Coronavirus

[dime]

I wonder if one could make reasonably accurate predictions on how good or bad things will go in countries and the world in aggregate based on these statistics plus information on lockdown duration and such. Probably wallstreet is doing calculations like this to "beat" the markets.

[Aytundra]

It is taking me time to trace information locally in my city, within 50 km radius.
They give pieces of information but not the complete picture. Probably to protect the privacy of those that had it.
I have to piece together different news sources from the radio, tv, and newspaper, to figure out where COVID19 victims are.
I want to gauge how much it is going on around in my local community.
At the current moment.
I know it is contagious. It is transmittable, at hospital level. Stay away from those places unless you really have an emergency.
I know it is in my city.
I know it is being imported too from other countries. Notably UK is flagging on my radar, along with USA.
and that is from current data that I got from Tuesday and Wednesday.
Most reliably I am getting the reliable info from my state's website. Local news is harder to piece together, but it gives anecdotes with real human names.

[RRM]

Probably wallstreet is doing calculations like this to "beat" the markets.

Yes, i cannot imagine that they dont.
Anything that could be remotely useful

how good or bad things will go in countries and the world in aggregate

Im afraid that things will get really ugly in India and African countries; anywhere where the informal economy is big.
In most African countries people simply cannot afford to stay home; they will starve under lockdown. Hence food riots will emerge. So, i think that many countries will opt for herd immunity instead.

[RRM]

The USA is doing better now. They ramped up testing and some major cities are in lockdown. This is reflected in the stats; the deaths/recoveries rate plunged down to 0.7 per 26/03. Lockdown is not nationwide though, and testing vigor varies greatly locally, so we may still see serious spreading in non-locked states / areas.
The UK rate is 4.2, and for Switzerland 1.5
There must be something wrong with the dutch numbers, as they still report only 3 recoveries (vs 434 deaths)

[RRM]

There must be something wrong with the dutch numbers, as they still report only 3 recoveries (vs 434 deaths)

Another 112 netherlands deaths added to the total, and no recoveries.
That does not make any sense. So, the dutch stats are totally unreliable.

[RRM]

For italy, the gap between newly infected and newly recovered peaked at 21/3, and has more or less stabilized since. That peak was 3 weeks after exiting the volatile phase B. The daily new deaths are plateauing also (just a slight increase still)
The USA has exited the volatile phase on 25/03. Projecting the italy data on the US, it might be that the newly infected vs newly recovered will peak somewehere near april 15. Daily new US deaths may stabilize a week later.
That means that in The US, we may see 3 more weeks of increases in daily death rates. In italy, over the course of the last 2 to 3 weeks, those increases were 500 to 2400%. Projecting those numbers on the USA, the daily death rates may peak at something like 2,000 to 7,000.

[RRM]

it might be that the newly infected vs newly recovered will peak somewehere near april 15.

Some scientists concluded similarly;
"Forecasters at the University of Washington's School of Medicine believe the peak of the outbreak may come in mid-April".

[RRM]

There are a few more countries newly on the list of countries with vey bad testing reports (and >100 deaths);
The accumulated death rates vs recovert rates

213 Netherlands (probably unreliable data)
19 Brazil
7.5 UK
6.6 Sweden
2.3 Portugal
1.8 Indonesia
1.5 Turkey

Bolsenaro (Brazil) has, for economic reasons, dismissed Corona as just the flu.
Sweden plays the liberal (citizen's responsibility) card (as do the netherlands)

[dime]

With those numbers I can imagine that Netherlands is testing only dead people posthumously (or people in extremely bad condition). If we assume the real ratio across countries to be similar, then Netherlands probably has similar number of cases as Germany.
According to the tv news, things are already bad in India, Africa, Venezuela, etc. as lockdowns are being put in place.

"Forecasters at the University of Washington's School of Medicine believe the peak of the outbreak may come in mid-April".

I'm wary of any forecasts. Just look at the optimism in this one from Feb 12, it's so off: https://www.economist.com/graphic-detai ... irus-curve We have no idea how it will develop in India and Africa, it's not inconceivable that it will spread to ~100% of the populations.

[RRM]

With those numbers I can imagine that Netherlands is testing only dead people posthumously (or people in extremely bad condition).

No, the active cases number was always much higher. somehow, they didnt/dont report recoveries properly.
This morning, coincidentally, (some of) that underreporting has been corrected; now the Netherlands has 250 recoveries. still a bad number though; the death/recoveries rate is 3.1 now.

it's not inconceivable that it will spread to ~100% of the populations.

Actually, it is inevitable, i think. just as the flue reaches everybody. Not everybody gets sick, but the virus cannot be stopped.
Also, it is inevitable that poorer countries get hit harder. Not just because of less developed healthcare system,, but also because it is virtually impossible to lock everybody down since majorities rely on informal income; most will not survive if they have to stay at home if the governments dont supply food to everyone.

[RRM]

For italy, the gap between newly infected and newly recovered peaked at 21/3, and has more or less stabilized since. That peak was 3 weeks after exiting the volatile phase B.

For Italy, the gap between newly infected and newly recovered is narrowing since 28/3, one week after peaking.

Some scientists advisors to the UK government think that the UK epidemic is slowing, based on stats showing exactly that (as the one below)

UK-new-deaths.jpg

But i beg to differ. Based on their poor testing regime (as shown by the high rate of accumulated deaths vs recoveries), im guessing they are just guessing. Im predicting more than doubling death rates within the next 3 days.
This is a good test for the predictive use of deaths/recoveries rates.

UK-deaths-vs-recoveries.jpg

[Aytundra]

For recoveries I wonder what the standards are?
Some places need to have 2 consecutive negative tests on an individual person, before they will consider the person recovered.
Some places have 1 week quarantine for those who test positive (see UK), while others have a 2 week quarantine from onset, before they can consider recovered.

[RRM]

others have a 2 week quarantine from onset, before they can consider recovered.

Which would cause a delay in reported recoveries, indeed

[RRM]

Im predicting more than doubling death rates within the next 3 days.

Here we go...

UK-deaths-vs-recoveries-02-04.jpg

[dime]

I see that they have added information about the number of tests here, there's new columns in the table with countries: https://www.worldometers.info/coronavirus/
UK, France, Netherlands, Sweden have quite a bit less testing in comparison to Norway, Switzerland, Germany, Italy.

Fresh paper about the death rates: https://www.bmj.com/content/369/bmj.m1327 Overall it's 0.66%
They are starting to talk about slowly relaxing the measures, and mentioning herd immunity in Germany.. which at that rate is almost 0.5 million deaths. It could be significantly reduced I guess if the elderly are properly protected from getting infected, but still many people will die..