Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[MetalSlimeHunt]

The dead were not necessarily infected weeks ago - minimum time from infection to symptom was reported as three days, and there have been plenty of reports of death within a couple days of becoming symptomatic.

[martinuzz]

From symptomatic to death can be as little as a few hours even.
Case diagnosis is flawed unless you also know how many tests were done, and the test groups compared were homegenous.
For example, 'double the amount of people diagnosed compared to yesterday' can mean that the infection rate has spread a factor 2 on one side of the spectrum, or that twice as many people have been tested and the infection spread rate has actually stayed the same.

[ChairmanPoo]

From symptomatic to death can be as little as a few hours even.

Well. That's not normal. I mean I dont doubt you have outliers. But the reported median times from symptom onset to admission are around a week.

Case diagnosis is flawed unless you also know how many tests were done, and the test groups compared were homegenous.
For example, 'double the amount of people diagnosed compared to yesterday' can mean that the infection rate has spread a factor 2 on one side of the spectrum, or that twice as many people have been tested and the infection spread rate has actually stayed the same.

Yes, that's what I said. But you only have around 5-7 days of delay with  case diagnosis. Hospital admissions would be around two weeks. Deaths around three. You can't really figure out where you are basing yourself on deaths.

For that matter I wonder how trustworthy deaths are too. We all know that the testing criteria (and thus the number of tests done) have changed several times in every country, depending not least on availability.  But are we sure they're counting deaths with the same criteria? Because my understanding is that they vary a lot as well. Today I read France is not counting those who die outside the hospital. I dont know if its true or not.



I expect it could go either way: bullshit news from the media here to pretend things are better organized than they are, or bullshit counting from France to pretend their numbers are not as bad as they could be. Like I said, I dont trust western goverments anymore. Turns out during a real crisis many things we took for granted wear off like glossy paint

The real question and conundrum is how long it takes Italy to recover, if ever. And how representative that is of qhat expects the rest  of us

[McTraveller]

MURICA! Now number one in confirmed cases!

[Inarius]

or bullshit counting from France to pretend their numbers are not as bad as they could be

well, the governement said it officialy (the real cases were around 40-50 k), so it's not as if they lied.

[Quarque]

Deaths are a delayed statistic. Bear in mind those who die got infected weeks ago.

A delayed statistic, yes. And some deaths have been wrongly attributed to the flu or other causes. It's not a perfect measure.

Even so, it is a far more reliable indicator of misery than "cases". Test density varies so much between nations that those numbers are fictional. South Korea probably has a pretty reliable figure, but otherwise -- useless.

In fact, an even better way to compare countries is to look at deaths per million inhabitants. The USA is not nearly at the top yet at this scale. But if Trump has his way, it will certainly get there.

[misko27]

Now see deaths per million is also flawed, because then you would see the country that is most suffering is San Marino. But San Marino can be supported by its neighbors because it is tiny, so a crisis there is less a crisis.

I think Italy remains the most hard-hit country in the world right now; it was just about to surpass China to take the top spot for infections when the US bloomed past, but Italy also has a fifth of the population of the US and having anywhere a similar number of infections as the United States and China is really god-damned scary. So it has a huge caseload, both in absolutes and in terms of infections per million.

[ZBridges]

My favorite part of the apocalypse is definitely the irony.

[MrRoboto75]

My favorite part of the apocalypse is definitely the irony.

The mexican border wall wasn't about keeping mexicans out

It was about keeping americans in

[ZBridges]

Oops, repost.

[ChairmanPoo]

Tbh covid19 is probably vastly underdetected in Mexico

[wierd]

Well, at least corporate america is still hard at work keeping those virus numbers high!

https://www.buzzfeednews.com/article/amberjamieson/spectrum-workers-coronavirus-gift-cards

[Quarque]

Now see deaths per million is also flawed, because then you would see the country that is most suffering is San Marino.

You have to filter out the mini-states, because on such a small population even a single death (let alone a small cluster) can completely distort the figure. Too much statistical noise.
If you do that, you see that the hardest hit states in the world at this moment are:

1. Italy
2. Spain
3. Iran
..
13. USA
..
19. China

Italy also has a fifth of the population of the US and having anywhere a similar number of infections as the United States and China is really god-damned scary. So it has a huge caseload, both in absolutes and in terms of infections per million.

I'm telling you again, "number of cases" are pretty much bogus numbers.

Italy is a good example. Over 8000 people died there. If the "number of cases" was correct, that would mean the death rate is 10%. Italy has great healthcare. Yes, that healthcare system is under terrible stress, but not so much that 10% mortality can be anywhere near the truth. The reason is simple: Italy stopped testing anyone who does not need hospitalization.

Multiplying deaths by 100 gets you a good estimate for the minimum number of real cases. At least 800.000 people have been infected over there, not 80.000.

Mind you, the multiplier can be much greater than 100 because of the delay. I am not saying you can reliably calculate infections that way. In countries where they have the disease under control like South Korea, number of cases is roughly about 100 times the death toll. In countries where it is spreading without any checks at all, the multiplier should be about 800 according to this model: https://medium.com/@tomaspueyo/coronavirus-act-today-or-people-will-die-f4d3d9cd99ca

However, Italy has been in lockdown for a few weeks and hopefully, the disease mostly stopped spreading there by now.

[misko27]

While I get what you're saying, deathrate also has a number of flaws as a statistic due to variations in the vulnerable groups as a percentage of the population, which is hardly constant. Even small differences would mean adding or subtracting hundreds of thousands from that number.

Moreover, the second, and perhaps more pertinent one, is that a 1% deathrate assumes that hospitals are able to care for the roughly 20% who require hospitalization. As cases mount and overwhelm localities, this no longer holds up. Taking the New York numbers of 8% require the ICU, and if we assume those who need respirators to breathe die without, we'd be at maybe 6% death rate (accounting for those who do get hospitalized). Which does not quite cut down your numbers by an order of magnitude, but gets to a pretty decent fraction of it. That's just a spitball guess though.

I guess the point here is that we just can't know, can we? There's so many assumptions... too many moving parts, not enough data.

[Max™]

Plus: we don't even know how many people actually have it right now in the US, and the 1200~ deaths are just from cases where we knew they had it first.

Extrapolating that to either number puts you at a guesstimate of 120,000 to 960,000 cases in the US right now and unfortunately that sounds pretty fucking low. If we aren't in the low millions of infected right now I'll be goddamn amazed at how lucky we've been.