Replies

Andy Jackson explains it better than me.

“Shouldn’t the size of the country’s population be taken into account?”

No! That is the point of exponential increases. The important number is the number of newly infected / day divided by the number of presently infected. If it doubles every three days, it doesn’t matter where you start. Soon most of your population will have been infected. The difference between infecting all of Italy and all of the US is about 1/2 of a week.

By plotting the numbers on a log graph you can determine the alpha for each country and as Travis’s plot shows it is pretty much the same country to country.

Now, if you were to show plots for Taiwan or Singapore or China you would see very different patterns because they have broken the chain of multiplication through effective public health measures.”Knowing that Italy, as an example, has 4 times the rate of infection as we do here in this country”

- - - -

You obviously have no scientific training, because you would realize that when you scale effects to the population, there are only two relevant numbers: In calculus it is written as 1/N dN/dt - the fractional increase in cases day over day, and the other is percent of population infected.

The former is about .13 for the US and about .05 for Italy - in other words the “RATE” [rate scientifically means change of a quantity per unit time] of infection in the US is almost 3 times as high as Italy. The second is the fraction of the population infected which is .16% for Italy and .04% for the US. Now why is this latter relevant - because saturation effects will not happen until a significant fraction of the population is infected. Even if both numbers are off by a factor of 100 it would barely effect the rate of spread of the disease.

That Italy has a higher fraction of population infected is because the epidemic has been going in Italy longer than the US - a fact we all know. With cases increasing in the US as they are, it is only a matter of time before we reach the same level of penetration as Italy.

- - -

BTW I am not a panic monger, except to the extent that panic is necessary to get people to jump off the track before the express train hits them in the dark of the night. I am very encouraged at the very preliminary indications over the last couple of days that we have broken out of an exponential increase.

But that does not mean, "pheww this is over" and we can go back to business as usual because you will just start up the same old exponential again, but now staring with a much larger more broadly seeded number of initial cases. Relax now and it could come at us a whole lot faster.

In other words we need to get the number of active cases way down, get them isolated and be able to track new cases aggressively.

And at the same time control entry on our boarders.

I disagree.

The point of 'it's exponential, bro' is to scare people. Exponential = lily pads covering the pond. That doesn't happen. It's a logistic pattern:

Separately, while plotting in log space ensures that the slope of each curve is comparable, it doesn't correct for size differences...or, more to the point, differences in timeline. That was my point about 10 cases being a bad equalizing 'starting point' for the plotting. If you wanted to make it 10 cases per 1MM citizens, then we're talking.

Finally, thank you for your commentary on my knowledge base. I bow to your superiority, my liege: only those members of the self-designated Bourgeois possessing a firm grounding in calculus can discern rate of change from a level for us prols. I'm sorry that the concept of normalizing for population differentials and putting metrics on a common footing put power into the hands of little people. It should be left only to experts, just like guns. That way, the smart people like you can control the narrative. I'm sorry. Maybe I'll endeavor to get your approval before I post.