Did Dominion Systems defraud voters in Antrim County, Michigan — and by extension the entire nation? A court ordered the release of an outside analysis of the election operation yesterday, and its conclusions have Donald Trump and his supporters claiming to have the smoking gun that should overturn the election. Dominion’s system has intentional design errors “to create systemic fraud and influence election results,” declares Russell Ramsland.

Not so fast, say Michigan election officials and Dominion:

State officials are disputing a report on Antrim County’s voting equipment — signed by a consultant who confused Michigan and Minnesota voting districts in an earlier election analysis — that says the county’s equipment “is intentionally and purposefully designed with inherent errors to create systemic fraud and influence election results.”

Michigan Elections Director Jonathan Brater said in a weekend court filing the report “makes a series of unsupported conclusions, ascribes motives of fraud and obfuscation to processes that are easily explained as routine election procedures or error corrections, and suggests without explanation that elements of election software not used in Michigan are somehow responsible for tabulation or reporting errors that are either nonexistent or easily explained.”

And Dominion Voting Systems, the company whose equipment is used in Antrim, issued a statement saying it is the subject of a “continuing malicious and widespread disinformation campaign” intended to undermine confidence in the Nov. 3 election.

Contrary to the impression some have, this report is not an independent analysis. Ramsland has been working on Team Trump’s legal challenges for some time. This appears to be a more formal version of the affidavits that Ramsland and attorneys have been submitting all along. And, as the Detroit Free Press points out, Ramsland’s efforts have not been entirely credible — especially in Michigan. His first attempt to prove Dominion was corrupt in Michigan analyzed the results in nearly two dozen towns and cities, but it turned out that all of them were in Minnesota, not Michigan.

Despite this error, Ramsland and Team Trump allowed that affidavit to continue uncorrected for at least a week. Before it got corrected, Sidney Powell was still basing her claims in several states in part on Ramsland’s erroneous initial affidavit. That matters when it comes to weighing credibility in court, as John Hinderaker explained at the time that he uncovered the errors:

Evidently a researcher, either Mr. Ramsland or someone working for him, was working with a database and confused “MI” for Minnesota with “MI” for Michigan. (The postal code for Minnesota is MN, while Michigan is MI, so one can see how this might happen.) So the affidavit, which addresses “anomalies and red flags” in Michigan, is based largely, and mistakenly, on data from Minnesota.

This is a catastrophic error, the kind of thing that causes a legal position to crash and burn. Trump’s lawyers are fighting an uphill battle, to put it mildly, and confusing Michigan with Minnesota will at best make the hill steeper. Credibility once lost is hard to regain. Possibly Trump’s lawyers have already discovered this appalling error, and have undertaken to correct it. But the Ramsland Affidavit was filed in Georgia just yesterday.

More to the point, Ramsland’s claims had been heard already in challenges to the election in Michigan. They didn’t convince courts at that time, nor does it appear that the court is yet convinced now. This challenge actually relates to a down-ballot county election, not the presidential election itself, although it’s clearly intended as a political argument for the latter. Don’t forget, too, that the Department of Homeland Security has specifically rejected Ramsland’s theory about Dominion.

As with all affidavits, the test is how the claims fare in court. To put it mildly, they have not produced a track record of success. Dominion and the state of Georgia point — with good reason — to the hand recounts performed in that state as affirmative rebuttals to Ramsland’s claims. As has been explained ad nauseam, the Dominion systems produce a paper ballot that voters check before submitting them to the precinct. If Ramsland’s analysis were correct, either the hand recount would have shown a massive shift in the results, or hundreds of thousands of voters would have testified as to getting an incorrect ballot from the machine. Neither happened; the recounts have all corroborated the original election results. In Michigan, election officials testified to the incorrect assumptions and hearsay qualities of this and other affidavits in support of Trump’s desire to throw out the results. In all cases, courts have dismissed the cases as unsupported by actual evidence.

Dominion should make their rebuttal public too, if for no other reason than to help people balance the two claims, just as courts have. That would certainly be helpful to people who keep hearing claims like Ramsland’s but aren’t clear on what the responses have been. They should be clear, however, about Ramsland’s track record and the track record in courts on these claims.

Update: There were a number of problems with Ramsland’s affidavit in Michigan even after correcting the city data, as Politifact pointed out three weeks ago:

For instance, Ramsland claims that Detroit saw a turnout of 139.29%. The city’s official results show that turnout in the city was actually 50.88% of registered voters. …

It is difficult to imagine that a turnout rate above 100% — let alone 782% in the City of North Muskegon or 461% In Zeeland Charter Township — would have escaped election officials compiling the statement of votes cast. But beyond the implausible turnout rates Ramsland alleges, there are other glaring problems with the list. Shelby Township is named twice. So is Zeeland Charter Township, with two vastly different turnout rates: 90.59% and 460.51%. Ramsland lists “Fenton” without specifying Fenton City or Fenton Township. But the turnout Ramsland lists for Fenton does not match the turnout in either jurisdiction.

The actual turnout statistics reveal the inaccuracy of Ramsland’s numbers. His figure for North Muskegon is off by a factor of 10: The actual number is 78.11%, not 781.91%. For Zeeland Charter Township, he inflated the turnout nearly sixfold. For Grout Township and the City of Muskegon, his number is more than triple the correct number.

And there is more besides that. Ramsland is at best very sloppy at his tradecraft. Adjust your estimations of credibility of this analysis accordingly.