A very cheap trick (that made a lot of people very rich)

Lies, damn lies, and statistics revisited

Just a short one today, but I want to make sure people see this if they haven’t. Bret Weinstein and Heather Heying discuss this on their most recent episode of Dark Horse, and I would recommend everyone listen to that full episode for several reasons—it’s one of their hardest-hitting episodes in recent memory. From Paul Offitt’s limited hangout to the most recent research showing significant myocardial injury in ONE IN 35 boosted individuals, it covers a lot of ground. But their discussion of the “cheap trick” analysis by Martin Neil and Norman Fenton is extra-galling.

You can go to 1:24:18 in the video for the discussion from B&H, but the essence of what is going on here is something many of us observed from the beginning and is now confirmed. In study after study—including those oft-cited “gold standard” studies showing such impressive efficacy in the NEJM and the Lancet—the authors deliberately redefine “vaccinated” and “unvaccinated” according to whatever goddamn rules they want. If including someone who is vaccinated gets in the way of your preferred outcome, you simply find a way to either recategorize them as unvaccinated or just exclude them altogether. Someone who is vaccinated is thus categorized as unvaccinated until…whenever the authors decide. 7 days, 14 days, 21 days, forever. Whatever! It’s super flexible! Science!

As Neil and Fenton discuss, this can be achieved in a variety of ways. If someone is vaccinated but you can’t verify it (again, according to your chosen verification parameter(s), which are entirely up to you as the author to define) you can just exclude them or categorize them as unvaxxed. Vaccinated and infected within the “problematic” window of 7/14/21 days are usually categorized as unvaxxed, because it’s impossible to catch Covid if you’re vaccinated. Or, you can just exclude them as well since they are so troublesome.

Turns out when you do this kind of thing, you exaggerate efficacy to the point of absolute absurdity. As B&H point out, with this kind of categorization scheme you create a statistical artifact that guarantees impressively high efficacy for ANY arbitrary intervention (tap your nose three times and wait three weeks!)

This is not just one or two or five important papers. It’s all of them. And it isn’t in nothingburger pay-to-play journals. It’s in the NEJM, the Lancet, Nature, JAMA, BMJ….this dog shit was all peer-reviewed. Many of us saw public health agencies doing this with their hospitalization data (remember all those charts showing how the hospitals were clogged with the unwashed unjabbed and this was ThE PAndEmIC oF thE UNvaXxinAteD?). But you expect that kind of blatant manipulation from public health agencies (parroted by idiot “journalists.”) You hope it doesn’t make it through peer review. But as Neil and Fenton demonstrate in their compiled list, it sure did!

And like our discussion of the Lancet’s climate sleight of hand on a recent BYG episode, it’s also not even all that subtle or clever or sophisticated. It’s just right there in our faces, censoring and demonetizing and accusing anyone who doesn’t fellate the narrative of being anti-science and spreading misinformation. It’s appalling. We are beyond hypocrisy. It’s just sheer evil.

I might be feeling a little disagreeable today.