Walshy Ball

Robert's got a new blog, Shot Clock. It's all about sports — namely looking for statistical models one could use to predict outcomes, similar to Michael Lewis's Money Ball. So you know I'm in.

In particular, Walsh and I were talking a couple weeks ago about a possible model for basketball (what with all of our extensive knowledge of methodology). Does anyone know of one? For my final project in my high school statistics class I actually did a simple regression analysis that tried to correlate an NBA player's performance with his salary. I couldn't find one. I'm sure you're all devestated.