LOOK AT THEIR MODELS. 1M TO 100,000 TO 55,000 TO < 20,000. THE FUCKING FLU KILLS 50,000 A YEAR! FUCKING BULLSHIT !
(i.maga.host)
🐂 Bullshit 💩
Comments (60)
sorted by:
DEFAULT COMMUNITIES
•
All
General
AskWin
Funny
Technology
Animals
Sports
Gaming
DIY
Health
Positive
Privacy
I can't go back and look at the previous IHME model, but actual data is currently tracking the median prediction closely:
http://www.healthdata.org/sites/default/files/files/Projects/COVID/figure_8_0405.jpg
If Cuomo was using this model, he was likely considering the top of the uncertainty interval (the purple shaded area on the graph). You can see that's above 40,000.
We're going to need less than 20K ventilators for the entire country. I knew his 40,000 request wasn't even close to accurate the minute I read it.
Even his threat to take them from upstate New York was ass covering to make it look like there is a shortage
Sorry, I misread your comment. I thought you were referring to hospital beds, not ventilators. The graph I linked includes hospital beds, ICU beds, and ventilators. But, I cited hospital beds.
Ventilators and ICU beds are essentially the same (both real data and projections). It looks like the peak of the median line is only about 6,000 ICU beds and 5,000 ventilators. The peak of the uncertainty interval is about 10,000 and 11,000.
Given how cases are currently distributed across the country, I think you are right -- or at least close enough that it doesn't matter. NY has about 1/3rd of the cases nationwide, and if you simply multiply the peak of the median projection by three, that's 18,000 ICU beds and 15,000 ventilators. Even if you use the top of the uncertainty interval, that's still only 30,000 ventilators.
The presentation of the IHME model results still has a big problem: it's not showing the actual results for hospital beds, ICU beds, and ventilators for all states. The datapoints on the graph that I cited above should be in the projection of hospital beds for all states. But, maybe they don't have that data.
They are showing actual data for the daily and total number of deaths. And if you look at this PDF:
http://www.healthdata.org/sites/default/files/files/Projects/COVID/comparing_dailydeathpredictions_byrelease_all_0.pdf
You can see the model predictions for all three published versions: 3/26, 4/1, and 4/5. For my state, the first release was too pessimistic, but it's looking much better.
I'll note that there's a rookie mistake in that PDF. The individual graphs (top left and right, bottom left) have different scales for the Y-axis. That's a big no-no, because it makes them hard to compare them (and often is used to misrepresent results). But, they partially redeemed themselves by overlaying all three graphs on each other in the bottom right, using the same scale.