Magnitude comparison under polar coordinates is harder than in Cartesian

What it is

Render the same data as a radial chart and as a Cartesian chart, ask readers to compare magnitudes, and the radial version is almost always harder: radial bars extend in different directions, demanding mental rotation before comparison; annulus comparisons mix angle with radius; a radial line's height differences wrap around the circle, so "whose peak is higher" requires chasing across arcs. In the Cartesian version all of this lands on "position compared along one common scale" — the top tier of the graphical-perception accuracy ranking. This leaf is the polar family's summary verdict: outside periodic structure (third leaf), choosing radial for magnitude-comparison tasks is downshifting on purpose.

Why it happens

A radial chart's magnitude cues are radial length and sector area, both ranked below position-along-a-common-scale in graphical perception: lengths pointing in different orientations require orientation normalization (mental rotation) before comparison, adding error and effort; area reading stacks the second leaf's quadratic distortion on top. Cartesian bars place all lengths in parallel orientation, and shared baseline plus parallel alignment turn comparison into a one-dimensional judgement. Radial layouts add centre occlusion and outer crowding: inner elements small and dense, outer ones large and sparse — precision resources distributed unevenly. These penalties compound, so "radial looks nicer" pays off only in display settings; in analysis it is a net loss.

Studying it

The evidence sits in the channel-accuracy ranking tradition of graphical perception: the elementary-task experiments of Cleveland and McGill (JASA 1984) established position along a common scale > length > angle/area, and radial charts' magnitude encodings (radial length, sector angle, annulus area) all fall behind position; the crowdsourced replication (Heer and Bostock, 2010) preserved the ranking's robustness at much larger samples. Direct radial-versus-Cartesian contrasts follow the same paradigm: time and error readers comparing magnitudes on twin versions of the same data, the radial deficit being this leaf's effect size.

Where it stops holding

"Generally harder" is not "never usable": periodic data (third leaf), directional data, and single-value display (one progress ring, no comparison task) still leave room for radial layouts; in decoration-forward, low-density communication settings, radial aesthetics may legitimately buy some precision away. The decision flow is fixed: first ask "is the main task magnitude comparison?" — if yes, Cartesian; then "does the data have a real cycle?" — if yes, radial is viable; neither, and radial enters only with the precision price explicitly accepted.

Applying it

• Decision table: magnitude comparison → bars/dots; periodic structure → clock/rose; multidimensional profile → parallel coordinates or faceted bars; radial reserved for display.
• Retrofitting existing radial charts: keep the shape as the cover, let a same-data Cartesian version in the body carry the comparison duty.
• Verification: give readers "rank the three largest values" on radial and Cartesian twins; radial simultaneously slower and more error-prone demonstrates the downshift — usable as a selection criterion.

Related

• Same group: U3.08.1 Polar coordinates map one variable to angle, the other to radius · U3.08.2 Equal radius increments give unequal areas; outer rings are systematically enlarged · U3.08.3 Polar coordinates suit inherently periodic variables · U3.08.4 Radar-chart axis order changes the polygon's shape and area
• Nearby: U2.07.1 Angles read imprecisely · U2.02.1 Bar charts suit magnitude comparison across categories
• Search terms: radial vs cartesian · elementary task accuracy · position encoding · graphical perception