Saul Loeb / AFP / Getty

Here’s a paradox of the Trump era: The president is the most overexposed figure in American political history. You can turn on your television, open any newspaper or magazine, or log on to practically any social-media site at any time and get a strong dose of Donald Trump. The man is everywhere.

Yet unless you’re a political junkie, you can easily get by without watching Trump at length or really paying deep attention. He may seem inescapable, but he’s just part of the atmosphere—an acrid presence you eventually get used to, like the industrial stench in a factory town. Some people are constantly focused on him: the Trump superfans who feast on his every word; the “resistance” super-sharers who howl at and mock it all; and, most pathetic, political reporters. All three groups are loud, but they are not the majority of Americans.

That brings us to the first presidential debate, last month. A debate is a chance to see both candidates relatively unfiltered. And given a rare chance to see the president at length, without filters or interruptions (except many of his own), many Americans looked at Trump and recoiled in horror.

The debate was the start of a bad week for the president. First, viewers panned his performance; then he fell ill with COVID-19. The president’s polling fell and it can’t get up, and the debate seems to be the major culprit. Consider the latest USC Dornsife poll. Any individual poll can offer only so much information about the race, which is why polling averages are so helpful. But the Dornsife poll is a tracking poll, which repeatedly surveys the same group of voters. That makes the up-and-down direction of the poll useful for spotting changing dynamics in the race.

Before the debate, the pollsters asked panelists to rate their expectations of the candidates on a zero-to-100 scale. The guesses came in the middle of the pack—51 for Trump and 50 for Joe Biden. The pollsters resurveyed after the summit. The former vice president didn’t have a great night, and his average rating was 46, a slight underperformance.

But Trump nosedived. His postdebate rating was a putrid 33, 18 points below expectations. While every ideological group rated him poorly, his worst showing came among Republican-leaning independents, with whom his rating dropped 24 points. That’s a group that is especially important to the president’s reelection chances. Although his base remains solid, he needs to get GOP leaners to vote for him, rather than going for Biden or simply sitting out the election.

The debate hurt Trump’s standing versus Biden in other categories as well. The Democrat’s edge on mental fitness for the presidency increased seven points, to 19, after the debate, with half of voters saying they don’t believe Trump is mentally fit. On a basic presidential-preference question, Biden extended his lead over Trump from nine to 13 points.

For regular Trump watchers, his performance wasn’t all that surprising. The president was rude, shouty, and racist. He spoke in fluent conspiracy theory, unintelligible to most members of the public, and was generally unpresidential. In other words, it was maybe on the worse end of his normal range. But most people aren’t watching Trump at length—they’re consuming him in small bites, whether that’s sound bites on the news or decontextualized tweets.

It’s true that Fox News gives Trump a great deal of attention, but Fox’s average prime-time viewership in the third quarter was a record 3.5 million people. By contrast, 73 million Americans watched the first debate on TV alone, leaving out those who streamed it on a device or listened to it on the radio. Meanwhile, interest in Trump’s rallies—once spectacles that would lead TV networks to preempt regular programming—has subsided.

When you get Trump only in small doses, it’s possible to overlook his incoherence, to dismiss his espousal of conspiracy theories as simple trolling, or to claim that the media are misrepresenting him. But over the course of 90 minutes, with no commercial breaks, those excuses start to dissolve.

“It’s not clear that seeing more of the president is necessarily a help to his campaign,” the Republican pollster Whit Ayres told The New York Times.

Coming into the debates, the conventional pundit wisdom was that Trump needed the debates more than Biden did, because the candidate who trails needs to change the race. Even after the poor reviews of the first debate, this logic still held. But after Trump’s COVID-19 diagnosis, the Commission on Presidential Debates announced that the debate would go virtual. Trump angrily pulled out. If it were true that Trump couldn’t afford to miss a debate, this would be weirdly self-defeating. But the Dornsife poll suggests something different: Maybe he couldn’t afford the exposure.

With the debate canceled, Trump and Biden are both participating in individual town halls tonight, on NBC and ABC, respectively. According to the Daily Beast, “Trump has told close associates that he wishes to counter-program the Biden town hall and score higher TV viewership numbers, and then use such a contrast in ratings to humiliate his Democratic opponent.” Perhaps he would be wiser hoping that his ratings tank.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.