Can someone run the projections on these numbers? (exponentially)

[Quixote1818]

Thanks in advance

This doesn't include China. I am using the graph in the bottom right hand corner: https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6

Past 10 days (thousands):


3-1 8.5
3-2 10.3
3-3 12.7
3-4 14.9
3-5 17.5
3-6 21.2
3-7 25.2
3-8 29.1
3-9 32.8
3-10 37.8

Day 1 + 1.8
Day 2 + 2.4
Day 3 + 2.2
Day 4 + 2.6
Day 5 + 3.7
Day 6 + 4.0
Day 7 + 3.9
Day 8 + 3.7
Day 9 + 5

[Dem2]

1. Like this?

https://imgur.com/slRSOy1

Except not projected...

[Quixote1818]

2. Thanks, I should have said projected

It looks like around a half million by April 1st, 2 million by April 10, 8 million by April 20. 32 million by May 1st, 120 million by May 10, 480 million by May 20th etc I am just eyeballing it

[JCMach1]

3. Assuming rate won't change with weather like the flu

We may have a Summer lull followed by a hellish Fall.

[Quixote1818]

4. Good point. It will probably start picking up in Australia and extreme Southern South America

but those places have very limited world population.

[JCMach1]

16. And Sub-Saharan Africa

There have been 0 cases last check in Kenya.

Weird fact, some African countries have a better response team than we do due to having been scared poopless about an Ebola pandemic...

[Ms. Toad]

5. Are you looking for the equation?

Number of cases = 7536 x e^(1.665 x D), where D = number of days from March 1 (That's a very close fit, with r^2 = .9956)

So, for example, on March 30 (30 days from March 1), the number of cases is projected to be 1,314,474

Again - that's assuming the currently reported cases are accurate & the rate of transmission doesn't change.

(I didn't trace the numbers back before your 10 days.)

[Quixote1818]

6. Thank you! nt

[Ms. Toad]

7. If I run a power curve, and I use Jan 24 as the starting date,

I get # cases = .0000007 x D^6.4964 (where D=# days since January 24). That's also a close match at r^2 = .9931

That would put the total on March 31 at 463,002 (pretty close to your eyeball).

[Quixote1818]

8. Interesting.

I imagine people are taking extra precautions right now slowing things down a bit which is really good compared to what we could be seeing. However, the unreliable testing numbers in a number of countries creates a hell of a lot of unknowns.

[Ms. Toad]

9. Yup.

Trying to match a curve - that may well change over time is pretty sketchy business. But I was within 3 days of predicting when COVID 19 would surpass the SARS deaths.

[Quixote1818]

10. Why did SARS die out so quickly? nt

[Ms. Toad]

13. I wasn't really tracking SARS

When I was trying to wake up the administration of the school I work at yesterday by giving them accurate information, they wanted to know how I knew so much. My response was that I have an at risk daughter - if something might kill her, I stalk it to death.

So I've been tracking this since mid-January.

SARS was after she was diagnosed with ulcerative colitis but before she was diagnosed with primary sclerosing cholangitis (a chronic liver disease that will ultimately require a transplant). So I didn't view SARS as a particular threat.

I think part of the difference is that COVID 19 has a high transmission rate but a death pattern that does not kill its host off before the host can infect others (contagious before symptoms, long drawn out death = more opportunity for transmission). SARS was more deadly (so it killed the host quickly) and an average transmission rate.

[Maru Kitteh]

11. 463K on March 31?

Holy hell.

[Quixote1818]

12. That is worldwide. It will probably be higher than that since testing has been inconsistent. nt

[Maru Kitteh]

14. Yes, I knew worldwide. Still.

Holy hell.

[Ms. Toad]

15. China only maintained a similar curve for about 20 days -

then it tapered off.

They instituted pretty draconian measures to make it taper off.

There are a couple of things that can taper it off - running out of uninfected bodies to infect (with 33 million, that's not likely to happen that quickly) or we implement measures to slow transmission (that's what China did - and what Italy is attempting to do)

[The_jackalope]

17. I ran the non-China case numbers from Feb 9 to Mar 11

I projected out to Mar 31, or ~3 weeks. A third-order polynomial gave the best fit, with an R^2 of 0.9987. On March 31 the projection indicates ~250,000 cases outside China. That implies a doubling of cases every 8 days or so, if the curve remains intact.