It wouldn't change the number you know, it would just change who it is. Its more like 'you're pretty unlikely to know a youngish person who dies of covid19' (of course, we have large anecdata sample sizes on the internets.... therefore its very likely that many people on here know somebody youngish who died from covid19 complications, but still a small percentage).

"I know a few people who know somebody who died" sounds consistent with what you might expect from 0.026% if everyone communicates with about a hundred people... crudely (and wrong, but order of magnitude): 0.00026 * 100 * 100 = 2.6 = "a few".

I phrased it pretty carefully, but I agree it didn't immediately clarify that mostly the victims will look pretty old.... just bebcause "people who are likely to die within several years" is going to be dominated by people over 70. FWIW a few different measures of "likely to die within 5 years" looks like one of the simplest single-index predictors of covid19 survival, much better than just age. If you're 40 and likely to die before 45 due to a serious health condition, Covid will be roughly as dangerous to you as if you're 82.

???

The number includes everyone who has tested positive, what do you mean by ‘applying it across the whole population’ ?

You sort of address it here:

> An extra person will die this year (beyond the usual 3 who die every year). It could be anyone, but you're told its most likely to be somebody who would die in the next 5 years.

The death rate is not an equally weighted distribution over the whole population. In fact, it's very weighted towards middle age and up, with younger people dragging it down. So in your example it's likely to be more than an extra person in the typical age of older parents and grandparents.

I see, applying it across the population age distribution, yes.

This virus is dangerous to the elderly, but not nearly as relatively dangerous as the flu is to the young:

https://freopp.org/estimating-the-risk-of-death-from-covid-1...

If the young had political power, the response would be a lot different, but because the wealth/power lie with the elderly, the current response is predictable. Fascinating times

Not to go off on a tangent, but that article looks at deaths per capita and then ignores the R0. The reason the COVID deaths per capita are as low as they are (and still bad) is because of lockdowns. Pointing to what happened with lockdown mitigations in place to claim that a lockdown isn't needed is misleading at best.

The article also uses 60k flu deaths as the average baseline which is incorrect. The 60k number was a particularly bad flu season. Last season flu deaths were ~34k [1]. This season could be much lower because of the lockdowns.

Comparing COVID to the flu is hard, and really should have never been done. Without taking into account the respective IFRs, CFRs, R0, etc... any comparison is going to be severely lacking. It's also hard to get accurate numbers while in the middle of the pandemic. I hope as things open up that COVID just becomes a memory. Unfortunately, that doesn't look like the case so far.

[1] https://www.cdc.gov/flu/about/burden/2018-2019.html I'm not really sure how the author got to a 60k average unless 2007-2009 was really bad. The years on the CDC site only have 1 season greater than 60k. Errors like this do bring all of the statistics into question.

Exactly.