2020-08-03 – We didn’t really need the Axios interview with Donald Trump to know that he was getting the his COVID response all wrong. It’s just one of many opportunities we’ve had to see this jamoke flounder in his “role” as President of the United States.
Statistics are hard for people who have been assiduously trained to think they are hard. So let’s consider the statistic Trump has been harping on. I don’t think he particularly has a name for it other than “it’s-the-lowest-so-that-means-we’re-the-best.” I (and most of the scientific community) call it the “fatality ratio.” Here’s what that looks like in “math”:
This statistic tells us nothing about how successful a nation has been in slowing the spread of the virus. NOTHING. Here’s why.
Let’s consider fourcountries that have IDENTICAL fatality ratios. Assume the countries are the same size.
The fatality ratio in all of these countries is 0.5%. Which country would you want to live in? Country C of course. It has the fewest number of cases and the fewest dead. The country you don’t want to live in is Country D.
But let’s say Country D is worried about the catastrophic number of cases and the catastrophic number of dead. So Country D starts testing like crazy (but does nothing else). The other countries don’t change their testing. This alters the picture of the statistics we just looked at. (Or does it?)
Country D is now number 1 in testing. Does that change your opinion about which country is the safest to live in? Not in the least. Country D still has 150,000 dead while the other countries have far fewer. The only thing that has changed is that Country D has identified a lot of people who are infected but may not have known about it. (If you increased testing in the other countries, you’d probably find a similar increase in the number of cases.)
It’s not worthless to use testing to identify asymptomatic cases. It’s a good thing—but only if you use the information to improve the quarantine. The idea is not to win some prize about testing. The idea is to stop the spread of the disease.
A country’s fatality ratio is not a measure of the effectiveness of its anti-COVID program. All countries probably have similar fatality ratios (not exact, of course, but similar—it depends on the effectiveness of their hospitals).
But we really don’t know a country’s TRUE fatality ratio. That’s because it is hard to count asymptomatic cases. Country D is counting more asymptomatic cases than the other countries. That doesn’t mean the cases aren’t there in the other countries; it just means they haven’t been counted. The reported fatality ratio (as opposed to the true fatality ratio) is really a measure of your system of counting.
Of course, there’s far less error in counting the number of dead.
Trump gets down in the weeds to confuse us about the failure of his anti-COVID programs. Country D in my examples is an approximation of the USA. Of course, Trump lives in the weeds. He probably doesn’t understand any of this and just homes in on the one statistic that puts him first—even though it doesn’t mean anything about the health of the American people.
What do you think? Scroll down to comment.
Like what you read? Share with your friends.
If you are new to EightOh9, check out the site and Follow Me by clicking on the Follow button that appears on the screen (lower right).
Thanks for illustrating the concept they talked about in the Axios interview. For lots of us, those types of visuals clarify. It seemed as if Swan was emphasizing the percentage of deaths relative to total population, indicating that was a more useful measure. What are your thoughts?