Cont: 2019-nCoV / Corona virus Pt 3

[zooterkin]

Still well off the high.

I'm keeping an eye on UK numbers, because they aren't as bad as I would have expected, given the idiotic early response and very late shutdown.

Early days, but both new infections & deaths seem to be falling when any modelling says they should be rocketing. I'll accept severe under-testing and slow reporting might be an issue, but it will be a big tick in the seasonality stakes of Covid if they continue falling.

I'm not seeing them falling yet. Latest prediction from the IHME is that we'll see two and a half thousand deaths a day before the numbers start to fall.

https://www.theguardian.com/world/2...ropes-worst-hit-by-coronavirus-study-predicts

 

[marting]

Not so much an attempt to reducing the scatter as a way to present the trends in a different way. My data is from the daily WHO situation reports, for US cumulative deaths.

I estimated the number of days for deaths to double.

For example, the 21 March number is 201, and the 23 March one is 402. So the number of days to double is 2.0. The next doubling? That would be when deaths (cumulative) were 804. 25 March: 673; 26 March: 884. So I did a linear interpolation, and the number of days to double is 2.6.

Here are my results:

Date|Deaths|x2|
Date|Cum|days
21 Mar|201|x
23 Mar|402|2
24 Mar|471|
25 Mar|673|2.6
26 Mar|884|2.5
27 Mar|991|2.2
28 Mar|1243|3.2
29 Mar|1668|2.7
30 Mar|2112|
31 Mar|2398|3.2
1 Apr|2850|3.5
2 Apr|3846|3.5
3 Apr|4793|3.4
4 Apr|5854|3.1
5 Apr|7020|3
6 Apr|8358|2.9
7 Apr|9559|3.5

Still rather a lot of scatter, obviously. However, the doubling time is in a farily narrow range.

The best news would be when it takes forever for the cumulative total to double; clearly we are not there yet.

A hint that good news is coming would be a clear trend that the time taken to double is increasing. Sadly, even if you squint hard, I don't think you can see such a hint in the data presented in this form.

Thanks I like the idea of presenting the numbers in terms of doubling or halving. It's pretty clear that most folks don't have an intuitive idea of exponential growth/decay re different exponents but doubling/halving is more intuitive.

Scatter noise is always a problem in these models. Especially in the earlier days whith small numbers. And then there are actual changes in timeline that complicate things. The WHO data, for instance, changed between 10:00 CET and 23:59 CET at various times messing up projections that assume a 24 hr delta between reports.

 

[sir drinks-a-lot]

I am looking forward to new excuse why COVID-19 is not big deal when he loses that bet. He seems to be really invested in that for some reason.

Looking forward to higher death numbers. Keep it classy.

 

[EHocking]

China has lifted lockdown restrictions in Wuhan and neighbouring provinces, allowing travel in and out.
It will be interesting to se if there is a second wave of infections or if, indeed, the Chinese got it right.

 

[The Atheist]

I'm not seeing them falling yet. Latest prediction from the IHME is that we'll see two and a half thousand deaths a day before the numbers start to fall.

The numbers are looking a bit patchy, too. That famed British efficiency at work?

China has lifted lockdown restrictions in Wuhan and neighbouring provinces, allowing travel in and out.
It will be interesting to se if there is a second wave of infections or if, indeed, the Chinese got it right.

Jury is definitely out on that still: https://www.newshub.co.nz/home/worl...rds-surge-in-asymptomatic-covid-19-cases.html

 

[Belz...]

China has lifted lockdown restrictions in Wuhan and neighbouring provinces, allowing travel in and out.
It will be interesting to se if there is a second wave of infections or if, indeed, the Chinese got it right.

Hard to tell. I don't trust China's numbers one bit.

 

[Roger Ramjets]

Hard to tell. I don't trust China's numbers one bit.

Why not?

 

[sir drinks-a-lot]

Why not?

Because they said there were zero deaths at Tiananmen square.
(stolen from twitter)

 

[Dr.Sid]

Why not?

No new cases is impossible. China at very best should see something like South Korea.

 

[Doghouse Reilly]

I am looking forward to new excuse why COVID-19 is not big deal when he loses that bet. He seems to be really invested in that for some reason.

I wouldn't have an excuse, I would just admit I'd been wrong. Pretty anticlimactic, I'm afraid.

 

[Belz...]

Why not?

Because it's China.

 

[Belz...]

I wouldn't have an excuse, I would just admit I'd been wrong. Pretty anticlimactic, I'm afraid.

Well in that case I'm still waiting for you to admit that the prediction that UK deaths would not pass 8k was wrong.

 

[trustbutverify]

Why not?

Body of work?

 

[lomiller]

I said fewer than 85k US deaths, combined flu and coronavirus by the end of 2020. Roboramma says more than 85k. The stakes are an avatar for 1 year.

Are flu deaths even counted in a way that would allow you to settle that bet?

 

[Doghouse Reilly]

Well in that case I'm still waiting for you to admit that the prediction that UK deaths would not pass 8k was wrong.

I never said it was right, I was sharing that guy's opinion. I didn't claim the UK would have fewer than 8,000 deaths.

 

[Doghouse Reilly]

Are flu deaths even counted in a way that would allow you to settle that bet?

Actually I was mistaken, we aren't doing combined, we are doing 85k 2020 total US coronavirus deaths (I say <85k, he says >85k) based on this website:

https://coronavirus.jhu.edu/map.html

Flu deaths are not considered in our bet.

 

[The Atheist]

Scientists in Australia making some serious claims:

A vaccine likely won’t be available in large amounts for another 18-24 months, and countries need to do more frequent testing, Peter Collignon, a professor at the Australian National University Medical School, told Bloomberg News.

People also aren’t likely to travel abroad for at least the next six months as nations try to contain the virus’ spread, Collignon said. Eradication of the virus is unrealistic, he said.

https://www.bloomberg.com/news/arti...deadliest-days-virus-update?srnd=premium-asia

 

[Ulf Nereng]

So what do rheumatologists actually say?
https://rheumatology.medicinematter...on-on-chloroquine-hydroxychloroquine/17862288

And the Lupus people?
https://www.lupus.org/blog/are-people-with-lupus-protected-against-covid19

I will listen to them before any mad TV quack out to make a buck.

And are you really happy to promote nonsense that could kill people? Really?

I saw one of his YT videos. He didn't mention the side-effects or how easy it is to overdose on it. Also failed to explain what a pilot study is. Left me wondering if he really is a MD. I blocked the channel so I won't have to watch such rubbish again.

 

[EHocking]

Hard to tell. I don't trust China's numbers one bit.

Regardless, China's economy must be hurting enough for them to lift travel restrictions with 55,000 "escaping" Wuhan by train on the very first day.
Many restrictions still apply, but workers are being given priority to travel to work by the looks of things.

Also they seem to be letting tourists to leave as they open their airport and a number of Chinese medical workers who were brought in at the start of the outbreak are also leaving.

Whether China has learned its lesson regarding covering up infection numbers after today and report any resurgence of infections is yet to be seen.

 

[Pixel42]

Top 15 Causes of Death in the U.S.

Covid-19 is currently killing about 8 times as many people as flu/pneumonia normally does.