Perhaps it’s a function of the era in which he owned casinos in Atlantic City. Perhaps, instead, it’s just a holdover from the always-winning, always-deal-making persona that he carefully cultivated over the course of his lengthy career in business.

Whatever the reason, President Trump has a habit of claiming to have predicted upset victories despite fairly obviously having not predicted them. It’s like the kid on the playground who can never be wrong, the one who insists that he does know what the vocabulary word means, he just wants to see if you do.

Next week, a new book by Politico’s Tim Alberta will start shipping, meaning that members of the media have by now gotten a sneak peek. Journalist Yashar Ali was one who got an early look at “American Carnage,” and relayed on Twitter one exchange Alberta had with Trump during an interview.

Trump, Alberta writes, predicted the primary victory of Rep. Alexandria Ocasio-Cortez (D-N.Y.) in late June 2018.

“Watching television in the White House earlier that summer with some of his political advisers, the president says he caught a glimpse of the Democratic insurgent on a cable news program,” Alberta writes in the excerpt shared by Ali. Trump was “starstruck” at the challenger to longtime incumbent Joseph Crowley.

“Trump says he told his team to call Crowley ‘and tell him he’s got himself a problem; he better get off his fat ass and start campaigning,'” the excerpt reads. “The president says they laughed him off, promising him that Ocasio-Cortez had no chance. Later, when she won, he took the opportunity to remind everyone that they had similarly underestimated him.”

Wow. Trump, political wunderkind, spotted something special in the emerging Democratic star and, for some reason, wanted to warn the “slob” Crowley that a political iceberg was looming. But, woe unto Crowley, Trump’s shortsighted advisers said not to, and the president uncharacteristically listened.

Or so Trump told Alberta. But there’s no reason at all to think that this anecdote is true.

For one thing, Ocasio-Cortez wasn’t on national cable news programs in the spring of 2018. A search of the Internet Archive’s cable news closed-captioning database from Jan. 1 to June 30, 2018, returns no hits before her June 26 primary victory.

A search of the news indexing site Nexis shows only one article from a major cable news network during that same period: A CNN article from the day of the primary.

Ocasio-Cortez did make several appearances on NY1, the local cable network on New York City’s Spectrum cable provider. On May 17, she was interviewed by the network’s Errol Louis. On June 15, the network hosted a primary debate. On both dates, Trump was in Washington, where it’s safe to assume he doesn’t have access to NY1. And even if he did and even if he were watching television on those dates, would he have interrupted his steady diet of Fox News to watch a potential challenger to a lower-tier member of the Democratic House leadership?

Over the course of her primary, Ocasio-Cortez did attract national media attention — though largely from more liberal outlets like The Intercept or Splinter, which Trump and his team seem unlikely to have seen. A campaign ad her team released in May went viral, but it’s not clear that it would have permeated Trump’s media bubble.

Bear in mind, Trump has lied about similar things before. He still insists that, while in the United Kingdom in 2016, he predicted the success of that country’s referendum to leave the European Union. He didn’t.

Shortly before the vote, he said that he thought Britons should vote to leave but that no one should listen to him because “I haven’t really focused on it very much.” The day after the vote, he celebrated the result, seeing in it a lesson about his own campaign’s long odds. Predicting something after it happens though, doesn’t really count.

Nonetheless, Trump has repeatedly claimed that he had predicted the success of the referendum. It’s not clear what value Trump sees in this claim or his claim to Alberta about Ocasio-Cortez. The simplest answer may be just that he likes to be seen as politically omniscient, as the guy who can keep proving the pundits wrong.

There’s just one other problem with this particular assertion of his own prescience. It’s undercut a bit by a tweet he sent shortly after Ocasio-Cortez was declared the winner.

Of course, that was well before Ocasio-Cortez became the media sensation that she is today. It’s more useful now for Trump to claim to have predicted her rise since people know who she is and because she’s become a potent foil for his politics.

When she actually won the primary, though, that wasn’t the case. Her win was useful as a way to goad Democrats, a stunning result that could be used as a way to suggest that Democrats should treat him more favorably.

It wasn’t a result that no one saw happening, to be fair. But Trump was almost certainly correct that it was not a result that he saw happening.