JOKES ARE FREE.
CLAIMS NEED
RECEIPTS.

We can call something ugly, annoying, spiritually exhausting, or profoundly stupid without convening a tribunal. The moment we call it false, manipulated, synthetic, stolen, or deceptive, the burden changes. Show the receipts.

START AS CLOSE TO THE THING AS POSSIBLE.

We prefer primary sources, direct examples, original screenshots, public records, platform documentation, first-hand testing, and the actual material being discussed. When another publication did the reporting, we credit and link to it. A screenshot of somebody saying a thing happened is not the same as evidence that it happened.

AI IS A TOOL, NOT A WITNESS.

Automated tools may assist with research, transcription, organization, coding, or production. They do not become evidence merely because they produced a confident paragraph. Factual claims should be checked against sources that can actually be held accountable for them.

SAY WHAT YOU KNOW. SAY WHAT YOU DON’T.

Facts, analysis, opinion, satire, and inference are different things even when the same article contains all five. When evidence is incomplete, the language should remain incomplete too. Certainty is not a design treatment.

CORRECTIONS SHOULD LOOK LIKE CORRECTIONS.

Material factual errors should be corrected promptly. When an update meaningfully changes a story, the article should display an updated date and explain the change when the context requires it. Quietly replacing a bad fact and hoping nobody noticed is its own small form of slop.

THE SLOP SCORE IS AN ARGUMENT, NOT A LAB RESULT.

The Slop Score is an editorial framework. It considers manipulation, originality, mass producibility, deceptive presentation, algorithm bait, and actual informational or cultural value. It is not scientific measurement, and the decimal point should not trick anybody into believing we own a laboratory.

Our standard is simple: if the story asks readers to be skeptical, the story has to survive skepticism too.