Steroids and Batting Averages

Light blogging this week as I'm on vacation, but I had a chance to run through some numbers based on last week's Mitchell report. Now that we have a better sense of who used steroids (instead of identifying Bruce Banner to Hulk transformations), let's see if there are any differences between the players named in the Mitchell report and the average ballplayer.

The blue line in the chart above shows the batting averages -- by age -- for the suspected players. The gray line shows the same for ballplayers who played between 1985 and 2006. Both lines only take into account players that had at least 100 at bats in a season. There were not enough players mentioned in the Mitchell report who had at least 100 ABs after the age of 37 for a meaningful comparison with the average ballplayer.

The obvious takeaway is that if the players named last week did use steroids, then they received a significant boost in their performance on the order of 7 extra points per year in their batting average over the 23 to 37 age period.

But it's also worth mentioning that the stars-to-duds ratio in the Mitchell report seems to be higher than you would typically find in baseball, so some of the apparent steroid-effect could be more related to selection bias.