A Certain Uncertainty

by Rant on March 2, 2007 · 7 comments

in Doping in Sports, Floyd Landis, Tour de France

Ever skipped voting because you thought your vote wouldn’t matter? Imagine, the 2000 US Presidential election was decided by just one vote. Most likely Sandra Day O’Connor’s, as she was often the swing vote in close cases during her tenure on the Supreme Court. In the end, their 5-4 vote decided the election for George W. Bush. Imagine if she’d called in sick that day. What a mess that would’ve caused.

Some time after Floyd Landis’ arbitration hearings end in May, just one vote is going to decide the first part of his case. It’s most likely to be the one cast by Patrice Brunet, a lawyer from Montreal. Reason being, he was chosen by the arbitrator picked by Landis (Christopher Campbell, a San Francisco lawyer) and the arbitrator picked by USADA (Richard McLaren, a lawyer from London, Ontario). It’s a pretty easy guess that the warring sides would pick someone most likely to vote in their favor, so I’ll give each of those votes to the side who picked `em.

So what are the likely issues that the case will turn on? Among the items I think will be decisive:

  • Whether or not the A sample test, especially the T/E screening test really showed a positive result,
  • Whether the carbon isotope ratio test performed on the A sample really shows the presence of synthetic testosterone,
  • Whether the B sample was degraded or contaminated (and therefore should not have been tested), and
  • Whether the B sample carbon isotope ratio test confirms that synthetic testosterone was present

It’s important to note that one thing that won’t be an issue of contention is whether the science behind the tests is correct. The way this process is structured, the science behind the tests is deemed to be correct, so that is not an angle of attack for the defense.

And it’s also important to note, despite what Travis Tygart might say to the contrary, the system requires the accused to prove his innocence. It does not require the prosecution to prove he’s guilty. The operating assumption is that if you’re charged, you must be guilty — unless you can prove why you aren’t.

This is exactly the reverse of what we are used to here in the US, and many view this approach as patently unfair. So that’s why Tygart tries to spin the innocent until proven guilty yarn. It sounds much more palatable to most people, and when he wins cases it sounds emminently fair. Except, the system is already rigged in his favor. Fair it’s not.

Other issues will probably come into play, especially as they impact the lab’s reported results. So, for instance, if the lab didn’t follow protocol when conducting a test, the argument could be made that the test’s results are invalid. That can look like getting off on a technicality, assuming the case is decided on that basis, but following established protocols and procedures is necessary in order to ensure the integrity of a lab’s results.

Chain of custody will very likely be an issue, if the defense can convince Mr. Brunet that there is enough doubt about whose sample was being tested or whether the sample was properly secured or handled.

But whether these other issues become deciding factors is hard to say. The data and its interpretation are fundamental to the case. If the data doesn’t support the original finding that led to the B sample test, or the B sample data can be shown to be invalid, the case falls apart.

And right now, given all the problems at LNDD coming to the fore, it appears USADA’s case may be falling apart. But that’s not a guarantee. Nothing is guaranteed until the arbitration panel makes their ruling, and until the CAS backs that ruling or overturns that ruling sometime late this year or early next.

Another thing that needs to be kept in mind is that this is an arbitration hearing, not a court case. And while precedents set in other court cases may be applied to an on-going case in the judicial system, a determination by one arbitration panel doesn’t have to be considered in a case being decided by another. So, the Landaluze case certainly seems to apply if Team Landis can show that lab technicians involved in analyzing the A sample were also involved in the B sample. However, it’s not a foregone conclusion that the arbitration panel will reach the same conclusion as the CAS panel did in the Landaluze case.

Two areas of uncertainty have been discussed at length lately. One is how many metabolites with a positive result does it take to have a positive finding for synthetic testosterone. Unfortunately, the answer is vague, at best. The World Anti-Doping Code refers to “metabolites” while the WADA technical document for the testing refers to “metabolite(s).” Nothing like clarity in language, eh?

Why this is important is both simple and complex. The carbon isotope test looks at four metabolites. In Landis’ case, two were definitely below the threshold. And one was significantly above the threshold. Another was above the threshold, but if you take the margin of error into account, it was lower than the threshold value plus the margin of error.

And this is where the other question of uncertainty comes in. Is the margin of error already factored in to the value that must be exceeded for a metabolite to be positive? Or can/should one take into account the margin of error before declaring a result positive?

How this applies to the “marginal” metabolite is this: If you go to the lower end of the range of possible values (measured value minus the margin of error), the result would not have been a positive reading. So one question is whether that particular reading should be counted as a positive or not. If I were looking at the data, I wouldn’t count it as a positive. Many people have debated whether it should or shouldn’t be. But the only person whose opinion will ultimately matter is Mr. Brunet.

If you’re interested in a good discussion of margin of error (also known as uncertainty), Trust But Verify has an article that covers the concept of uncertainty as it applies to the data in the Landis case.

Circling back to the question of whether one metabolite can be enough for a positive result: If only one metabolite turning up positive is enough, then Team Landis has to find a way to explain the reading that was elevated beyond the margin of error. This is going to be tough. They’re going to have to show that either the test was performed improperly, or the data has been misinterpreted. I’m not sure how they can do the latter, but perhaps that’s one of the cards their holding close to their chests.

One thing perhaps working in their favor is the fact that different labs use different standards.

At least two anti-doping labs — at UCLA and in Australia — require more than one metabolite to be positive in order to declare the test a positive result. LNDD, on the other hand, appears to go by the standard that any metabolite coming up positive is enough for the result to be declared a positive for synthetic testosterone. Again, it’s going to be up to Mr. Brunet’s vote on how that issue gets settled.

With all the issues that have been raised, only one thing is really certain about the Landis case. As much as we might like to see a particular outcome, the outcome is far from certain.

Will March 2, 2007 at 1:35 pm

What possible defense(s) can Floyd argue using his samples taken after stage 17?

I wonder how many randomly chosen people out of 1000 would turn up positive on this test using the one metabolite benchmark.

ORG March 2, 2007 at 1:56 pm

Better question, if one metabolite was the standard, how many false positives would you get? I see to remember Tom Fine do an analysis and coming up with 30%. If it was two 7% and three was less than 1%.

Do I have this right?

randy March 2, 2007 at 4:04 pm

Thanks for the clarifications, very helpful.

You seem to assume that this case will necessarily go on to the CAS in Switzerland. Is this necessarily the situtation? Could it possibly end with the USADA hearing? I assume if Landis “loses” at the USADA they will appeal (?is this the right word) to CAS. But if Landis “wins” at the USADA who can appeal and force them to the CAS? Of course, they still have to deal with the French…..

Rant March 2, 2007 at 4:11 pm

Randy,

USADA can appeal to the CAS, WADA can appeal to the CAS, and the UCI can appeal to the CAS. My assumption that it will automatically get appealed goes like this: If Landis loses, he’s going to take it to the CAS to try and clear his name. If USADA loses, one or all of the other players will go to the CAS to overturn the judgment, because they’ve invested too much political and personal capital in it to allow the judgment to stand. Sure, USADA/WADA/UCI might throw in the towel if Landis wins, but consider the personalities involved: Travis Tygart, Dick Pound, Pat McQuaid. I can’t imagine any one of them letting it go at after an initial Landis win.

– Rant

randy March 2, 2007 at 4:46 pm

And they call Economics the “Dismal Science”!

swimyouidiot March 4, 2007 at 5:33 am

Rant, on this “guilty until proven innocent” issue, we are all used to thinking in terms of a grand jury indicting someone. If they decide there are grounds for indicting someone, then when the case goes to trial, everythig is back to even. But here, the test is not a like grand jury. Everything doesn’t go back to even. That seems to me to be a major fault in the system, at least to my American way of thinking.

Given the fact that character, reputation, career and in some cases (like FL’s) big money are all at stake, it really seems the test ought to operate like grand jury that sets the innocent until proven guilty process in motion.

To me, the question is, did the athletes ever agree to this system, or was it simply imposed? Athletes have to believe the system is fair or it is never going to work.

Rant March 4, 2007 at 8:29 pm

Swim,

My impression is that the system was imposed on the athletes after the UCI signed on to the World Anti-Doping Code. So whether they believe in it or not, they’re stuck with it now.

– Rant

Previous post:

Next post: