Some thoughts on Braun

So all told, it’s probably best to reserve judgment on Braun’s issue until everything is fully sorted out by MLB. The smart money says he probably used some form of banned substance, whether intentionally or not, but maybe he will be exonerated of wrong doing. There’s no harm in waiting a few weeks to bring the righteous outrage, and if the appeal fails you’ll get to deliver it with even more vigor!

That said, the media really needs to cool down on these matters. Ken Rosenthal, in particular, seems to have won the race to write the most forceful hand-wringing of PED usage in baseball based on the test, with the old “baseball will never be clean” chestnut. Obviously Ken didn’t take enough philosophy classes in college (or go to law school) or he’d be familiar with the false-positive paradox, which basically undercuts the entire premise of his column. But in general, what is supposed to be the point of this worrying? Say Braun did use a banned substance. Then…what? He got caught, he’ll be suspended, and it would be incredibly hard to mount a credible conspiracy theory about Major League Baseball covering up steroid use by their star players. What is MLB supposed to be fretting over in the big picture?

At the end of the day, the reason the steroid issue will never go away is because it’s like concentrated liquid crack to national sports writers, who just can’t resist the urge to return to the highest peaks of moral preening they reached at the end of the Steroid Era. Which is why even an isolated case involving the newly minted N.L. MVP that’s in dispute (and represents one of a relatively few cases of usage by established MLB players in the first place) can inspire this sort of hand-wringing and outrage. The fact of the matter is that, whatever the outcome of Braun’s appeal, the real lesson of this test is that baseball’s testing regime is working pretty much as it should.

The real threat is that leaks like this will sour the MLBPA on the process. Though if I had to put money on the matter, I’d wager the leak came from the WADA lab.


About Brien Jackson

Born in Southwestern Ohio and currently residing on the Chesapeake Bay, Brien is a former editor-in-chief of IIATMS who now spends most of his time sitting on his deck watching his tomatoes ripen and consuming far more MLB Network programming than is safe for one's health or sanity.

36 thoughts on “Some thoughts on Braun

  1. Brien, terrific piece! I started to write essentially the same piece this weekend, and I'm glad you wrote it instead.

    The facts as revealed so far simply do not add up. That's not unusual in the early stages of a big sports doping story. The best possible outcome is that we all take three steps back and let the participants (Braun and his camp, the lab, MLB) sort this one out. I'm reasonably certain that the ultimate story will differ from the one we have before us in important respects. No predictions, but Braun could emerge as innocent, or at least as not proven guilty.

    For those who do not know me, I write here and elsewhere about doping issues, and I know an absurd amount (for a non-scientist, that is) about how the labs test for testosterone doping. Anyone interested should post questions in the comments section. When more (and more reliable) facts emerge about this case, I'll certainly post an article here, but in the meantime I'll be happy to answer any questions that anyone might have at this point.

    • Am I being overly cynical to think that maybe Braun's appeal was making some headway, causing one of the righteous zealots at the WADA to leak the story to the press?

      • No, that's not overly cynical. My personal guess is that the sources are within MLB. But I'm just guessing.

        • It seems to me that MLB has no incentive at all to leak the result, and an individual within MLB would be fired if they're discovered to have leaked it. A source at the lab seems much more likely from a purely incentives-based standpoint, but that's obviously just pure speculation.

          • You're assuming everyone that works for MLB would put their self interests behind the interest of "the company"? I'm guessing one could get a pretty big ego trip leaking the story to the media…..

          • Well…maybe. It seems unlikely that anyone in a position to know this would need that sort of ego boost though.

          • The content of the leak makes me think that the leakers could be within MLB. The leak was very accurate about the MLB rules, and fuzzy about the lab testing. Normally I'd blame the reporters for getting the lab stuff wrong, but the ESPN reporters on this story (Quinn and Fainaru-Wada) are two of the very best when it comes to PED stories.

    • Like I said on Twitter yesterday, no one brings the righteous outrage to every possible story like Passan.

      • Yeah, I called him out on it too. But of course he didn't respond. He's too important for that.

    • Pedro Gomez is also a d-bag about steroids too. I sent he and Passan a message the other day about throwing out the steroid eye test, but of course, I got nothing in return.

      The one guy I truly love when it comes to steroids? Buster Olney.

  2. Absolutely agree, Brien. Anonymous leaking and questionable journalistic ethics have become the real blight of the PED issue. If it wants the players to stay clean and to keep making the testing policy stronger in each successive CBA, MLB really needs to get its own house in order. I'm sure Larry can speak to this in greater detail, but everything I've read suggests that false positives are an inevitable part of the process and, like the supposedly anonymous results of 2003, these players should be protected from slander until confirmations have been made.

    • Someone said yesterday that the false positive rate of the "A" test is something like 9%, but the "A" test is only a trigger. The "B" test is the one that tests for synthetic testosterone, and since no one knows what test MLB uses, we don't know the error rate.

      • Brien, correct, we don't know exactly what protocols MLB uses for drug testing. But it appears that the lab performs two tests: an initial test of the ratio of testosterone to epitestosterone (T/E ratio), and a second carbon isotope ratio (CIR) test on various testosterone metabolites. This would follow the best science known and the WADA protocols. Baseball is not tied to the WADA rules, but the lab that does the MLB testing is a WADA-certified lab, and there's no reason to think that the lab would develop different testing protocols for baseball than the one it uses for all other sports.

        The T/E ratio test is cheap and relatively easy to perform, but it's also notoriously inaccurate and is suitable only as a screening test. The CIR test is more accurate — some in the drug testing world consider it to be close to foolproof, though I personally (a non-scientist and non-expert) can make arguments against the CIR test.

        By the way, you'd be best advised not to refer to these two tests as "A" and "B". In drug testing parlance, an athlete's sample is typically divided into two parts, an "A" sample and a "B" sample, with the "B" sample tested separately if the "A" sample tests positive.

        As for not knowing the error rate: the WADA protocols for validating drug testing have a specified maximum error rate, but WADA has been criticized for not doing enough to validate a test's error rate in practice. But what Hippeaux said, EVERY lab test known to man generates a certain number of false positive results. It's inevitable.

        I'm purposely not providing as much detailed information as I could, so keep asking questions if you have them.

    • Whats ESPN supposed to do? The story is news. All they did was report the facts. Braun tested positive for a banned substance. Braun is appealing the results. No player has ever succesfully appealed the results. If the results are upheld, he will be suspended for 50 games. These are the facts. Your asking ESPN not to report news?

  3. IF braunn is guilty, should his MVP go to Matt Kemp? Let's face it, Kemp had a better season and only reason why Ryan won the award was the fact that brewers went to the postseason.

    • Well there's no precedent for that, though that seems irrelevant to me. There's no precedent for anything…until there is, and the factors would be straight forward enough to justify a re-vote if BBWAA wanted to do so, IMO.

      Though whether Kemp deserved it in the first place is immaterial, IMO.

  4. Looking at the other side, Moiuz – if it turns out that this IS just a false positive, WHAT can be done to erase what will be a cloud on Braun's rep for the rest of his career?

    It is pretty much an acknowledged fact that the dirt of an accusation sticks longer and better than any number of followup reports trying to clear the reputation. Even when the original accusation is completely without merit.

    • Well if the media reaction is clear about it being a false-positive, it shouldn't be much of an issue. The "accusations last longer than the truth" axiom is true because, in the cases of random people being accused of media frenzy crimes that don't come to fruition, the media reporting of the subsequent facts, or lack thereof, tend to be much more muted than the original accusations.

    • Jay, the sad truth is that any successful defense will probably be regarded as a "technicality" by the court of public opinion. There's no way Braun can prove he did not use PEDs, any more than you or I could prove that we don't use PEDs.

      • My view is sort of the opposite. Public opinion about issues they know nothing about tends to be shaped largely by "elite" opinion, so if the media writ large accepts the notion that the positive test was an error and matter-of-factly reports as much, most of the public likely will as well.

        • I hope you're right. What you describe is the way it turned out for Diana Taurasi, for example.

          • Admittedly I'm applying Poltical Science research here, the effect could be different with sports involved.

  5. The false-positive paradox only applies if the actual positive rate is very low; then the error rate times the negative population will be larger than the positive population. So for a disease like HIV, with a positive rate of .3% overall in the US, and lower in low-risk groups, a 5% or even 1% false-positive rate would lead to many more false than true positives. It seems to me that steroids are different, because the true positive rate could be high, like 5% or 10%, or higher. In that case even a lousy false-positive rate of 5% would mean that a majority of positive tests were real. I'm no expert, just a (painting dots on bugs) biologist, but the other issue is the source of error. False HIV positives can be largely dealt with by retesting the sample. This works if the source of the error is the test itself, not individual variation. If the source of error is natural variation (some people have different hormone ratios without PEDs), then retesting the sample won't help. Great site, just a comment, don't know anything about the actual false-positive rate or retesting protocol in MLB

    • Well my invocation of the paradox was meant for a hypothetical world in which baseball was "clean," meaning the usage rate was 0.0%. In that world, you'd still get positive test results, but 100% of them would be false positives.

    • Mick, you've got it right. But even if you assume that 10% of baseball players are using a particular PED, a test purported to be 99% accurate for that PED would result in about 1 in 10 positive results being false positives. We baseball fans would not accept PED testing if we believed that 1 out of 10 adverse analytic findings was false.

      Baseball provides for samples to be split into "A" and "B" portions, with the "B" portion tested separately at the athlete's request if the "A" sample tests positive. As you pointed out, retesting the sample should reduce the lab error rate somewhat, depending on the source of the error.

      • I'm not sure I follow your math. If we have 1000 players (yes, I know it's really 750) and 10% or 100 are users we would expect to get 100 positive tests. If the test is 99% accurate wouldn't that mean that 1 of those 100 positive tests is probably false, in other words 1 in 100 not 1 in 10?

        • No. In the test of 1000, you would get 100 "real" positives and 10 "false" positives. Therefore, 10/100 positives will be false (10%).

          • Joseph, your 100 "real" positives assumes 100% accuracy against false negatives. The true figure is much lower.

          • Yes, my understanding is that testers are happy if they get a positive result in cases where the subject was truly positive 50-75% of the time (so if there are 100 real positives, only 75 would show up, making the fraction of total positives that are false even higher). I think our sense of the meaning of the test depends on what we think is just, or worth having suspicion of a player. Some might feel that a 10% chance of an incorrect conviction means we should assume innocence (there is no trial by jury here, no resolution). Others might feel (and not even say it) that a 70% chance that someone is guilty is enough to judge them guilty. I don't even want to admit how I feel. Happy it's just about baseball.

          • Mick, no one knows the false negative rate, otherwise we'd know how many people are doping and getting away with it. But the false negative rate has to be higher than 25% or even 50%. How many times do you figure you'd have to test a doper before he's found out? 5 times? 10? 50? Marion Jones never tested positive, neither did Barry Bonds.

            Computing a false negative rate is tricky, since a doping athlete may not have any amount of PEDs in his system when he's tested. If the PEDs have cleared the doper's system at the time of the test, or if the PEDs are masked, or if they're undetectable, is that a false negative?

        • I understand your confusion. Consider that lab tests measure the accuracy against false positives separately from the accuracy against false negatives. It's commonly believed that PED testing has a relatively low accuracy against false negatives; in other words, lots of athletes are going to be able to take PEDs and pass a drug test.

          So when we're talking about 99% accuracy, we're talking shorthand about 99% accuracy against false positives.

          Next: we need to measure the accuracy of a lab test in a way that's independent from the prevalence of the condition we're trying to measure. This requirement is not easy to understand, but consider it this way. If you want to determine the accuracy of a new PED lab test, you would try out the test against two populations, one known to be using the drug and one known (or at least suspected) of being clean. That's really the only way you can measure the accuracy of a test — if you try out the test against a real world population of people that may or may not be using PEDs, how could you know whether the results you were reaching were any good?

          So when we say that a PED lab test is 99% accurate against false positives, we're saying that when the test is used against a population that has not taken the PED, the test will produce a positive finding (obviously, a false finding) 1% of the time.

          Now we get to the paradox. Let's say that we test 1000 athletes for exogenous (artificial) testosterone. Remember that our 99% accurate test is going to throw out a false positive 1% of the time against a population of clean athletes. So, if we perform the test 1000 times, we're expecting the result to be positive 10 times regardless of whether the subject tested was clean.

  6. How much of a PED must be ingested or in a person to cause a positive result. I remember reading about certain beef in South America having some substance that may cause positive results. So what would stop a zealous RedSox fan who is a server at a restaurant in NY from spiking the food of a NY Yankee (a sprinkle of female fertility drug) in the hopes of him failing a test and being suspended 50 games. Sounds stupid but is it plausible?

    • The simple answer to your question is that most PEDs have no minimum threshold for finding a doping positive. If they find the PED in an athlete's system, then the athlete is presumed guilty. But things can get more complicated than this. The tests used to detect PEDs may have built-in threshold limits to distinguish between doping and what the human body can produce naturally. Braun is suspected of having taken artificial testosterone, and of course every human alive has testosterone in their system. So for testosterone, the substance has to be measured above certain limits in order for the measurement to be a doping positive — the limits are set above what a human being can (supposedly) produce without doping.

      As for the case you pose of PED sabotage? Yes, that is possible too. The rules of the World Anti-Doping Agency (WADA) expressly acknowledge this possibility — if someone spikes your cocktail with a PED, you're not liable for punishment if you fail a doping test as a result (though you're still supposed to be disqualified from the event where you tested positive). Of course, good luck proving that! There HAVE been doping cases that have fallen under this rule — the two I know about involve athletes that (allegedly) kissed a cocaine user in a bar and subsequently tested positive for cocaine.

      The contaminated beef question you mention involved the PED clenbuterol. This is the drug that cyclist Alberto Contador tested positive for in the 2010 Tour de France (his case is still pending). I've written about the clenbuterol question if you're curious — see for example