Read on Twitter
1/ I'll admit I stopped worrying about the geniuses at @IHME_UW for a while. I assumed after April they had given up embarrassing themselves.
Nope. Seems they're still here. Somehow after six months of real world data their forecasts are worse than ever. https://covid19.healthdata.org/united-states-of-america?view=total-deaths&tab=trend
Nope. Seems they're still here. Somehow after six months of real world data their forecasts are worse than ever. https://covid19.healthdata.org/united-states-of-america?view=total-deaths&tab=trend
2/ So they are now projecting 410,000 #Covid deaths in the US by the end of 2020 (up from "181,756.26" on Aug. 30 - I guess someone was 26% dead.)
That's a huge increase - a horrifying projection. It translates into almost 60,000 new deaths a month.
Really?
That's a huge increase - a horrifying projection. It translates into almost 60,000 new deaths a month.
Really?
3/ Since the end of May, the US has had roughly 75,000 deaths (let's not get into the counting/PCR issues, we'll pretend they're all real for the purposes of this discussion). That's 25,000 a month.
Despite the fact that the virus spread UNCONSTRAINED THROUGH THE ENTIRE SUNBELT.
Despite the fact that the virus spread UNCONSTRAINED THROUGH THE ENTIRE SUNBELT.
4/ From Miami to LA, Houston to Atlanta - almost 150 million people. Outside of California, most states had limited restrictions - bars and masks (which weren't used all that widely, more on that in a bit).
Yet monthly deaths this summer were less than 1/2 the IMHE forecast...
Yet monthly deaths this summer were less than 1/2 the IMHE forecast...
5/ But maybe the epidemic is accelerating in those states?
NO, NO, NO. All across the Sunbelt, hospitals have emptied. Deaths are dropping too. The daily average in Arizona is down 2/3s from the peak. Nationally, the US is now back under 900 deaths a day, down 20% in a month.
NO, NO, NO. All across the Sunbelt, hospitals have emptied. Deaths are dropping too. The daily average in Arizona is down 2/3s from the peak. Nationally, the US is now back under 900 deaths a day, down 20% in a month.
6/ So where is this new epidemic going to come from? The Northeast/Midwest belt that got hit first and hardest? Sorry, those states have seen no evidence of an uptick, even though their lockdowns are long over. (Have you herrrd of immunity?)
7/ The last belt that might get hit might be the Bay Area/Washington/Oregon (thanks super-hard lockdowns) - and then east into the Midwestern/Plains states that for whatever reason didn't have many cases to start (though maybe not, since many of those didn't have lockdowns)...
8/ But those areas have maybe 50-60 million people in all - assuming they have Sunbelt rather than Northeastern mortality (and they should, New York was exceptionally good at killing people), their death rates will never even get close to 1,000 a day.
9/ Basically, the @IHME_UW forecast assumes that a virus that has shown NO seasonality (a surprise, admittedly) will suddenly start behaving like the flu. This forecast would actually make more sense if deaths and hospitalizations had been LOWER this summer...
10/ Then we would have to worry: Is #sars-cov-2 hiding from us? What's going to happen in the states with huge populations that didn't get hit in March?
Instead, now we know. Not much.
Instead, now we know. Not much.
11/ One last point. These folks used to love lockdowns. Now they love masks. Oh, how they love masks. The only way to SAVE US ALL is masks, see? (The green line is death projections if we all wear masks; the scaaary red line is if we misbehave and stop social distancing.)
12/ But according to their own data, mask use has basically not changed at all since May - and in fact has recently fallen slightly. That’s probably right. If you can square it with the Team Apocalypse position that MASKS SAVED THE SUNBELT, you’re smarter than I am...
13/ But seriously: I don’t know how people who claim to be interested in science can keep making the same mistakes over and over. They’re not just wrong, their forecasts willfully ignore the current real-world data and trends. Why? I don’t know. They don’t return my calls.
