A tiny Texas addition

For those of you who didn’t see this illustration in my last post because you were too busy watching the video, here it is again.

When I made yesterday’s video, I wanted to mention something I hadn’t figured out yet. I stopped the video, intending to solve it, then finish the video. However, the part of my mind responsible for making videos seems to work in a different gear than the part responsible for solving algorithms. I felt the mental gears clash and decided to finish the video without solving the problem.

This morning, I figured it out.

It may sound like a minor detail. But there's a difference between proving that a crime was committed and proving that every single element of it was deliberate and designed. I prefer the latter.

The Setup

On February 18, 2026, voters checked in for the Bexar County, TX Republican primary. The next day, the county posted the check-in file online. I obtained a copy through Dr. Walter Daugherity, who received it from candidate Westin Martinez and investigator Lori Gallagher.

The file contained 4,110 fraudulent voter entries — fake identities built on real registered voters, each carrying a State ID number that does not exist in the Texas voter database. I documented how I know they are fake in my previous two posts. The short version: when you sort those 4,110 records by ID number and divide the total span by the spacing between records, you get exactly 4,109 — a perfect integer, zero remainder. Random or accidental data cannot do that. Only deliberate computation can.

I also established that the algorithm had to be run after voting concluded. It required the complete list of everyone who checked in that day before it could generate a single fake record.

The Structure of the File

When you sort all 4,845 February 18 records by State ID number, something immediately visible happens. The first 633 records are real voters with legitimate integer IDs. Then, starting at record 634, the fractional IDs begin — and they run without interruption through record 4,743. Then the integers resume for the final 102 records, through record 4,845.

Real — Real — Real … [633 records] … Fake — Fake — Fake … [4,110 records] … Real — Real … [102 records]

This is not a coincidence of layout. It is a direct consequence of where the algorithm placed the fake IDs. The 4,110 fraudulent records occupy a dead zone in the Texas ID number space — a range of approximately 90 million consecutive ID numbers containing no legitimate Bexar County voters whatsoever. The algorithm selected this empty band deliberately; placing fake IDs there required prior knowledge of which parts of the ID space were unoccupied. That is reconnaissance.

The 633 real voters before the fake block simply have ID numbers that fall below the dead zone. The 102 real voters after it have ID numbers above it. Together they account for all 735 real voters in the file — 633 plus 102 equals 735 exactly. The fake records slot into the middle of the ID-sorted list not because the algorithm interleaved them, but because it parked them in a gap between two clusters of real IDs, and the real IDs on either side sort naturally around them.

The Question

When sorted by ID number, the fraudulent sequence begins with a man namedAbel and ends with Braswell. These two names define the minimum and maximum of the fake ID range — and from those two numbers, every other fake ID in the sequence can be derived mathematically.

But here is the thing: Abel and Braswell do not have the lowest and highest real ID numbers among the 735 voters whose identities were used. They don’t even define the range of last names used. That list extends into the C’s. They are not special by any criterion I could see at first. So why them?

The Answer:

The algorithm sorted all 735 real voters alphabetically by last name.

Abel gets the minimum fake ID because ABEL sorts first. He is position 1.

Braswell gets the maximum fake ID because BRASWELL sorts 435th — and 435 is not an arbitrary number. It is the mathematically unique solution to the system of equations governing how many fake records each real voter receives. With 735 anchor names and 4,110 total fake records, where each real voter gets exactly 5 or 6 copies, there is only one possible split: 300 names get 5 copies, and 435 names get 6. No other combination works. The algorithm needed a stopping point for its final pass, and that stopping point is position 435 in the alphabetical sort — Mr. Braswell.

His neighbor on the same street, also named Braswell, but a different first name, sits at position 436. She gets only 5 copies. The cutoff falls between two members of the same household.

The Dead Zone

The script that mapped the Texas voter ID database gives us, for the first time, precise coordinates for the void the algorithm chose as its hiding place — and the scale of what it found there is larger than previously stated.

The Texas statewide voter roll contains 18.3 million records spanning ID numbers from roughly 1.0 billion to 3.4 billion. Within that range sits a gap: the last legitimate voter in the entire state of Texas before the void is a voter named at ID 1,222,380,331. The next legitimate voter after the gap is at ID 2,000,050,898. Between those two numbers — a span of approximately 777.7 million consecutive ID numbers — there is not one registered Texas voter. Not in Bexar County. Not anywhere in the state.

The 4,110 fraudulent records were placed inside this void, running from 1,253,115,467 to 1,343,862,000 — a span of roughly 90.7 million ID numbers sitting entirely within the empty band.

The dead zone contains no legitimate Texas voters at all. The algorithm didn’t find a gap in one county’s ID distribution. It found a gap in the entire state’s ID space — across 18.3 million registered voters in 254 counties.

Identifying that gap required querying, or having prior access to, the complete Texas statewide voter database. The local Bexar County poll book would not have been sufficient. The reconnaissance was not county-level. It was state-level.

Why This Matters

Both endpoints were selected by the same mechanism — last-name alphabetical sort — but each position was determined by a different mathematical requirement. Position 1 comes from the alphabet. Position 435 comes from solving an integer constraint system. The algorithm had to do both before it could assign a single fake ID number.

The placement of the fake IDs in a dead zone of the ID number space — producing exactly 633 real records before the block and 102 after — required advance knowledge of which ID ranges were empty. The selection of Abel and Braswell as endpoints required solving a two-equation system and sorting a voter list that could only exist after voting concluded.

A glitch does not sort voter lists by last name. A display error does not solve systems of equations. A database migration artifact does not place its endpoints at positions 1 and 435 of an alphabetical sort because those happen to be the unique solutions to two different mathematical constraints simultaneously.

This was purpose-written software, executing a complete specification. And now we know the specification down to the last detail.