Roger Clemens’s agent released an 18,000 word report that presents and analyzes a number of statistics (OK, a LOT of statistics) based on Roger Clemens’s 24-year career. The purpose of the report is to “prove” that Roger Clemens did not use performance enhancing drugs. I didn’t have time to read the entire report, but the gist of the “proof” is that all of his abnormally good statistics are just random deviations that we would expect to see. Although feigning randomness is a legitimate defense, I don’t buy it, and neither does anyone else. Tim Keown of ESPN concurs, writing that
The bulk of the report is a skull-crushing dissection of nearly every start the man ever made. It is placed in the context of run support and other factors I think are supposed to make you believe Clemens is … not a steroid user… Comparisons with Curt Schilling, Randy Johnson and Nolan Ryan are central to the report. They all had abnormal strikeout seasons at ages when even the greatest pitchers are retired, which proves … what, exactly?
This is the most mathematical defense against using steroids that I have seen, but I’m not impressed. However, I’m not one of those people that analyzes baseball statistics, so I could be wrong.
Read better sports and statistics applications in The Journal of Quantitative Analysis in Sports.