At a briefing two weeks ago, President Trump articulated the baseline he used to measure the devastation wrought by the novel coronavirus.

“You can’t lose a million people,” Trump said on April 20. “That’s more than — that’s almost double what we lost in the Civil War. I use that as a guide. Civil War: 600,000 people died. So it’s not sustainable. But it could have been much more than a million people.”

The coronavirus, he insisted, would be more like the flu.

“They compare 50 [thousand deaths] to the 35 of the flu — because it averaged 35, 36,000 over a 10-year period. It’s a lot. Who would think that?” Trump said. In the case of the flu, there were no broad efforts to limit the spread with social distancing, he noted, although with the coronavirus there are — and “we’re still going to lose between 50 and 60.”

“If we didn’t do anything,” he added, “the number wouldn’t be 50 to 60,000. The number would be a million people dead. It would be a million-five, a million-two.”

For Trump, those were the poles: tens of thousands dead like the flu or hundreds of thousands — even millions dead — like the Civil War or worse. At both ends of the spectrum, things got fuzzy, as needed.

On Sunday, sitting in the shadow of the president who occupied his position during that war, Trump nonchalantly increased his estimate for the covid-19 death toll by up to 100 percent. Fox News’s Martha McCallum, one of the moderators of the virtual town hall event in which he was participating, asked if he thought the social distancing measures maybe “went too far.”

“No, we did the right thing,” Trump said.

“Look,” he continued, “we’re going to lose anywhere from 75,000, 80,000 to 100,000 people. That’s a horrible thing. We shouldn’t lose one person out of this. This should have been stopped in China. It should have been stopped. But if we didn’t do it, the minimum we would have lost is 1.2 million, 1.4 million, 1.5 million, that’s the minimum. We could have lost probably higher than, it’s possible, higher than 2.2.”

It’s important to note that the increase in Trump’s estimate of how many might die of the virus was inevitable. For weeks, he unrealistically insisted that the toll would be well under 100,000, perhaps as low as 50,000. Even on April 20, that was extremely unlikely, as we wrote at the time. The country passed 50,000 deaths on April 24 and 60,000 deaths on April 27, according to data from Johns Hopkins University.

Trump’s assertion that the toll might land at 75,000 or 80,000 is similarly unlikely. The country will likely pass 70,000 deaths this week — and that’s recorded deaths, not actual deaths, a distinction that’s important to keep in mind particularly when comparing the toll with the flu. As The Post’s Chris Ingraham detailed this weekend, the numbers of deaths associated with the flu are based on after-the-fact estimates compiled by the Centers for Disease Control and Prevention. The CDC goes back and looks at epidemiological patterns to establish likely ranges of infections, hospitalizations and deaths for each flu season, what it calls the “disease burden.” The number of recorded deaths from the flu — deaths that follow flu tests and are documented as flu-related, for example — is always smaller than the burden. The same will hold true with covid-19.

It’s also unlikely because of how the number of deaths has shifted, day over day. From mid-March to early April, the number of daily deaths in the United States quickly ramped up to about 2,000. Since then, with some variation, it has largely stayed in that range. The median three-day average of new deaths in the United States has been about 1,900 since April 4. Last week, that daily average was over the median on three of seven days.

Interestingly, the current cumulative death toll is remarkably similar to a prominent model the White House relied on in March to set social distancing guidelines. That model, from the Institute of Health Metrics and Evaluation at the University of Washington, estimated in early April that more than 90,000 people would die from covid-19 in the United States by June 1. When the model was revised downward later that month, it was seen as a sign that the social distancing efforts were working. In reality, though, the number of cumulative and daily deaths has tracked with that original projection.

This may be a case of a broken clock being right twice a day, certainly. But it may also be the case that the distancing efforts weren’t as effective as it seemed they were likely to be.

“Mitigation hasn’t failed; social distancing and other measures have slowed the spread,” former FDA administrator Scott Gottlieb wrote on Sunday. “But the halt hasn’t brought the number of new cases and deaths down as much as expected or stopped the epidemic from expanding.”

As we wrote last week, the period after the peak in new deaths nationally and in many states hasn’t been one of decline but, instead, a plateau. That has happening nationally in part because the surge in cases seen in New York has been replaced by smaller surges in other states. On April 4, the three-day average in New York made up 45 percent of the national total. That surged to nearly 6 in 10 fatalities nationwide. Now, New York is back down to about 25 percent of deaths — although the number of deaths is about where it was a month ago.

This is the fundamental question for the governors in each state. Given the political push to scale back distancing measures, can they be confident that they can even hold new deaths at a plateau? Or will the numbers nationally continue to sit in the 1,000- to 2,000-per-day range? If they do, of course, we’ll hit 100,000 deaths in another month, if not much sooner. Recorded, not estimated.

Update: Shortly after this article published, the New York Times reported that estimates from the CDC put the daily toll even higher in the coming weeks, with daily new deaths at or near 3,000 by June 1.

If we go back to Trump’s answer on Fox News, we see again how he’s approaching this calculus. He offers an unrealistically idealistic lower bar — 75,000! — and an exceptionally large upper bar of 2.2 million or “probably higher.” He’s crossing his fingers and moving the goal posts at the same time.

When he unveiled the social distancing guidelines in March, based in part on that IHME model, the boundaries were straightforward: 100,000 to 240,000 dead with distancing; 1.5 million to 2.2 million without. For weeks, Trump embraced the idea that we would come in well below 100,000, something that was almost certainly out of reach — as it likely was even at the time he offered those predictions. Instead, we’re where the models suggested, and hopefully falling under that 240,000-death marker.

During that town hall on Fox News, Trump also predicted that there would be a vaccine for the virus by the end of the year. America has to wonder then: Is this Trump explaining what’s actually likely to happen, or is this just another example of the president trying to wish a positive outcome into existence?

The disease burden of the coronavirus in its first year will still probably yield a substantially smaller number of deaths than the Civil War. But the toll from that conflict may also end up being a much closer benchmark than Trump likely thought it was back on April 20.