#### ZWK

##### Footballguy

Inspired in part by Coach Potato's Bruce Hammond's use of Quality Years Remaining (QYR) in his dynasty rankings, I've taken a look at the aging patterns for RBs and WRs.

The basic idea is to look at players who are coming off a good year and see how long they last. It would be easier to just look at the age distribution of the top performers (how many RB1s are 26 years old vs. 29 years old), but I wanted to track individual players instead of just looking at a cross section (which can be misleading). But I needed to focus on good players - plenty of players don't last long in the NFL simply because they aren't good enough, and they don't tell us much about aging patterns.

So what I did is take every RB who had a RB1 season at age 26 and look at how many productive years they had left. And I did the same thing with age 27 and every other age, and with WRs, and plotted a graph.

The graph is here, with RBs in orange and WRs in blue. The x-axis is age and the y-axis is number of years left, and the graph shows how many good years a player had left, if he had a RB1/WR1 season at that age.

I used Pro Football Reference's data, looking at every RB who entered the league from 1983-2003 (and WRs 1983-2002). I defined a RB1 season as 50+ VBD, since about 12 RBs made that cutoff each year (and a WR1 season as 40+ VBD). There were 33 RBs in the data set who had a 50+ VBD season at age 26, and I averaged together how many good years each of them had left and plotted that on the graph in orange at age 26. And the same for other ages and for WRs.

You'll notice on the graph that there are 4 different orange lines, and 4 blue lines. That's because I wasn't sure how to count how many years a player has left so I came up with 4 different ways to do it and graphed them all. I'll explain what they are below, if you're interested, but the important thing is the pattern (which is basically the same for all 4 lines) and the differences between RBs and WRs.

You can see that WRs last longer than RBs - the WR lines are above the RB lines, and they don't run into the x-axis until age 35 (vs. 31 for RBs). The WR lines are also a lot flatter than the RB lines - they're not nearly as steep.

The steeper lines for RBs mean that a RB's remaining performance is heavily dependent on his age; a WR's age is less important. A WR who is coming off of a very good season at age 33 probably has less left in the tank than a WR who is coming off a very good season at age 28, but the difference between them is not all that big - it's smaller than the difference between a 25-year-old RB and a 28-year-old RB. There is a narrow age window when RBs hit the wall, around 28-31, which means that even a RB who has a great season at age 30 is living on borrowed time and more likely than not doesn't have another great season in him. But WRs hit the wall at many different ages, from 29-35. If an older WR is coming off a great season and his play hasn't shown signs of decline, then he's likely to have another year or two left (unless he's 35+).

You can see a similar pattern on another graph I made here, which shows total VBD remaining at each age. Again, RBs are in orange and WRs in blue. The lines with square data points show the average remaining VBD for all players who had a RB1/WR1 season at that age. I was concerned that fluke players who had one big year might be skewing the data, so I looked separately at the players who were coming off at least 2 consecutive RB1/WR1 seasons; their average remaining VBD is shown by the lines with triangle data points.

This again shows that WRs age more gradually than RBs. According to Bruce Hammond's QYR, a WR who is turning 32 (coming off his age 31 season) has about as much left as RB who is turning 30. I said something similar here ("WR age 33 is like RB age 30"), based on a different data set (how many RBs & WRs had 1,000-yard seasons at each age). But it looks like that's wrong, if we're focusing on individual players who haven't shown much sign of decline yet (and have just put up a RB1/WR1 season). A 32-year-old WR who just had a WR1 season (at age 31) probably has more left in the tank than a 30-year-old RB. So does a 33-year-old WR, or a 34-year-old WR. A 30-year-old RB is more like a 35-year-old WR. But a 27-year-old RB (who is coming off his age 26 season) is a lot like a 27-year-old WR. In between, RBs all hit the wall while WRs gradually thin out.

More on Methodology

As I said, I wasn't sure about how to count how many years a player has left so I came up with 4 different ways to do it. Let's look at Eddie George to see the 4 ways of counting. Here is what he did from age 26 onward:

26: 124 vbd

27: 133 vbd

28: 17 vbd

29: 74 vbd

30: 7 vbd

31: 0 vbd

32: out of football

Eddie George was a RB1 at age 26. How many good years did he have left after that?

One answer is that he had 3 good years left. That's the answer if "good years left" means how long he had until his last RB1 season. The last time he had 50+ VBD was at age 29, and that came three years after his age 26 season, so the answer is 29 minus 26 equals 3.

Another answer is that he had 2 good years left. Just count how many good years he had after his age 26 season: one RB1 season at age 27 and another at age 29. So, two.

These two answers both take "good years" to mean RB1 years, but maybe any year as a fantasy starter should count as "good." That gives us two more methods for counting good years left, modifying each of the two that we already used.

How long did he have until his last season as a fantasy starter? Four years, since age 30 was his last year with any VBD, which was four years after his age 26 season.

Or, we could count how many years he had as a fantasy starter after age 26. That would also give an answer of four, since he was a fantasy starter at ages 27, 28, 29, and 30.

Those are the four ways I came up with. In the graph, the two ways that use 50+ VBD (or 40+ VBD) as the standard for good years have the data points represented as triangles, and the two ways that use starter as the standard have square data points. Comparing them in the graph, you can see that even late in their careers, players usually have about 1 more starter season left than they have RB1/WR1 seasons.

If someone can convince me that one of these four options (or some other option that I can calculate) is the right way to do it, then that'll make things simpler for me if I do any more work with this.

Edit: here are my data spreadsheets with the raw data for RBs and WRs.

The basic idea is to look at players who are coming off a good year and see how long they last. It would be easier to just look at the age distribution of the top performers (how many RB1s are 26 years old vs. 29 years old), but I wanted to track individual players instead of just looking at a cross section (which can be misleading). But I needed to focus on good players - plenty of players don't last long in the NFL simply because they aren't good enough, and they don't tell us much about aging patterns.

So what I did is take every RB who had a RB1 season at age 26 and look at how many productive years they had left. And I did the same thing with age 27 and every other age, and with WRs, and plotted a graph.

The graph is here, with RBs in orange and WRs in blue. The x-axis is age and the y-axis is number of years left, and the graph shows how many good years a player had left, if he had a RB1/WR1 season at that age.

I used Pro Football Reference's data, looking at every RB who entered the league from 1983-2003 (and WRs 1983-2002). I defined a RB1 season as 50+ VBD, since about 12 RBs made that cutoff each year (and a WR1 season as 40+ VBD). There were 33 RBs in the data set who had a 50+ VBD season at age 26, and I averaged together how many good years each of them had left and plotted that on the graph in orange at age 26. And the same for other ages and for WRs.

You'll notice on the graph that there are 4 different orange lines, and 4 blue lines. That's because I wasn't sure how to count how many years a player has left so I came up with 4 different ways to do it and graphed them all. I'll explain what they are below, if you're interested, but the important thing is the pattern (which is basically the same for all 4 lines) and the differences between RBs and WRs.

You can see that WRs last longer than RBs - the WR lines are above the RB lines, and they don't run into the x-axis until age 35 (vs. 31 for RBs). The WR lines are also a lot flatter than the RB lines - they're not nearly as steep.

The steeper lines for RBs mean that a RB's remaining performance is heavily dependent on his age; a WR's age is less important. A WR who is coming off of a very good season at age 33 probably has less left in the tank than a WR who is coming off a very good season at age 28, but the difference between them is not all that big - it's smaller than the difference between a 25-year-old RB and a 28-year-old RB. There is a narrow age window when RBs hit the wall, around 28-31, which means that even a RB who has a great season at age 30 is living on borrowed time and more likely than not doesn't have another great season in him. But WRs hit the wall at many different ages, from 29-35. If an older WR is coming off a great season and his play hasn't shown signs of decline, then he's likely to have another year or two left (unless he's 35+).

You can see a similar pattern on another graph I made here, which shows total VBD remaining at each age. Again, RBs are in orange and WRs in blue. The lines with square data points show the average remaining VBD for all players who had a RB1/WR1 season at that age. I was concerned that fluke players who had one big year might be skewing the data, so I looked separately at the players who were coming off at least 2 consecutive RB1/WR1 seasons; their average remaining VBD is shown by the lines with triangle data points.

This again shows that WRs age more gradually than RBs. According to Bruce Hammond's QYR, a WR who is turning 32 (coming off his age 31 season) has about as much left as RB who is turning 30. I said something similar here ("WR age 33 is like RB age 30"), based on a different data set (how many RBs & WRs had 1,000-yard seasons at each age). But it looks like that's wrong, if we're focusing on individual players who haven't shown much sign of decline yet (and have just put up a RB1/WR1 season). A 32-year-old WR who just had a WR1 season (at age 31) probably has more left in the tank than a 30-year-old RB. So does a 33-year-old WR, or a 34-year-old WR. A 30-year-old RB is more like a 35-year-old WR. But a 27-year-old RB (who is coming off his age 26 season) is a lot like a 27-year-old WR. In between, RBs all hit the wall while WRs gradually thin out.

More on Methodology

As I said, I wasn't sure about how to count how many years a player has left so I came up with 4 different ways to do it. Let's look at Eddie George to see the 4 ways of counting. Here is what he did from age 26 onward:

26: 124 vbd

27: 133 vbd

28: 17 vbd

29: 74 vbd

30: 7 vbd

31: 0 vbd

32: out of football

Eddie George was a RB1 at age 26. How many good years did he have left after that?

One answer is that he had 3 good years left. That's the answer if "good years left" means how long he had until his last RB1 season. The last time he had 50+ VBD was at age 29, and that came three years after his age 26 season, so the answer is 29 minus 26 equals 3.

Another answer is that he had 2 good years left. Just count how many good years he had after his age 26 season: one RB1 season at age 27 and another at age 29. So, two.

These two answers both take "good years" to mean RB1 years, but maybe any year as a fantasy starter should count as "good." That gives us two more methods for counting good years left, modifying each of the two that we already used.

How long did he have until his last season as a fantasy starter? Four years, since age 30 was his last year with any VBD, which was four years after his age 26 season.

Or, we could count how many years he had as a fantasy starter after age 26. That would also give an answer of four, since he was a fantasy starter at ages 27, 28, 29, and 30.

Those are the four ways I came up with. In the graph, the two ways that use 50+ VBD (or 40+ VBD) as the standard for good years have the data points represented as triangles, and the two ways that use starter as the standard have square data points. Comparing them in the graph, you can see that even late in their careers, players usually have about 1 more starter season left than they have RB1/WR1 seasons.

If someone can convince me that one of these four options (or some other option that I can calculate) is the right way to do it, then that'll make things simpler for me if I do any more work with this.

Edit: here are my data spreadsheets with the raw data for RBs and WRs.

Last edited by a moderator: