Hey guys, just a few quick thoughts before I head out for a walk to fight the quarantine-fifteen I've packed on.
@Parley thanks for linking the covidusa.net site. I think I saw it long ago but it is much fuller now. About that site:
I find it alarmist and in my view it portrays a situation much worse than I believe to be the case. Their data isn't wrong (they apparently harvest from the same places everyone else does) but some of their projections seem way off.
My first beef with them is their comparison between the flu and COVID-19. They say COVID-19 is 34 times as deadly as the flu, but they are comparing the COVID-19 CASE fatality rate (CFR) to the longstanding flu fatality rate which is more or less an INFECTION fatality rate (IFR). While they state upfront that there are lots of asymptomatic or undiagnosed COVID cases, they don't factor this into the math because they probably goofed and don't understand the flu metrics which take some digging to get at. But, again, flu mortality rates are statistical estimates that are much more akin to infection fatality rates than to CFRs.
About yesterday's 271 deaths: remember it was a Sunday on a holiday weekend so lots of yesterday's deaths haven't hit the books. I prefer the seven day moving average deaths per day chart at worldometers; it shows we're in the 500's of deaths per day smoothed by 7 day rolling average. We've been coming down but I suspect a gradual uptick because a lot of the intermediate states like NC and MD have already stopped dropping from their highs and have stabilized at low numbers; with few states dropping in daily deaths, the rise from the GA/FL/TX/AZ cluster will likely become visible within a week and go up from there.
Anyway, back to it being alarmist: it suggests an estimate of 15,890 deaths in the next 7 days. Nope, not happening. We're at ~520/day or so right now, and it will slope up, but we're not going to average more than 2000 deaths per day over the next week. They are running a model that is projecting a mortality rate based on historical data through an algorithm that is grabbing the recent spike in cases per day and prematurely factoring them into deaths. It's well established that many of the early part of the spike were youthful and that cohort will have a very low death rate except in high risk groups like severely obese, immunocompromised, or "too stupid to live Alabama coronavirus partiers."
That youthful group will infect others and the reaper's a-comin', but not that quickly.
Its math about hospital ICU usage is way off; it says we run out of space tomorrow. Seriously. Things are getting worse for sure and some states are well into emergency plans, but as a country we're not out of beds (and the site's statistics are average for the country).
Any idea who runs the site? I'd say it was pushing an agenda from the extreme wing of the COVID-19 terror campaign team, but I'm at all those meetings and this isn't one of ours.
covidusa.net
In April, the virus was just getting started. Now SH!T's got real. (pardon my colloquialism) 
