Are older players playing better this season?

320x

With players like Jason Kidd, Tim Duncan, Kobe Bryant, and Ray Allen posting excellent numbers to start the season, I’ve been wondering: are this season’s older players playing abnormally well?

To answer this question, we first have to determine the definition of “older player”. After doing a little bit of research, I settled on players who are 35 years old or older by the end of the season. Why? Because if that age was any lower, there would be too much data for me to go through in a reasonable amount of time. Besides, 35 is a nice round number. Although I should note that by this definition, Kobe Bryant is not included in this season’s sample. We also have to decide on a time-frame. I chose to include data from this season, as well as the past ten seasons.

Given these parameters, we can use Wins Produced to determine if older players are playing better than they have in the past. But first, let’s take a look at the number of players aged 35+ across the last eleven seasons:

players35+_byseason

The number of older players fluctuates from season to season, but despite better medical technology, it doesn’t seem like there are more older players playing in the league today than there were ten years ago. This is especially true if you account for the fact that the league added an extra team starting at the 2004-05 season, which means that there are more roster spots available to older players. This season’s group of older players is smaller right now than it will be at the end of the season, so this will need to be revisited once the regular season is over.

Getting back to the main question: how do this season’s older players compare to previous seasons’ older players? Using players’ birthdates, I calculated each players’ age on the last day of the regular season — although for this seasons’ players, I calculated their ages based on yesterday’s date. Then I calculated the average age, weighted by minutes played, of each season. Finally I used TheNBAGeek to calculate the total Wins and WP48 of each season sample. The results are summarized in the following table:

players35+

Now, using this table, it is a bit difficult to discern any patterns. There are certain seasons where it is clear that the older players are more productive, but at first glance that seems like it might be related to the ages of the players. If we really want to find out if this season’s older players are playing any better, we should graph WP48 vs weighted age to see if the productivity of older players is related to their ages.  The graph looks like so:

players35+wp48_age2

Well, so much for this season being any different — it falls into the same pattern as the rest of the seasons. It turns out that the age of the oldest players explains 26% of the difference in productivity; as we would expect, the older the sample, the lower the productivity, and the younger the sample, the higher the productivity.

Another interesting thing to point out is that in three out of the eleven seasons, the oldest players in the league produced at an above average level of productivity. In the three seasons the oldest players didn’t hit the average mark, they were still quite close. How could the oldest players in the league be so productive? The answer is survivorship bias: typically only the very best players managed to stick around until they are 35 years old or older, which means the less productive older players have already been weeded out.

So, to answer the question: no, today’s older players aren’t playing better than in previous seasons. Believe it or not, they’re just younger than usual.

- Devin

Comments are closed.