Ryan Braun’s positive test for testosterone showed a level that was extremely elevated, and likely the highest that has been recorded in Major League Baseball, according to sources with knowledge of the NL MVP’s test. – David Epstein, Sports Illustrated
Before I get too far into this I want to make a few things clear:
- MLB’s substance abuse policy is pretty air-tight, and
- Ryan Braun is very likely to get hit with a 50 game suspension.
That is the most likely outcome. You do not have a lot of recourse under the policy, and once you test positive the odds just aren’t in your favor, and it doesn’t matter if you made an innocent mistake at GNC or had Jose Canseco shoving a needle into your buttock.
That said, the situation is not totally hopeless. Look at the quote above. The widely-reported fact that Braun may have set some sort of record in his test is excellent evidence that he may get off. MLB has done a lot of testing and they have seen a huge sample of positive tests, spiked hormone levels, etc. To set a record is a special thing. A “normal” PED situation probably isn’t good enough to accomplish the feat. Braun’s test is an outlier, and outliers are very interesting things precisely because they are not ordinary.
Outliers happen, of course. In any scenario with a lot of random variation and a large enough sample you are going to get a few readings which are off the charts, but outliers are also rare, and we often use outliers to spot things outside of the norm. Think about Freakonomics and Steve Levitt looking for cheating in Sumo Wrestling (or on standardized tests). In both scenarios he established expectations of what should have happened and found data which clashed with those expectations.
Braun is only one guy, he’s not a systemic problem with lots of examples, but as a test gets more out-of-line with what we expect, the more questionable the test itself becomes. These are perhaps not the best examples because in both cases Levitt found cheating and here I’m attempting to drum up evidence against cheating, but think about it this way. In Levitt’s Sumo example the expectation is for honest matches, and he found evidence of dishonesty. In Braun’s case our expectation among cheaters is for elevated testosterone, but Braun had higher testosterone than every single cheater. His test is outside the norm for cheaters.
There’s a Simpsons episode where Professor Frink is attempting to reverse engineer the “Flaming Moe”. It goes something like this:
Frink: Brace yourselves gentlemen. According to the gas chromatograph, the secret ingredient is… Love!? Who’s been screwing with this thing?
He was expecting to find something more like cardamom or bacon or buttermilk. “Love” clearly indicates that the test itself was flawed. It is an impossible answer. The more a test result is like “love”, the more likely you are to have a bad test. A “record level” of testosterone is not as ridiculous as “love” but you get the point.
The other point I want to mention is that everyone who says that no player has ever won an appeal is wrong. The fact of the matter is that this process is supposed to be secret and it is entirely possible, and even quite likely that some players have won appeals. Testing does not always go according to plan. Lab techs are not perfect, doctors are not perfect, and those who take the samples are not perfect. Some player has probably won an appeal and managed to keep the proceedings confidential, as it should be. We even know of at least one minor leaguer who tested positive for a high testosterone level and was not suspended. It is very possible that we only know about those who lost appeals, and if your formula for figuring out that “no player has successfully appealed” has as its denominator only those players who have lost appeals, your formula is garbage.
It’s possible that Braun’s suspension is being announced as I write this, but just keep in mind that the one way that a player at this stage can win his case is if something was wrong with the test, and one of the best indicators that something was wrong with the test is an abnormal test result.