Color, Pattern, and Direct Labels

About 1 in 12 men cannot reliably tell certain colors apart. If your chart says red means loss, those users see no difference at all. 🎨

Encode every series with a second channel besides color: a pattern, a shape, or a direct label. Color becomes a bonus, not the only cue.

Give bars and areas distinct patterns like stripes or dots. The fill tells the series apart even in pure grayscale.

On line charts, mark points with distinct shapes: circles, squares, triangles. Each line stays distinguishable without relying on hue.

A separate legend forces a color-matching game. Put the label right on the line or slice so the name sits next to the data.

Some hue pairs, like red and green, are the hardest to tell apart. A color-blind-safe palette such as Okabe-Ito avoids those clashes.

Adjacent series and their text labels need real contrast. WCAG asks for at least 3 to 1 between meaningful graphical parts.

Quick check: view your chart in grayscale. If you cannot tell the series apart, neither can a color-blind user. Fix it before shipping.

This stacking of cues is called redundant encoding: color plus pattern plus label. Lose one cue and the chart still reads clearly.

Balance matters. Heaping on every pattern at once gets noisy. Add just enough so each series is distinct and the chart stays calm.

Direct labels and patterns also rescue charts viewed in sunlight, on cheap screens, or printed in black and white. Everyone wins. ☀️

Why is color alone a poor way to distinguish chart series?

You learned to never lean on color alone: add patterns, shapes, or direct labels, choose safe palettes, and test in grayscale. 🖤