Every news story has a core event: the small set of things that actually happened. Everything wrapped around that core is a choice. We show you the core event first, and then we show you the choices.

The core objective event

When news breaks, we reduce it to its bare, objective core: who, what, when, where, each fact confirmed across outlets and attributed. We record when each event was born and how it forks into new events over time. That core event is written by us, hosted by us, and held to one standard: every word must be defensible as non-loaded.

Bias becomes visible, not argued

Around each core event we hang the actual coverage: left-leaning, right-leaning, and neutral. We don't tell you an outlet is biased; we show you the evidence: the loaded phrases, who gets quoted, which facts get headlines and which get buried, and what a story omits entirely. You can drag the Dial and watch the same event re-framed across the spectrum.

The Rating Method: eight dimensions

Every article, and every outlet's coverage choices, are scored on eight dimensions. The exact weighting and detection methods are ours, but the dimensions themselves, and what they look for, are public:

What we host, and what we don't

We host the facts and the analysis. The framing belongs to the outlets that wrote it, so we always link you to the original articles rather than republishing them. Reading the spin at its source is part of the point.

Our promises

Lean and factual reliability are separate measurements, never blended into one number. Our rating dimensions are published, versioned, and open to challenge; every score has a "disagree" button. We claim a process, not omniscience. And the interface will never use the tricks we measure: no outrage bait, no infinite scroll, no loaded headlines.

See what happened before the spin. Then decide for yourself.