This is an official ODERSA website. Here’s how you know
The official domain

The address of this site ends in odersa.org. Every service the association runs sits on a subdomain of odersa.org and nowhere else. If the address in your browser’s address bar does not end in odersa.org, this site is not ours.

Free, and no account

Everything is open straight away. No sign-up, no account, no password, no subscription, no advertising. Nothing is held back for those who pay, because there is nothing to pay for.

No data collected

This site does not follow you: no tracker, no tracking cookie, no measurement tool built into these pages, and nothing measured on your device. Our host counts requests in aggregate, as any server that answers does: a total, never a profile. You do not have to take our word for it: open your browser’s developer tools, go to the Network tab, and reload the page. You will see the full list of what the site asks for. Everything comes from odersa.org, nothing goes anywhere else.

Free to reuse

The content is published under the CC BY 4.0 licence. You may copy it, translate it, print it and pass it on, for your classes as much as for the people around you, on one condition only: credit ODERSA.

FrançaisEnglishEspañolالعربيةहिन्दीবাংলা中文

← Back to the blog

Published on 2026-07-25

Why does a tiny change look like a cliff on that chart?


A graph's vertical axis can turn an 8 percent bump into a dramatic cliff. The trick is old, well studied, and easy to catch once you know where to look.

Picture two bar charts, side by side. Same company, same year, the same eight percent rise in revenue. In the first chart, the bar barely lifts off the floor. In the second, it looks like the company's business just tripled overnight. Nobody touched a single number. Somebody just moved where the chart starts counting.

The cut you never see

Most bar and line charts put zero at the bottom. Slice that bottom off, start the axis at 90 instead of 0, and every small wobble in the data suddenly has far more room to stretch. A move from 97 to 98 fills almost the whole chart. The statistician Darrell Huff gave this move a name back in 1954, in his book 'How to Lie with Statistics': the gee-whiz graph. Seventy years on, the trick hasn't aged a day, because the underlying weakness it exploits is ours, not the software's.

The UK's Office for National Statistics, whose whole job is presenting numbers honestly, puts the rule plainly for bar charts: they should always include zero. The reasoning is about how eyes actually work, not about etiquette. We don't read the little numbers printed along the axis. We compare the length of the bars. A bar twice as tall reads as twice as much, even when the real figures sitting behind it are 97 and 98.

You can read the numbers right and still fall for it

Here's the part that should make you a little more humble about your own eye for detail. The visualisation researcher Michael Correll and his colleagues tested how people judge truncated bar charts, including versions with a little zigzag mark at the base warning 'this axis has been cut'. The warning didn't help. Neither did switching chart types. People could still tell you the actual numbers accurately when asked directly, but their gut sense of how big the difference was stayed just as inflated as if no warning had been there at all. The distortion doesn't attack your arithmetic. It attacks your first impression, the one that forms before you've consciously read a single label.

A truncated axis doesn't lie about the numbers. It lies about how the numbers feel.

When starting away from zero is the honest choice

None of this means every chart must crawl back to zero, and it's worth saying clearly: a line chart tracking a narrow, meaningful range is a different animal from a bar chart comparing lengths. A fever chart that only matters between 36 and 40 degrees, a central bank's daily interest rate, a stock index moving in fractions of a percent: forcing those onto a zero baseline would flatten everything into a dull, unreadable line and hide the real movement the chart exists to show. The problem sits specifically with bars and area, the shapes our eyes measure by size. A line chart, read as a trend rather than a stack of lengths, can often afford to zoom in. So the honest version of this habit isn't 'reject any chart that skips zero'. It's 'notice when it happens, and ask whether the shape being shown matches the story being told'.

Two seconds before the shape convinces you

A chart is meant to save you time: one glance instead of a column of figures. That's exactly why it works so well as a trick when someone wants it to. The fix costs almost nothing next to the deception it prevents. Before the shape of a chart convinces you that something exploded or collapsed, glance at where its vertical axis actually begins. Most of the time it will be a perfectly ordinary chart, zero at the bottom, nothing to see. Once in a while, you'll catch the cliff for what it is: a hill, with the bottom cropped out of frame.

Topics: charts statistics media literacy

Sources

Keyboard shortcuts

?Open this help
EscClose the panel or this help
1-3Pick an answer during the quiz
EnterNext case, once you've read the answer
TabMove from link to link: the whole site works by keyboard