Ryan Braun and Type I vs. Type II Error
I’ll echo Dave Cameron and start by saying I do not know if Ryan Braun cheated. What we do know is that he will not be facing a suspension based on his October 2011 drug test. The independent arbitrator determined that the irregularities in the process were serious enough to warrant tossing out the apparently positive test. It is worth noting that the arbitrator did not declare Braun “innocent,” rather he simply refused to uphold the “guilty” result. In social science terms, the arbitrator decided the risk of making a Type I error was greater than a Type II error. A Type I error occurs when a null hypothesis –- in this case that Braun was clean –- is rejected despite being true. The flipside is a Type II error where a null hypothesis is maintained –- again Braun is clean –- when rejection of the null is warranted.
The fact that the arbitrator decided to potentially commit a Type II error is certainly good news for Braun and the Milwaukee Brewers, but is this good for Major League Baseball? I would argue yes. Our society has a long history of preferring Type II errors to Type I errors. The best example is our criminal justice system. Defendants are assumed innocent until proven guilty. A defendant does not have to prove that he or she is innocent of the crime he or she has been charged with, he or she simply has to raise enough reasonable doubt to prevent the state from proving that they did in fact commit the crime. This bias towards Type II errors is often controversial, as there are cases where many people believe that a guilty defendant was freed through the trial process (i.e. e.g. Casey Anthony, O.J. Simpson), but our society still supports a system that attempts to minimize the extent to which innocent individuals are falsely convicted.

