Open a media bias chart and you get a tidy picture. Outlets are plotted left to right, sometimes up and down for reliability, each one a labeled dot on a grid. Ad Fontes Media's January 2026 chart placed 137 sources that way. AllSides keeps its own ratings, updated through blind surveys and editorial review. These tools are useful. They give a fast read on where an outlet tends to land, and that read is usually defensible.
But a bias rating describes an outlet. It does not describe a story. And most of the time, the story is the thing you actually wanted to understand.
What a media bias rating actually measures
A bias rating is an average. AllSides and Ad Fontes both arrive at a placement by sampling many articles, across many topics, over a stretch of time, and summarizing the result as a position on a chart. That average is real, and it is stable enough to be worth knowing. It tells you that Fox News leans right and NPR leans center-left as a general matter.
What an average cannot do is tell you how either outlet handled the specific thing you read today. A center-rated outlet can still bury a story. A right-rated outlet and a left-rated outlet can land in the same place on one story and miles apart on the next. The rating smooths all of that out, because smoothing is what averages do. That is the limit baked into the format — and it is why a chart is a starting point, not an answer. (We made the same case from the other direction in why source count matters more than any single article.)
A rating can't tell you a story was missing
The clearest limit is the one a chart structurally cannot show: the story that an outlet never ran.
Consider the arrest of Adys Lastres Morera, a Cuban national and lawful permanent resident whose green card was revoked over alleged ties to the Cuban government. Her brother runs GAESA, the military conglomerate that controls much of Cuba's economy. In the Signal/noise source breakdown, every outlet that covered it sat in the center or right band — Reuters, Newsweek, the NY Post, Fox News. No left-leaning outlet touched it. A bias chart can confirm that Reuters is center and the NY Post leans right. It has nothing at all to say about the outlets that simply weren't there.
The same gap showed up on a far smaller story. When the House committee advanced the Sunshine Protection Act to make daylight saving time permanent, four outlets ran it — Fox News, the NY Post, TIME, and The Hill. No left-leaning outlet appeared in the coverage at all. That absence is the story. A rating chart, which only plots the outlets that exist rather than the ones that showed up, will never surface it. Detecting it is its own skill, one we walk through in what bias by omission is and the other types of media bias.
Two outlets with the same rating can frame a story in opposite directions
Here is the limit that surprises people most. Take that Cuban arrest again and look only at the two center-band outlets, Reuters and Newsweek. A bias chart places them within a hair of each other. They covered the same arrest on the same day.
They did not cover it the same way. Reuters ran it as a national security item, identifying Lastres Morera by her family connection to GAESA's leadership. Newsweek led its headline with "green card holder" and "ICE," foregrounding her immigration status rather than her family ties. Same event, same rough chart position, two genuinely different stories about what the news was. A bias rating cannot register that split, because the split happens below the resolution a rating operates at.
This is the difference framing analysis is built to catch. A rating asks where an outlet sits. Framing analysis asks what an outlet chose to put first — and "what went in the headline" is often where the real editorial decision lives.
Full-spectrum coverage is not the same as balanced coverage
A chart can also flatter a story. When the DNC released its 2024 election autopsy, ten outlets across the spectrum covered it — the Guardian, Vox, and NYT on the left, PBS, Axios, and AP in the center-left, Newsweek in the center, RCP and National Review center-right, the Daily Caller on the right. Run that through a bias chart and the sourcing looks balanced. Box checked.
The framing was not balanced, and it was not really trying to be. The Guardian led with the report's omission of Gaza, framing it as a failure of political honesty. Vox argued Democrats "don't need an autopsy" because the report was too incomplete to be useful. National Review called it an "unfinished shambles." Each outlet picked a different thing to be the news. "Balanced sourcing" and "balanced coverage" are not the same measurement, and only one of them is visible on a chart.
You can see the same pattern in the dropped obstruction charges against Illinois anti-ICE activists. CNN, AP, and the NY Post all ran the story — left, center-left, and right, one outlet each. A chart calls that a clean spread. But the NY Post built its piece around the fact that the dismissing judge was a Biden appointee, while CNN and AP treated the grand jury misconduct as the story. The spread of dots was even. The spread of frames was not.
Ratings are slow; framing moves story by story
There is also a tempo problem. Bias charts update on a cadence — Ad Fontes released its big revision in January 2026, AllSides refreshes through periodic reviews. That cadence is fine for a stable property like an outlet's general lean. It is too slow for framing, which can shift inside a single news cycle. The qualifier Fox News added to its daylight saving coverage — noting the push "is actually bipartisan" — is the kind of move that no quarterly chart will ever capture, because it lives at the level of one sentence in one article on one day.
Use ratings as a starting point, not an answer
None of this means bias charts are wrong. AllSides and Ad Fontes do something specific and do it well, and we said as much in our roundup of the best media bias tools in 2026. Knowing an outlet's general lean is worth knowing. It just answers a narrower question than most people think it does.
The fuller question — did this story get covered, by whom, and what did each outlet decide was the news — needs a per-story, per-headline lens. That is the gap framing analysis fills, and it is the work behind every story page on Signal/noise. If you want the method itself, our five-step framing analysis guide walks through how to do it on any story without a tool at all.
A bias chart tells you where a newsroom tends to stand. It will not tell you where it stood today. For that, you have to read the coverage — all of it, side by side — and watch what each outlet put first.