In its wild news conference Thursday, President Trump’s legal team promoted a very simple-sounding theory that seems likely to be central to its voter fraud allegations: that many precincts in the key states had more votes than actual voters, particularly in Michigan and Wisconsin.

“Well, in Michigan and Wisconsin, we have over-votes in numerous precincts of 150 percent, 200 percent and 300 percent,” Giuliani said.

Sidney Powell alleged that it was “up to 350 percent in some places.”

Within hours, the evidence for that claim began to crumble.

The source of Powell’s number is an affidavit from a security consultant in Texas named Russ Ramsland. Ramsland appeared on Lou Dobbs’s Fox Business Network show on Tuesday and detailed his claims.

“The things that you find in Michigan are amazing,” Ramsland began. “There are over 3,000 precincts where the presidential vote cast compared to the estimated voters … is 99 percent all the way up to 350 percent. Those kind of numbers don’t exist in the real world. So where did all those votes come from?”

But as the Powerline blog first reported, the affidavit made a major mistake. Its data wasn’t actually from Michigan; it was from Minnesota. What’s more, its conclusions about over-votes even in those Minnesota locations aren’t backed up data from the Minnesota secretary of state.

Here’s the applicable part of the affidavit, which Trump campaign lawyer L. Lin Wood has now filed as part of a lawsuit in Georgia:

Ramsland affidavit

Ramsland’s affidavit later focuses on alleged over-votes specifically in Wayne County, Mich., where Detroit is located:

Ramsland affidavit Part 2

The list of precincts next to their alleged over-voters is striking. But what was also striking to my fellow Minnesotans at Powerline was that those sound a lot like cities and towns in Minnesota. And indeed they are. It even lists the Minnesota precincts not just as if they are in Michigan, but specifically in Wayne County. The text states that, in Wayne County, “25 of those 47 precincts/townships show 100% turnout.” But it then lists 25 precincts from Minnesota.

The fact that an affidavit with such a glaring error would be not just cited by the Trump campaign but filed as a part of an actual lawsuit might be the biggest takeaway here. It certainly speaks to a lack of due diligence. And it casts doubt on virtually everything else in the affidavit.

Of course, errors can happen. And just because they got the state wrong doesn’t mean that the data might actually show something nefarious — albeit not where intended.

Giuliani’s falsehood-filled news conference, in less than 4 minutes
Trump campaign lawyer Rudolph W. Giuliani claimed at a news conference on Nov. 19 that President Trump lost the election because of a baseless conspiracy. (Zach Purser Brown/The Washington Post)

So what about the idea that it shows something amiss in Minnesota? On that count, too, the data is highly suspect, and it most certainly doesn’t speak to a plot to actually help Joe Biden.

It’s not exactly clear where Ramsland was getting his data on “estimated voters” in each precinct. (I have reached out to Ramsland for more information.) But his affidavit cites “SOS Est. Voters” — an apparent reference to the secretary of state. Yet the Minnesota secretary of state also has a detailed breakdown of its registered voters, and its numbers don’t show over-votes.

The biggest alleged over-vote in Ramsland’s affidavit is from Benville Township — the place where actual voters were allegedly 350 percent of estimated voters. According to the Minnesota Secretary of State’s election results, 63 people voted in the township. But its data show there were precisely that many registered voters in the township at 7 a.m. on Election Day, and then eight more registered on Election Day (Minnesota has same-day registration).

The story is the same for the second-highest alleged over-vote: Monticello’s 1st Precinct. There, 3,693 people were registered when polls opened, and another 385 registered on Election Day. The number of votes was 3,211.

The number of votes in each of the next three biggest alleged over-voting precincts (Monticello-2 and two precincts in Albertville) also fall well shy of the number of registered voters.

While it’s not clear what numbers Ramsland was using for “estimated voters,” one logical conclusion presents itself. As Max Hailperin, professor emeritus at Gustavus Adolphus College in Minnesota, theorized to me, the affidavit might have been relying upon incomplete “estimated voters” data from the Minnesota secretary of state in the days after the election.

In Benville, for instance, 18 of the 63 voters were recorded as casting absentee ballots. Absentees a type of vote that might very logically have been included in the secretary of state’s “estimated voters” numbers before others were, given they would have arrived the earliest. If the results were updated more quickly than the “estimated voters” numbers, there might have been a time in which the results showed 63 votes but the “estimated voters” number showed 18. And what’s 63 divided by 18? Exactly 350 percent.

Even sub-100 percent numbers in the current secretary of state data suggest very high turnout. But that’s one area in which transposing Minnesota data with Michigan data matters. That’s because Minnesota routinely has among the highest turnout rates in the country — and often the highest. The idea that we’d see 79 percent turnout in Monticello’s 1st Precinct is hardly remarkable. In fact, it’s not that dissimilar to past elections, including 2016, and 2020 was the highest-turnout election since 1900.

Another key reason the difference between states matters: the voting equipment used. Minnesota secretary of state spokeswoman Risikat Adesaogun confirmed to me that the precincts identified above didn’t use the Dominion software that the affidavit cites as being problematic and that the Trump campaign has baselessly linked to voter fraud. And even in Michigan, the Princeton University professor cited in the affidavit as raising concerns about Dominion has clarified that the state doesn’t use the specific software he warned about. He noted Michigan uses paper ballots which can be recounted by hand — “the state-of-the-art most-secure-known way of conducting elections.”

And then there’s one final key point. Let’s say, for argument’s sake, that the affidavit’s data are accurate based on numbers from somewhere else. What if the Minnesota secretary of state has just somehow done a really good job at covering its tracks by rigging its voter registration numbers too?

Well, even in that case, this would be a pretty harebrained plot to throw the election to Biden. While Ramsland’s affidavit appears to suggest these were precincts in Wayne County, in fact they are from rural Minnesota — i.e., Trump country. In Benville, where actual votes were allegedly 350 percent of total voters, Trump won 75 percent of those 63 votes. Trump also won about 6 in 10 votes in Monticello and more than 6 in 10 in Albertville. The results in each precinct are also similar to 2016, suggesting this wasn’t a bunch of votes being added to Biden’s tallies.

Giuliani at Thursday’s new conference repeatedly attacked the media for not reporting on these affidavits, saying they are real, bona fide evidence of fraud.

“It’s your job to read these things and not falsely report that there’s no evidence,” Giuliani said.

And it’s the Trump campaign’s job to actually do some research on the claims they spout in public and in court.

This post has been updated.