How we rate outlets: Left, Center, Right
Left, Center and Right labels show where published media-bias ratings (AllSides, Ad Fontes Media, Media Bias/Fact Check) place each outlet, not who its audience is. They're approximate, and they aren't a judgment of any single article.
We don't pick an outlet's column ourselves. It comes from three published media-bias ratings: AllSides, Ad Fontes Media and Media Bias/Fact Check. Each rating counts as Left, Center or Right like this:
- AllSides: Left or Lean Left counts as Left. Center counts as Center. Lean Right or Right counts as Right.
- Ad Fontes Media: a bias score below −6 counts as Left. A score from −6 to +6 counts as Center. A score above +6 counts as Right.
- Media Bias/Fact Check: Left or Left-Center counts as Left. Least Biased counts as Center. Right-Center or Right counts as Right.
The column most of an outlet's available ratings agree on is its column. If they split evenly, AllSides decides; if AllSides hasn't rated the outlet, Media Bias/Fact Check decides. An outlet that none of the three has rated isn't used. Affiliates and wire copies take their network's column.
A channel with no political rating of its own, such as a weather or sports channel, goes in the Center. It never takes its parent network's column.
We use only US national outlets: no foreign or local ones. TikTok creators appear only if they have a published bias rating.
An outlet needs at least 3 graded pieces before it is ranked on the Outlets page.
How we grade articles and videos
Each article or clip gets a letter grade for how it reports, not for whether we agree with it. The standard is Jim Lehrer's rules of journalism (Aspen Institute, 1997). We grade only text or transcript we actually read.
The letter comes straight from the score: A is 90 to 100, B is 80 to 89, C is 70 to 79, D is 60 to 69, and F is under 60. We never grade on a curve or move a piece to fit a share.
What each lapse costs:
| Lapse | Points |
|---|---|
| Central lapse: the core claim rests solely on anonymous sources, with no named official or document backing it in the piece | −45 |
| A claim, analysis or opinion in the outlet's own voice (or a guest's, inside a news segment) that isn't labeled as such, including treating people charged as guilty | −24 |
| A factual error, or a claim the piece can't support | −24 |
| An unnamed source (or anonymous sourcing on a core claim that named officials or documents also support), or a figure with no stated source | −18 |
| Leaving out a fact or voice that the other pieces report, including airing only one side | −18 |
| A loaded word or phrase in the outlet's own voice | −12 |
Fabrication, or letting an unnamed source attack someone, is an automatic F. Every deduction is backed by a quoted phrase (articles) or a timestamped quote (video), and the popout lists each one. Every quote is word for word.
Pundits are graded on what they actually said, not on what someone assumes they meant.
Opinion pieces carry an Opinion tag. We grade the facts they rely on, not the opinion.
Center picks
Center picks are placed by published bias ratings of the show's network. When no Center-rated podcast has a new episode, we show a labeled news clip instead.
Disagree with our approach or a rating? Have other feedback?
Tell us what you think we got wrong, or anything else. We read every message.