whydoitweet

Today I realized that there are perhaps too many Dutch-style blunt opinions in the Netherlands, and not enough of them elsewhere. Time to fix that a little ;).

Let me start with Covid-19. I'll talk a bit about Trump, a bit about nonsense studies/posts, and a bit about elevators. I might cover other stuff in a later post separately.

Trump

I'm not sure what is messier: the way the White House gave updates about Trump, or the way Boris Johnson gave updates about the current lockdown measures. But at least for Trump there is a good timeline of his condition, namely on Forbes here.

My take on it? I'm pretty sure he's had Covid and I'm pretty sure he's still infectious (nobody has seen any negative test results, and I cannot think of any incentive for the POTUS to hide such results). So if you see him talking near any other person without a mask, you can be assured that he's deliberately spreading sickness without even turning on the volume. Last time around, I expected Trump to win the election: this time around I frankly find it too cynical for my own taste to be expecting that...

Covid-19 Superforecasters

The research field of Covid-19 forecasting took a big blow this week.

First there were headlines nationwide for a study form Edinburgh, which has issues. In the study, Rice and his colleagues claim that lockdown actually increases the number of Covid-19 deaths, using simulations of the CovidSim code. The detailed paper is here, and a Guardian article here.

The paper has several dangerous passages which can be utterly misleading when taken out of context, such as the following one:

“We confirm that adding school and university closures to case isolation, household quarantine, and social distancing of over 70s would lead to more deaths compared with the equivalent scenario without the closures of schools and universities. “

They decided to conclude this after having tested the CovidSim code (which I reviewed back in the days here) under a few different situations. Now simulation codes are useful, but they really are not the same as physical reality. I think we must be MUCH harsher on those who pretend that they are.

In the case of CovidSim there is one huge simplification the researchers chose to ignore: the code has the same death rate per 100,000 infections over the whole period if you run a simulation over two years. But in reality the death rate has dropped by more than a factor two (!) due to improved facilities and treatment. It sounds simple perhaps, but it's this little assumption that completely underpins their outrageous conclusion.

And indeed, the clumsy sentence was indeed predictably taken out of context by the Daily Mail among others, leading to train-wreck headlines such as:

Source: *https://www.dailymail.co.uk/news/article-8817075/Coronavirus-lockdowns-kill-herd-immunity.html*. (Btw. I love how beautifully this header renders in my browser)

The fact that *any* researcher is able to push this kind of crap into an Impact Factor 30 journal means only one thing:

We Covid-19 spread forecasters are facing a huge credibility problem, and crap peer review is to blame for it.

More bad forecasting news: Youyang Gu quits

Anyway, this scientific insult was not the only blow the community had to endure. Additionally, one of the most skilled forecasters in the US (perhaps even the most skilled one), Youyang Gu, decided to call it a day and focus on other Covid-19 related data issues. It's a sensible decision, but let's hope others can step up to the plate to do decent forecasts from now on.

PeakD Covid-19 Nonsense

As some of you may know, I play Splinterlands these days now, and one of the platforms I'm exposed to is the Peakd blogging platform.

The platform, arguably a competitor to the Coil blogs, has some merit and potential, but for me its credibility was instantly flushed down the loo when I saw this Covid-19 post. There's no need to click/read it, the tl;dr is that it describes a Covid-19 conspiracy theory involving the Rockefeller Foundation, and that it has a whopping 176 upvotes.

I was the first to actively downvote it (it still has only 1 downvote), because apparently I am the only sane person who has read it so far.

Covid-19: the elevator edition

Lastly, a little bit about elevators. I will be short on this one: don't get stuck in an elevator with a Covid patient!

Credits

Header image courtesy of AssociationsNow.

Edits

Fixed some language issues at 15:06 BST with thanks to CloudyTheChihuahua. Apparently it's hard to type good English with a cast ;).

Continue reading with a Coil membership.