# Post ts_116091294028607026

- Post ID: `ts_116091294028607026`
- Platform: Truth Social
- Posted: 2026-02-18T11:02:47.579Z (UTC)
- Deleted: no
- Repost: no
- Canonical URL: https://trump.fm/post/ts_116091294028607026
- Analysis page: https://trump.fm/post/ts_116091294028607026/analysis

## Post text

_No text of its own. The post is its media, below._

## What the media says

**Video transcript** (6:13), transcribed automatically, so names and numbers may be wrong. Speakers aren't identified, so a voice in it may not be his:

> - If these ballots were run through multiple times, as you've alleged, I think you said possibly up to 30,000 times, definitely thousands of times, being counted, when we examine the poll book and see who voted, and we count those numbers, would they then not line up with the total number of votes coming out of Detroit? - Just later, later, the poll book is completely off, completely off. - Off to 30,000? - I'd say that poll book is off by over 100,000. - Where are you from? - Oh, I'm from down river. - Okay. - I'm here because of the 38,000 ballots that arrived here at 3.45 AM, they're just now sitting over there and getting counted, supposedly. - Okay, okay. - And that's why I arrived. - Okay. - To an eye-opening of hypocrisy, let me tell you, something I want you to see, they're taping up the windows, the security's taping up the windows so that people who are challenging cannot see in here, and he can't see out, so it's an interesting development. Interesting development. - I began observing the process of ballots at several counting boards. I quickly noticed that none, and I mean none, of the names who were being scanned from the ballots were in the electronic poll books. All of the names that were found were found in the supplemental sheets. We know that the poll books, the electronic poll books, were updated on Sunday, November 1st. That meant that all of these new ballots had to have been brand new registered voters on Monday, November 2nd, or Tuesday, November 3rd. I found it very hard to believe that tens of thousands of people had registered in just two short days, so I began writing down the names on the ballots immediately after they were scanned by the computer poll worker. As soon as I began, I was severely obstructed by five different persons that I could no longer, such that I could no longer continue my legal responsibility as a poll challenger. These five persons comprised of a poll worker, a supervisor, an uncredential person, a Democrat challenger, and one of the top leaders of the AVCB. The details can be found on my affidavit. Does anyone believe that 30 or 40,000 new voters were actually signed up legally in two days? - Yeah, and what have you seen? - I've seen ballots that don't belong to anybody. They don't throw up in the voter registration. There's no name attached to them. - And what are they doing with those ballots? - I'm challenging them, and they're supposed to put them in a box to be evaluated, but then multiple ballots that have several different numbers, one person with different numbers. We have whole lists of them. We're writing everything down all day long. - So are they improperly filled out, and you're looking to disqualify them? - Yes, they're improperly filled out if they have no name, if they don't have a home, or if the person was born in 1921, but registered to vote in 1900, that kind of doesn't make sense. - But that's not on the ballot. When you vote, your ballot is just the balls in which you have on there. What are you seeing? Are you seeing people's registration forms? So on the tables, you have the ballot, and then you are able to see next to them on a computer screen whose ballot that is. - Sometimes the ballot itself has no name, but the ballot is filled out. And then if we go into the registration file, we find the person that was born in 1921, and there's a lot that's not right. - What's wrong with being born in 1921? - But registering to vote in 1900, 21 years before you're born. - Only a few of the most egregious items witnessed. I was not allowed to perform my job, and from what I observed were any of the GOP challengers in accordance with Michigan laws. We were constantly prevented from getting close enough to see what was going on with respect to the ballots. Procedures were not being properly followed when we brought these to the attention of the supervisors. Our concerns were dismissed and ignored. We observed ballots being backdated from 11-4 to 11-2, and data being entered into the computers for these mail-in ballots. Poll workers were entering ballot information from new mail-in ballots, including dates of birth of 1/1/1900 into the electronic poll books and processing those ballots for tabulation. The information wasn't in either the electronic poll book or the paper supplement book prior to that. Dead people were voting. I sent a website to that effect, and the list that they had up to that date was 17,327 randomly checked Michigan voters that were registered for voting and were above the age of 80. Every name on this list has voted in the 2020 election, check for yourself, and has also been found in one or more of the obituaries online. This is the most important election in my lifetime, and I think the most important election in the history of our republic. - Yeah, poll book, why don't you look at the registered voters on there? How many registered voters are on there? Did you, do you even know the answer to that? - No, I guess I'm trying to get to the bottom of this here. - Zero, zero, there's zero. - So my question then is if the-- - Guess how many, wait, what about the turnout rate? 120%. - Let's let Representative Johnson ask his question a little bit.

## Media

- video: https://static-assets-1.truthsocial.com/tmtg:prime-ts-assets/media_attachments/files/116/091/290/724/781/294/original/9f87bb767ed915ca.mp4

## Engagement

- Likes: 9,784
- Reposts: 2,930
- Replies: 363
- Views: unknown
- Metrics collected: 2026-02-19T17:30:41.183Z (UTC)

# Analysis

_Machine-generated by trump.fm on 2026-09-09T08:56:31.526Z (UTC): a model's reading of this post, not his words. Its psychological terms describe the language, not a clinical assessment of him._

## Summary

This wordless video repost amplifies archival 2020 Detroit/Wayne County poll-challenger testimony alleging ballot-count fraud, observer obstruction, and implausible voter-registration numbers — claims litigated and rejected in 60+ court cases and unsupported by subsequent audits. With no caption, the analyzable behavior is selection: choosing, six years later and during a 2026 midterm election year, to recirculate settled fraud allegations to his audience.

This fits a pattern of grievance maintenance rather than acute narcissistic injury — there is no proximate trigger (no recent electoral defeat or criticism in the surrounding context) — suggesting either habitual rumination on 2020 or a preemptive function ahead of upcoming elections. Paranoid features predominate: persecutory framing of election administrators, grudge-holding across years, and implicit reality distortion by treating an adjudicated matter as unresolved. Denial and projection are evident in the content choice, though these defenses originate in the amplified material rather than the subject's own words.

No explicit violent or eliminationist language appears, but recirculating unresolved-sounding fraud claims during a midterm year carries a documented capacity to erode confidence in election administration, warranting an "elevated" rather than "none" danger rating.

Authorship attribution rests solely on early-morning timing (~6 AM ET) since no text exists for stylometric analysis; confidence is capped at low, and this reflects who chose to post the clip, not who authored any language within it. Longitudinal tracking of similar reposts as the 2026 midterms approach would clarify whether this reflects habitual grievance recall or a deliberate delegitimization strategy.

## Post Overview

This post carries no original text from the subject — it is a bare video repost. The clip is archival poll-challenger footage from Detroit/Wayne County, Michigan concerning the 2020 election count, featuring an unidentified challenger and bystanders alleging ballot-count irregularities, obstruction of observers, and implausible new-voter-registration numbers. None of this speech can be attributed to the subject; the analyzable behavior is his choice to select and amplify this specific clip roughly six years after the 2020 election, during a 2026 midterm election year.

## Levels 1-3: Traits, Motives, Narrative

No first-person text exists to score dispositional trait facets directly. At the level of characteristic adaptations, the selection reveals a persistent schema: elections producing unfavorable outcomes are inherently fraudulent, and election officials/poll workers are adversarial conspirators. Narratively, this continues a long-running contamination sequence (a legitimate 2020 process rendered illegitimate by alleged corruption), with the featured challenger occupying a victim/persecuted role by proxy and unnamed election administrators cast as the contrasting-other antagonist.

## Level 4: Clinical Indicators

**Paranoid features:** The clip's content (windows taped over, observers obstructed, ballots allegedly "backdated") matches longstanding persecutory ideation about electoral processes. Amplifying this material years after the underlying claims were litigated and rejected in 60+ court cases, and after multiple audits (including Republican-led ones) affirmed the certified Michigan results, is more consistent with grudge-holding than with responsiveness to new evidence.

**Reality distortion:** Resurfacing unadjudicated 2020 fraud allegations without acknowledgment of the extensive judicial and administrative review that followed functions as passive revisionism — presenting settled, rejected claims as though still live and unresolved.

**Defenses:** Denial (of the certified, litigated outcome) and projection (attributing procedural corruption to election administrators) are implicit in the clip's chosen content; selecting it for a 2026 audience suggests these schemas remain active rather than resolved.

## Rhetorical/Behavioral Act

The act here is selection and amplification, not composition. The clip deploys anecdotal generalization ("if 30,000 ballots... why not 100,000") and insinuation-by-suspicious-detail (birth-year/registration-year mismatches, taped windows) — techniques inherent to the source video, not authored by the subject. By posting it without commentary, he neither endorses nor disavows specific claims explicitly, but the act of continued amplification in a new election cycle is itself a data point: recirculating stolen-election material during a midterm year plausibly serves to pre-seed doubt about upcoming election administration, a pattern with documented downstream effects on public trust and, historically, on mobilization (cf. January 6, 2021 context).

## Timing and Authorship

Timestamp 2026-02-18T11:02:47 UTC converts to approximately 6:02 AM Eastern Time (no DST in February), whether the subject was at the White House, Mar-a-Lago, or another East Coast location. This falls at the early edge of his documented early-morning posting pattern. However, with no written text, there is no stylometric evidence (no typos, syntax, capitalization, or grammatical idiosyncrasies) to examine. The only available evidence is timing and the fact that video-only reposts of this kind have historically been posted personally rather than by staff (aides more typically post polished statements, event announcements, or scheduled content during business hours). This is weak evidence and speaks only to who chose to post the clip, not who found, edited, or captioned it (there is no caption). Confidence is capped at low accordingly.

## Danger Assessment

No explicit calls to violence or mobilization appear in the transcript, and the clip itself is presented as first-hand testimonial/interview content rather than incitement. However, in a midterm election year, recirculation of previously adjudicated and rejected fraud claims carries a documented capacity to erode confidence in upcoming election administration — a pattern flagged by researchers as a precursor condition (not a direct trigger) for election-related unrest. Rated "elevated" rather than "high," given the absence of named targets, explicit grievance-to-action linkage, or eliminationist language in this specific post.

## Summary of Confidence

Given the total absence of first-person text, this analysis necessarily rests on lower-confidence inferential grounds (content selection as behavior) rather than direct linguistic evidence. Longitudinal comparison — has the subject posted similar 2020-fraud content repeatedly and increasingly as the 2026 midterms approach? — would substantially strengthen or weaken the "priming for election delegitimization" hypothesis offered here.

## Authorship Analysis

**Uncertain** (score: 55%)

### Indicators

- Timestamp converts to ~6:02 AM Eastern, at the edge of his documented early-morning solo-posting window
- No caption or commentary accompanies the video, consistent with impulsive one-tap reposting rather than staff-drafted content
- No written text exists, so no stylometric evidence (typos, syntax, capitalization) can be assessed
- Historically video-only reposts with no caption skew toward personal rather than staff origin, but this is a weak, non-linguistic signal

## Psychological Profile

### State

**Vulnerable State**

**Trigger:** Maintenance (Recirculation of settled 2020 Michigan election-fraud narrative absent any new precipitating event in the provided context)

Sentiment: -0.40

### Clinical

**Malignant Narcissism:**
- Narcissistic: 30%
- Antisocial: 10%
- Paranoid: 60%
- Sadism: 0%

**Defense Mechanisms:**
- denial (immature)
- projection (immature)
- splitting (immature)

**Cognitive Complexity:**
- Complexity: 0%

**Parasocial Techniques:**
- Amplifying eyewitness-style testimonial content to lend affective credibility to a contested claim
- Repetition/recirculation of a familiar grievance narrative to reinforce in-group belief without new argument

## Danger Assessment

**ELEVATED**

### Indicators

- Recirculation of unresolved-sounding election-fraud narrative during a midterm election year, which carries documented potential to erode public confidence in upcoming election administration, though no explicit target, grievance-to-action linkage, or violent language is present in this post

### Gaslighting

- Recirculating rejected 2020 fraud claims without acknowledgment of the extensive judicial/administrative review that found no widespread fraud, implicitly treating a settled matter as still open

## Tags

- election-fraud-narrative (90%)
- 2020-election-relitigation (85%)
- paranoid-ideation (60%)
- media-amplification-no-caption (70%)
- midterm-year-priming (50%)
- grievance-maintenance (55%)

## That day

_From trump.fm's machine-generated digest of the day, not his words._

**The 6 a.m. hour went to six-year-old ballot-fraud video; the afternoon went to telling Starmer not to lease away Diego Garcia**

Trump posted twenty times, in an evening batch and a dense hour just after six in the morning, then went quiet for most of the working day. Nearly half of it was 2020 election material — old Detroit ballot-fraud testimony, a Dominion segment, a Georgia post crediting an ally with beating Fani Willis — none of it prompted by anything that happened that week, and almost all of it posted without a word of his own. The one thing he actually wrote was an afternoon message telling the British prime minister not to lease out Diego Garcia, dismissing Mauritius's internationally recognized claim as fictitious and raising the possibility of needing the base to strike Iran; it contradicted his own State Department. He said nothing about the Syria withdrawal, the herbicide order he signed that day, or the Geneva talks. No anger anywhere in the day, a normal night's sleep gap, and an unusually early stop in the late afternoon.

Full digest for 2026-02-18: https://trump.fm/date/2026-02-18/analysis

## Citation

- APA: Trump, D. J. (2026, February 18).  [Social media post]. Truth Social. trump.fm. https://trump.fm/post/ts_116091294028607026
- MLA: Trump, Donald J. "" Truth Social, 18 Feb. 2026. trump.fm, https://trump.fm/post/ts_116091294028607026. Accessed 9 Oct. 2026.
- Chicago: Donald J. Trump, "," Truth Social, February 18, 2026, archived at trump.fm, https://trump.fm/post/ts_116091294028607026.

## For agents

- Site overview: https://trump.fm/llms.txt
- API specification: https://trump.fm/openapi.json
- MCP server: https://trump.fm/mcp (search and fetch tools, no auth)
- This post as JSON: https://trump.fm/api/posts/ts_116091294028607026
- Analysis as JSON: https://trump.fm/api/analysis/ts_116091294028607026
- All citation formats: https://trump.fm/api/cite/ts_116091294028607026
- Same day: https://trump.fm/date/2026-02-18
- The record alone, without the analysis: https://trump.fm/post/ts_116091294028607026.md?analysis=false

_Markdown view of a trump.fm page. Post, analysis, date, feed, contradictions, search and about pages answer in markdown at their URL with `.md` appended (`/index.md` for the home page), or when sent `Accept: text/markdown`._