

In 2020 the democrats were calling the border wall racist and they won, then in 2024 it was “we’re the ones who are actually gonna build the wall, Trump’s all talk.” They literally tried to position themselves to the right of Republicans on the issue in order to win over the mythical centrists, and predictably what happened was that their support among Latinos broke down.
A lot of these people are religious and conservative, but were willing to vote for Democrats as long as there was substantial differences on race/immigration. But even if they were the “lesser evil” on immigration from a pro-immigrant perspective - something which they denied as hard as they could, by the way - if the difference didn’t appear substantial any more, if it was framed in technical arguments about how to do it rather than moral arguments about what to do, then many of them no longer saw it as damning and voted based on other issues where they’re more aligned with Republicans.
This is often what swing voters actually look like, by the way, and why pivoting to the right to capture them is often counterproductive. It turns out pivoting right on an issue where doing so directly harms millions of people so you can appeal to the dozen or so people who like Dick Cheney loses elections. Swing voters are a lot more complex than the idiotic “conventional wisdom” that just has everyone at a different point on a one dimensional left-right scale.
One easy trick that makes you immune to propaganda - simply respond “not sure” to every question you’re ever asked. It doesn’t really even save you though because they’ll just lump you in with the people who chose the wrong answer. The site repeatedly uses the phrase, “failed to identify as false” to group the “not sures” in with the incorrect responses.
There’s an almost endless way to present poll numbers and survey results to support whatever conclusion you like, you could say that “fewer than half the respondents were able to identify this claim as false,” or you could say, “80% of respondents avoided incorrectly labeling this claim as true,” depending on what narrative you prefer. And that’s assuming that the raw data itself, which comes from an internet survey, is reliable and representative.