Re: Re:
WillemS said:
carton said:
(if Froome is currently doping, the bayesian odds for Dumoulin doing what he did clean are next to nil).
Care to elaborate on that? Not the theoretical background of Bayesian statistics, but how you're able to update your posterior probabilities with such determination based on a single performance comparison. I would say that beating a doper might make it more likely that you dope, increasing the odds ratio in the favour of doping, but a "next to nil" estimate seems pretty far fetched based on this piece of introduced evidence alone. The only thing you're doing is
rationalizing a gut feeling with scientific terms.
I think most opinions on internet forums are rationalizations on the gut feelings that compel us to blab to strangers about the minutiae of topics that we're not paid to research or discuss.
But in any case, I admit it was a hastily written, hastily arrived-at conclusion. I have not actually managed to even postulate a proper hypothesis in such a way as to actually be able use Bayesian inference as it is commonly understood to determine anything about Dumoulin's performance. In any case such an estimation would likely rely on assumptions so strong as to render the results suspect (because then again who actually was a clean cyclist, after all). However, what I actually meant to say and what I will still assert is that however you want to go about calculating it the
conditional probability that a rider of his characteristics could manage these kinds of performances clean given an otherwise dirty peloton is IMHO very likely to be minute. As always, YMMV (but if it does, and you think you can adequately support it, I'd be very interested in how).
Just to double down on it, his was at the very least by far the most
remarkable ride today from someone who doesn't ride for one of the two teams the UCI has tried to delicense out of the sport for systematic doping. To be clear, I don't know if Dumoulin is doping. Also, (and I've been ridiculed for this ad nauseum on this forum) I don't know if Froome is doping. Going into this GT, I found it much more likely that Dumoulin was clean than Froome. Given what has transpired, however, I find it very hard to imagine how the Dutchman could be doing this clean while guys he left cracked like peanuts like Froome and Quintana (who climbed better than Froome did in the Tour) were doping. I really don't know how you can rationally believe one and not the other.