https://www.medrxiv.org/content/10.1101/2020.04.24.20078717v1
I call Bullshit. They said statistically significant disagreement. Here’s the website into which you can plug your numbers and prove them wrong. I don’t know who is right but you certainly haven’t proven your case by sighting numbers that clearly have the “% move off a low base” problem. So here it is: https://www.easycalculation.com/statistics/statistical-significance.php The doctors didn’t present their evidence but you didn’t argue against their actual argument. You get an F.
I’ve read their denouncement. Its exactly that. I’m tired of argumentation by authority. I listened to the whole thing 48 hours before they pulled it and it doesn’t argue anything they say it argues. This whole thing is psycho-drama.
I had a reaction that was in the ballpark of a Gell-Mann shock: I listened to the whole thing and came away thinking that they were making one important claim and backing it with the relationship between their data and the serology studies. The claim that’s important is that there’s no statistical significance to the variance in deaths across mitigation regimes. That is a very specific claim I have seen nobody attempt to argue. Its a formula. There’s a webpage to plug in your numbers. Let’s see some numbers there, please, or this is just psycho drama. They discuss their large sample size and illness trends and state they believe the data is stable enough to multiply by the range of serological estimates to come up with a decent guess at infection fatality rate. When I saw people freaking out that they were committing a logic and category error I was astounded.
Iowa isn’t anywhere near .5% or 1% IFR using the lowest end of USC’s blood work. At the lowest end of USC’s range Iowa is at 2/1000 IFR. Everyone’s mileage seems to be varying. I’ll be having pizza out on Friday night.
Matt, this is exactly how I heard it. I was blown away to see that people thought he was committing that logic error. Scott Adams disappoints again.