FanPost

What NFL Aging Curves Tell Us About the Chiefs' Contract Decisions


A few months ago, I started seeing a pattern in Kansas City's decision making regarding players to keep and pay, players to keep at low risk contracts, and players to let go. It was seemingly a familiar pattern, as I'd watched Andy Reid play the "it's better to let a player go a year early than a year too late" game with Philadelphia. However, this time it was slightly different. In Philly, Andy might let someone like Jeremiah Trotter go when Trotter wanted more money than the franchise would give, but he welcomed players back (like Trotter) a few seasons later. Aside from signing players to one year contracts, I've not seen Brett Veach bring back older players for substantial contracts.

It was about that time I learned about aging curves.

A few different sites have done statistical analysis, often using Pro Football Reference's Approximate Value metric, but a few others using things like Pro Football Focus's WAR metric or specialized metrics like 538.com's True Receiving Yards measure for receivers. Maddeningly, many of the articles out there on the subject either focus on a specific position group only (or in the case of most fantasy sites, just on offensive positions) or they have at least partially disappeared (in a few instances, I found text articles missing their graphs or with broken links to promised data).

If I could afford to retire, I would tackle the project of taking the latest data and calculating curve fits for players at each position (probably based on AV), but alas, I do need my salary and the lost hours of fun that go with it. So instead, I found as many sources as I could and gathered a rough composite of the curves that they collectively described for each position (where applicable). Yes, this is a survey of other people's analyses, which is problematic at best, but it's the best I've got right now. Fortunately, enough of the sources roughly agree on the key milestones defining the aging curves (within a year or so of each other) that a representative curve can be defined without it being too out of whack.

In particular, I tried to capture the age at which players at each position begin their steepest ascent toward their peak performance, the point where they reach that peak and begin a plateau or gradual degradation in performance, and the age at which the average player begins a steep decline in value due to on field performance. Here are the rough milestones I found:

Position Rise Peak Decline
OT 25 28 31
C 26 28 32
OG 26 27 29
QB 26 28 30
WR 23 26 27
TE 22 25 30
RB 22 24 28
DL 22 29 30
LB 23 24 27
DB 22 24 27

So why are there these milestones marking specific spans of a player's career, and why are they different for different positions? My speculation is that players come into the NFL largely still needing improvement in many of their technical on-field skills, whether it's reading defenses or polishing (or developing) techniques in blocking. Some positions also may require additional strength and conditioning by the players in comparison to college. I think developing the technical skills implies why QBs see their rise relatively late and the combination of both push the rise to peak for offensive line later as well. As a player progresses in their careers, experience allows them to do their jobs better and more efficiently, but it comes with the price of the wear and tear the body takes with the repeated stress of playing. That's probably why running backs tend to have a very early peak. Injuries also pile up, which can affect speed, lateral quickness, and agility as a player gets older, which is probably why we see the decline of more speed-oriented players setting in earlier than many other positions.

A couple things to note:

First, these milestones are relative to the performance of the player. That is, an All-Pro linebacker at age 28, even if he's on the decline, is still going to outplay a journeyman who's at his peak at 24. What it does mean is that at age 28, we can expect most linebackers to see their performance dwindle at an accelerating pace so that maybe that All-Pro is playing much closer to a peak journeyman when that superstar is at 31 or 32. It also means that the journeyman isn't likely to make many leaps in playing ability beyond that age 24 year.

Second, annoyingly the only material I could find on DEs and DTs had them lumped together as "Defensive Line". If I were to do the study directly from data, I would separate out edge rushers from interior defensive linemen and develop the aging curves for them separately. Unfortunately, that wasn't in the cards for this exercise. In addition, I wasn't able to find data separating corners from safeties among the DBs.

Finally, there are obviously always going to be exceptions. The models I looked at were aimed at coming up with the best fit for the overall population they studied and weren't aimed at the exceptions. Tom Bradys exist in the world, but there aren't many of them.

So what does this mean for KC's decision making? Let's take a look at some specific situations.

One contract that has been very unpopular after the fact was Frank Clark's contract. It's easy to judge the deal harshly in retrospect. But let's take a look at what we knew at the time. When the trade was made and the new contract was signed, Clark was just coming off a 13 sack season in his age 25 year. Given that age, we would normally expect him to improve or at worst stay about steady for the next few years, meaning paying him like a 13+ sack rusher wasn't a horrible idea in projection. Unfortunately, between illness, injuries, off field considerations, and who knows what else, Clark's performance has declined instead.

The Chiefs signed Tyrann Mathieu to be their defensive leader starting in HB's age 27 year. This actually was a bit of a risk, but HB proved to be one of those exceptions to the rules. It's logical that safeties probably age slightly better than corners (again, wish I had time for a hands on analysis rather than a survey), but HB's contract still took him up to his age 30 season. At this point, I think it's logical that even if HB holds his on-field value for another year, a multi-year contract would start to look pretty bad as we went through the out years.

The contract offers to Orlando Brown, Jr. have of course been a hot topic today. He turned down what many of us saw as a reasonable contract offer, but many other commenters declared he wasn't worth that given his play to date. I think many of those comments put too much weight on his performance early in the season while still learning KC's offense versus later in the season when he started playing much better. But putting that aside, those comments also ignore the fact that, at age 26 this season, he should keep getting better, at least over the next couple of seasons. Analysis of the second half of his season put him in the top 10 or so LTs in the NFL during that span (after he had adjusted to the KC offense and better gelled with his teammates, but also after injuries hit some of his competition in the top ranks of LTs), so improvement over the next couple of seasons should theoretically get him at least among the top performers even if he doesn't reach the actual top that he is aiming at. As such, paying him as one of the expected top performers isn't completely insane. That expectation of improvement is also why I think the Chiefs keep around younger players that fans would otherwise like to kick to the curb, only to finally cut ties at seemingly arbitrary points (and while Dan Sorensen lasted through his age 31 season, Ben Niemann was let go just as he reached 27).

Finally, the Chiefs also let Tyreek Hill go after the receiver market reset and the asking price for Hill's service went up dramatically. While the average wide receiver decline starts at age 27 and Tyreek just completed his age 27 season, Tyreek would seem like a likely exception to the rule given he is seemingly superhumanly fast. Interestingly enough, his age 26 and 27 seasons showed lower AV than his age 23 and 24 seasons (age 25 was somewhat interrupted after that shoulder injury against Jacksonville, but the team won the Super Bowl, so who's to complain?). At least part of the lower AV is due to defenses figuring out how to better defend him. The question is whether a 28 year old Tyreek will figure out new approaches to avoid getting penned in like he has the past couple seasons. It's unclear if a veteran that far along has the same will to learn and adapt that a young kid trying to make the team will have. It's also unclear how many more largely uninjured seasons Tyreek has in him. With those questions, it makes a lot of sense that the team didn't break its salary cap bank for him.

Before ending this, I'd like to highlight the most obvious exception to all of this aging discussion, Travis Kelce. According to AV, Kelce hit his peak in 2018, his age 29 season. in the intervening years, he's had a very gradual decline in his AV measures while putting up gaudy counting stats. We don't know how long he can keep putting off that more precipitous decline, but the fact that his brother Jason is also defying the NFL aging curve and maintaining a high level of play into his advanced years bodes well for Travis to keep it going for at least a little while longer. Plus I think you always have to cut the heart and soul of the team a little slack in making these judgments.

My apologies for the long screed, but hopefully there's been enough useful information in here to make slogging through this worthwhile. What I'd most like you to come away with is appreciation for the thinking that goes into both remaining patient with a player early in their career and refusing to give out an expensive long term contract to an older player, even when it's a position of need and when the player's performance to now would make it seem like the salary makes sense.

This is a FanPost and does not necessarily reflect the views of Arrowhead Pride's writers or editors. It does reflect the views of this particular fan though, which is as important as the views of Arrowhead Pride writers or editors.