Republicans must accept the truth. Trump can’t actually win a general election.

Insanity is doing the same thing over and over again and expecting a different result. And yet, for the fifth election in a row, Republicans are abandoning sanity and putting their faith in the same nonsense polling that has been wrong over, and over, and over again.

To believe Donald Trump is winning Republican voters—in any state—by 40 percent is as fruitful as staying up all night for Santa to come down the chimney. Republicans are blindly stuffing their stockings full of garbage primary and general election polling with as much value as a Hunter Biden painting. Even the Wall Street Journal, which decried traditional polling in the past, claims the 2024 race is now Trump’s to lose—based on the exact same, still-flawed polling practices.

I voted for Trump in 2016 and 2020. If he actually wins the Republican primary, I would vote for him again, unlike the “Trump-only” cultists out there. But I do so knowing the cold, hard truth: While Trump may win the Republican nomination (a still-dubious proposition), Trump cannot win the general election.  The former president simply does not appeal to the vast majority of independents—not anymore—who may not like Biden, but loathe Trump. His appeal is limited to a very loud, but very narrow base, driven largely by paid influencers and X bots.

What worked in 2016 didn’t in 2020, and still won’t translate eight years later—despite the garbage polls claiming otherwise. They don’t make Trump likable to people outside of his MAGA base. Nor does getting booted from the Colorado and Maine primary ballots, however unconstitutional and likely to be overturned.

Most polling is junk. It’s not even valid enough to be called “junk science” because you cannot replicate results—the hallmark of validity. Even polling averages are junk because, no matter how many polls you use, when you start with garbage, you end with garbage—and the average of all that garbage is still garbage.

Why can’t we get it right? Because modern polling largely only reaches and gets responses from political zealots who are desperate to convince themselves—and the pollsters—they are in the right. No amount of “weighting,” applied by pollsters to fix their wild oversampling of the extremes, will work—it just makes the results even more unreliable. But that won’t stop the mainstream media from eating it up for clicks or campaigns from raising money off it, perpetuating the “value” of junk polls.

And that’s how Republicans will lose again in 2024. Entrusting the wrong data can’t predict the millions of people—and not just Democrats—who will come out and vote against Trump.

The more accurate measure of U.S. public opinion is “anti-polling,” which monitors actual behavior of very large populations over time—what they say and what they do, without them even being aware of the monitoring. Anti-polling measures sentiment through observational analysis, without the biases and errors that make traditional polling meaningless. Its purpose is not to frame media narratives or drive clicks; it is simply to reveal the truth, based on robust sample sizes of social media activity.

Observing over 40,000 swing overs across America, third-party data from Eyesover confirm that, while there are almost always more social media posts about Trump, his supporters are more likely to post, and post far more frequently. But that doesn’t make Trump popular among swing voters, or even right-leaning ones. Not does it make the Left’s fervent hatred subside or moderates interested in a 2020 redux—that ship has sailed.

A separate eight-month study of nearly 20 million home IP addresses makes the “Trump Prison” all too clear—a tiny sliver of the electorate (17 percent to 22 percent) that doesn’t grow, but from which his now-former supporters slowly get paroled. Tens of millions of Trump loyalists still exist, but they alone are not enough to swing a general election.

In Nevada, where more than 8,000 likely Republican primary voters were monitored with anti-polling, there is a similar story. Trump remains at 55 percent soft positive, but also 45 percent hard negative. Then there’s Iowa, where Trump can claim 31 percent of likely Iowa Republican caucus voters, while 34 percent are flat-out against him and fully 35 percent are neutral—but, if you’re not pro-Trump already, you’re most likely not changing your mind now.

The only question that matters is this: Do Republicans actually care about winning? If so, it’s time to put the polling and political failures of 2018, 2020, and 2022 behind us, and choose a winning candidate.

The clock is ticking before the first primaries. If Republicans keep hitting snooze on the truth, America will wake up to a nightmare.

Dan Backer is a veteran campaign counsel, having served more than 100 candidates and PACs, including the pro-DeSantis Ready to Win. He practices law as a Member of Chalmers, Adams, Backer & Kaufman LLC.