Welcome to the inaugural post of The Bar Stool GM, a blog that will cover sports from a data-driven and statistical perspective. As it’s fantasy football season, I wanted to start off with some analysis to help you get an upper hand on your league mates. This first series of posts will cover age and age related factors, and their impacts on player performance. While it’s meant to offer an insight into the fantasy universe, I believe much of it can be applied to the real world as well.
NFL Players have very short shelf lives, and this has very real implications on your fantasy team. To understand exactly what these were, I gathered some data from Pro-Football-Reference and analyzed it to see how player production varied with age. Today, we’ll walk through what’s commonly thought to be fantasy football’s most important position: running backs.
My data set included all running backs from the 1970 season through the 2013 season. I cut this down to only include running backs who averaged more than 10 touches (defined as carries + receptions) per game. Admittedly, this is an arbitrary cutoff, but I wanted to ensure that we are only looking at the fantasy relevant players; we aren’t really interested in the career progression of the Roy Helus of the world, after all.
Next, I determined each player’s production in relation to that player’s career best production. Production is defined here as average rushing yards per game plus + average receiving yards per game. Then I grouped the players by age and averaged their productivity as a percentage of their career best productivity. Taking this relative approach (rather than simply looking at absolute product) helps minimize selection bias, as the values are normalized and ensure that players are compared on an even playing field (no pun intended).
Data and Analysis
Here’s the first graph of my findings:
The first takeaway from this graph is that running backs seem to peak at age 24, and have a “prime” from ages 23 to 26. After that, the decline is sharp and consistent. Don’t worry about the odd peak at age 35. The sample size for players aged 34, 35, and 36 is extremely small (9, 3, and 2 players , respectively to be exact, while most other ages have over 50 players with qualifying seasons at that age), so I would ignore that portion of the graph. These findings are rather intriguing, as they imply that running backs peak a little bit earlier than conventional wisdom believe.
But that’s scrimmage yards. We, as fantasy players, are most interested in fantasy points, right? For the below graph, I used the same methodology as above, but instead used fantasy points per game rather than average total yards per game, where fantasy points per game = total yards per game* .1 + touchdowns per game* 6 + fumbles per game * -2 (the standard scoring system used in most leagues). Let’s see how fantasy points vary with running back age:
Nope, that’s not a mistake. The two graphs are almost identical: a quick rise to the peak performance age of 24, a two year window of steady performance, followed by a precipitous drop after age 26.
You may have noticed by now that all of this analysis is done on a per-game basis. I did this to control for injuries, but it is of course very important to explore injury risk as well. I thought initially that older players would be more likely to get hurt than younger players, and so would play fewer games:
However, there doesn’t appear to be much of a relationship between age and the number of games played for NFL running backs (correlation = -0.14). This is pretty unfortunate, as injuries are some of the most volatile and unpredictable parts of fantasy football; it’d be really useful to have some insight regarding what causes them.
What does all this mean for your fantasy team? It means to place a premium on younger players and discount the values of older players. But you probably knew that already, right?The biggest advantage you can get from this analysis is with those players who are between 27 and 29 years of age. Commonly thought to be in their physical primes, the data actually suggests that these players are actually at significant risk of decline. Yes, this includes the likes of the 29-year-old Adrian Peterson (though admittedly, the model does only apply to human beings).
For easy reference, here are the top 40 running backs based on the ESPN’s Rankings, as of 7/26/2014, and what age they will be by the end of the 2014 regular season:
|Adrian Peterson||29||Trent Richardson||23|
|LeSean McCoy||26||Chris Johnson||29|
|Jamaal Charles||28||Ray Rice||27|
|Matt Forte||29||Steven Jackson||31|
|Marshawn Lynch||28||Rashad Jennings||29|
|Eddie Lacy||23||Shane Vereen||25|
|Doug Martin||25||Joique Bell||28|
|Arian Foster||28||Stevan Ridley||25|
|Zac Stacy||23||Bishop Sankey||22|
|DeMarco Murray||26||Pierre Thomas||30|
|Le’Veon Bell||22||Knowshon Moreno||27|
|Alfred Morris||26||Toby Gerhart||27|
|Montee Ball||24||Maurice Jones-Drew||29|
|Giovani Bernard||23||Chris Ivory||26|
|Reggie Bush||29||Fred Jackson||33|
|Ben Tate||26||Danny Woodhead||29|
|Ryan Mathews||27||Darren Sproles||31|
|C.J. Spiller||27||DeAngelo Williams||31|
|Frank Gore||31||David Wilson||23|
|Andre Ellington||25||Bernard Pierce||24|
Green indicates a young player poised for improvement, yellow indicates that the player is in his prime, and red indicates the player is over the hill.
Of course, the implications of this analysis are not limited to the fantasy universe. The Chiefs just gave the 27-year-old Jamaal Charles a massive extension that will keep him with the team through the 2017 season. He’ll be 30 at the time, and given this analysis, he’ll be getting worse each year until that time. I haven’t estimated the value of a win yet, and Charles has certainly played at an elite level the past few years, but on the surface, this contract seems questionable at best.
While the curve above may apply to the population of running backs as a whole, any individual running back can easily buck the trend and have a career year at any age (I’m looking at you, 32 year old Fred Jackson!) One potential reason Jackson is having such success late in his career is that he came into the NFL when he was 26. He’s got fewer yards on him than other running backs his age. In my next post, I’ll look at some of these other factors that may explain running back performance, namely yardage, touches, and years in the league. After that, I hope to do analysis on quarterbacks and pass-catchers before moving on to other topics. I hope you enjoyed my first post; any feedback would be appreciated!