That’s reading people’s mind. I do my own mind reading and I say that compliance to lockdown is down because people are fed up, and because the virus turned out to be nowhere near as dangerous as initially claimed by the media (I remember a headline dated from end of April from The Times claiming covid was as deadly as ebola...)
No the headline was a very misleading statement based on the fact that people showing up in ICUs with severe forms of covid had about the same death rates than Ebola.
Just because the media did a horrendously bad job at reporting on this virus does not mean it isn't a serious threat. We see now in the US how it can spiral out of control if we let it.
And meanwhile some people actually took the time to look at the statistics. A lot of countries with lockdowns had more trouble than countries without lockdowns. e.g. Argentinia with masks and lockdowns has one of the worst curves of the planet... while Sweden is doing fine.
And COVID-19 is strictly following Gompertz-Mathematics - just like a seasonal corona-virus. For any usual outbreak (f'/f) is falling exponentially all the time... There are no visible trend-changes around most policy-changes.
And there's no statistical epidemiological evidence that masks did anything good. In some countries with masks things got a lot worse. (more efficient catch+inhale? Or catch+transmit without being infected?)
Most of the hype around COVID-19 is just cheating by using "reporting date" instead of "date of death" to make Gompertz-Functions look like seriously dangerous exponentials...
If you have some datasets that allow to compare "reporting date" with "date of death" it is astonishing to see how many reporting-anomalies appear around policy-changes...
I think you should take another look at Sweden's data. Sweden (and a lot of countries) were doing fine, likely because the weather is nice and people were spending their time meeting outdoors, where infection risk is way lower. Now that it is getting cold, their case counts are on the way back up as are their deaths. It is worth noting that most of my Swedish friends say that many people were staying home anyways, just because they did not have a government mandated lockdown doesn't mean many people did not decide to avoid the risk themselves.
There are plenty of datasets that include date of death instead of reporting date and they show the count of deaths increasing again as well, on a lag, just like you would expect.
The low fatality rate we have enjoyed recently has been partially a result of the availability of improved medical care. Unfortunately that is a stepwise function where once you have surpassed the amount of available care, the improvements we have made in care revert to some extent. We don't know how much.
I see that you only really post about COVID stuff on hackernews, and I don't know why anyone would bother, but what truly blows my mind is the amount of hubris it takes to sit at home, read a few news stories, and say "I know better than the majority of people who have spent their lives studying disease because I know some statistics and read a few articles that disagreed with the mainstream."
The mathematics of COVID-19 is not that hard - and has been known since pretty much the beginning. No need for hatred and personal attacks..
Lets do some real mathematics.
1) Get good data (e.g. death by date).
2) Take a look at the Logarithm of the growthrate ln(f'/f)
3) Spot all those straight lines - those are the curves of outbreaks
4) Turn those lines into a working formula for an outbreak
5) Calculate predictions - evaluate trend-changes - look when new outbreaks happened.
It's just that simple. Results won't give you insight about the mainstream, but it will give some insights about reality.
Some countries are easier (Mexico, Germany, ...).. Some have more outbreaks and complications. But the basic principles of the mathematics of COVID-19 outbreaks around the globe are the same.
Calculating the new outbreaks of Sweden is more advanced because the start of the new curve is hidden behind another outbreak. But now there's at least enough visible to give good estimates.
I just looked at the data for Sweden today. Looks like a seasonal outbreak starting mid-September (infection date).
Parameters of the curve look like the first wave (Gompertz-Exponent ~0,06) with less force.
Start is hidden under other smaller outbreaks, so it's not so easy to measure it exactly. But soon it should peak (if it hasn't already...).
If they don't do any nonsense that changes the mathematics of COVID-19 completely Sweden will be fine. The curve has normal size of a seasonal flu. Other countries in Europe are doing a lot worse.
What puzzles me is those serology studies in Spain and Italy that show that people who were locked down were infected at the same rate than essential workers who were not locked down (except, understandably, for healthcare workers). It seems to be a major data point on whether a lockdown actually has any impact and it seems to have been completely ignored.
