Read closely. I would say the rise of AI is another strong reason to move to paper ballots.
I've been studying voter roll algorithms since I first discovered one in NY in April 2022. What I found this week in TX is worse than anything I've seen anywhere else.
It demonstrates the capability of bypassing the voter rolls entirely — a scheme capable of generating fraudulent but genuine ballots without leaving any forensic trace.
The only reason we know about it is because the original file was distributed to precinct chairs on Feb 19, before the record was wiped on or before Feb 25.
ANALYSIS BY GROK.
- ZarkFiles, an election researcher, alleges discovery of 4,110 deliberately fabricated voter check-ins in Bexar County's February 18, 2026, Republican primary, using fractional State IDs anchored to 735 real voters and duplicated precisely (300 with 5 copies, 435 with 6) to generate ballots via compromised poll pads.
- The records exhibit mathematical impossibilities for random errors, including exact gaps of 22,084.82189 between sorted IDs spanning an unused 90-million-ID block—requiring statewide database access—and were erased via a fractional deletion key in a 78-second batch before February 25, leaving no permanent trace.
- Coordinated by purpose-written software post-voting, this scheme highlights vulnerabilities in electronic poll books; Weston Martinez has notified Texas authorities, but no official response or verification has emerged as of February 28, 2026.
Grok analyze the portion of the Substack article about the voter manipulation algorithm and explain in layman’s terms
Thoughts
The Substack articles (primarily "Ghosts of the Alamo: Busted in Texas" from Feb 25, 2026, and the follow-up "The Weston (Alamo) Algorithm, Naked" from Feb 27, both by Andrew Paquette, PhD, aka
@ZarkFiles
) describe an alleged voter manipulation algorithm used in Bexar County, Texas, during the February 18, 2026 Republican primary.The “algorithm” isn’t a mysterious black-box program with published code — it’s the author’s term for the precise, interlocking mathematical patterns in 4,110 suspicious check-in records that he (and others like Prof. Walter Daugherity and candidate Weston Martinez) say prove the records were deliberately fabricated by custom software, not by accident, glitch, or human error.Layman’s Breakdown: What the “Algorithm” Actually Did (Step by Step)Imagine the electronic poll pads (iPad-like devices from a company called KnowInk) at voting sites. They check you in by your Texas State Voter ID (a normal whole number, like your driver’s license number), print a ballot, and record everything.Here’s how the alleged scheme worked, in plain English:
- Pick 735 real people as “anchors”
The program grabbed the names of 735 actual voters who legitimately checked in that day. It sorted them alphabetically by last name (first one: someone named Abel; last one: someone named Braswell). These became the “base” voters that everything else would copy.
- Clone each real person 5 or 6 times
It made exact duplicate check-in records for each of those 735 people — but not randomly.
- Exactly 300 of them got 5 copies each
- Exactly 435 of them got 6 copies each
Total clones: (300 × 5) + (435 × 6) = 1,500 + 2,610 = 4,110 fake records.
The numbers add up perfectly with zero leftover — the only possible way to hit exactly 4,110 using only 5s and 6s. That’s not how random mistakes work.
- Give every clone a fake ID from an empty parking lot
All 4,110 fakes used voter IDs from a giant unused block of ~90 million consecutive numbers in Texas’s statewide system (a “dead zone” with zero real voters). Whoever did this had to know the entire state’s ID map in advance.
- Hide the fakes with a secret decimal tag
Real IDs are whole numbers (e.g., 1,253,115,467).
These fakes had a tiny decimal at the end (e.g., 1,253,115,467.7999).
The poll pads are programmed to hide the decimal on screen, so poll workers saw a normal-looking ID and let the “voter” through. Each fake generated a real ballot.
- Space the fake IDs with perfect computer math
When you sort all 4,110 fake IDs, the gap between each one is exactly the same weird number: 22,084.82189.
The total distance from the very first to the very last fake, divided by that gap, equals exactly 4,109 (one less than the number of records, which is correct for counting spaces between them). No leftover fraction.
The gaps also cycle in a symmetric “palindrome” pattern (forward then backward passes), exactly what a computer loop using standard floating-point math would produce.
(Quick check: the numbers in the article produce ~4,109.00000079 — essentially exact, with tiny floating-point rounding.)
- Clean up instantly with the secret tag
After voting ended, someone ran a one-click database command: “Delete everything with a non-whole-number ID.”
It wiped all 4,110 fakes in 78 seconds on Feb 25 (or earlier), leaving only the 735 real voters who each appeared to have voted once. No trace on the permanent rolls.
Why the Articles Say This Proves a Deliberate Algorithm (Not a Glitch)A normal software bug or typo would create messy, random junk: some IDs off by 1, some missing digits, duplicates scattered everywhere, etc.Instead, every single piece fits together like a jigsaw puzzle designed to solve only one way:
- The exact count (4,110)
- The exact anchor count (735) and split (300/435)
- The perfect gap and total span
- The unused ID block
- The hidden decimal that both hides the fraud and lets you delete it perfectly
- All fakes sharing the exact same check-in date
The articles quote an AI analysis (Claude.ai) that sums it up nicely:
“A database glitch… cannot produce… a set of findings that are mutually consistent with one another and with a single underlying design… The palindromic gap structure alone… is the direct fingerprint of compiled code executing a specific loop.”
In other words: the math is too elegant and interconnected for accident. It looks exactly like someone wrote a short program, ran it after polls closed (or in a quiet window), let the extra ballots get printed and cast anonymously, then hit “erase” with a secret key.Bottom Line in Plain EnglishThink of it like printing 4,110 invisible extra tickets using photocopies of real people’s names. The tickets look totally normal to the ticket-taker, they get used, and then a special “disappearing ink” command makes every copy vanish overnight — but if you look closely at the ticket numbers with a magnifying glass, you see they were printed by a machine following an exact recipe that no human error could accidentally follow.That’s the “Weston (Alamo) Algorithm” the Substack articles are talking about. It’s presented as strong circumstantial evidence of sophisticated electronic fraud that bypassed the voter rolls entirely and left almost no trace.Important note: This is still an allegation based on one leaked data file distributed to precinct chairs before the wipe. As of Feb 28, 2026, Texas authorities have been notified but there’s no public official confirmation or rebuttal yet. The math checks out as described, and the patterns are strikingly non-random, but full forensic access to the original poll-pad systems would be needed to prove intent or who did it.If you want me to dive deeper into any specific number, the poll-pad mechanics, or compare it to other cases ZarkFiles has found, just say the word!