Round And Longevity: Another Look At NFL Career Length
ROUNDS AND LONGEVITY: ANOTHER LOOK AT NFL CAREER LENGTH
In the twelve year span from 1995-2007, about 2,600 players that heard their names called in the NFL Draft have seen the field on Sunday somewhere and for some length of time. The question we are examining in this particular work is, “How long?”
The answer may not shock many people - it might very well depend on where you were picked in the draft. We know from anecdotal evidence that the typical NFL career is on the short side, perhaps even fleeting. Indeed, in the sample I used – 1995-2007 – regardless of round, you could expect to spend 5.21 years in the NFL on average with a standard deviation of about three years. Here’s the broad distribution of players by career length:
The distribution is very, very skewed – in a sample of 2,624 players, 30.7% of them played 8 or more years, and only 12.8% of them played 10 or more years. Of more note, 49.3% of those players played five years or fewer, so you can see where that particular average comes from. Speaking of average career length, it began to take a tumble towards the end of the studied period.
Let me change this up for a moment and explain why I used 1995-2007 as a time frame. The primary reason is that I wanted to use a time period where there would not be very many careers still in progress. Using data too close to the present would drag down the averages for the first few rounds and probably do little to the remaining rounds. Also, it is a relatively recent timeframe, so the game was in a state at least close to what it is at the moment.
Bearing that in mind now, let’s look at average length of service by round overall:
Here, you’ll note the bars which show plus or minus one standard deviation. Essentially, the average career length of a first round pick is pretty atypical in the NFL, and you can see this tumble rather nicely to the also atypically brief career of a seventh round pick. The line crosses the average career length between the 3rd and 4th rounds, so basically your chances for a career of some length are greater in the first three rounds. That shouldn’t surprise you, but the numbers back it up.
Here’s the same data, but broken out by round and year. It will appear cluttered at the bottom (which you should expect as the separation decreases rapidly in later rounds), but you’ll note that the first and second rounds more or less fly high above the others.
Another way to look at this is to see the individual rounds distributed by length of service and player counts so that you can see the profile of each round in this time period. The first and second rounds are fairly normal, although they both have some multimodal moments:
The third round’s profile is very different, heavily skewed towards briefer careers:
The fourth through seventh rounds therefore should not be surprising:
So, we can see in this sampling that your draft round definitely can be a factor in how long you manage to stay in the league. In another part of this particular set of diaries, we will zoom in and look at the first two rounds and see if there is a potential impact from one’s spot in the draft order and later on we will tackle this by position to examine tendencies by round.
Oh, and in case you're ever confused...
Everything seems kind of what I would expect...except the "glitch" in your first plot where it is more likely to be a 7 year player in the NFL than a 4 year player. I will take a stab at a theory.
Generally speaking, the plot should simply decrease as time goes by (age + injury potential). But at year ~4, some guys "figure out" how to be NFL players and stick another so many years. Those guys who figure it out last until the age/injury curve makes them leave and the curve goes back down. So the glitch is the guys who figure out how to be a professional.
That being said, it might just be monetary or maybe based on contract length!
The other funny little bump - and perhaps I should have put this in the diary itself - is the median length of career for each round.
1st Round - 8 years
2nd Round - 4 years
3rd Round - 3 years
4th Round - 5 years
5th Round - 4 years
6th Round - 3 years
7th Round - 3 years
That small bump in the later rounds is something I hope is made more clear when I do the breakdown by position in a couple weeks.
April 17th, 2015 at 11:42 AM ^
It's contracts. 7 years is the max length of a rookie contract, or how long he'll be restricted as a free agent. Guys who don't have it will peter out as teams give up on them, but teams will tend to hold onto assets they control; a guy who's just hanging on at the conclusion of that is less likely to be retained.
April 26th, 2015 at 10:40 AM ^
My last pay check was $9500 working 12 hours a week online. My sisters friend has been averaging 15k for months now and she works about 20 hours a week. I can't believe how easy it was once I tried it out. This is what I do...www.globe-report.com
Correct me if I am wrong, but your sample only includes players that actually played in a regular season NFL game. If that is the case, a bimodal distribution makes sense. Players drafted in rounds 1-3 are more likely to see the field early on and less likely to be put on the practice squad. Hence, they are less likely to be project players and will get the benefit of doubt over later picks. Day 2-3 players are much more likely to be on the practice squad and can be considered project players. They will only see the field if they show a predisposition for success in training camp or on the practice field. My gut says that if you were to show average career length for all draftees instead of just those that saw the field you would see a steady decline from rounds 1-7.
but also wanted the answer to this question.
This was taken from straight dumps by year from Pro-Football Reference, so everyone they had data on that was taken in any round is in this data set. There were probably only about 100 names over the course of 12 years of data that were name only and draft year but no other data on how long they were in the NFL.
clears that up.
April 15th, 2015 at 11:12 AM ^
It seems that the first two round career lengths have been trending down the last few years. Is that because teams are getting less effective at identifying eleite talent (doubtful) or is it because there has been a change in the top position groups taken, position groups that have a generally shorter career.
When I do the position / position by round version of this, I am hoping to find this out actually. Part of why I stopped short of present time in this diary was, as I said, to get averages that were as unmuddled as possible, but hopefully we get to see if part of the reason for this is buried in the NFL's draft behavior, if you will.
As lots of people have noted, the main problem for inference here is that first round players are better than second round players, so the "round effect" might be a result of prior factors rather than the draft process itself.
If you want to improve causal identification of the "round effect", you could do what folks call "regression discontinuity": essentially, you compare players who fell right next to important cut points. For instance, compare the three last players of the first round to the three first players of the second round. In principle, there shouldn't be much fundamental difference in quality between those two sets of players. This could give you a better estimate of the causal effect of being selected in an early round.
http://en.wikipedia.org/wiki/Regression_discontinuity_design
Comments