Kellyanne Conway speaks to the media after the third presidential debate in Las Vegas. (Andrew Harrer/Bloomberg)

It was an inauspicious launch to what’s semi-affectionately become known as Trump TV: The campaign’s nightly broadcast of what’s going on in Trumpland without the “spin” of the lamestream media. Nothing but pure, unfiltered truth from Team Trump, delivered live on Facebook from the bright beige inner sanctum of the campaign’s Trump Tower headquarters. The opening seconds of the broadcast included a boom mic dipping into the shot and a bit of confusion about when, exactly, the hosts were live (about 15 seconds before they realized it). Then the inaugural guest was introduced — Donald Trump’s campaign manager Kellyanne Conway — and the real-keeping got underway.

We can jump ahead to the last thing Conway said during her brief segment (before the broadcast went to a sort of commercial, for some reason).

“We will win,” Conway said, looking into the camera with a smile. “Let me just say unequivocally: We will win.”

How was she so certain? She’d just explained:

Of the nine major swing states that were competitive in 2012, Mitt Romney only won one of them. We’re winning — we're up or within the margin of error in about six or seven of them beyond the one that he won. It’s very exciting. You’re looking at Florida, North Carolina, Ohio, Iowa. Then we add Nevada, we add New Hampshire, we’re in Colorado, we’re in Pennsylvania. We’re competitive in many places.

We’ll come back to that. Before we do, we’ll note that Conway had gone into even more detail about Trump’s path to victory in an email sent to supporters earlier Monday. (The email took the form of a private background memo and culminated in a request for contributions for voter turnout work.)

Conway’s detailed walk-through:

Donald Trump must reach 270 electoral votes to win the presidency. We’ve outlined a map to show where we currently stand as well as the possible paths forward to win.

While we are not taking any state for granted, those colored red are states that we can safely assume will vote for Donald Trump. Blue states favor Hillary. Blank states are too close to call.

Over the past month, polls have shown us winning Iowa, Ohio, Maine, Florida, Nevada, and North Carolina. If we maintain our leads in those six states, we can reliably claim 266 electoral votes. Hillary can claim 193. But we’d still have 4 electoral votes to go.

Polls show us close in New Hampshire, Colorado, and Pennsylvania. Winning just any one of those states would lead us to victory.

First of all, the math Conway presents on that map is broadly correct. Conway spots Hillary Clinton 14 states and sets aside 22 for her candidate. If you give Trump the six states where “polls have shown [them] winning,” Trump gets to 266 electoral votes. Just need one of those other four states!

The only problem is that the foundation of Conway’s argument — that map — is so bafflingly distant from reality that it’s hard to think this isn’t somehow an inside joke. A more accurate version of that map would look like the one below, with any state that’s got a margin of between 2 and 5 points in the RealClearPolitics polling average colored lightly for one candidate or the other. Gray states are toss-ups.

This map comports with the state of the race, according to public polls. I mean, look. Clinton has a 6.2-point lead in Pennsylvania. She’s up 6.5 points in Wisconsin. She’s up 10 in Michigan. She leads in Colorado by 8 and Virginia by 10. Those are margins that put those states easily out of reach. Put another way, polls do not show Trump close in Pennsylvania, New Hampshire and Colorado — they show Clinton leading by 6, 7 and 8 points, respectively.

Using that map, Clinton has 323 electoral votes, and a victory. Move all of her light blue states into the “maybe” column and she is at 273. She has won.

Conway says that over the past month, polls have shown Trump winning Iowa, Ohio, Maine, Florida, Nevada and North Carolina. That’s true except for Maine, where there are far fewer polls and Clinton has lead in eight of eight. In Florida, Clinton has led in 19 polls in the past month; Trump has led in two. In Ohio, Clinton has led in 11, Trump five. In Nevada, Clinton’s led in 14; they’ve been tied in two. In North Carolina, Clinton’s led in 19 and Trump two. In averages of recent polling, Clinton leads in Florida, Ohio, Nevada and Maine. Leading in a poll over a month doesn’t mean you’re going to win on Election Day.

But that’s why Conway used the language she did in that Trump TV segment.

It’s true that Romney won only one of the nine closest states four years ago. (In fact, he won only one of the 11 closest.) But he was “within the margin of error” in the other eight, assuming a polling margin of error of 3.5 points. Margins of error are bell curves; you might have a result that’s at the low end of the range of possibilities, and your opponent might be at the high end. If your support is 3.5 points too low and his is 3.5 points too high, there are scenarios in which you’re 7 points closer than the overall poll suggests. But it’s more likely that you’re both closer to the fat middle of the curve — and that the result will be closer to those top-line numbers than a surprise result. In 2012, most of the final results matched what late polling showed.

The numbers don’t show a close race. They show a race that it will be very hard for Trump to win without a massive, collective underestimation of his levels of support. This was the dream that powered Romney backers in 2012: In that (much closer) race, observers assumed that the polls overrepresented how much President Obama's base would turn out. As it happened, those voters were underestimated.

This isn’t the argument that Conway — who runs a polling shop — was making. She was arguing that Trump had a viable path to victory, using, in the email, a map that resembles some other race than the one that we’re in. In reality, the race we’re living through shows Trump is at risk of losing Arizona, Texas and Georgia — all states that Romney won. If Trump’s “within the margin of error” in Pennsylvania, he’s even closer to losing the three states just mentioned. At this point it seems almost probable that he'll lose Utah, a state that could hand its electoral votes to independent Evan McMullin. The map Conway sent in the email was the best semi-believable map they could think of, and even then, Trump struggled to reach a tie.

But on Trump TV, viewers didn't get any of this “spin.” They just got the unvarnished truth, straight from the people responsible for getting Trump into striking distance: Trump’s in striking distance. On Nov. 8, I fully expect Trump TV to say that Trump actually won.

After all, Conway said they would. Unequivocally.