Why the spider chart is a powerful, yet deeply flawed, tool for data visualization

Why the spider chart is a powerful, yet deeply flawed, tool for data visualization
Image: Intellect (generated)

In short: Radar charts, or spider plots, offer a striking way to visualize multivariate data by mapping variables onto radial spokes. While they excel at highlighting outliers and comparing performance across small datasets, they often distort reality, exaggerate large values, and create artificial connections between unrelated metrics. Use them with extreme caution.

First introduced by Georg von Mayr in 1877, the radar chart—also known as a Kiviat diagram or star plot—displays multivariate data by arranging quantitative variables on equi-angular spokes. Each observation is represented as a polygon, with the distance from the origin proportional to the variable's magnitude. This format is particularly effective for small-to-moderate datasets where the goal is to identify clusters or compare a single observation against a benchmark, such as evaluating a basketball player’s stats against league averages or assessing the performance of sorting algorithms.

However, the chart’s geometry introduces significant cognitive and mathematical pitfalls. Because the area of the polygon scales with the square of the linear measures, the chart inherently exaggerates the impact of large numbers. For instance, a value of 90 creates a polygon area significantly larger than one with a value of 82, despite the linear difference being relatively small. Furthermore, the radial layout imposes an artificial cyclic structure, forcing the first and last variables to sit adjacent even when they share no logical relationship. This can lead to spurious correlations and visual clutter that makes comparing multiple samples nearly impossible.

Critics argue that radar charts are poorly suited for complex trade-off decisions, as radial distances are notoriously difficult for the human eye to judge accurately. When datasets exceed a few hundred points or include too many overlapping samples, the visualization becomes overwhelming and prone to misinterpretation. While algorithmic reordering of variables can sometimes mitigate these issues by grouping related metrics, experts often suggest alternatives like parallel coordinates, small bar charts, or principal component analysis for more robust statistical insight.

Ultimately, the radar chart is best reserved for scenarios where the data is naturally cyclic—such as a 24-hour cycle—or when all variables are ordinal and share a common scale where 'more' consistently means 'better.' When these conditions are not met, the chart risks distorting the very patterns it aims to clarify, turning data analysis into a game of visual guesswork rather than objective assessment.

Opens the Intellect+ feed with this card first. Swipe up for the next one.

Sources

To cite this card, link to https://intellectplus.co.uk/app/science-radar-chart.

How we source, check and correct cards: Editorial standards.

General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.

Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?

Related

Finding something worth knowing…