What follows is a look into the research that has been conducted on the aging curve for NFL defensive players.
In my previous post we looked at the aging curve for NFL offensive players, and in the post before that, which serves as an introduction to this research, we reviewed the (1) methodology, (2) value metric, and (3) population that was used for the project as a whole.
For people that would like the readers digest of these three points, (1) this research uses the delta method to minimize the survivorship bias inherent in attempts at aging curves (a survivorship bias is still present in this method, but it is smaller than through other means), (2) we use Pro Football Reference’s approximate value metric as a way to judge player performance, (3) and our population includes all NFL players from 1980 to 2014.
As an update to points two and three, we will use Pro Football Focus player grades as an additional value metric. Pro Football Focus player grades give a score to a player for every single snap that they are on the field. The advantage of this metric is that player grades ascribe value to players independent of the other players that are on their team (i.e. a good player isn’t discredited because he’s on a bad team). Player grades are also able to give value to positions that are void of conventional stats—like offensive lineman—and replace rudimentary statistics—like interceptions for cornerbacks—that aren’t an accurate depiction of a player’s true talent level.
I previously looked at the aging curve for NFL defensive players as a group with Pro Football Focus player grades, but was unable to come up with an answer to the question around the aging curve for specific positions. Upon further efforts, adjustments were made to answer that question, and you will see those findings in what follows.
To account for the fact that Pro Football Focus only has publicly available player grades from 2007 to 2014, along with the previously mentioned time period of 1980 to 2014, we will also look at the aging curve of Pro Football Reference’s approximate value for the time period of 2007 to 2014. We need to do this to make sure that any difference between the aging curves for player grades and approximate value are symptomatic of the metric itself and not the time period.
While Pro Football Focus player grades and Pro Football Reference’s approximate value both look to put a single number on the value of a NFL player’s performance, the aging curve that emerges for each specific NFL defensive position changes depending on which value metric you look at.
Aging Curve for NFL Defensive Players
This chart seems to suggest that defensive tackles peak later in their careers and take time to develop, with their peak coming in their age 28 season. To go along with what we found in our last post, this evidence seems to reinforce the idea that positions that rely more on strength than speed peak later in their careers.
For these graphs and all other graphs in this post, don’t pay attention the units of measurement that are on the vertical axis of each chart when looking to compare the two charts; player grades and approximate value are not on the same scale; you’d be comparing apples to oranges. Instead look at the shape of the curve and where players peak and drop off.
Pro Football Focus’ player grades seem to agree with what approximate value has to say as far defensive tackles peaking in their age 28 season, but this chart suggests that defensive tackles may decline more quickly than approximate value thinks.
We can see that by approximate value, defensive ends peak at 26 and sustain a long period of peak performance until their age 30 season. Defensive ends peak before defensive tackles and are closer to their peak when they enter the league, and because defensive ends rely less on strength and more on speed than defensive tackles, this makes sense.
At 25, the aging curve for player grades suggests that defensive ends peak slightly earlier than approximate value implies.
Approximate value shows that linebackers peak in their age 26 season and stay around their peak into their late 20s. Once a linebacker turns 30, we can see that their performance starts to decline rapidly.
With pinnacles of 23 and 24 respectively, the PFF player grades for both outside and inside linebackers show an earlier peak for the position than approximate value. While both approximate value and player grades show large drop offs for linebackers at certain points in their careers, player grades seem to think that it happens at around 27 and 28, much earlier than approximate value indicates.
According to approximate value, we can see that cornerbacks are at their peak in their age 27 season, and they are closer to their peak in their age 22 season—when most players enter the NFL—than any other position, narrowly ahead of linebackers and, as you will see next, safeties.
Cornerback and safety are where approximate value and player grades disagree the most. While the former metric thinks that cornerbacks peak at 27, the later would say 23.
Along with linebacker and cornerback, we can see that safeties fall within the group of players that are closest to their peak when they enter the league when it comes to approximate value. This graph suggests that safeties age better through their age 35 season than other positions and have a sustained period of peak performance until their age 30 season.
Player grades seem to think that safeties peak at age 22, age nicely until their late twenties, and fall off a cliff at 29, while approximate value believes safeties peak at age 26.
Tying It Together
Through this graph, we can see the same narrative appear for defensive players that we saw for offensive players when it comes to pre peak development; positions that rely on their strength—defensive end and defensive tackle—are further from their peaks when they enter the league than positions that depend on their speed—cornerback, linebacker, and safety.
While Pro Football Focus player grades agree that speed dependent players—cornerbacks, linebackers, safeties—peak earlier than speed independent players—defensive tackles and defensive ends—the two metrics are in a disagreement about which positions age better.
Pro Football Focus suggests that linebackers age worse than other positions, while Pro Football Reference considers defensive tackles, defensive ends, and cornerbacks as the positions that mature the worst.
We will get into which stat you should trust in a moment, but as far as which metric to believe when it comes to positions that age the worst, the author has to agree with Pro Football Focus. When you think about players that not only take the hardest hits, but get hit most often, you think of linebackers. When you look at the 30 players that lead the NFL in tackles in 2014, 23 were linebackers. This is purely conjecture, but one would assume that the fact linebackers take more hits than any other defensive position, and that those hits seem to be more violent (a tackle by a linebacker on a running back up the middle seems to be harder on the body than when a cornerback tackles a wide receiver in open space), would cause them to age worse than other positions.
Generalization: with the exception of defensive tackle, which both metrics are in agreement about, the consensus of approximate value seems to be that defensive players peak around age 26, while Pro Football Focus’ player grades imply that defensive players peak much earlier in their careers.
Who do we believe?
While we have to keep in mind that the sample size for Pro Football Focus player grades is much smaller than the population for Pro Football Reference’s approximate value, and that the peak ages for player grades may change slightly as the population grows, the true question of who to believe comes down to an introspective look at the methodology of each of the metrics.
To come up with their value on player performance, Pro Football Focus grades every single player and every single play to come up with a numeric assessment of how well a player performed independent of the other players on their team and the inherent bias that exists in the outside world. They look at the actual performance of the players, not the name on the back of the jersey, as they are self proclaimed “performance based scouts.”
Pro Football Reference’s approximate value does none of what player grades accomplish. Approximate value is calculated different for each position, but common, key components in the calculation of each position’s approximate value formula is the performance of the offense or defense (i.e. good players are discredited for being bad defenses, and bad players are overrated for being on good defenses) a player played on and whether or not the player made a Pro-Bowl or was voted to an All-Pro team.
To be more specific on how approximate value rewards or discounts players based on the performance of their team, rather than an individual’s true talent, this article, titled Approximate Value: Methodology, on Sports Reference says (in reference to approximate value),
On the defensive side of the ball, things just aren’t so clear. If we split a defense’s production into rushing defense and passing defense, to what extent do [we] then divvy up the passing defense points between pass rushers and pass defenders? I really don’t have a clue how to answer that question in general. How much credit do linebackers get for pass defense versus run defense? I don’t know. How do we account for the fact that some teams use three linemen and four linebackers and some do the opposite?
Absent of stats that look at the actual performance of players independent of their team (what Pro Football Focus player grades actually do), we may unfairly ascribe more credit or blame to a player than they actually deserve. There may be some correlation between individual player performance and team performance on the defensive side of the ball, but the correlation is sure to be small with numerous exceptions scattered throughout the data (e.g. the Redskins were 30th in the NFL by weighted DVOA in 2014, but that doesn’t mean that Ryan Kerrigan wasn’t a great player; approximate value holds his team’s performance against him, while player grades don’t).
On the point of players being rewarded based on Pro-Bowl and All-Pro honors, I find this is equally as troublesome as assuming that a defensive player’s ability correlates with his team’s performance. Currently, players are voted into the Pro-Bowl by players, coaches, and fans, and while the first two groups may be largely biased (e.g. would a player vote for division rival or a teammate at the same position to make the Pro-Bowl?) through fraudulent opinions (they have something to gain by having players on their team make the Pro-Bowl), we can all admit to be skeptical of an average fan’s ability to objectively asses the performance of players, especially when they may not be familiar with all the relevant players outside of their own team’s roster. All-Pro voting is equally as deceitful with writers pretending to be analysts of football just because they write about football; the skills that are needed to be good at writing and analyzing are not necessarily general to both domains.
While the aging curve for NFL defensive players changes depending on which metric you look at, because of the more objective and robust methodology, there must be greater validity and attention given to the aging curves that have been generated based off of Pro Football Focus’ player grades. The previous article that I wrote looked at the aging curve for individual offensive players based on approximate value, and this research will eventually be redone to look at the aging curve with player grades used as the value metric. While the data is still somewhat noisy, because of the small sample of players that have publicly available player grades, we can definitively say that, as a generalization, defensive players age earlier than most would expect.
Photo Credit: SteelCityHobbies
Latest posts by Devin Jordan (see all)
- Week 13 NFL Breakdown: Joe Flacco Shouldn’t Make You Think Twice - December 7, 2016
- Week 12 NFL Breakdown: Mark Ingram Rolls Against the Rams - November 30, 2016
- Week 11 Breakdown: DeVante Parker Could Be Turning the Corner - November 23, 2016