The Econophysics Blog

This blog is dedicated to exploring the application of quantiative tools from mathematics, physics, and other natural sciences to issues in finance, economics, and the social sciences. The focus of this blog will be on tools, methodology, and logic. This blog will also occasionally delve into philosophical issues surrounding quantitative finance and quantitative social science.

Wednesday, August 30, 2006

Footballmetrics

In less than a week, another season of professional (American) will begin. Football fans (me included) -- especially devotees of 'fantasy football' -- are no doubt pouring over the statistics of their favorite players. Ironically, from a statistical standpoint, this devotion to numbers may be misleading.

For example, writing in the 'Keeping Score' column in the New York Times Sports section(Numbers Often Lie When It Comes to Football, Aug. 27, 2006), economist Martin B. Schmidt argues that individual stats are so interlinked with factors beyond the inherent athletic prowess of the individual player (e.g., the individual stats are also a function of team effort, good or bad coaching, the quality of the opponents, etc.) that they are highly misleading. A good example of this is the high variance in a quaterback's passer ratings from season to season:

How much of a quarterback’s rating is predictable? Not much. If we look at six seasons of quarterbacks who threw at least 224 passes in successive seasons — the minimum to be ranked in the N.F.L.’s quarterback rating — a past rating has a poor predictive value.

Take Peyton Manning. He produced a quarterback rating of 104.1 last season. This season, based on the six seasons of ratings, there is a 95 percent chance that his rating will be between 73 and 111. In other words, he is going to be either pretty bad or pretty good.

Some innovative thinkers have tried to come up with alternatives to the currently popular statistics used by football fans, sports broadcasters, and the teams themselves. Football Outsiders is a website devoted to applying Moneyball style statistical analysis and thinking to football (Aaron Schatz, the founder of Football Outsiders, also writes a book of football statistics and analysis called Pro Football Prospectus 2006: Statistics, Analysis, and Insight for the Information Age.)

My opinion on this is that, at the very least, improved football (and sports) statistics must incorporate team effects as opposed to crediting (or blaming) the individual for too much of the success (or lack of success).

BTW, although I avoid making predictions about the stock market or the economy on this blog, I will make a prediction about a far more important issue ... football. My guess is that the Miami Dolphins will do relatively well this season (at least a playoff appearance). I should note that I am NOT a Miami Dolphin fan nor have I any ties to that region. Why am I making this bold prediction? Nick Saban, the coach of the Dolphins, is a disciple of , the highly successful coach of the New England Patriots (Belichick, an economics major in college, is a supporter of applying quantitative thinking in football), and Saban will be into his second year at Miami. The Dolphins also have a new quarterback, Daunter Culpepper, who brings a lot to the table in terms of his skills as a passer and on-the-field leader of the offense. Add all of that to a defense that has been consistently good even during the lean years, and it makes for a compelling case to 'go long' (so to speak) on the Dolphins.

So if the Dolphins do well this year, you heard it here first!

PS: October 24, 2006 -- I don't know if dolphins eat crow, but I am over picking them to have success this season!!! ;)