Updated October 2026
Radar charts are easy to build and easy to build badly — the geometry doesn't stop you from plotting misleading data, it just draws whatever you give it. This is a working checklist of the mistakes that most commonly undermine a radar chart's credibility, and how to avoid each one.
The single most common radar-chart mistake. Plotting "revenue in millions" on one axis and "customer satisfaction out of 10" on another, even rescaled to fit the same visual range, produces a polygon shape that looks meaningful but is actually an artifact of arbitrary rescaling choices, not a real signal. Fix: every axis should measure something on a genuinely comparable scale — ideally the literal same scale (all 1–10, all percentages, all percentile ranks). See axes explained for more on choosing axes.
A radar chart's whole value is instant shape-reading — that breaks down past roughly 3–4 overlapping datasets (this tool caps it at 4 for exactly this reason) or past roughly 8–10 axes. More isn't more informative here; it's just harder to read. See how many axes is too many.
A radar chart's center point should represent the true zero of the scale, not an arbitrary minimum. Starting the scale at some non-zero floor (to make small differences look dramatic) is the radar-chart equivalent of a truncated bar-chart y-axis — a well-known way to visually exaggerate differences that misleads anyone reading the shape at a glance rather than checking the actual axis values. This tool's radius scaling always runs from a true zero at the center to the max at the outer ring.
Randomly ordered axes make related criteria scatter around the circle rather than cluster, which makes the shape harder to interpret meaningfully at a glance. Group related axes together, or order by priority/importance, so a viewer's eye reads structure rather than noise. See axes explained for the full guidance.
A radar chart implicitly claims "these items are directly comparable across these criteria." Comparing a budget product against a premium product on identical axes with no context about price or category can produce a technically accurate but misleading impression — the premium product will usually win on quality axes without the chart ever showing why. If context matters (price tier, target audience, category), state it explicitly next to the chart rather than letting the polygon imply a comparison it doesn't fully support.
Radar charts compare categories, not moments in time. "Revenue by quarter" plotted as 4 axes (Q1, Q2, Q3, Q4) technically works but reads worse than a simple line or bar chart, because a radar chart's polygon shape doesn't naturally communicate "trend" or "direction" the way a line moving left to right does. Reserve radar charts for genuinely categorical, non-sequential criteria.