This article was produced by Voting Booth, a project of the Independent Media Institute.

"Truth is truth and numbers are numbers," said Arizona Senate President Karen Fann on September 24, as she summarized the most important finding in the long-awaited report from the pro-Trump contractors hired to assess the accuracy of the 2020 election results in Maricopa County, where two-thirds of Arizonans reside.

That bottom line -- yes, Joe Biden won, and his vote totals had increased while Donald Trump's totals had fallen -- was noted by almost everyone following American politics except for the people who arguably needed to hear it the most: Trump and his base of true believers.

"Yesterday we also got the results of the Arizona audit, which were so disgracefully reported by those people right back there," Trump said at a Georgia rally, pointing to the press as attendees cheered. "We won on the Arizona forensic audit yesterday at a level that you wouldn't believe!"

"We call on each state to decertify… Decertify… DECERTIFY… [their 2020 presidential results]," yelled Republican Arizona state Sen. Wendy Rogers, at a pro-Trump rally outside legislative chambers after the Senate's contractors reported that Biden won Maricopa County by 45,469 votes. (The official results showed Biden beating Trump by 45,109 votes countywide and 10,457 votes statewide.)

Despite the unexpected affirmation of the accuracy of Maricopa County's voting system, other parts of the reports revived and expanded previous conspiratorial claims. There were claims that more than 20,000 ballots might have come from wrong addresses—making the ballots uncountable. Or additional thousands might have come from voters who might have moved away, or people might have voted twice. Another Senate contractor, CyFIR, a cybersecurity firm, said that county election employees were seen on video in what might be images of them erasing key computer records from the 2020 presidential election.

Fann concluded the hours-long hearing by releasing a letter calling for Arizona's attorney general to investigate the alleged data erasure (which county officials deny) and other alleged problems. Fann said that further hearings would be held on the 2020 election.

Schism Between Reality and Fantasy Grows

The reaction by Trump and his base to the Senate's 2020 review, which has sparked copycat efforts in other swing states, underscores that these exercises have always been more about cultivating doubts about unpopular election results for partisan gain than about settling the lingering questions held by the most loyal supporters of a losing candidate.

One need look no further than coverage of the base's reactions by anti-Trump Republicans such as Charlie Sykes, editor of the Bulwark, whose newsletter on September 27 said, "If you have been living in a bubble of naivete or denial, you might have imagined that the results of the Cyber Ninja[s] audit in Arizona would usher in a New Era of Sobriety in our politics. Fat chance."

Still, there are some corners in the world of politics and elections where facts matter and conducting transparent audits where the methodologies and findings are fully released is the standard for credibility. The Cyber Ninjas still have not released their full data sets (they still are fighting public records requests in court), which has led many experienced election officials to comment that the public cannot trust anything they claim—including saying that their results from their controversial hand count was as close to the official results as they reported.

"Cyber Ninjas has no expertise in election audits, so it's no surprise that the methodology of their report makes it impossible to validate their findings," said Matthew Weil, Bipartisan Policy Center elections project director. "Real auditors show their work. Despite finding almost no change in the overall vote totals from 2020, they have succeeded in degrading faith in the results of a free and fair election and delaying discussions of real reforms to improve the voting experience."

Voting Booth, along with a team of experienced election auditors, obtained a draft copy of the Cyber Ninjas' report three days before its presentation in the Arizona Senate and worked on a section-by-section analysis that debunked its false claims and evidence. That analysis was shared with numerous reporters in Arizona and nationwide and election policy analysts as a baseline for their ensuing coverage.

The Cyber Ninjas' draft report insinuated that tens of thousands of voter registrations and paper ballots might have been illegitimate, forged, or even illegal. (In some cases, their final report rolled back or increased the number of voters and ballots that they alleged were questionable by several thousand, but they didn't change the evidence cited.)

The attacks on voter registrations, for example, were based on imprecise commercial data, not on government records used in elections. The forged ballots accusation indicated that Cyber Ninjas didn't know that ballots are printed for voters after they arrive at vote centers on Election Day. Under a microscope, the lines on those ballots appear less crisp than the lines that appear on mailed-out ballots, which are printed by an industrial press weeks ahead of an election.

Nonetheless, the Cyber Ninjas included recommendations for legislative action that are consistent with decades of GOP efforts to put partisan constrictions on voting and intimidate Democratic Party voters, based on clichéd false claims of fraudulent voting. Their legislative recommendations also included authorizing a new private election review industry, which would perpetuate their business model.

"The real reason the GOP is abetting Trumpist conspiracy theories is to justify restrictive voting rights laws, keep the base fired up for the [2022] midterms and lay the groundwork for letting partisan actors step in to influence the outcome of close elections," said Marc Elias, one of the Democratic Party's top lawyers, in an email touting his analysis of Arizona's 2020 review.

Whose Cover-Up?

While most Americans will not delve into the election administration details of Maricopa County's 2020 presidential election or the claims and evidence cited by the Cyber Ninjas, one of the foremost takeaways by pro-Trump contractors was the accusation that Maricopa County was caught destroying key evidence in February 2021. Election officials replied that their staff was archiving data, one of many responses and explanations offered via live tweets.

However, it appears to be the Cyber Ninjas who have been covering up their work and data—even after they issued their report. There were filings and hearings in two Arizona courtrooms on September 24 and 27 over the Cyber Ninjas' refusal to provide public reports, including the complete ballot and vote counts, to the Arizona Republic and public-interest groups. Their refusal is important, because that data will likely reveal the extent of the Cyber Ninjas' incompetence and underscore that Arizona-style "audits" should not occur elsewhere.

"They should never be hired again to do this by anybody," said Benny White, a longtime election observer for the Arizona Republican Party, lawyer, and part of the team of experienced auditors who have been using 2020 public election records to debunk the Senate's review. "They're incompetent, and they lie about what they've done."

White's comments come after reviewing a handful of the tally sheets included in the Cyber Ninjas' report. His team has spent months to identify how many ballots and votes are in each batch and storage box from the election.

"It's very difficult to discern where they got their numbers from," he said, pointing to several columns where spreadsheet fields are blank. "My question is: Why is there not better data there for everything?"

What unfolded between late April and mid-August was a pattern in which the Cyber Ninjas changed the review's focus—moving the goal posts—from retallying the presidential and U.S. Senate election totals to attacking voter rolls and mailed-out ballots and flagging possible cybersecurity issues.

Their early blunders are briefly noted in Volume II of the Cyber Ninjas' report, in a discussion of "quality controls." The report said that "all [handwritten] Tally sheets originally aggregated in the first three weeks of counting were re-entered in the new forms," meaning they had to be redone. Those sheets, which grew to more than 10,000 pages, then had to be entered into an Excel spreadsheet at computer terminals. The report said overhead video cameras were used to catch data entry typos. "The primary function of these cameras was to… demonstrate irrefutable evidence that the data entered was accurate."

By late June, the Cyber Ninjas knew that the hand count's results had differed from the official results by thousands of votes, Voting Booth was told at the time by insiders. The contractors never released the hand-count results and, throughout the summer, went to court to oppose releasing their records to the press. In early July, the state Senate purchased machines to count the number of paper ballots—not their votes—as a way to try to understand what was wrong with the hand count. Until they presented their report on September 24, the contractors never discussed the machine count results.

Meanwhile, White and his colleagues, who had been working for months to hold the Cyber Ninjas accountable, believe that the Cyber Ninjas panicked in late June. That was why they began a machine count of the number of ballots (not votes) in hope of finding new pro-Trump evidence, he said. Instead, that tactic backfired as it confirmed the number of ballots and votes and left no room for speculation about Biden's victory.

At that point, the Cyber Ninjas announced that they had to expand their investigation, which the Senate president allowed—and they revived the longstanding GOP strategy of attacking voter rolls, by alleging that there were tens of thousands of illegitimate voters and thousands of forged ballots. These claims and their specious evidence, all debunked on the eve of the final report's release, involved volumes of votes larger than Biden's margin of victory.

Above all, perhaps one statistic from the Cyber Ninjas' report stands out as an indicator of their lack of expertise as auditors. In the presidential election, they reported counting 2,088,569 ballots. In the U.S. Senate race they reported counting 2,088,396 ballots in the U.S. Senate race—a difference of 173 ballots.

This is a basic auditing mistake; there should be no difference in the number of ballots counted in the same election.

Steven Rosenfeld is the editor and chief correspondent of Voting Booth, a project of the Independent Media Institute. He has reported for National Public Radio, Marketplace, and Christian Science Monitor Radio, as well as a wide range of progressive publications including Salon, AlterNet, The American Prospect, and many others.