Data Rules Baseball

Any person familiar with American sports will have noticed the claims by the Mitchell Report stating that a number of baseball athletes did take performance enhancing drugs some years back. These drugs are the Human Growth Hormone and steroids that are banned by the Major League Baseball. From the perspective of economics, it is easier to understand the incentive for an athlete to take banned substances weighed against the possibility of capture and suspension. Financial rewards that come even with a moderately successful sports career in the MLB today is substantially high in my view to tempt a number of athletes into taking that risk.

One of the athletes that has been named in the report responded by appointing statisticians to supply statistical evidence showing that the accusations in the report do not apply to him. This NYT piece put together by a group of statisticians and the Economist Justin Wolfers have looked at the data and drawn very interesting conclusions. It makes for an interesting read but is clear that there was clearly selection bias in the data and the metrics used in exonerating the athlete. In addition, it states that Roger Clemens had a very unique late career when considered against the historical record of peers. To their credit, they state that evidence of drug use cannot be gleaned from an examination of the data but that it is clear that Roger Clemens had a very unique career history and that the reason cited in the report in defense of his record is not supported by any data. However, the argument remains open and Ray Clemens may yet provide the exculpatory data and analysis but I do not bet on it. Teachers of economic theory and emprirical economics may cite this case to make classes fun while demonstrating the value of all that data crunching.