Saturday, November 18, 2006

The Meaning of Statistics

La di da, look at me, a Saturday night blog! I'm sitting here munching on bread at the Panera on White Bear taking advantage of free Wi Fi. If my shirt said "I'm blogging this" rather than "Pacifist Viking" (yep, it really does say that) I would order a latte and say, "La di da, look at me" out loud.

I love sports statistics. I enjoy studying them and I think they can help us to understand the events on the field.

However, last week I got into a debate at Football Outsiders because I expressed skepticism over the ability of a statistical system to comprehensively explain reality.

Here's the simplest way I can explain my position:

I accept math as a way to make sense of reality; I do not accept math as the inherent truth of reality.

My background is in literary theory, not mathematics. Studying literary theory has made me extremely open to different approaches to making sense of reality, while making me very skeptical of claims to exclusive method of perceiving reality correctly. In grad school, in a paper on Marxist literary theory, I wrote about why I find Marxist criticism limiting, while recognizing reasons I might think this way. My professor wrote a comment about this point that has stuck with me: "recognition that this is A POSITION and not THE TRUTH is an important act of intellectual honesty."

This also takes me back to a long debate I held with my freshman roommate about whether quality could be ranked in an objective "cosmic" sense (to determine, objectively, whether one girl is prettier than another, whether one song is "better" than another, etc.) as he argued, or whether such distinctions of quality have no objective criteria and that quality is determined almost solely on subjectivity and taste, as I argued. This has resonance of some old philosophical debates. The point is I've always had difficulty accepting claims that a single inherent truth has been found or revealed, always recognized the limitations of such claims, and always believed that multiple methods and perspectives are possible for making sense of reality.

I think nearly any position is tenable, AS LONG AS one recognizes it as a position and not THE truth. So I become skeptical when one claims that a particular statistical system can give us THE TRUTH about events in the NFL. If a statistical system is a tool to help us perceive reality, then I am all for it. If a statistical system attempts to define reality in a singular way, then I have little choice but to reject it.