I have this friend I’ll call Tom, and when we each watch TV, we notice different things. Recently, Tom had been noticing white men like him — or rather, noticing a lack of them. He felt white men were increasingly absent from his screen, and he surmised that a push toward gender and racial diversity had shoved them out. He mentioned this on Facebook, which was the only place I knew him; we’d never met in person.

I felt Tom was watching TV from another planet. (Had he accessed a secret cable package sponsored by Emily’s List that I hadn’t known was available?) But Tom also seemed to exemplify a fear expressed, in ways both articulate and vague, by a lot of men in the post-#MeToo era. There are Reddit threads full of guys claiming they’re now treated as pariahs. Op-eds asking whether all this social awareness has gone too far. The terrifying panic, whispered at dinner tables and in conference rooms, that white men are being erased.

In Tom’s case, the fear manifested in such a concrete way it seemed simple to solve. So I tried to solve it. By sending links to studies showing he was wrong. One, from San Diego State University, found only 24 percent of 2017’s films had a female protagonist. A different SDSU study found only 21 percent of TV shows had more women than men.

It didn’t work.

Tom diligently read these stats but was convinced that they didn’t represent what he saw.

“I’ve noticed it,” he told me. “I’ve noticed white men aren’t there. I’m not making this up.”

It was mostly in TV commercials that the white men had gone missing, he clarified. So I tried to solve it again. By sending more links to more studies, focusing specifically on ads. A 10-year analysis by the Geena Davis Institute on Gender in the Media found men appeared in ads across the English-speaking world four times more than women.

It didn’t work. He responded with polite disbelief; I responded with polite bafflement.

“I know you’re not making up that you’ve noticed it,” I told him. “But I don’t think we’re seeing the same world.”

Here's what I learned while talking to Tom: He was divorced several years ago and is now estranged from his kids, whom he loves and misses deeply. He felt the courts had unfairly prioritized his wife in their custody battle. After that, he started to notice other ways women seemed to get preferential treatment. In 2016, he felt Hillary Clinton's supporters liked her only because she was a woman. He told me he thought there were a lot of men who felt the same way; I said I thought he was right about that part.

Here’s the other thing I realized while talking to Tom: Our worldviews are shaped by our experiences. We all obsess over our own scars until we start to think they’re symbols for broader injustice. We believe what we feel. And then we believe our feelings are facts.

So as Tom and I talked, during the same week that Donald Trump suggested he found Vladi­mir Putin more trustworthy than our American intelligence establishment, here’s what I wondered: How do you address beliefs when they’re not rooted in reality? How do you tell someone, I’m trying to treat your fears seriously, but your facts don’t exist? How, as individuals, and how, as a country?

A scientific experiment was Tom’s idea. He’s a data guy; he likes experiments. In this one, we’d make our own data by watching an hour of the same channel, coding ads by gender, by voice-over vs. on-screen appearance, main actor vs. extra. (Tom suggested that white men set up as the butt of a joke be assigned the special coding, “HB,” in honor of the ridiculous Harry Bentley from “The Jeffersons.”)

And as we planned, we veered into confessional. I told him that, to be honest, I might have noticed a slight decrease in white-man ads recently, too — but that I found it elating. There was a new motorcycle ad where the rider is revealed at the end to be a woman, and it made me cry. Could we actually be noticing the exact same thing: a slight decrease? But whereas I saw it as balance, he felt it only as loss?

It’s possible, he said. Anyway, he suggested the experiment would be a win-win for him. “Either I get to prove you wrong, or I’m wrong, and the world is less scary than I think it is.”

So, a Nissan commercial (one speaking white man, one nonspeaking white woman). A Kotex ad (one white woman). Realtor.com. Aveeno. Home Depot. Febreze.

Tom emailed the next day, attaching a spreadsheet and a note:

“In reviewing these statistics, I have to change my opinion. They do represent the population fairly well. There are some commercials that don’t have any white males in them, but this is to be expected.”

His spreadsheet matched mine, but the email was still stunning. Had his mind really changed? Over an hour-long experiment?

A few days later I phoned him, certain I’d missed something. Nope. “Give yourself a win,” he said. Something about that analytical-viewing hour had been cathartic. He’d begun reminding himself, for example, that if there was no man in one ad, there might have been in the previous one.

“That’s great,” I told him, unnerved by the tidiness of this ending and not sure what to do next.

The real question, and one he couldn’t answer, was why he’d been seeing a skewed version of television. Why even a potential slight decline had felt so personal, and made him so upset.

And what it meant that we had spent a night squinting at our televisions, trying to guess whether an extra in the background was a man or a short-haired woman. And why this small, strange act suddenly seemed to matter so much.