Hacker News Reader

Comment by air7 | Hacker News Reader

"During the 15th week of the year, we saw the highest death rate in Sweden this millenium. " [0]

Statistics are just so hard to interpret and translate into a narrative that headlining a single data point is often misleading. I'm not interested enough to dive into this particular issue so I'm still not sure which claim is "true".

[0] https://emanuelkarlsten.se/number-of-deaths-in-sweden-during...

View on HN

They may both be true. The 15th week of the year is in late March while the OP is talking about September.

For reference, September is roughly weeks 36-40.

Kinda makes perfect sense that the same year would have a month with both the highest and lowest no?

Yup. Redistribution, people dying in March didn't die in September.

In France as well we had severely negative excess mortality in June. The combo of preloading deaths that would have happened later and lockdown measure preventing other diseases/road accidents, etc.

I means that people who would have died later this year (e.g. September) already died in March.

The article title and first paragraph also say that 2020 has been the third least deadly year on record (to date).

All three things can be true -- bad week in March, good September, good 2020. But March explaining September is more tenuous.

> But March explaining September is more tenuous.

I don't see why? Covid killed a lot of old people early on really quickly who might have died sometime later in the year.

This millenium is 20 years in. That's an example of hyperbolic language implying a much greater time horizon.

They aren't contradictory. 2019 had unusually few deaths, in part due to a mild flu season: https://www.icuregswe.org/globalassets/dokument/influenza-in.... So there were lots of vulnerable people who likely would have died in the next year, but COVID compressed many of those deaths into a few weeks.

Reply on news.ycombinator.com