Football Sabermetrics Are A Lie!
For a few weeks, ESPN.com has been running a series of articles about the most overrated and underrated players in the NFL at their respective positions. Not terribly original, but it's a quiet time in the NFL right now, so it's understandable.
The essential problem with these articles are their use of sabermetrics in defining who the players are. K.C. Joyner, who calls himself "The Football Scientist," is the author of these pieces and one of the first people to try and use sabermetrics in football. The thing is, it does not work in that sport.
Using metrics in baseball is completely natural. One of the most engaging aspects of baseball is the myriad of numbers and different combinations of such that creates tantalizing, unique analysis of each player. The reason this works is because of the one-on-one nature of the sport. It is always pitcher vs. batter, with fielders as the support system. This is why pitchers get wins and losses next to their names.
Football is a completely different beast. It is the quintessential team game, twenty-two players on the field, eleven on each side, and every moving part must be in sync with their assignments to achieve success. This means that using sabermetrics in football must correlate with watching game film, and Joyner admitted as much on his website by saying, "The Football Scientist bridges the gaps ... by combining stat research with game film analysis."
Unfortunately, the use of sabermetrics was designed with the idea that players and teams could be evaluated objectively, strictly using numbers to judge performance, which is virtually impossible in football. For example, here are some of the categories Joyner uses:
- Weighted bad decision percentage
- Bad decisions
- Total Yards Per Catchable Attempt
- Catchable pass
- Near interception
- Not at fault interceptions
And while not all of his metrics use personal judgements as factors in the numbers, you can see that many do.
The question remains, who gets to make the assessment on what is or is not a bad decision or a catchable pass? It's strictly in the eye of the beholder, which is brazenly anit-sabermetric. It defeats the purpose of using metrics. Really, this is just one person's opinion being passed off as statistical fact. It is misleading to the general public, who does not understand these numbers are not based on factual data, they are based on one person's opinions.
Developing new statistical analysis in football is something that should certainly be explored to try and broaden the knowledge of the game for coaches, player and fans. But it has to be done right, and done objectively.
2 Comments:
Some people are doing it right: check out Football Outsiders. They do rigid statistical analysis, but in the explanations recognize that their analysis does not separate individual players from their team contexts.
Baseball is a lot easier--the team doesn't have a great effect on individual performances. But you can still do the thorough metrics for football: you just have to recognize that you are analyzing the player as part of a team more than you are in baseball.
Yeah, Football Outsiders does good work, but even they admit that there is still issues using sabermetrics.
By the way, I like your site. It's good reading.
Post a Comment