What This Project Is

This is a pattern recognition database. Not a belief system. Not entertainment. Not monetized.

We collect encounter reports from people who saw something they couldn't explain and chose not to talk about it — sometimes for decades. We're specifically interested in accounts from before 1990, before pop culture contaminated the language people use to describe these experiences.

Why Pre-1990?

In 1991, David Icke published his first book connecting reptilian entities to conspiracy theories. After that, every account became suspect — was the witness describing what they actually saw, or what TV told them it should look like?

Before 1990, there was no cultural script. A witness in 1968 had no framework for what they saw. No YouTube, no Ancient Aliens, no Reddit threads. Their description was raw — unfiltered by media. That's signal.

After 1990, every report carries contamination risk. We still accept post-1990 submissions, but they're scored differently. The AI credibility system accounts for media exposure, cultural contamination, and whether the witness actively rejected pop culture versions of their experience.

What We Look For

The strongest signals come from witnesses who:

• Stayed silent for years — Hoaxers talk immediately. Real witnesses shut up because they know how it sounds.
• Give pre-emptive exclusions — "It was NOT a scuba diver." "It wasn't a bear." They proactively eliminate what people will assume.
• Reject pop culture framing — "That's not what they look like" when shown media depictions. The witness corrects the narrative instead of joining it.
• Report physics violations — Objects entering water with no ripple. Entities appearing or disappearing. Impossible speed. Electronics malfunctioning. These are harder to fabricate than visual descriptions.
• Describe with hyper-specificity — "Under Armour mannequin build but MORE muscular" is a real memory. "It was big and scary" is not.

The Hypothesis We're Testing

If genuine encounters occur at specific locations, those locations should share environmental characteristics. We're testing whether encounter reports cluster near:

• Moving water (creeks, rivers, lakes)
• Granite and quartz geology (piezoelectric properties)
• Electromagnetic infrastructure (power plants, military installations, high-voltage transmission)
• Magnetic anomalies

If independent witnesses from different decades describe similar encounters in locations with similar geology and EM profiles, that's a pattern worth investigating. If they don't, the hypothesis fails and we say so.

How Credibility Scoring Works

Every submission is automatically scored on three axes:

• Credibility (0-100) — Based on years of silence, pre-emptive exclusions, rejection of pop culture, physics violations reported, narrative detail, and multiple witnesses.
• Contamination Risk (0-100) — Based on media exposure, familiarity with Icke/Paulides, post-1990 timing, and whether media consumption preceded the encounter.
• Detail Specificity (0-100) — How many descriptive fields the witness filled with substantive detail. Vague reports score low.

The Intel Dashboard lets you filter by these scores. Crank the credibility slider up and the contamination slider down — what remains is the cleanest signal in the database.

What This Is Not

• Not monetized. No ads, no subscriptions, no data sales.
• Not affiliated with any TV show, podcast, book, or media personality.
• Not a community or forum. There are no comments or profiles.
• Not trying to prove anything. The data either shows patterns or it doesn't.

This is an unfunded science project built by someone who heard a story that was too specific, too reluctant, and too consistent to ignore. If you have a story like that, we're listening.