DataslopePie Charts and Why They Are Controversial
When pie charts are reasonable, when they fail, and what to use instead
The pie chart is the most controversial chart type in data visualization. Some people use them constantly; many serious practitioners would happily ban them. This page explains both positions and gives you a clear rule for when a pie chart is — and isn't — the right tool.
What a pie chart shows
A pie chart shows parts of a whole. The full circle = 100% of the total, and each slice's angle represents the share contributed by one category.
The chart says "Organic is the biggest chunk." With 5 slices and a dominant one, it works reasonably well.
Why critics dislike pies
Recall the Cleveland-McGill ranking from earlier:
- Position along a common scale ← most accurate
- Length
- Angle / slope
- Area
- Volume
- Color intensity ← least accurate
Pie charts encode quantity using angles and areas — both near the bottom of the ranking. That's the technical core of the criticism: the eye is bad at comparing angles. Try the next chart:
On the pie, ranking the slices takes real concentration. On the bar, it's instant. For comparison tasks, bars beat pies almost every time.
When are pies acceptable?
Pie charts aren't universally bad. They are reasonable when:
- You have just a few categories (2-4 slices).
- One slice is dramatically larger than the others.
- The story is part-of-a-whole, not precise comparison.
- You need the audience to immediately recognize "this is a share of something."
Example: "60% of users are on mobile, 30% on desktop, 10% on tablet." The story is composition, the categories are few, and the dominant category is clear. A pie works.
When are pies bad?
- Many slices (more than ~5).
- Slices of similar size (humans can't rank similar angles).
- Comparing two pies (a famously hopeless task).
- 3-D pies, exploded pies, donut-with-icon pies — any decoration that makes it harder to read the angles.
Better alternatives
When a pie isn't a good fit, the alternatives are almost always:
- Sorted bar chart — for comparing magnitudes.
- Stacked bar chart (single tall bar) — for parts-of-a-whole with named segments.
- Treemap — for nested hierarchies.
A horizontal stacked bar carries the same "part of a whole" message without forcing the eye to compare angles.
A balanced rule of thumb
Use a pie when there are ≤4 slices, one slice dominates, and the story is composition. Use a sorted bar chart in every other case.
That single rule will keep you out of trouble.
Why this matters: charts as arguments
A pie chart, when used carelessly, makes data harder to read than the raw numbers. That undermines the whole point of visualization. The reason this chapter exists is not to win an internet argument about pies — it's to make you think about which encoding earns its place on every chart you make.
The deeper lesson: every chart type is a tool, and tools have appropriate uses. Pie charts have a small but real niche. Don't use them outside it.
Check your understanding
Why are pie charts often criticized for comparing values?
Pie charts cannot be interactive.
Pie charts are mathematically incorrect.
Pie charts encode quantity using angles and areas, which the human visual system judges much less accurately than position and length used by bar charts.
Pie charts are too colorful.
In which situation is a pie chart most reasonable?
Comparing the populations of 30 countries.
Showing a trend over time.
Showing a part-of-a-whole composition with 2-4 slices where one slice clearly dominates.
Showing the distribution of customer ages.
A colleague hands you a chart with two pies side by side, asking you to compare them. Why is this a difficult task for the viewer?
Pies are forbidden in modern reports.
The colors might clash.
Comparing the same slice across two different pies requires comparing two angles, which the eye is poor at — and any subtle change in slice composition makes the comparison harder still.
Two pies cannot fit on one page.
The best alternative to a pie chart for showing parts of a whole when comparison matters is usually:
A 3-D pie chart.
A scatter plot.
A sorted bar chart, or a single stacked bar.
A histogram.