casual obsever wrote:
rekrunner wrote:
The reason all those improbabilities combined are underwhelming is because these improbabilities are the wrong ones to argue against a likelihood that the source of a nandrolone was ingestion of the bad pig parts.
Not at all - that's exactly what this doping case, or rather Shelburrito's ridiculo excuse, is all about.
Au contraire. The AIU argued the probability, given a burrito, it could produce a positive. We are interested in the probability, given a positive, it could come from a burrito. These are not exactly the same probabilities. One can seem overwhelming, while leaving the other underwhelming. One probability can be 1 in 10 million, while the other can be 1 in 10, without contradiction.
The AIU essentially “proved” that most burritos would cause an athlete to test negative, but some would test positive. The CAS agreed with such a possibility, but that the likelihood would be small. 121 million pigs slaughtered per year give many opportunities for this unlikely event to occur much more frequently than is intuitive. Supply issues during the pandemic, conceded by the AIU expert, created conditions for this specific positive test to become increasingly likely.
As one of the givens is a positive test result, and we are interested in determining the source of the positive, Houlihan should not be compared with the much larger population of burrito eaters who would test negative, but rather among the population of all athletes who would test positive, from all sources. An ideal evaluation would compare all the possible sources of a positive test against each other, with likelihoods adjusted and uncertainties reduced based on complete evidence.
In Houlihan’s case, the evidence is incomplete, as the late notice, 1 month after the event, did not permit preserving the primary evidence she needed, and the likelihood of other alternatives are simply not required to be considered. For example, we simply cannot determine whether Prof. Ayotte’s alternative suggested source is more likely, however plausible it may seem to “experts” and “fans” alike, because the adjudication process itself does not require collecting the necessary evidence to permit such a determination. All we can really say is that, with limited evidence, Houlihan could not “prove” that the burrito was the source to a CAS panel on the balance of probability, and therefore doesn’t qualify for any reduction or vacating of the sanction of the guilty verdict.