I’m reading Superfreakonomics (which is infuriating in that the assumptions the authors make in order to justify their arguments are leaps of faith off very shaky platforms), and I started thinking about the 99% accurate test that they put forward.
If a test is 99% accurate, most people would agree that’s really, really good. If any of the kids I tutored got 99% they certainly wouldn’t need me to tutor them.
But consider a test that is 99% accurate. In the book they imagine a test that is 99% accurate in identifying terrorists. But, what about a test that is 99% accurate in identifying dopers.
Out of 100 people that are dopers, it would correctly identify 99 of them as dopers. Out of 100 people that aren’t dopers, it would incorrectly identify 1 of them as a doper. But what if that 1% error is out of lots and lots of innocent people. Out of 100,000 non-dopers, it would incorrectly identify 1,000 as dopers.
For identifying terrorists this doesn’t really work, obviously, because most, most people are not terrorists, so you get an overwhelming number of people wrongly identified, which means for any one person identified as a terrorist there’s a pretty good chance they’re not.
Whether or not it works for doping kind of depends in part on how prevalent you think it is, but also on what we’re doing with all these false positives. Since the terrorist test is imaginary I wasn’t really interested in it. But, doping tests are not imaginary. They are real and they’re pretty high stakes and they are not 100% accurate.
It depends obviously on the particular test and newer, better ones are being developed all the time, but even if they have 99.9% accuracy those tests are incorrectly identifying some people as having used performance-enhancing drugs. And when you multiply that by the hundreds, if not the thousands, of tests that professional (and now, amateur) athletes have to take that’s a scary number of false positives.
And, that’s assuming no human error. [Kind of like how condom use’s effectiveness in the real world is actually much lower than 98%.]
And, that’s not even talking about the athletes who get positive doping tests because they fail to fill out the right forms to get their asthma medication approved or didn’t know certain medicines on a long list of scientific names would trigger the tests. (On the email I got about banned substances, it listed a bunch of things that are banned “in competition,” but not out of. It took me a while to figure it out, but then it became clear you’re not allowed to be on Sudefed, Nyquil or Tylenol during a race, which means it has to clear your system before the race. How long does it take to clear your system? I really don’t know. Maybe just don’t take it 3-4 weeks before a race. But you’re racing every other weekend, so, just don’t get sick.)
And, we’re not even mentioning the fact that the World Anti-Doping Agency’s online tracking system is supposedly impossible to understand.
WADA is under fire – and rightly so – for an ill conceived system that now forces athletes to provide three months’ notice of their location an an hour each day for seven days a week between 6 a.m. and 11 p.m. for testing.
Even ignoring all of those issues (because WADA and USADA certainly are), I’m still concerned about the 1% error. Or .1% error. Because .1% out of 100,000 tests is still 100 people.
So, let’s assume, of course there are mistakes in drug testing. Of course. There’s technician error. There’s biological differences between people that would mean a test may fall outside the range of “normal.” There’s people who have filled out the forms wrong or didn’t mean to take performance-enhancing drugs with their over-the-counter medication. And, then there’s downright false positives. So, sure, of course, mistakes happen.
Except, according to USADA, they pretty much never make mistakes.
Which is why it’s totally ok that the high-stakes process operates completely outside of the confines of the pesky legal system. Who needs trials and evidence and defenders when you have “arbitration.” Why give the athletes access to legal assistance and their own test results; that seems so unnecessary.
It really shouldn’t be a surprise that few athletes appeal a sanction once it’s handed down and that on appeal the USADA seldom loses. USADA didn’t lose its first appeal until 2007 and that was only after a law professor and four law students worked pro bono on the case.
(USADA also shortened a hammer-thrower’s suspension earlier this week because she had taken a diuretic after being horribly depressed. It was a “rare loss.”)
So, if you experience human error in your test or make a mistake in the bureaucracy or are simply one of the false positives, then you are faced with an overwhelmingly biased, complicated, and opaque arbitration system, which makes it nearly impossible to prove your innocence and which almost never finds in your — the athlete’s — favor.
I would have more faith in a system that admitted it makes mistakes. Justice that claims to be all-knowing is seldom just or all-knowing.