I live not far from Johns Hopkins Hospital and between the helipad there and this dystopian thing where BPD has a helicopter just circling the city 24/7 surveilling (in addition to the spy plane circling doing the same thing), there’s a helicopter overhead basically every 20-30 minutes or so.
Anyway point is maybe you’ve just had lots of murders recently.
Can we just bring out the guillotines already and start with these fucking idiot democratic politicians that are telling their constituents how important it is to stay home while at the same time going to dinner parties and taking vacations?
We’re taking shit seriously but even dumb ol me can see how hard it is to convince people how serious this is when political leaders don’t appear to care.
WTF is wrong with these people? It doesn’t take above average intelligence to figure out that these actions are a bad look.
“This is the worst crisis we’ve ever faced. It’s imperative that you stay home and stay safe. Alright, gotta run, my plane is about to leave for Mexico. Remember, stay home! Adios!”
ETA: @Riverman, can you do a multi paragraph, super angry rant on this? I think that would express my feelings better than I can.
ETA2: it’s even worse. The dude recorded a video from fucking MEXICO telling people it’s not the time to relax.
“Hey guys, this isn’t the time to relax. Stay home, stay vigilante! This is very serious (whispers off screen for another mai thai)”
Model update
We seem to be fitting a 4% growth curve, so here it is out through the end of the year. A little bit of noise in the last week, but seems to be holding pretty well. Its based on 2% 28 day cfr.
Historical 28 day cfr
Finally, the positivity is going up and has been for two months. Clearly our ratio of known/total cases is going down. We are now passing the positivity from the summer surge.
Nothing headed in a good direction. The deaths through the end of the year are pretty much baked in. I don’t see how we don’t see 4,000. If we are lucky the average won’t get there. But spikes over 5,000 and average over 4,000 is probably as likely by Dec 31 as not.
Playing right into Trump’s narrative too with a lot of this shit coming out right after the election.
This is going to move people the other way even. I see these news stories and I have to remind myself that these fucks all know they have a hospital bed for sure if they need it, maybe Regeneron too. For them, if Regeneron is available early, it basically is the flu.
This many are getting caught, how many more are doing it and getting away with it?
holy shit
I hope that the growth slows down. We will never measure 500,000 or more in a day, the positivity will just shoot up to 25%. The 28 day cfr has been in a pretty tight window for (1.8-2.2) for over 3 months now.
Seems to have hit a plateau at ~200k, but maybe Thanksgiving gives us a turbo boost, and who knows how long the tail will be.
What kind of a turbo-asshole doesn’t visit his family for Thanksgiving or X-mas but does go on vacay to Mexico? The fucking priorities on this guy.
I mean I would totally rather go to Mexico than see my family on the holidays…
If the positivity rate indicates we’re missing more cases, isn’t the 28-day CFR likely to go up? I guess it may only express itself as excess death.
Can you do one out to 1/21/21? Maybe 2/17/21? (Deaths baked in before Biden takes over)
Like by 1/1/21 we’re on pace for 25M cases which probably means a minimum of 50M actual cases with untested/asymptomatic included. Fair guess? If we’re cranking out 700K pozzes a day, that’s gotta be at least 1.4M actual cases, right? (I won’t even adjust for testing capacity since a lot of this model is based off before or just as we to through capacity. So likely the official results undershoot the model but reality is closer to the model.)
So even if we plateau on New Year’s Day that’s another 48M actual cases by inauguration, bringing us to 98M. 70% of the population is 231M, 85% is 281M. At 1.4M cases a day, we’d be hitting it around mid May…
In reality we’re unlikely to plateau by 1/1 (we’d have to mitigate enough in the next two weeks to get Rt <= 1), and just in a very general sense without running any math on it, it’s hard to imagine we mitigate much in the next couple weeks and even if we do, it seems like the Christmas surge would offset that to a large degree if not entirely.
Combine those charts with anecdotal stuff like my blue county in a state with a Dem governor having hospitals tip over and sticking with the statewide rules (25% indoor bars/restaurants, 50% if you certify your business is following guidelines LOFuckingL), indoor recreation 50%, massage/spas/saunas/salons/tattoo parlors 50%, entertainment and casinos 50%).
Like yo the mask mandate ain’t cutting it those are the same guidelines that got us exactly this fucked. But on the other hand a great weakness of the halfway decent Commonwealth of PA is that the state constitution mandates a flat tax not a progressive tax. As a result, and with a GOP gerrymander on the legislature, we have virtually no flexibility to increase revenue.
Thus we can’t do anything at the state level to offset the lack of federal aid, which I think makes it impossible to shut down again.
Gotta hand it to Trump and Mitch, leaving this to the states then cutting off all aid was a pretty great way to make us all Die for the Dow. Show them what they’ve won: money and a massive pile of dead bodies.
Maybe later. My projection of cases assumes we are counting cases in a similar ratio all the time, no adjustments for % poz. So for the model I will just leave the cfr at 2.
Otherwise Just assume a doubling every ~3 weeks without a slowdown.
But lol the real world, yes we expect to see the cfr go up. Wouldn’t be surprised if we climb back to 3-4 range. 2.5 only would be good. Not shocked at 5.
I feel 1.3 acres isn’t the very best site for the latest info (its main strength is sorting thru long term trends/geographic areas with a great interface), but is it true as they indicate that the positivity rate has climbed to 15% Monday and 17% yesterday? This seems very alarming if it becomes a trend.
In daily cases, right?
This was similar to nun’s argument several months ago - he was convinced that reported testing and case numbers were completely bogus starting in August or something, so that we were increasingly undercounting true positives. You’re talking about something similar - increasing positivity rate implies that we’re missing more and more positives.
If true, I think the possible outcomes are:
Cases lead to hospitalizations and deaths regardless of whether those cases are detected early or not. So the increased cases (identified plus unidentified) will lead to more deaths. When you compare those greater deaths to just the identified cases to calculate a CFR, you’ll see an increasing CFR.
The distribution of identified vs. unidentified cases is non-random because symptomatic people are much more likely to get tested than asymptomatic people. If you believe that symptomatic people are more likely to end up in hospitals and dying, then observed CFR won’t really change because the increased number of unidentified cases are largely asymptomatic and less likely to die.
After going through the discussion with nun, I’m in camp 2. CFR (I use a 21 day lag) just hasn’t really changed much in the last 4 months. That suggests to me that if there is an increase in unmeasured cases, those unmeasured cases are generally asymptomatic people who do not end up suffering severe health outcomes.
I think the highest average death count we get will be much closer to 3,000 than 4,000, at least for the rest of 2020. Peak 7-day average deaths of 3,000 is my over under by 12/31.
It’s camp 2 to an extent. At some point when even hospitals run out of testing capacity it shifts to camp 1. But of course hospitals being overrun likely increases CFR on its own even before that.
We haven’t blown past the capacity in the last four months, at least not at this scale. Maybe in specific locales.
I hope you’re right, but this seems optimistic. (Fuck this hellscape where 3K deaths a day is optimistic.)
Yes. Our fidelity in determining that will become a guess pretty soon as long as the positivity continues to climb.
I mean depending on where you look we hit between 2800-3200 deaths yesterday and hospitalizations are still going up. 3,000 is on the very low end of where I think we peak out at as far as a 7dma. Although you have to keep in mind our current 7dma isn’t even up to 1,700. It’s possible probably only if large amounts of people self lockdown. My guess would be closer to 4,000 but somewhere between 3,000-5,000 seems like a reasonable range.
Spider makes a good point in that there is a bias in the sample of people going to get tested (towards symptoms) and that as soon as testing becomes scarce, that bias will be even bigger as “contact” based testing without symptoms folks won’t bother to deal with the house.
Yeah, my co-worker has it, but I feel fine and if takes more than 10 minutes I’m not doing it
As for the net math, purely a guess. The growth terms seems pretty consistent so my lean is that the portion of undercounting and thus the cfr will go up, but hopefully a good deal cancelled out by the nun/spider theory. If the selection bias is really that strong, then it could go the other way but I’d be surprised if the cfr dropped more than a couple of tenths in that case.
I guess we will find out the hard way.
