5 ms·

the 25 million just does not seem realistic.

cycrutchfield 7y ago
It's pretty realistic in an unconstrained spreading scenario. That's exponential growth for you.
ajross 7y ago
The H1N1 epidemic in 2009 was objectively less contagious and less dangerous than COVID-19. In comparison, it grew mostly uncontained and eventually infected about a billion people. If COVID-19 was allowed to do exactly that we'd be looking at 10M fatalities, and realistically it would infect more people (being more novel than a mere flu). The 25M number doesn't seem realistic because every country that has had a COVID-19 outbreak has freaked out trying to control it, and either succeeded via very expensive means or (in most cases) hasn't had the outbreak grow far enough to know whether or not they've failed.
- wolco 7y ago
  Using the billion figure best case is 40 million not accounting for which countries would be hit the hardest which would increase the rate as the systems get overwhelmed.
stevenwoo 7y ago
Do you reject the SIR mathematical model for all infectious disease or the numbers chosen for the constants for covid-19 or some other reason?
wolco 7y ago
I thought so too then I did the calculations and realized it could be larger if Italy's death rates transfer.
- maxk42 7y ago
  The US death rate is 1.49% and dropping rapidly.
  - cycrutchfield 7y ago
    Only because the number of new cases is rising rapidly. "Death rate" is not a meaningful metric. What you want is "case fatality rate".
  - therealdrag0 7y ago
    We haven't overwhelmed our healthcare system yet.
Animats 7y ago
It's just outside most people's experience. Look at the graph.[1] Those are actuals, not predictions. This gets 10x bigger each week. The US is 12 days behind Italy. US at 10,000 now. 100,000 next week. 1 million in two weeks. The lockdown may slow that down. Probably not enough. [1] https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06441 https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06...
- usaar333 7y ago
  I don't see any scenario where you could hit 60% of the population that fast. I mean sure, if you held up the same R0, yes, that math works out, but people aren't trees. They'll voluntarily socially distance, dropping R0 significantly.
dredmorbius 7y ago
I wrote a basic model in Awk for US and global "RoW" (ex-China) spread of COVID-19. Both sets of confirmed-case projections (for US see comments for the mathematically correct model) are tracking remarkably close. Lagging slightly below, but only just. (Deaths have trended higher than projection for RoW, lower for US, but the delta remains only a few days.) https://joindiaspora.com/posts/73d4e930421d0138028e002590d8e506 https://joindiaspora.com/posts/73d4e930421d0138028e002590d8e... Exponential growth boggles the mind. Script source: https://pastebin.com/raw/j89ZUdVq https://pastebin.com/raw/j89ZUdVq

It's pretty realistic in an unconstrained spreading scenario. That's exponential growth for you.

The H1N1 epidemic in 2009 was objectively less contagious and less dangerous than COVID-19. In comparison, it grew mostly uncontained and eventually infected about a billion people. If COVID-19 was allowed to do exactly that we'd be looking at 10M fatalities, and realistically it would infect more people (being more novel than a mere flu). The 25M number doesn't seem realistic because every country that has had a COVID-19 outbreak has freaked out trying to control it, and either succeeded via very expensive means or (in most cases) hasn't had the outbreak grow far enough to know whether or not they've failed.

wolco 7y ago
Using the billion figure best case is 40 million not accounting for which countries would be hit the hardest which would increase the rate as the systems get overwhelmed.

stevenwoo 7y ago

Do you reject the SIR mathematical model for all infectious disease or the numbers chosen for the constants for covid-19 or some other reason?

wolco 7y ago

I thought so too then I did the calculations and realized it could be larger if Italy's death rates transfer.

maxk42 7y ago
The US death rate is 1.49% and dropping rapidly.
- cycrutchfield 7y ago
  Only because the number of new cases is rising rapidly. "Death rate" is not a meaningful metric. What you want is "case fatality rate".
- therealdrag0 7y ago
  We haven't overwhelmed our healthcare system yet.

Animats 7y ago

It's just outside most people's experience. Look at the graph.[1] Those are actuals, not predictions. This gets 10x bigger each week. The US is 12 days behind Italy. US at 10,000 now. 100,000 next week. 1 million in two weeks. The lockdown may slow that down. Probably not enough. [1] https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06441 https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06...

usaar333 7y ago
I don't see any scenario where you could hit 60% of the population that fast. I mean sure, if you held up the same R0, yes, that math works out, but people aren't trees. They'll voluntarily socially distance, dropping R0 significantly.

dredmorbius 7y ago

I wrote a basic model in Awk for US and global "RoW" (ex-China) spread of COVID-19. Both sets of confirmed-case projections (for US see comments for the mathematically correct model) are tracking remarkably close. Lagging slightly below, but only just. (Deaths have trended higher than projection for RoW, lower for US, but the delta remains only a few days.) https://joindiaspora.com/posts/73d4e930421d0138028e002590d8e506 https://joindiaspora.com/posts/73d4e930421d0138028e002590d8e... Exponential growth boggles the mind. Script source: https://pastebin.com/raw/j89ZUdVq https://pastebin.com/raw/j89ZUdVq