Radar Chart
A radar chart, also known as a spider or web chart, plots multivariate data on a series of axes radiating from a central point - one axis per variable. Data values are plotted along each axis and connected to form a polygon, making it easy to see the overall “shape” of a profile at a glance.
Radar charts excel at comparing performance or characteristics across multiple dimensions simultaneously.
An example of an embedded radar chart
Creating an Effective Radar Chart
Recommended data types for each axis:
- X-Axis Categorical data (each category becomes one spoke of the radar)
- Y-Axis Numerical values
- Breakdown Axis Categorical data (each unique value becomes a separate polygon)
Description
- Spokes - equi-angular axes radiating from the center; each spoke represents one variable
- Polygons - values along the spokes are connected to form a shape; a larger, rounder polygon indicates uniformly high values
- Center - represents the minimum or zero value; outer ring represents the maximum
- Colors - when multiple series are plotted, each polygon uses a distinct color with semi-transparent fill
- Legend - maps each color to its corresponding series
Ideal Data for Radar Charts
Radar charts work best when your data has:
- 3–10 variables - fewer than 3 variables doesn’t benefit from the radial layout; more than 10 spokes becomes crowded
- Comparable scales - variables should be on similar scales or normalized so no single spoke dominates visually
- Related metrics - the dimensions should belong to the same conceptual group (e.g., skill categories, product attributes)
When to Use a Radar Chart
- Performance profiles - compare strengths and weaknesses across multiple attributes for one or more subjects
- Skill assessments - visualize competency scores across different areas for a person or team
- Product comparisons - contrast specifications or feature ratings across multiple products
- Before/after comparisons - overlay two polygons to show how performance changed across all dimensions
When to Avoid a Radar Chart
- Precise value comparisons - angular area is hard to read accurately; use a bar chart when exact differences matter
- More than 2–3 series - overlapping polygons become indistinguishable with many series
- Variables on very different scales - without normalization, one spoke will overwhelm the rest and distort the shape
- Ordered or time-series data - use a line chart when the X-axis has a meaningful sequence