A Look at Floyd's Case

Hm, I sense another high traffic day coming on. This is getting interesting in a hurry...

I've downloaded the slide show offered by Landis on his website and parsed a few arguments from it. At least from a cursory review of the slideshow, which itself is a cursory review of his case, it looks like he cites the lab for a seemingly unending chain of typos concerning the ID number of the subject rider, making one wonder who exactly was being tested. These aren't to be lightly dismissed, but it seems like there are three bigger points to his case.

On the flip...

First point: The sample was contaminated and therefore unreliable.

Remember: More than 5% means the specimen is contaminated or degraded and should not be used. Just like food with mold or maggots, such a sample should not be used. The table shows the math: 7.7% degraded epitestosterone. According to WADA protocol, since the epitestosterone level exceeds 5% (it is 7.7%) the specimen should not have been evaluated for an adverse analytic finding. It should have stopped here.

Pretty straightforward. Sounds like a decent case right there, at least until we hear from the prosecution.

Point two: there is a huge disparity in test results.

When the sample was screened for T:E ratio, the calculated ratio was 4.4. When the sample was tested to confirm the ratio, it was 11.8.

Hm, 4.4 is itself a problem, I seem to recall, so I'm not sure they want to pin their hopes on this point too much, but the threefold increase suggests some serious inconsistency in what (we're led to assume) should be a very consistent process.

Point three: the Carbon radioisotope test was applied four times, with three negative results.

Considering the criteria for positive (3.0) and stated accuracy of the lab (±0.8) isotope absolute values must be higher than 3.8. Only one of Floyd's four breakdown products examined even arguably met the criteria to determine a positive result.

OK, so here a 3.0 or higher is a positive test, with 0.8 variable, making 3.8 or higher a clear positive. Floyd's four results were 2.02, 3.51, 2.65, and 6.39. The next slide explains that for some reason the 2.65 result was from the application that should be the most reliable. Making the score one positive, one maybe that looks bad, one maybe that looks good (if the variable is 0.8, then you have to score under 2.2 for a definitive negative), and one clear negative. Now, Jacobs' statement that only one result "even arguably" is positive is overblown -- arguably three of them raise concerns, but only one of them is inarguably positive. Conclusion:

[chirp... chirp...]

For all I know the arguments I skipped over here (e.g. the low testosterone conundrum) are actually more powerful, but these three excerpts are fairly simple cases drawn from empirical data, all of which suggest the case against Landis is somewhere between foggy and complete bullshit. I hate to get sucked in by an accused rider, but this is a far cry from a "disappearing twin" hypothesis. I had wanted to tune out the Landis litigation, but that's not possible anymore.