skill tiers. While separating individual talent from supporting casts to determine who fits in which tier is difficult and subjective, the tiers themselves should not be. After a few variations I settled on a 5 tier system; the labels indicate the approximate quality of the QBs and each tier has a maximum number of QBs: 3 All Pros, 7 Pro Bowlers, 7 Above Average, 10 Below Average, and all others in Replacement. The intention is each tier represents a tangible benefit over the one below, and the size of the tiers match the reality that the drop off from the 2nd to the 6th best quarterback is much more significant than the marginal difference between the 18th and 25th best QB.

The last step (for now) is giving each combination of age and talent an overall Outlook. I originally had 5 different Outlook categories – Fantastic, Great, Good, Acceptable, and Poor – before adding a “Why” category for the worst possible situations as a joke. At the time, I didn’t imagine any team would bother wasting its time starting an Old, Replacement level quarterback when there are plenty of Young, Replacement quarterbacks that might improve. I was wrong!

YoungBelow AvgGood
YoungAbove AvgGreat
YoungPro BowlFantastic
YoungAll ProFantastic
PrimeBelow AvgAcceptable
PrimeAbove AvgGood
PrimePro BowlGreat
PrimeAll ProFantastic
AgingBelow AvgPoor
AgingAbove AvgAcceptable
AgingPro BowlGreat
AgingAll ProFantastic
OldBelow AvgWhy
OldAbove AvgPoor
OldPro BowlGood
OldAll ProGreat

The general philosophy is a Young Pro Bowl QB is Fantastic because of the expectation that they will be an All Pro in their Prime years, which is also Fantastic, while an Old All Pro is only Great because they aren’t likely to stay an All Pro for long. You may also notice there’s no way to have a Poor outlook with a Young quarterback under center – you only need to look at Oakland and Jacksonville to see the eternal optimism around Young QBs no matter their level of play.

I also placed a heavy emphasis on higher quality, which is why I made it impossible to have a Good Outlook if your QB is Aging, or Acceptable if your QB is Old. The aging studies indicate that if your early 30s QB isn’t a borderline Pro Bowler every year then they aren’t ever going to be, so unless your team is stacked everywhere else you won’t find much success. And teams often compound the problem by paying them exorbitant sums – I’m talking about you, Jay Cutler and Alex Smith.

So, readers of Football Perspective, what do you think of my reasoning up to this point? As I said above it’s still a work in progress so I’d appreciate any advice as I continue to refine it. One improvement I’d like to make it moving to a more continuous system as the increase from age 30 to 31 and 35 to 36 can be extremely harsh, but quantifying age is very easy. The problem is creating a simple and continuous judge for talent, although shortcuts like ANY/A or a fantasy point style system would get most of the way there.
The big kicker to distinguish this from something you’d find on a dynasty fantasy website would be to include contracts, but that’s an even larger obstacle. How do you easily quantify the Length, Average Annual Value, and Guaranteed/Dead Money of such disparate contracts as Matt Ryan’s 4 year, $20.8M AAV, and $34M DM with Andy Dalton’s 6 year, $16M AAV, $10M DM, much less against Andrew Luck’s presumably massive future contract in a straightforward way?
But before I go, here’s how my system ranks the notable quarterbacks in the NFL. The best part is if you disagree with my Talent evaluations, which I did during the preseason, I’ve provided you with the framework to make your own Outlooks. It’s a fun exercise, and can be quite difficult to decide between the last QB in a tier!

Aaron RodgersGNB32AgingAll ProFantastic
Tony RomoDAL35AgingAll ProFantastic
Andrew LuckIND26PrimeAll ProFantastic
Philip RiversSDG34AgingPro BowlGreat
Ben RoethlisbergerPIT33AgingPro BowlGreat
Matt RyanATL30PrimePro BowlGreat
Russell WilsonSEA27PrimePro BowlGreat
Teddy BridgewaterMIN23YoungAbove AvgGreat
Peyton ManningDEN39OldPro BowlGood
Tom BradyNWE38OldPro BowlGood
Drew BreesNOR36OldPro BowlGood
Matthew StaffordDET27PrimeAbove AvgGood
Ryan TannehillMIA27PrimeAbove AvgGood
Cam NewtonCAR26PrimeAbove AvgGood
Joe FlaccoBAL30PrimeAbove AvgGood
Marcus MariotaTEN22YoungBelow AvgGood
Jameis WinstonTAM21YoungBelow AvgGood
Eli ManningNYG34AgingAbove AvgAcceptable
Sam BradfordPHI28PrimeBelow AvgAcceptable
Colin KaepernickSFO28PrimeBelow AvgAcceptable
Andy DaltonCIN28PrimeBelow AvgAcceptable
Nick FolesSTL26PrimeBelow AvgAcceptable
Tyrod TaylorBUF26PrimeBelow AvgAcceptable
Blake BortlesJAX24YoungReplacementAcceptable
Derek CarrOAK24YoungReplacementAcceptable
Johnny ManzielCLE23YoungReplacementAcceptable
Robert GriffinWAS25YoungReplacementAcceptable
Geno SmithNYJ25YoungReplacementAcceptable
Carson PalmerARI36OldAbove AvgPoor
Ryan FitzpatrickNYJ33AgingBelow AvgPoor
Jay CutlerCHI32AgingBelow AvgPoor
Alex SmithKAN31AgingBelow AvgPoor
Brian HoyerHOU30PrimeReplacementPoor
Kirk CousinsWAS27PrimeReplacementPoor
Ryan MallettHOU27PrimeReplacementPoor
Josh McCownCLE36OldReplacementWhy

Footnote Caveat: aging curves are not perfect as no quarterback improves or declines exactly as expected, so this system won’t be perfect. At the start of the 2012 season Matt Schaub was entering his age 31 season after four years as one of the most efficient quarterbacks in the league, and his combination of Aging and Pro Bowl would have given the Texans a Great outlook, and he would go on to earn his 2nd Pro Bowl. Obviously that didn’t work out as Schaub was finished by mid-2013.

  • PatrickH

    Kaepernick ANY/A worst season is better than Tannehill’s best season. His worst season QBR is better than Tannehill’s best season. Yet Tannehill is “Prime/Above Average” and Kaepernick is “Prime/ Below Average”.

    • That’s a good point. I am pretty down on Tannehill, but I only say that because it seems so many people are high on him. I’d prefer to say I’m neutral but everyone just goes ga-ga over him for no season. I’d say Kaepernick has a higher ceiling than Tannehill, although I wouldn’t fight you if you said he also has a lower floor. I’m not necessarily a huge Kaepernick fan anymore, either.

      • Adam

        I don’t understand the Tannehill love, either. He reminds me of Andy Dalton – guaranteed to be at least competent, but unlikely to ever be great.

        • Richie

          I can’t argue with either of you.

          The Dolphins were kind of stuck in no man’s land with him and kind of had to give him a new contract.

    • James

      Based on today both look like they should be in the Replacement category…

      There’s no way for you to tell but I view Tanny and Kaep about the same, it just happened that they fell on the edge of the two categories. I ended up breaking the tie in favor of Tanny because he is a year younger and played QB less in college. But that decision gives the two of them very different outlooks when ideally it wouldn’t.

      It’s one of the reasons I want a more continuous system. I’m already working on it and it has Kaep and Tanny essentially equal.

  • David Morrison

    Certainly Old/Pro Bowl looks to be the worst of the ‘Good categories’ ie action plan essential in 2016 (as they haven’t done it for 2015). Analysis also needs to consider quality of back ups, particularly for teams with old/aging quarterbacks as they are more injury prone, and could as is quite rightly been mentioned, fall off the cliff at any point in time-although with current teams bad Brady/Brees/Peyton Manning still far better than Jimmy Garoppolo/Luke McCown/Brock Osweiler. Effectively Denver, New England, New Orleans and potentially Dallas need solutions to find their 2017 (and probably 2016 for Denver based on this season) starting QB, other teams in the ‘Good’ bracket almost certainly don’t.

    • James

      I agree on Old/Pro Bowl, but that’s the high risk/reward that teams are willing to take.

      However, I don’t know what to consider about the backups. We have little to no information on them and what we do know indicates they aren’t any different than your typical Replacement level player. The plan for teams seems to be ride the QB as long as possible and then flail around looking for a QB like everyone else.

  • WR

    Most of the ratings here are fine, although I think there’s an argument for putting more players in the All-Pro category. But what has Andrew Luck done to deserve that rating? Since he came into the league, Luck is 13th in the NFL in any/a, and 18th in passer rating. His totals are a little better, but he’s still only 6th in most passing yards, and 9th in touchdowns. I just don’t see anything in his performance to suggest he’s at an All-Pro level.

    • Well, Luck was really good last year, the first time his stats really began to feel commensurate with his talent level. Some of the Colts 33 wins the last three years are due to schedule, but I can’t imagine too many QBs could have done that with the talent on the rest of the Indianapolis roster. Luck probably isn’t at an All-Pro level yet — and certainly his 2015 performance hasn’t helped — but I suspect he will be pretty soon.

      • WR

        I hear you, Chase. 2014 was certainly Luck’s best season, and he did lead the league in passing TDs, which is really good. But this exercise rates him as the 3rd-best QB in the league, and I just don’t see that. The reason I bring it up is not because of the ranking in this article, but because it’s an example to me of the phenomenon where Luck is treated like he’s already a great QB, when he really isn’t. It reminds me of the Peyton Manning effect. Manning was treated like he was an all-time great long before his performance was at that level, and I think it’s affected the way Manning has been perceived in a way that isn’t really accurate. I’m worried the same will be true for Luck down the road, and that’s based on the assumption that Luck’s performance improves.

        • Michael

          I understand your point on Manning, but I disagree. He was actually elite early on. He was around league average his rookie season (26 TD,28 INT, 18th in DVOA, 12th in DYAR). Then his next two seasons were great.

          99: Improved team from 3 wins to 13 wins. 26 TD, 15 INT, 4135 yards, 2nd in ANY/A, 2nd in DYAR (just 5 DYAR behind Warner), 2nd in DVOA
          00: 33 TD (lead league), 15 INT, 4413 yards (lead league). #1 in DYAR, #1 in DVOA

          His next 2 seasons could be considered down seasons by conventional standards but he finished in the top 10 of ANY/A, DVOA, and DYAR both of those seasons.

          • WR

            You’re right about Manning’s stats, but my point is that pre-2004, Manning wasn’t performing at an all-time great level, but was being talked about as though he was. In his first six seasons, Manning had three that were really good-99, 00, and 03. This earned him two all-pro selections, though it’s not clear that he was the league’s best QB in any of those seasons. In 98, 01, and 02, he was good, but not spectacular.

            I just don’t buy the narrative that Manning has been head and shoulders above his contemporaries, since he took a few years to reach elite level. The same is true for Brady and Brees. But there’s been a perception, since very early in Manning’s career, that he’s a statistical standout. I think this is one of the reasons why he’s won so many awards, not all of which were clearly deserved.

      • James

        “I suspect he will be pretty soon”

        This was my reasoning – I guessed it would be this season. Two of my All Pros haven’t worked out so far, but in my defense I seriously considered putting only Rodgers up there. He’s that much better than everyone else right now.

        • So I guess the “reasoning” of the “other side”, is: What it is about Luck that makes you give him the benefit of the doubt? Are you rating anyone else based on how much you guess they will improve? Sure seems like a lot of people are taking his ascension to HOF greatness for granted.

          Sure, Indy’s QB future is brighter than NE’s, but I just can’t figure out a definition of QB play by which Luck has been better than Brady, this season OR last.

          • James

            Generally speaking the whole idea is to rate people considering how I expect them to improve or decline, although so far this season it seems the biggest movers are not the ones I guessed but Carr and Dalton.

            The main reasons why I give Luck the benefit of the doubt is everything he’s done up to this point: he was the best prospect in a long time both physically and mentally, he immediately stepped in and played well, he elevated his rather poor supporting cast to great heights, and he’s still relatively young. His offensive line has been bad his whole career and I haven’t been impressed with Pagano, yet Luck’s still done remarkably well.

            In particular with Brady, I think he benefits greatly from the greatest coach in NFL history who has designed a quick, short passing scheme that doesn’t work anywhere else. Brady’s deep ball is clearly below average and that generally has been one of the defining characteristics that separates the good QBs from the bad: http://archive.advancedfootballanalytics.com/2010/09/deep-vs-short-passes.html. I think it’s very informative that Belichick has surrounded Brady with Welker, Edelman, Amendola, and a plethora of tight ends.

            • WR

              James has touched upon something here that I think is very significant, and often gets discussed here at FP. There seem to be a lot of people on this forum who view Brady as a system QB. I strongly disagree. I don’t understand how a “system” QB can post the kind of career and single season numbers that Brady has produced. If a qb can consistently gain a lot of yards with YAC, doesnt that indicate he’s good at reading defenses? Montana mostly played the same way, was Joe a system guy, too?

              • James

                Joe Montana is obviously a great QB, one of the best in NFL history. But he played in a system that made Ken Anderson, Montana, and Steve Young all produce like Hall of Famers. So either Bill Walsh is the luckiest coach in NFL history, or his system was a significant advantage that elevated all of them to greater heights than their talents alone could.

                I’ll also share this analysis from Brian Burke that shows YAC is consistent among receivers year to year, but not quarterbacks, which indicates WRS are almost exclusively reaponsible for YAC: http://archive.advancedfootballanalytics.com/2007/08/who-gets-credit-for-yac-follow-up.html

                • WR

                  Ok, if you’re willing to say that Montana is also a guy who benefitted from a system, I’m with you. But we’re talking in Montana and Brady about the 2 most successful QBs of the last 35 years. So my question would be, which of the all-time great QBs would you say is not a system quarterback? I suspect a lot of people might say Peyton Manning, and I’ll grant you he’s never played in a system like that of Walsh or Belichick. But Manning has had a set of circumstances that have really helped his career.

                  He’s played for two franchises that have worked hard to put the best team around him. He’s played for a hall of fame coach in Dungy, and his performance was much better under Dungy than Mora. He’s had a lot of talent around him at the skill positions, including Harrison, Wayne, Edgerrin James, Addai, Dallas Clark, the Thomases, Eric Decker, Knowshon, and Emmanuel Sanders. He’s had 6 defenses that were top 8 in fewest pts allowed. And he’s played nearly half his career indoors, and unlike Brady, has rarely had to play in bad weather, factors which have undeniably helped his passing stats.

                  My point is that while Manning may not be a system guy, all great QBs are the beneficiaries of favorable circumstances. Even Marino played most of his career for Shula, arguably the best coach ever. So if you’re going to deduct points for Brady and Montana because of the system, that’s fine. But you also need to subtract points for other players, for the reasons I’ve mentioned here. So I’m not sure how useful the system argument is, when properly put into context.

                  • James

                    Peyton is a fun example because he breaks the mold – you could say he’s an *extreme* system QB because he struggled with both Fox’s and Kubiak’s systems. But does that matter when he immediately solved the problem and found success with his old system? I don’t know of any other QB that brings an offensive system with them.

                    As far as entanglement goes, we can be very certain that Peyton is a great QB because he’s been a great QB with so many different pieces around him – two different teams, a laundry list of skill players, offensive linemen, and many head coaches. Brady has been in the league long enough he isn’t far behind with sustained success despite rotating OCs and skill position players, but it’s nearly impossible to separate what’s him from Belichick.

    • Adam

      Agree with you on Luck. I don’t think he’s reached the status of All-Pro yet, but for some reason everyone has been trying to anoint him as a future HOFer since the moment he was drafted. To me Luck is clearly below Rodgers and Brady, and possibly behind Roethlisberger, healthy Romo, and healthy Peyton.

    • James

      Do you really think there should be more All Pros? My reasoning for fewer is if everyone is an All Pro then no one is, and honestly Rodgers is so much better than everyone it seems unfair to put anyone else in the same tier.

      That said, a few years ago it was clear Brady, Brees, and Peyton joined Rodgers as the big four and it would have been tough to leave any of them out.

  • Adam

    James, I like what you’ve done here; it’s a different perspective on QB rankings than we’re used to seeing. The one adjustment I would suggest is to incorporate an injury risk component. Looking at Romo and Brady, for example, I think your system has them backwards. Romo may be younger, but even before this season he was certainly at higher risk for injury than Brady, which IMO makes Brady more valuable. Same deal with RGIII – he’s young but that doesn’t matter because he seems so unlikely to stay healthy.

    • James

      That’s a good idea, although I think it would be hard/unreliable to implement. Griffin is a clear high injury risk, but what about others with injury histories (Stafford, Bradford) and other runners (Cam’s superman TD last week is the exact type of play Griffin is criticized for). Do I factor in things like bad OLines killing Luck, even though this is supposed to be independent of supporting casts? Also, as far as I understand Romo’s surgery last offseason “fixed” his back problems, so I don’t know if he’s any more likely to get hurt than others. I’d consider the broke clavicle a fluke, but others may feel differently.

      Personally I think injuries are more random than repeatable, as Brees certainly would have been as low risk as possible. Maybe make it correlate with QB hits and carries?

  • sacramento gold miners

    Not sure I buy into the Teddy Bridgewater hype, he looked overwhelmed on Monday Night Football. I also think Tony Romo is closer to the end of his career than many think. At 35, with the back issues, will he finally be able to succeed in the playoffs? He’ll have to prove it.

    • Richie

      Yeah. Hopefully he can put those 2014-style (125 passer rating) playoff performances behind him.

      • sacramento gold miners

        In the playoffs, timely plays outweigh overall stats, and Romo just hasn’t made enough of those games during his career. I’ll take the QB with the 85 passer rating who delivers a winning performance. Two playoff wins lifetime is horrible.