Sunday, January 07, 2007

Pro Football Analysis Methods

While I can't seem to find an electronic version of the article, I'd point you to an interesting article in the January 8th New Yorker magazine on the current state of Professional Football, written by Adam Gopnik, titled The Unbeautiful Game.

The article is mostly a resume of the current culture of Pro Football, the comings and goings of the players, the lack of 'guaranteed money', and the league's strict policy on dress code and uniforms - flowing all the way down to the dress that the coaches wear on the sidelines. None of that is really Earth shattering stuff, but interesting.

There is an interesting section in the article about football talent evaluation, and lamenting the fact that there hasn't really been a 'Bill James of football' yet. The author makes some interesting arguments about the truly team nature of football vs. the greater role of the individual contributions in baseball.

I would look for example at the fact that while there are parallels in football, the notion of the game winning home run or complete game shutout are not as strong in football. Sure a player can return a kickoff for a touchdown, or break a run for 80 yards, but there is a full team's worth of things that has to go right for that to happen - blocks for example - where in baseball, the one-on-one match up between a pitcher and a batter is a more discrete, self contained event. (Granted the positioning of the fielders may effect that match up, but I would argue that connection is not as strong as the blocking of an offensive line and full back to support a half back.)

You could argue that in football, if you look at each play from scrimmage as an 'at bat' that forms the basis for an analysis method. However the objective of the offense plays a very strong role in what an individual play from scrimmage is going to look like, with the most obvious factor being whether the team is ahead (ball possession paramount - run) or if the team is behind (move the sticks, improve scoring position - pass). This is also true in baseball - a batter may be asked to move a runner along, or swing for the fences - but these 'specialized' at bats seem to be fewer and further between then a football team 'running the ball for the second half'.

I would argue that these factors are a major reason as to why a good predictive analysis of football is yet to emerge. But the debate in any case is an interesting one, as is the article.

If anyone out there finds an online version of the article, feel free to post a link here. I will do the same if it is released once the January 8th New Yorker is off the News Stands.

No comments: