Coronavirus - Excellent Data-Based Analysis

Dan Kasten

Supporter
(DELETED per his request),
What part of Neil's post did you not get. Everyone is concerned about any deaths related to the virus, however your political views of our President is not needed here. The very last reason I come to this site is to talk politics.
 
Last edited by a moderator:
The IHME have said they will release modelling results soon for a projection of the level of infection to give some idea about when it might be safe to start opening up the country. I think you can get a rough idea by using a sliding window on the new cases data (Worldometer). The width of the window would equal the average time between being tested (symptoms presenting) and patient becoming virus free (non-shedding). The amount of active infection in the population at any time is then the sum of all cases within this window; the window is incremented on a daily basis.
Based on a couple of recent research articles, a window width of about 18 days seems reasonable.

The main difficulty is trying to work out the rate at which new cases will be dropping from the peak (on Apr 10th). Unfortunately IHME don't include that information on their current site, but you can get a rough idea from the deaths per day curve. The attached curve uses a combination of the 4 days of actual data since the peak plus 2 different daily case reduction rates; i.e. 600/day (purple line) and 800/day roughly deduced from the IHME curve.

Although the extended projections have a lot of uncertainty, the near-time (til the beginning of May) have more accuracy and a couple of observations can be made.

The amount of active infection at the beginning of May is about at the 80% level compared to what it was on the day of the peak number of cases. Any widespread opening within the next 2 weeks would be a really bad idea.

The other thing is people may start to feel its safe to relax social distancing measures as the peak daily increase of new cases has seemingly already past. However this is exactly the time to be more careful as the active total infection is reaching its peak (day 5 on the plot).
 

Attachments

  • Estimated Relative Infection - 18 day window.png
    Estimated Relative Infection - 18 day window.png
    77 KB · Views: 237
One gem in this mornings news here in NZ... The UK apparently do not include those that die at home in the published death numbers, that may also be the case in other countries, which could explain the wide difference in numbers in charts above.
 
Hi Jac,
Yes in the detailed description background notes from the IHME site, they state that they only use data from within the hospitalization system. They explain that the primary role of their modelling/analysis is to estimate peak demand for resources being placed within that system.
 

Terry Oxandale

Skinny Man
Took a random pick of which of the two current strings to place this comment:

So I think I got a stimulus check deposited in the bank for me. Not to look a gift-horse in the mouth, but why the hell would someone that hasn't "suffered" financially from this situation, be sent tax-free money. Yeah, I get that it would be onerous to determine who needs (gets) what, in a timely need, but this just plain don't feel right. And I'm not even a conservative...did I say that out loud?
 

Randy V

Moderator-Admin
Staff member
Admin
Lifetime Supporter
Took a random pick of which of the two current strings to place this comment:

So I think I got a stimulus check deposited in the bank for me. Not to look a gift-horse in the mouth, but why the hell would someone that hasn't "suffered" financially from this situation, be sent tax-free money. Yeah, I get that it would be onerous to determine who needs (gets) what, in a timely need, but this just plain don't feel right. And I'm not even a conservative...did I say that out loud?

The IRS did not send you some of their money. They sent you some of your money...
 

Howard Jones

Supporter
Thanks for the deletes do mine also. NY just included at home covid deaths yesterday I believe. And if you believe you don't need the stimulus money yourself then stimulate someone else that might need it. Maybe buy some takeouts for a few weeks of your favorite restaurant.
 
Thanks for the deletes do mine also. NY just included at home covid deaths yesterday I believe. And if you believe you don't need the stimulus money yourself then stimulate someone else that might need it.[[[

[[[ Maybe buy some takeouts for a few weeks of your favorite restaurant.]]]
If you are allowed out to visit one, no such luck here & nothing open, some looking at delivery!!
 
Just to give you a visual idea of where we're at right now.... If you go back to Neil's video of the mouse traps and ping pong balls (post 9 of this thread) and instead of using 900, remove 18 (2% of the population); then instead of throwing a single ball in to set off the traps, toss in about a dozen (as we are at >80% of peak infection), and imagine what happens next.

We are tickling the dragon's tail.
 

Jared V

Supporter
.19% of the US population infected?
.009% of the US population killed?
.19% of US population die of cardiovascular disease?

Maybe my math is off.
 

Jared V

Supporter
One only needs to read the first reference in the sources to not bother reading the rest of that paper that you refer to in your "estimate" of the undocumented carriers, the basis for your math.

"National Health Commission of the People’s Republic of China, Update on the novel coronavirus pneumonia outbreak; http://www.nhc.gov.cn/xcs/yqtb/list_gzbd.shtml [accessed 8 February 2020]."

Not to mention it is dated at this point.
 
If you look at the modeling I've done above; see posts 41 and more recently 53, you can see from the cases curve compared to the infection curve there is a ratio of about 10:1. The infection curve was derived in a completely different mehod as described in the text. As the modeled curves seem to be in reasonable agreement with the IHME projections, I believe they are a reasonable approximation. I would appreciate if you have a link to an alternative source that quantifies and explains the rational for using a different value. I am interested in improving my model.
 
Source: Covid Tracking Project? Is the one ran by Robinson Meyer? Far-left journalist? Another very independent source

Science doesn't have a political philosophy. Perhaps that's why our govt is getting rid of any decisions based on science.
 
Source: Covid Tracking Project? Is the one ran by Robinson Meyer? Far-left journalist? Another very independent source.

You skipped that the first source which is the CDC.
The great thing about data is that it is neutral and based on math and science, not opinion.
 
Back
Top