[morsch]

if I take a random American, what's the likelihood they're a person aged 50-64 and they're known to have had Covid by November 10, 2021 and then BP

Why would that be the calculation you're looking for? It's already off by an order of magnitude because the likelihood of any given American being in the age bracket is around 1:7. And it's another order of magnitude off because while not everyone has had a first infection or vaccination yet, everyone will, in the (not so) long run.

You're trying to prove you're less likely to get BP from Covid than from a vaccine while using a number -- the 0.08% -- that says the opposite, and yet you end up with a number you like. Because your calculation is just nonsense.

It's all kind of a moot point, since it's seemingly a relatively rare complication in both the disease and the vaccine; you're more likely to die from Covid than get BP if you're older than 45, as far as we know.

https://link.springer.com/article/10.1007/s10654-020-00698-1...

[nickthemagicman]

chance of infection * chance of some complication

Its a very simple intuitive calculation that is used widely throughout epidemiology.

It feels like, with Covid, people try to add on unnecessary calculations or ignore important factors to shoehorn the numbers to fit a narrative.

[morsch]

The thing is, you're not calculating the chance of infection correctly. You're taking the number of cases in an age group, and then dividing by the total number of residents. That doesn't work. Either start with the total number of cases, or divide by the number of residents in the age group. Here's the latter:

7.2 million cases aged 50-64 [1]

58 million people in the US aged 50-64 [2]

So based on those numbers, 12.5% people in that bracket were infected by early November. From that you can attempt to extrapolate an infection risk per annum of around 8%.

Of course, now we're extrapolating from a time with varying degrees of voluntary and involuntary NPIs such as mask wearing and social distancing to a time where those won't be practiced widely, some of the time range we're extrapolating from also had the virus localized to regional or social communities, while now the distribution is more and more homogeneous. On the other hand, it's possible that virus spread will measurably decrease now that more and more people have some resistance through vaccination or past infections. So it's not a very reliable extrapolation at all.

[1] https://www.statista.com/statistics/1254271/us-total-number-... whether or not that number is accurate is debatable, our estimates of the true number of infected vary widely

[2] https://www.wolframalpha.com/input/?i=age+distribution+usa

[nickthemagicman]

I think you may be wrong.

All age groups are equally INFECTIOUS once they have the disease, but are not equally able to BE INFECTED , otherwise people under 55 wouldn't have to wear masks. We wear the mask to protect others right? No matter what our age is.

So why would you reduce the population size to just an age slice?

The same assumption is made when calculating herd immunity. They use the entire populations infectiousness because chance to transmit doesn't change with age but chance to be infected does..

But you DO limit by age group on the infection side because that individuals chance of contracting it from the general population, is based on age/immune system .

[morsch]

Like I said, you can do this calculation (percentage infected) for an age bracket (#infected in age bracket / population size in age bracket) or you can do it for the whole population (#infected in whole population / whole population size).

But it doesn't make sense to do #infected in age bracket / whole population size. You might as well calculate #infected in age bracket / number of cattle in Texas.

Of course there are many reasons why one would want to look at the number of infected in certain age (or social or whatever) groups. Beyond that, I'd prefer not to get further into the weeds with you.

[nickthemagicman]

If I walk into a room of people with the black plague, I'm concerned about every single person in that room giving me the disease not just the people of my age bracket.

That's the chance of infection.

This same idea is used in mask policy and is used with herd immunity calculations and all over epidemiology to be honest.

I'm really sorry that you have to do mental gymnastics to shoehorn this subject into your narrative.

Have a good day!

[morsch]

I wasn't the one who brought up age brackets, you were ("infected 50 year olds"). But here's the calculation for the entire US population.

Total number infected until early November 2021: ~37 million [1]

Total population of the US: ~329 million

Infection percentage so far in pandemic: 11.2%

"Chance of infection" (per annum): ~7%

See above why this is an absurd extrapolation that nobody with any brains would take seriously. In the long run, pretty much everybody will get Sars-Cov-2. It's an endemic disease.

[1] https://www.statista.com/statistics/1254271/us-total-number-... per your source, NYT and Our World in Data refer to 47 million cases