AZ Voter Databases Have a Problem (ZarkFiles Research Thread)

Cleanup for duplicate embed

Wow, I figured for sure the usual suspects would've swooped in to screech incoherently about how the data doesn't prove what Zark says it does.

This probably deserves its own thread, but I don't want to junk up the board with election fraud stuff that's just going to draw the same reactions from the same, tired NPCs.

Raise your hand if you spot a pattern in this chart:

Loading Twitter/X Embed...
If tweet fails to load, click here.

Too much at stake for the establishment not to cheat.

With enough iterations, things become more predictable and eventually sure things.

I wanted to believe this guy's findings, but I ran the data through Claude and Gemini and it turns out he doesn't understand standard database architecture.

The "Diagonal Line": Real databases never look like a clean line because people move, change names, and update their info over time.

The 15-County "Algorithm": All 15 counties use the exact same setup because the federal Help America Vote Act (HAVA) of 2002 legally mandated one centralized state system.

"Clones" are just logs: He’s confusing active voters with historical logs, since the National Voter Registration Act (NVRA) of 1993 legally forces states to keep "inactive" records of people who move for two federal election cycles before finally purging them (and "purging" just means changing their database status to "Canceled" for mandatory audit trails, not actually deleting the row).

Pre-1990 Data: His own table shows a 15% duplicate rate prior to 1990, proving these are just 35-year-old legacy logs from long before the 2002 HAVA law he blames for creating them.

They are cleaning the rolls: His chart actually proves they are cleaning up the system, showing the duplicate rate dropping from 24% in 2010 down to under 8% today (AZ has canceled over 1.7 million inactive records since 2020, almost 30% of the roll).

Bottom line: He downloaded the raw database file, counted millions of legally mandated historical logs, and is trying to pass them off as active fake voters.

I wonder if by now, ol’ Joe even remembers he was “president” with the most votes ever!!

Most secure, transparent election in the world. Lol.

Smh. How embarrassing this republic has become.

If you look at the very first chart he posted, look all the way to the top right corner. You'll see a distinct, stair-step line moving perfectly up and to the right. That is the "clean diagonal" line he claims is missing.

The Diagonal = Brand New Voters: That line at the top right represents fresh, first-time voters registering recently. Because they are brand new, the system generates a new, high Voter ID and a new, high Registration ID at the exact same time. This creates a perfect 1-to-1 sequential line.

The Messy Parts = Historical Reality: The reason the rest of the chart looks like a scattered mess is because of time. When a voter from 1995 (low Voter ID) moves or changes their name in 2020, they get a new transaction (high Registration ID). This pulls their data point way out of that diagonal line and creates the horizontal bands.

The Irony: The author claims a normal database should just be one big diagonal line. In reality, a database only stays in a perfect diagonal line if no one ever moves, dies, changes party, or updates their info.