https://www.youtube.com/watch?v=L2eU96vCCFU



Dissects it with a wooden spoon...

Mentions that the plural of anecdote isn't data. Except it sort of is, because a single anecdote is data. Just not data you can generalize from--depending on the strength of the observation...

That time doctors injected insulin into a juvenile diabetic for the first time and the kid recovered--that's an anecdote. What sets it aside from bigfoot sightings is the quality of the evidence, not whether it's an anecdote.

He calls Mikhaila Peterson's experience implausible. Mm. I find her explanations for here experience a little implausible--when she gets too specific. Lectins, oxalates, effects on the gut microbiome. Hard to pin down that oxalates in spinach are doing any damage unless you do a trial of just oxalates. So getting specific, that's sketchy. The experience of distress or pain if spinach is taken is another thing, that could be repeatable. I personally had shoulder pain for years. When I went keto it went way down, when I did a trial of carnivore--with most of my carbs coming from heavy cream--it went away. Oxalates? Heck if I know. Plants contain countless compounds. There might be some we've narrowed down to likely candidates, but who knows? Low plant matter might just have me more consistently ketogenic--but being ketogenic might have just lowered the plant matter in my diet.

He mentions that there's no placebo for carnivore, since you know you're eating meat. This is disingenuous, and that's the nice way to put it, since it covers just about every possible diet, at least any that involves eating whole foods.

And he brings up the massive body of epidemiology. How many anecdotes make for data? Once you get up into the thousands, does it start counting?

Points out that every people in the world has had a varied, omnivorous diet. Which just isn't true--or not true enough to make his point. Long plant-less winters are a thing for some ancestral diets.

Says the people who show up in hospital with carnivory-induced scurvy aren't included in the anecdotes. Here's a place where an anecdote definitely would be data. Vegans sometimes show up with deficiencies. Personally I wouldn't do carnivore without some supplementation, not because I think it's necessary, but because I'm not out to prove that carnivory is self-sufficient, I'm willing to ruin my n=1 on this particular point with a multi-vitamin. The only anecdote I've seen that looked something like scurvy was a zero carber named Danny Roddy, some years back he did a stint on some pemmican--the meat only version. He described it as a little overcooked, or so my memory, that I can't fully trust of course, tells me. He developed some symptoms that sounded scurvy-like to me at the time, dark spots on his legs, etc. Maybe there aren't enough people experimenting with carnivore for a whole lot of anecdotes to be out there. But they should really be very common, at least if you agree with this doctor. When Steffanson was experimented on, they pretty much thought it would be universal. That study only gave two subjects, but that's enough to show that the need for plants to prevent scurvy is at the very least not universal.

He brings up self-selection bias... yes, but...

Here's the thing with data. Bob stopped eating wheat. He stopped suffering. Turns out Bob was celiac. His anecdote is useful to some subset of people suffering what he suffered. When they stop eating wheat, their results are spectacular. Some people may try what Bob tried, and it might work for other reasons--like, the weekly birthday cake at the office, they don't succumb to it anymore, they don't stop for a donut every day before work, etc. Some of the people who think they're benefiting from avoiding gluten might be kidding themselves about their response to the gluten itself, yes. But is it implausible, for somebody who has benefits, that it might be the gluten? Absolutely not--it's just not established. Big difference.