One of the perplexing things about the way President Trump uses numbers is the randomness. At times, he simply invents data, as with his repeated insistences that his approval among Republicans is at 96 percent. At others, he uses out-of-date or cherry-picked data to make his case. And sometimes it’s just a hodgepodge, numbers and comparisons picked seemingly at random, offered less as evidence of his thesis than as evidence-like tidbits meant to make his thesis seem more robust.

So we have this tweet, from Monday afternoon.

Few have mastered the art of jamming more information into fewer characters than Trump, so we’ll set aside for now the “China Virus” elocution and the phrase “Lamestream Fake News Media” beyond to say that each is formatted in the same way: a noun modified by pejoratives. Let’s instead consider the evidence-ish numbers Trump presents, that reduction of deaths by 39 percent and the low mortality rate.

It is categorically true that the daily number of coronavirus deaths has dropped since mid-April. As the pandemic has erupted anew in the South and the West, the number of deaths that result hasn’t spiked. Experts say it still may, but we haven’t yet seen the sort of death toll that was experienced when the outbreak was centered in New York City and state.

As a result of that decline, the mortality rate in the United States has dropped. If we compare deaths with new cases over the preceding two-week period for each day since early February, we see that we’re at a new low: The total number of deaths over the past two weeks is about 1 percent of the number of new cases.

That fall in the mortality rate has happened more rapidly than the fall in the number of deaths for the simple reason that new cases are shooting back upward.

So let’s evaluate Trump’s claim.

First, it is true that the rate of new deaths each day is down 39 percent. It’s also true that it’s down 1 percent or up 40 percent or down 79 percent. It all depends on when you start your comparison. Trump appears to be looking at the month-over-month change, though it was more important for him to use 21 characters to disparage the media instead of typing “month-over-month.” Had he gone back two months, the drop would have been even more impressive.

It is not the case, though, that the mortality rate is now the lowest in the world. Again, this depends on how you measure, but a review of the same metric used above for countries with populations of more than 10 million shows there are 29 countries with lower mortality rates over the past two weeks, including Greece, Thailand, South Korea and Australia.

Why exaggerate this point? Why not simply say “one of the lowest in the world?” Reader, are we still asking such questions at this late hour in Trump’s service as president?

What Trump’s data skips over, of course, is that the reduced mortality rate now plays a role similar to the recent uptick in testing. It’s good — but a lot of damage has already been done. The United States’ failure to contain the initial outbreak of the virus elevated our death toll to the highest in the world. On a per capita basis, the vast majority of countries reporting case and death data have seen fewer deaths per 100,000 residents than has the United States.

There’s an irony to Trump’s celebration of the mortality rate. The United States’ failure to test broadly in March and April meant many cases went undetected. Had we actually measured the full scope of the pandemic then, our mortality rate would have been lower, since thousands of people who contracted the virus and lived would be included in the denominator of the equation. Our mortality rate may be lower now than it was then in part because we were doing such a bad job tracking the pandemic three months ago.

The more important question is why Trump is focused on this point. Here, too, the answer is grimmer than he would like to admit. He wants to present the virus as something that’s mostly behind us and the death toll something of a side effect to life returning to normal. He’s betting Americans will trade a few hundred deaths a day for the ability to live their lives as they did last summer, and that the surge in cases that accompanies that resumption of normalcy won’t lead to a surge in the death toll.

Or, at least, that’s what he’s doing by picking out the snapshot of data he highlights in his tweet. Whether his confident (and erroneous) claims will withstand another few weeks of elevated case totals remains to be seen.