Could Utah go Democrat for the first time in more than 50 years? If this sounds like a political fantasy-league bet, you’re not far off. A purported new poll from the Salt Lake Tribune and the Hinckley Institute puts Hillary Clinton slightly ahead of Donald Trump after the national conventions, 36/35. The only problem with the poll is its composition, which we’ll get to in a moment:

Sounds bad for Trump, eh? Well, there’s one wee issue with this poll … it doesn’t exist. As in literally doesn’t exist, and not just a Bidenesque “literally.” The Washington Post’s Philip Bump wondered why he couldn’t get to the raw data, either at the Tribune or Hinckley websites, and got this very strange explanation:

The KUTV report indicates that Clinton gets 36 percent of support to 35 percent from Trump — leaving 29 percent of the electorate going somewhere else. This is actually quite similar to a SurveyUSA poll from June showing Clinton and Trump tied at 35 with Gary Johnson — whose campaign headquarters is in the state — pulling 13 percent. That leaves 17 percent of voters undecided. …

The Salt Lake Tribune doesn’t have a story about a new poll at its website. The Hinckley Institute at the University of Utah doesn’t have any mention of it either. In an email to The Washington Post, the Tribune’s Dan Harrie confirmed that there in fact wasn’t a new survey from the paper. He speculated that KUTV was simply picking up a convention post-mortem the paper ran over the weekend — though there’s no reference to a 36-35 result.

In short: Utah, like any state in any year, could break out of its long-standing voting pattern. There is no new post-convention poll suggesting that the race is close, though, and it’s not fair to assume that Clinton will convert a tie to a win once all of the state’s Republicans weigh in.

So … the Tribune just made up a poll? That’s certainly a novel approach to political reporting — maybe — but it’s not like we have a dearth of polling on which to report. Next time a newspaper feels a need to make something up for content, how about offering some coherent and consistent policy positions from both candidates? Naaah, no one would ever buy it.

Anyway, it’s true that SurveyUSA conducted a poll for the Tribune in June, but that’s 48 days old now. The result was 35-even, not 36/35 for Hillary (a small difference, but still a difference). Interestingly, RealClearPolitics doesn’t include the SUSA poll in its aggregation for Utah, and it shows no other new polling since mid-June. A Gravis poll in the beginning of the month put Trump up seven points at 36/29 — still not a terribly good result for a Republican in Utah — and a UtahPolicy poll put him up nine points two weeks later in a four-way race, 36/27.

It’s possible that Trump has torpedoed himself so badly in deep-red Utah that Hillary would suddenly become more popular, but it seems unlikely, especially since we haven’t seen that kind of sudden inflection in less-Republican states. Wouldn’t we be seeing that first in North Carolina, New Hampshire, or Florida?

One other key sign that this isn’t happening: Democrats aren’t spending any time or money in Utah. Instead, they’re aiming at neighboring Arizona, where Team Hillary thinks the Hispanic vote will give them a pathway to flip the state to the blue column for the first time since 1996. And Team Trump has taken notice, according to Politico’s Theodoric Meyer:

Prominent Arizona Democrats have been urging Bill Clinton to press the campaign to direct more money into the rapidly changing, Hispanic-heavy state, which he carried in 1996, the last time it went Democratic in a presidential election. They continued that push with officials during last week’s Democratic National Convention, making the case that Arizona should be on equal footing with a state like North Carolina, which President Barack Obama carried in 2008 and where Democrats are feeling increasingly confident.

There are already signs Arizona is on Trump’s radar: The businessman’s campaign has identified the state as a potential battleground, and recent polls have shown a close race between him and Clinton. Then there’s Pence’s rallies Tuesday with voters in Phoenix and Tucson. In 2012, neither presidential candidate held a public event in Arizona after the party conventions. …

Democrats here believe a confluence of factors put Arizona in play, four years after Mitt Romney carried the state with 54 percent of the vote. Trump is deeply unpopular among Hispanics, as is Maricopa County Sheriff Joe Arpaio, a prominent Trump ally. And Arizona has a higher percentage of eligible voters who are Hispanic than any other battleground state.

Democrats also say it’s easier to build a field operation in Arizona than in other states, because more than three-quarters of voters are concentrated in just two metro areas: Phoenix and Tucson. The state Democratic Party already has 100 paid staffers working on voter registration, as well as hundreds of volunteers.

This seems less of a phantom, but perhaps still in the realm of pipe dreams. It’s true that the Hispanic vote will weigh heavily on Trump’s campaign in Arizona, but so will immigration with other voters. The problem for Trump will be turning out the latter vote to compete with the former. So far the campaign hasn’t put much effort into traditional GOTV operations. If they don’t turn that corner in Arizona, they could well lose the state. Polling in June put this in the toss-up category.

However, don’t lose the forest for the trees here. If Arizona’s in serious play, that’s very bad news for the GOP. Fighting to keep a former stronghold (where it won 54% in 2012’s national loss) will draw resources that should have gone to competing in places like North Carolina, Florida, Virginia, and Ohio, not to mention Pennsylvania, which is the key to Trump’s Rust Belt strategy. That suggests that Trump’s redrawing the Electoral College map as promised, but that Republicans aren’t going to like the results.