Election Oversight Complaint: Error Pair Causes Undercount in Elections

A Williamson County municipal election on October 26, 2021, “produced end-of-day tape reconciliation reports that seriously undercounted the number of ballots actually cast.” If not for an astute poll worker and a small group of concerned citizens who formed a group called the Tennessee Voters for Election Integrity (TVEI), the anomalies and QR code failure errors would not have been investigated. The undercount and other issues were found in six Dominion scanners in three distinct voting centers in Franklin, Tennessee. While these were not the only issues found in Williamson, the undercounted ballots were of interest to two men investigating elections in Georgia because of identical issues discovered there.

In Georgia, two persistent citizens, Kevin Moncla and David Cross, who founded the Elections Oversight Group, discovered the same anomalies and QR code failure errors in 65 out of the 67 counties from which they were “able to obtain the requisite records.” The numbers are significant, representing “some 97% across the state’s” 159 counties.

Sadly, in both cases, in their attempts to investigate and resolve the anomalies, these citizens were met with outright obstruction and delays in obtaining records and answers at almost every turn. This column focuses on months of investigation that ultimately led to the complaint filed by Moncla and Cross on October 11 in Georgia.

In a lengthy conversation on Wednesday, Moncla told UncoverDC that his lawsuit is about more than anomalies found in election machines. Rather, it is the culmination of many months of investigation with the help of a small group of citizens in the state into the poorly monitored processes and procedures behind elections in Georgia. When business from his multimedia marketing company dried up during the pandemic, Moncla suddenly had a lot of extra time on his hands. While he had no particular expertise in the mechanics of elections, intuitively, he knew something wasn’t right about the 2020 election. It became his mission to investigate all aspects of U.S. elections to ensure the same mistakes would never happen again. He began to look at the many election processes, including the certification process, the checks, and balances at the polls, and ultimately issues with the machines themselves.

Moncla had long been investigating errors in Georgia, but when the Williamson County, TN election showed errors that were later referred to and confirmed by the Election Assistance Commission (EAC) in its March 31, 2022 report, Moncla’s findings in Georgia were substantiated. Incredibly, in the case of Williamson County, the EAC was able to replicate the errors, but they “admitted they do not know what caused the DVS machines not to count the ballots.” Moncla says the same errors had also been found in Pennsylvania, Michigan, and possibly Nevada. The Williamson County investigation resulted in the suspension of Dominion Voting Systems (DVS) equipment. DVS was replaced with ES&S—which isn’t much better.

Incidentally, when Moncla asked the author of the Williamson County EAC report whether this anomaly had occurred in other areas of the country—the author replied, “[the] EAC reported one report of the issue occurring. As part of the investigation process, the issue was communicated out to all affected jurisdictions. No additional reports were received.”

The Williamson County undercount event was a triggering event for the Moncla Cross complaint. While digging through Georgia’s elections, Moncla came upon information from the January 5, 2021, Senate run-off election in Coffee County. Astonishingly, the anomalies and QR Code failures in Coffee County were identical to the ones later found in the Williamson County election. A pattern began to emerge that Moncla and Cross could no longer ignore, and they began to expand their investigation.

The Williamson County Incident On October 26, 2021, in Franklin, TN, an “astute poll watcher meticulously documented the happenings at one of the polling locations as the polls closed. The next day on October 27, 2021, poll workers began their reconciliation process, which included hand-counting the paper ballots and comparing it to the number of ballots cast as reported by the two tabulators. One tabulator had 163 paper ballots, but the poll closing tape only showed 79 ballots counted. The second tabulator contained 167 paper ballots, and the corresponding poll closing tape showed only 19 ballots had been counted.”

Williamson County Poll Tape/Moncla Cross Complaint p. 3 The same scenario repeated in several polling locations, “with 7 of the 18 tabulators having scanned significantly more ballots than those counted.” Notably, those ballots that were not counted were, for some reason, “mistakenly identified by the machinery as provisional ballots” when an “error pair” was triggered. The assignment of the ballots to a provisional “folder” was the anomaly triggered by the error. It should be noted that these error events did not pop up or notify on the screen. Only after examining the tabulator audit files were errors and anomalies discovered. The full report and overview of the incident from TVEI can be found on their website here.

It is important to note, Moncla researched elections throughout the United States and found that not a single U.S. location allows for the assignment of a ballot to a “provisional folder” in a tabulator. Provisional ballots in America are set aside for manual review and are then cured. It is then they are scanned and tabulated. It bears repeating that in the U.S., there should be no option at all to “mistakenly identify” a ballot as provisional. (p. 4 of complaint). Keep all of this in mind when digesting the investigation results below.

The Investigation of the Williamson County Incident As a result of the incident, the Secretary of State was prompted to investigate. He found he could repeat the errors but could not identify a cause. The EAC then investigated and was able to identify the triggering error for the anomaly. According to the EAC report, “The investigation was conducted at the Williamson County Elections Commission facility on January 19 through January 22, 2022. This analysis was performed by both EAC-accredited Voting System Test Laboratories (VSTL), Pro V&V, and SLI Compliance. The EAC, Williamson County staff, TN SoS, and Dominion staff were present during the analysis.”

During their investigation of the audit log—also known as the tabulator System Log (SLOG)—EAC found “entries that coincided with the manifestation of the anomaly.” They found an “error pair” that consisted of a ‘security error QR code signature mismatch,’ and a warning message ‘Ballot format or id is unrecognizable’ indicating a QR code misread occurred,” similar to that found in DeKalb County, GA, below:

DeKalb/Moncla Cross Complaint p. 7 EAC also found that each instance of the error pair was always associated with the anomaly or the assignment of the ballot as provisional. Still, not every instance of the log entries resulted in the anomaly (provisional status). Furthermore, “the scanners correctly tabulated the ballots until the anomaly was triggered.” However, following the anomaly, ballots successfully scanned and tabulated by the ICP were not reflected in the poll reports on the affected ICP scanners.” In other words, once the anomaly was triggered, the ballots were not counted whether they were successfully scanned or not.

The EAC also found that “the anomaly suspiciously caused the tabulator’s protective counter not to increment. The protective counter is a legally required meter which counts every ballot scanned, including test ballots, for the life of the tabulator.” This resulted in the tabulators’ failure to count the ballots being scanned when the corresponding ballot images were hidden in the provisional folder.”

Moncla Cross Complaint/P. 5 The EAC report was frustratingly, ultimately “inconclusive” concerning the “direct cause of the anomaly.”

Dominion Says There is a Misread of QR Code EAC ultimately deferred to the Dominion investigation, whose explanation below begs credulity.

Dominion Explanation/Moncla Cross Complaint p. 4 QR codes are either read or not read. (p.4 of complaint) There is no such thing as a “misread” that “affects a certain part of the QR code.” And no, the Image Cast Precinct (ICP) scanner also cannot “mistakenly interpret a bit in the code that marks the ballot as provisional.” This can be confirmed by IT expert Jeffrey Lenberg, who spoke about QR codes in a discussion with Moncla and Cross on the September 29 CannCon Broadcast.

Oddly, per the complaint, the “fix” for the error/anomaly was to “modify the source code to reset the provisional flag presumably after each ballot is scanned. This does not address the cause, which has not been identified, and does not prevent a ballot [from] being erroneously flagged as provisional and then sent to the provisional folder.” With the source code patch, Dominion declared the “anomaly successfully fixed.” This “fix” is something akin to a mechanic resetting the check engine light on a car without fixing the root mechanical issue.

EAC Report/March 31, 2022/P. 4 Moncla and Cross correctly ask, “if the “erroneous code” was not due to malware and was a mistake by Dominion’s programmers, then how did it survive certification testing? This would also suggest that the “erroneous code” could have affected several past elections in these various locales unbeknownst to anyone. Dominion claims it only affected Democracy Suite 5.5B and 5.5C, but doesn’t state from what point in time.” They also point out that the Williamson County incident “established that a ballot can alter the behavior of the tabulator, including how and which votes are counted,” threatening the “integrity of the voting systems and our critical infrastructure.”

The Georgia Investigation Records obtained by Moncla and his team show that the two issues found in Williamson were found in 65 of 67 counties in Georgia, with a “rejected-then-accepted pattern” affecting “approximately 20% of all ballots cast in Georgia.”:

Moncla Lawsuit According to the complaint, “The deficiencies noted above are also associated with several instances in which ballots were found to be scanned by the tabulator but not reflected in the tabulator count. This, too, is consistent with the manifestation of the anomaly as found with the Williamson incident.” It is important to understand that both Williamson County and many counties in Georgia not only show the same error pair being identified but, just as was seen in Williamson, ballots were scanned but were not included in the tabulator results. Examples of this undercount in Georgia include:

Moncla Cross Complaint The QR code mismatch has also been confirmed with the higher-speed Image Cast Central (ICC) tabulator as well. Additionally, the same error pair was found in Coffee County for the 2020 general election.

Moncla and Cross Test Ballots With Third-Party Software “When the tabulator produces an error, the ICP ‘reverses’ or returns the ballot to the voter,” states the complaint. “Aside from a genuine mechanical or folded paper error, the ICP should reverse the same ballot for the same error no matter how many times the ballot is scanned (within acceptable tolerances). For example, A ‘QR code signature mismatch’ error should be reversed on the second, third, and 25th attempt; however, the logs and corroborating reports reveal that ballots are being reversed on the first attempt but accepted on the second or subsequent scanning attempts.” These issues were consistent with the Williamson County/EAC findings. This also happened “in county after county thousands of times” probably ruling out inaccuracy as a cause. As a result, “evidence strongly suggests that the error as initially returned is not really an error.”

Moncla and Cross decided to test the reversed ballots they obtained. They were able to identify the exact ballots because when they were “reversed.” Their images were all put in what is called the “Not Cast Images” folder. They show proof of the folder in log files on pages 11 and 12 of the complaint. Keep in mind that when ballots are tabulated, they are freshly printed, so there is little chance of corruption. If everything is in the right place and the ballot is not damaged, the ballot should scan properly. The explanation continues in the images below.

Moncla Cross Complaint/Pp. 11-13 Moncla’s team printed those ballots which were previously reversed because of errors. They then scanned them using a third-party QR code software available at zxing.org/w/decode.jspx that “Dominion tabulators are supposed to use.” Every QR code image they printed decoded “just fine,” according to Moncla. Moncla told UncoverDC that the decoder is “robust too because if you cover part of the QR code, the decoder still has the capacity to read it up to a 15% loss.” The question remains—Why did the ballots reverse and produce errors during voting when they did not do the same during their testing using the Dominion-sanctioned third-party decoding software?

Georgia Does Not Reconcile Counts for Early Voting During his investigation, Moncla found a complete lack of “basic election accounting” for early voting in Georgia that would permit poll workers to catch the discrepancies in the count. Unlike the reconciliation procedures in Williamson County, most election locations in Georgia lack a reconciliation process. “Specifically, the poll workers [in Williamson County] counted the number of paper ballots and then compared that number to the poll closing tape of the scanner, and the discrepancy was revealed.” In some counties, according to the complaint, the “early-voting tabulator closing process practiced in several counties was devoid of any reconciliation whatsoever and in violation of nearly all Rules and Regulations defining the same.”

Three Races Plus Coffee County Incident The complaint cites three incidents identical to the Williamson County incident, where the ballots were scanned by the tabulator but not counted by it. All were discovered “by happenstance.” The incidents detailed in the complaint were from the DeKalb County 2022 primaries, the Floyd County 2020 General Election, and the Gwinnett County 2020 General Election.

DeKalb The details of each are in the complaint; however, there are some notable entries. In the case of the DeKalb 2022 Primaries, the errors in the count were only discovered because of a demand for a hand-count by Michelle Long Spears, candidate for the May 24 DeKalb Commission 2 race. She came in third and was “seemingly out of the run-off.” In that race, “several precincts showed that she had received zero votes, including her own precinct where she and her husband had cast votes for her.” The hand count revealed she actually won. According to the complaint:

“The error in counting was purportedly caused by tabulators not being properly updated when a candidate had dropped out of the race—causing votes to be attributed to the wrong candidates. This same scenario was said to have caused the problem in Antrim County, Michigan, during the 2020 General Election in which Joe Biden erroneously received several thousand votes which voters had actually cast for President Trump. Yet this software deficiency remains.”

In the same DeKalb race, the hand count also revealed: “approximately 2,810 ballots that had been scanned by the tabulators, but not counted by the tabulators.” The race proved that “no effective reconciliation, accounting, or canvass process exists to protect the integrity of elections in Georgia.”

Gwinnett The Gwinnett County 2020 General Election is notable for its eventual connection between Marilyn Marks and an election supervisor named Misty Hampton Martin in Coffee County. Marilyn Marks was an authorized monitor of the audit process in Gwinnett County. Marks is associated with The Coalition for Good Governance and filed a Declaration in the Curling V. Raffensperger case. Curling v. Raffensberger “challenge[d] the constitutionality of the State of Georgia’s use of Direct Recording Electronic voting machines (“DREs”) and associated software systems in today’s era of heightened cybersecurity threats.” She would later use the case to falsely implicate a breach perpetrated by individuals responsible for the elections in Coffee County. Interestingly, Marks experienced some of the same ballot-counting issues during her audit oversight.

In her Declaration, Marks asserts explicitly that there were “significant problems with the Dominion server processing certain batches of scanned ballot images,” with technicians making “repeated unsuccessful efforts to process the ballots.” During the audit, a Dominion technical expert, David Moreno, was flown in to “remedy the vote tabulation problem.” Thousands of ballots were “caught up in the failure.”

Marks observed that it “appeared that Moreno made software code changes in real-time to circumvent the problem to force the system to process most, but not all, of the uncounted ballots.” She states, “after most of the ballots were processed and counted, Gwinnett quickly closed and certified the election. I estimated that at the time the election was certified, at least 1,600 ballots remained uncounted. I asked county officials repeatedly, in emails and on-site, for an accounting of these ballots but received no response.”

“A few days later, a statewide hand count audit of the presidential race was conducted. I was an authorized monitor of the audit process in several counties, including Gwinnett. According to the audit summary published by the Secretary of State, attached hereto as Exhibit 1, during the audit Gwinnett discovered 1,642 more ballots than were originally counted. This confirmed my belief that over 1,600 ballots had not been counted even after Dominion made real-time software changes and the Gwinnett Board of Elections certified the result.“

The complaint makes clear that “Ms. Marks’ Declaration indicates the alteration of software code within a previously certified voting system in real-time during its operation for an election in violation of Georgia election code. The actions Marks described clearly violated the voting system certification, and all use of that system should have been immediately halted and further use prohibited until such time as the system could be brought back into compliance and properly tested.” To date, nothing has been done to resolve the root cause of the anomalies and errors in Gwinnett County.

Coffee County Coffee County experienced issues with its elections beginning with the June 2020 primary. The county repeatedly asked for help, but “their pleas went unanswered.” According to the complaint, Donald J. Trump requested a recount following the November 3 election because Coffee County’s Board of Elections “unanimously refused to certify” the election, as referenced on pages 17 and 18 in the complaint.

Misty Hampton, now Misty Martin, is a seasoned Coffee County Elections Supervisor who explained some “important points” as “captured in the November 10, 2020, County Board of Elections Minutes.” Please see “Exhibit J” of the complaint. Briefly, Martin explained that ballots can be scanned multiple times and that elections often rely on honest people to do the right thing. However, she also said, “the honest person is not in every county.” Additionally, she explained, “not all counties have the same checkpoints in place.” She also stated, “during advance voting, the number on the scanner never matched the number of ballots voted.” Because she is a veteran Supervisor, Ms. Martin told the Board she always reconciled the number of “physical ballots with the number of ballots cast as reported by the scanner and that they never matched.”

Additional problems occurred in the January 5, 2021, Senate run-off election. Pursuant to a Trump lawsuit that was associated with his call to SoS Raffensberger and a litigation hold letter from the Smith and Liss Lawfirm, Martin authorized the imaging of the Coffee County election equipment in early January by an independent data firm Sullivan Strickler paid for by Attorney Sidney Powell. All manner of conspiracy theories and news articles have alleged that Martin hired “election deniers” to breach data on the equipment. However, according to Moncla, nothing of the sort occurred. Sullivan Strickler merely imaged the equipment and secured it with the proper chain of custody for potential review. In addition, it was well within the purview of Martin to authorize the imaging of the equipment, especially in light of the litigation hold letter requiring her to preserve all platforms in her possession, or she could be subject to “claims for damages.”

Litigation Hold Letter Weeks after the imaging took place, two investigators were asked to go to Coffee County to attempt to reproduce the ballot rejection/miscount issue. The investigators were Cyber Ninjas CEO Doug Logan, who ran the forensic audit for the Maricopa County 2020 election, and Jeff Lenberg, who analyzed voting equipment in Antrim County, Michigan, and Otero County, New Mexico. Neither Logan nor Lenberg examined the equipment or any of the logs. They merely scanned ballots per the complaint below:

“Mr. Lenberg had the Elections Supervisor run a mock election (Mr. Lenberg had the Election Supervisor control the machines). An equal number of ballots were created for President Trump and Joseph Biden (20 each), which were then scanned several times on an ICP. Out of approximately 480 ballots scanned, 15% of Trump ballots were reversed due to error as opposed to only 2.5% of those ballots for Biden. In other terms, ballots were being reversed at a ratio of 7:1, Trump to Biden.“

The complaint also states that Lenberg’s findings support “that which was witnessed in Coffee County by Cathy Latham on January 5, 2021, Senate run-off after the polls closed.” From Ms. Latham’s affidavit from “Exhibit I” in the complaint:

Cathy Latham Testimony Here’s where Coffee County’s Misty Martin and Marilyn Marks’ from the Gwinnett County audit intersect. Marilyn Marks had been involved in the ongoing Curling v. Raffensberger lawsuit. For reasons still unclear, Marks elected to request materials from Coffee County to add to her civil suit while suddenly alleging Martin “breached” the security of the equipment, even though she had known for over a year about the imaging of the equipment.

Holy smokes!!!

An equal number of ballots were created for President Trump and Joseph Biden (20 each), which were then scanned several times on an ICP. Out of approximately 480 ballots scanned, 15% of Trump ballots were reversed due to error as opposed to only 2.5% of those ballots for Biden. In other terms, ballots were being reversed at a ratio of 7:1, Trump to Biden.“

How can that be? if it was just a coding error, then it would apply to both equally.

Can you imagine the number of undercounted Trump votes??

One tabulator had 163 paper ballots, but the poll closing tape only showed 79 ballots counted. The second tabulator contained 167 paper ballots, and the corresponding poll closing tape showed only 19 ballots had been counted.”

Where did the errors happened?

In Georgia ... QR code failure errors in 65 out of the 67 counties from which they were “able to obtain the requisite records.” ... Williamson County, TN election showed errors that were later referred to and confirmed by the Election Assistance Commission (EAC)... the same errors had also been found in Pennsylvania, Michigan, and possibly Nevada..

What was the error?

ballots that were not counted were, for some reason, “mistakenly identified by the machinery as provisional ballots” when an “error pair” was triggered Furthermore, “the scanners correctly tabulated the ballots until the anomaly was triggered.” However, following the anomaly, ballots successfully scanned and tabulated by the ICP were not reflected in the poll reports on the affected ICP scanners.” In other words, once the anomaly was triggered, the ballots were not counted whether they were successfully scanned or not.

How come the error was not caught?

Briefly, Martin explained that ballots can be scanned multiple times and that elections often rely on honest people to do the right thing. However, she also said, “the honest person is not in every county.” Additionally, she explained, “not all counties have the same checkpoints in place.” She also stated, “during advance voting, the number on the scanner never matched the number of ballots voted.” Because she is a veteran Supervisor, Ms. Martin told the Board she always reconciled the number of “physical ballots with the number of ballots cast as reported by the scanner and that they never matched.”

This is horrible!! the testing should have been done at the company not with biases (and criminal) poll workers.. I can see that an error like this (once a particular error is trigerred, ballots after that are not counted) but how could it affect Trump ballets more than Democrat ballots?? And how come two years after the event and years before the machines were deployed and supposedly tested this was not caught??

This is evidence that should be presented at the January 6 trials and plastered all over the News media.

The first thing to do is to remove the machines as ways to count the ballots

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.