IF YOU MISSED 2020, GET READY FOR 2026.
GOVERNMENT
The sworn testimony. The voting-machine companies. The missing memory cards. The thousands of ballots. The election results that changed. And the evidence trail now demanding national attention.
MIND BEND THEORY | SPECIAL INVESTIGATIVE REPORT
November 14, 2026
In 2020, Congress heard warnings about election software, data transfers, voting-machine access, unexplained numerical changes, and the reliability of election records.
Six years later, the November 2026 midterm elections have produced documented incidents involving missing ballot results, unprocessed absentee ballots, reporting-system errors, and at least one legislative race in which the apparent leader changed after omitted results were discovered.
The companies are identifiable.
The witnesses are on record.
The numbers are documented.
And new allegations involving election-system access and personnel have brought the technology behind American elections back into the national conversation.
If you missed the congressional testimony in 2020, you need to understand what happened in 2026.
Because this time, there is a growing collection of specific incidents that can be examined against official records.
And the consequences of getting those records wrong are not theoretical.
In Georgia, a missing memory-card upload changed who appeared to be winning a legislative race.
In Idaho, results from 2,343 ballots were absent from election-night reporting.
In Iowa, approximately 2,000 absentee ballots remained unopened until officials discovered them after Election Day.
These incidents have different causes. They do not establish a coordinated operation.
But they demonstrate why every election result must be traceable to the underlying ballots and why the public deserves an explanation whenever that chain breaks.
I. DECEMBER 16, 2020: THE SENATE HEARING THAT DESERVES ANOTHER LOOK
The United States Senate Committee on Homeland Security and Governmental Affairs convened a hearing titled Examining Irregularities in the 2020 Election.
Chairman Ron Johnson and Ranking Member Gary Peters presided over testimony from six sworn witnesses.
Witness
Position and role
Kenneth W. StarrFormer federal judge; constitutional and election-law analysis
Donald Palmer
Commissioner, U.S. Election Assistance Commission
James R. Troupis
Trump campaign attorney in Wisconsin
Francis X. Ryan
Pennsylvania state representative and certified public accountant
Jesse Binnall Trump campaign attorney in Nevada
Christopher C. KrebsFormer director, Cybersecurity and Infrastructure Security Agency
Their testimony covered election laws, absentee ballots, data systems, voting equipment, cybersecurity, and allegations of fraud.
The hearing was contentious.
Campaign attorneys argued that serious irregularities had occurred.
Krebs defended the security of the election.
Peters challenged the allegations and pointed to court decisions rejecting election challenges.
The testimony is important because it preserves the competing claims and technical explanations in one official congressional record.
II. THE 2020 NUMBERS: WHAT WITNESSES ACTUALLY ALLEGED
Pennsylvania: Data changes and access-control questions
Representative Francis X. Ryan testified about discrepancies in Pennsylvania's election records.
His presentation included:
Issue described in testimony Figure
Philadelphia mail-ballot total displayed 508,112
Mail ballots Ryan said had been issued 432,873 Difference
75,239 Change between November 15 and 16 returned-ballot datasets 75,505
Alleged mailing-date irregularities 154,584
Alleged records involving voters over age with a count of 1001,573
Ryan acknowledged that the Philadelphia reporting figure had subsequently been corrected.
His questions concerned access permissions, transaction histories, and the ability to reconstruct changes.
These figures were allegations and observations presented by a witness, not findings that the corresponding numbers represented fraudulent votes.
Nevada: The allegations presented by Jesse Binnall
Binnall alleged approximately 130,000 unique instances of voter fraud.
He cited voter-list comparisons involving names, addresses, death records, residency, and citizenship.
He also described claims attributed to two Clark County technical employees concerning changes in data associated with USB drives.
The Nevada election challenge was rejected, and the testimony did not establish that 130,000 fraudulent votes existed.
But the allegation identified specific records that would be needed for a technical examination: device data, original logs, custody records, and authenticated transfer histories.
Wisconsin: More than 200,000 ballots challenged
Attorney James Troupis argued that more than 200,000 absentee ballots were affected by legal or procedural objections.
His categories included incomplete certificates, missing clerk initials, ballot collection procedures, indefinitely confined voter status, and application requirements.
The number represented ballots challenged under the campaign's legal theories—not 200,000 proven illegal votes.
Wisconsin's recount confirmed Biden's victory.
The distinction between an allegation, a documented error, and an adjudicated finding is essential to understanding this hearing.
III. THE COMPANIES AND INSTITUTIONS IN THE 2020 RECORD
The congressional hearing discussed a complex network of voting-system manufacturers, testing laboratories, government agencies, and election-data organizations.
Dominion Voting SystemsVoting-system manufacturer and supplier
Election Systems & Software (ES&S)
Voting-equipment and software manufacturer
Hart InterCivicVoting-system manufacturer
Edison Research
Unofficial election-night data reporting
Associated Press
Election-result collection and media reporting
U.S. Election Assistance Commission
Voluntary voting-system certification and oversight
Pro V&V Voting-system testing laboratory
SLI ComplianceVoting-system testing laboratory
NISTTechnical standards and laboratory-accreditation responsibilities CISA
Federal election cybersecurity coordination
Smartmatic also became a prominent subject of the broader 2020 election-technology controversy and subsequent litigation.
Additional individuals named in that broader history include John Poulos, Dominion's former CEO, and Eric Coomer, a former Dominion executive who denied accusations of election manipulation and pursued defamation litigation.
DLA Piper, attorney Charlie Cain, and political adviser Boris Epshteyn appear in newer public allegations concerning legal relationships and proceedings.
Those allegations are distinct from the six witnesses' sworn testimony and should not be presented as findings made by the Senate committee.
The same applies to allegations involving personal relationships with NIST employees: an alleged association is not evidence of institutional wrongdoing.
These distinctions matter because the companies and individuals did not all perform the same function, work on the same systems, or appear in the same evidence.
IV. THE TWO DATA STREAMS: WHAT CONGRESS DISCUSSED IN 2020
During the hearing, Senator Johnson described the distinction between official election reporting and unofficial media reporting.
Official results move through election-management processes.
Unofficial results are distributed to media organizations and election-data providers.
He named Edison Research and the Associated Press in explaining the second pathway.
Christopher Krebs also testified about voting-system connectivity.
He acknowledged that certain equipment could use communications capabilities for transmitting results while emphasizing security controls and the importance of paper-ballot audits.
This is the technical connection that matters when examining 2026.
A number can change because ballots were newly counted.
It can change because a previously counted batch was uploaded.
It can change because a media feed was corrected.
Or it can change because a reporting-system error was discovered.
These events are not equivalent.
To understand the difference, investigators must compare the source ballots, machine records, election-management exports, public feeds, and final canvass.
V. NOVEMBER 2026: THE EVIDENCE GETS SPECIFIC
The November 3, 2026 election produced several documented administrative failures and discrepancies.
LocationNumerical issue
What happened
Latah County, Idaho2,343 ballots
Previously tabulated results omitted from public reporting
Bannock County, Idaho333 ballots
Ballots initially left uncountedJohnson County, Iowa
Approximately 2,000 ballots
Absentee ballots discovered unopened
Seminole County, Georgia
Approximately 388-voter discrepancy
Missing memory-card results changed the apparent leader
Tarrant County, Texas1,016-ballot difference
Preliminary reconciliation under examination
Wyandot County, OhioMock-election data
Practice results improperly included in unofficial reporting
These are different kinds of problems, not a combined total of fraudulent ballots.
Georgia: A missing memory card changed the apparent leader
On November 6, Georgia Secretary of State Brad Raffensperger announced an investigation into Seminole County.
A post-election review identified approximately 388 more voters checked in at polling places than ballots reported as cast.
Officials subsequently determined that a memory card containing Election Day results had not been uploaded into the county election-management system.
Before the correction, Democrat Harry James had reportedly led Republican incumbent Gerald Greene by 58 votes.
After the omitted results were included, Greene reportedly led by 262 votes.
That represents a 320-vote swing in the apparent margin.
This was not merely a change in a television graphic.
It was an omitted results upload that changed the unofficial standing of a legislative contest.
Raffensperger opened an investigation into how the memory card was overlooked and whether required procedures had been followed.
This is one of the most consequential documented 2026 incidents because the reporting failure affected which candidate appeared to be winning.
It does not establish deliberate manipulation, and the final certified result must be distinguished from the changing unofficial count.
Idaho: 2,343 ballots absent from public reporting
Latah County officials discovered that results from 2,343 ballots had not been included in the initial public election totals.
The ballots had already been tabulated.
The failure occurred when their results were not transferred into the statewide reporting system.
Those ballots represented approximately 12.9% of the county's 18,104 ballots.
Idaho Secretary of State Phil McGrane oversaw the response.
Latah County Clerk Julie Fry acknowledged the error.
A full recount was conducted.
No election outcome changed.
But the omitted votes materially affected some reported margins.
In a county commissioner contest, Republican Dan Schoenberg ultimately retained a 129-vote lead over Democrat Laura Heinse.
The Moscow school facilities levy finished at 60.34%, narrowly above its 60% passage threshold.
A reporting omission involving nearly 13% of a county's ballots is significant even when a recount confirms the same winners.
Idaho: Another 333 ballots in Bannock County
In Bannock County, officials discovered 333 ballots that had not initially been counted.
The ballots came from Pocatello Precinct 2.
Bannock County Clerk Jason Dixon acknowledged the failure and said officials were working with the state to prevent a recurrence.
Unlike Latah County, these ballots had not merely been omitted from public reporting.
They had not initially been counted.
Officials reported that including them did not change any race outcome.
The two Idaho incidents therefore involved 2,676 ballots but different failures.
Iowa: Approximately 2,000 ballots still unopened
On November 5, Johnson County officials discovered approximately 2,000 absentee ballots in secure storage that had not been processed.
County Auditor Julie Persons announced that counting was completed on November 8.
The county reported a total of 36,138 absentee ballots counted.
Officials stated that the ballots remained secure and that their inclusion did not change any election outcome.
This was a processing failure involving ballots already in official custody.
The case raises questions about ballot inventories, processing checklists, and end-of-election reconciliation.
Texas: The Tarrant County reconciliation question
MBT's examination of preliminary Tarrant County records identified:
Preliminary measureCountReported voters734,858Counted ballots733,287Rejected or pending ballots2,587Preliminary reconciliation difference−1,016
A separate roster comparison identified a 1,079-record difference.
These are different measurements and must not be combined.
The preliminary 1,016-ballot difference is approximately 0.14% of the reported voter count.
The final county reconciliation is necessary to establish whether the difference reflects reporting definitions, ballot status, data timing, or another explanation.
Ohio: Practice-election votes appeared in unofficial results
Wyandot County officials discovered that mock-election results from an October 28 test had been left in the reporting system.
Those practice figures inflated unofficial November 3 totals in several contests.
Election Director Darcy Zimmerman acknowledged the error.
Officials said the problem did not change any election outcome.
This incident is particularly important because it demonstrates that erroneous figures can enter public results without corresponding to real ballots cast by voters.
It is a reporting-control failure, not proof that fraudulent ballots were inserted into the election.
VI. THE PEOPLE ON THE RECORD IN 2026
The 2026 incidents involve identifiable public officials and candidates whose statements, records, or election results form part of the investigation.
Name / Role
Brad Raffensperger Georgia Secretary of State; opened Seminole County investigation
Harry James Democratic candidate in Georgia House District 154Gerald Greene Republican incumbent in Georgia House District 154
Phil McGraneIdaho Secretary of State; oversaw response to reporting failures
Julie FryLatah County Clerk; acknowledged omitted results
Jason Dixon Bannock County Clerk; acknowledged uncounted ballots
Dan Schoenberg Latah County commissioner
candidate Laura HeinseLatah County commissioner candidate who raised concerns
Julie PersonsJohnson County Auditor; reported absentee-ballot discovery and completion
Darcy Zimmerman Wyandot County elections director; addressed mock-election data
These individuals are named because they have relevant official responsibilities, public statements, or candidate roles.
They are not being identified as participants in a fraudulent operation.
The significance of their involvement is that the incidents have an identifiable documentary record and accountable officials—not merely anonymous internet claims.
VII. THE MOST IMPORTANT 2020–2026 COMPARISON
Six years apart, the records describe several recurring categories of election-system questions.
Issue 2020 congressional testimony
2026 documented incident Changing public totals
Pennsylvania data discrepancies alleged by Francis Ryan
Latah County's omitted reporting batch Missing or incomplete result transfers
Discussion of memory cards and reporting controls Seminole County's omitted memory-card upload
Ballot-processing procedures
Wisconsin absentee-ballot challenges
Johnson County's unopened absentee ballots
Electronic record integrity Nevada USB-data allegations
Idaho result-transfer review
Election-night reporting
Official versus unofficial data streams
Wyandot County's mock-election figures
Reconciliation Testimony about election logs and controls
Tarrant County preliminary reconciliation questions
Independent verification
Krebs's emphasis on paper ballots and audits
Latah County's full paper recount
These parallels identify common investigative questions. They do not establish that the same actors, software, or methods caused the incidents.
But they demonstrate why the technical subjects discussed in 2020 remain relevant.
VIII. THE NEW ALLEGATIONS ARE BRINGING THE VENDORS BACK INTO THE CONVERSATION
New public statements describe a purported report titled Foreign Insider Access to United States Election Systems.
The claims concern developers, privileged access, vendor relationships, and alleged activity across:
2020 — 2022 — 2024 — 2026
Names and organizations appearing in the renewed allegations include Dominion, Smartmatic, ES&S, former Dominion executives, attorneys, and other individuals alleged to have relevant relationships.
The allegations must be evaluated according to the roles actually held by those individuals and the records supporting each claim.
A technical report that identifies access privileges would raise one set of questions.
Evidence of unauthorized access would raise another.
Evidence that ballots or certified outcomes were altered would require an even higher standard of proof.
The important point is that the allegations now have to confront a documented 2026 election record—not just a collection of election-night screenshots.
IX. THE NEXT PHASE IS ALREADY UNDERWAY
The 2026 election record is still developing.
Idaho's state certification process and post-election audits are continuing.
Georgia has opened an investigation into the Seminole County memory-card failure.
County canvasses, reconciliation reports, and supporting records are being examined.
Public-records requests are seeking additional documentation.
The next phase is not a hypothetical future event. It is the ongoing process of audits, certifications, official responses, and examination of the original records.
The public should prepare to read those findings carefully.
Some will explain errors.
Some may identify weaknesses requiring correction.
Others may produce additional questions.
And any new technical report alleging coordinated misconduct will have to be assessed against the same documentary standards.
X. MIND BEND THEORY: FOLLOWING THE RECORDS
Mind Bend Theory has been conducting an independent review of the 2026 election record, including county reporting discrepancies, election-night data, audit findings, and official public-records requests.
The investigation is comparing original election information with the technical questions preserved in the 2020 Senate hearing.
MBT is also tracking the status of requested records and subsequent official explanations.
The objective is straightforward: determine what the records actually establish, preserve the evidence, and identify any discrepancy that remains unexplained.
XI. IF YOU MISSED 2020, GET READY FOR 2026.
In 2020, Congress heard allegations about voting machines, absentee ballots, electronic records, and election-data access.
Some allegations were rejected by courts.
Others concerned real questions about system design and oversight.
Federal cybersecurity officials defended the election's security.
And the resulting testimony became part of the permanent congressional record.
Now, six years later, election officials have documented missing result uploads, uncounted ballots, unopened absentee ballots, and incorrect public reporting figures.
In one Georgia legislative contest, an omitted upload changed the apparent leader.
In Idaho, a full recount was necessary to verify results after nearly 13% of one county's ballots were missing from initial public reporting.
In Iowa, thousands of absentee ballots had to be processed after Election Day.
These are documented events.
They deserve examination without exaggeration or dismissal.
And the next round of audits, records, and official findings deserves the same scrutiny.
If you missed 2020, pay attention to 2026.
The hearing is on the record. The witnesses are named. The numbers are available. The investigations are underway.
Things could get messy as officials, investigators, and the public work through the evidence.
But the outcome of that scrutiny must be determined by what the records prove.
America should prepare for the next chapter of the 2026 election record.
And Mind Bend Theory will be following it.
MIND BEND THEORY
Independent Research | Election Forensics | Technical Accountability
The ballots are the evidence. The records are the trail. Accountability begins with verification.
No assumptions. No invented timestamps. No conclusions before the evidence.