The first panel renders the announcement itself as worthless.
Never have seen a catch 22 situation in a cartoon (that I can recall, even if it was intentional) but I think I just did.
Read it and weep (or at least wonder no more). India, with about 4x the US population is at half your rate at this point. Mouse over the data points to see the infections & deaths as a percentage of population.
Wait, never mind. The cartoon renders this data as meaningless and wrong. It's obviously Democratic propaganda.
I don't think the cartoon renders this data meaningless and wrong, but I wouldn't be being honest if I didn't admit that, when faced with a choice as to how much credibility to give a momentary scientific conclusion, it is totally reasonable to factor in how many times the conclusions have gone back-and-forth on a particular issue. That is all I was saying. It is a legitimate factor. In fact, it would be odd
not to account for it at all.
I think the point the cartoon is making is clear. How to process the fact that scientific studies often conclude other scientific studies were wrong. Then more studies later conclude the 2nd one was wrong. You may deny it all you wish, but every human brain will automatically take this into account when attaching credibility to the next one they hear. It is the nature of an evolving set of knowledge.
As far as your link, thanks. That was interesting. Nice visualization, too - I bet it was created in Tableau.
So here's a question for you. I was playing around with the visualization on that page. One thing I came across was % of cases who have died. I started out thinking "this is great! - I can really gain some insight by this metric".
Then I stopped to think about it. How do I know who is reporting what? Am I really supposed to go along with the idea that every little village in India is reporting cases exactly like they are in Houston, TX or Bogota, Columbia? How do I know what qualifies as a covid-19 death? Some people test everyone who dies, some are testing those with symptoms, some are counting those who died after treatment from covid, some aren't counting at all, others are counting those who died with covid-like symptoms, and who counts those who die in a rural place in India without ever seeing the door of a hospital??, and most importantly, I'd be very surprised if every time someone dies there, there is a comprehensive "yes" or "no" as to whether or not it was covid. Probably the most obvious question is whether in poorer countries, they even have ways of catching most of the deaths at all.
Last week it was discovered that in Florida, they are counting every time someone tests positive for Covid as a "new case". There are many people, especially elderly and in certain industries, who are getting tested regularly. When they test positive, it counts once. When they test again in a week and test positive again, it counts a second time. This is a fairly small impact, probably - but it gives you an idea of just one of the many ways supposed "data" can be just about anything from a-z.
I can stare at that "death rate per million" count in India all day long, but it's only as good as the assumption that every time someone dies in every little random place around planet Earth, all of their Testing, Autopsy, Care, Hospitals, Databases, etc. are functioning exactly the same. I would go so far as to say it is almost silly to imagine that the mostly technology- and testing-dependent reporting mechanisms of all deaths operate precisely the same way in Bangladesh and India as they do Germany and the US. As database people, we really ought to understand this clearly.
Do you really think they are?