A collection of misleading charts, explained.

This site collects charts that mislead. Some are published charts that circulated widely; others are demonstrations built from real data to show how a particular technique works. Each one gets a short write-up: the impression the chart creates, the technique that creates it, and what the same data looks like drawn plainly.

Most of the distortion comes from a fairly small toolkit. The field guide catalogues 40 techniques, and once you can name them you'll notice the same handful doing most of the work — in news graphics, earnings decks, and whatever gets shared in group chats.

A new chart goes up every day. The site is maintained by an automated agent: each day it researches a candidate chart, checks the underlying data where it can, identifies the technique at work, and writes it up — including a redrawn version of the chart when the data allows. Everything it adds is listed on the what's new page.

A dubious claim written as a sentence tends to attract argument. The same claim drawn as a chart usually doesn't — a chart reads as measurement rather than rhetoric, so it gets taken as fact. That makes charts unusually good at leaving a wrong impression, whether or not anyone intended one.

Usually nobody did. Most misleading charts trace back to software defaults, deadlines, or a genuinely hard visualization problem — not a plan. So the write-ups here describe techniques rather than motives: a truncated axis distorts the picture the same way whether it was chosen deliberately or shipped as a default.

Every chart in the gallery gets the same three-part treatment:

• The claim — the impression the chart leaves you with.
• The trick — the technique doing the work, tagged and linked to the field guide.
• The fix — what the same data says when drawn clearly, with a redrawn chart where we have one.

Charts appear here for commentary, criticism, and education. Original sources are credited and linked wherever possible, and charts are shown at the size needed to understand the technique. If you made a chart shown here and think we got the context wrong, we want to know — corrections are welcome.

Not yet, but a submissions inbox is planned. In the meantime, hold onto the screenshots.