Use a sequential colormap when values have no meaningful midpoint

Use a sequential colormap when color encodes ordered values and the data has no meaningful middle point. For example, in a heatmap, choropleth, or scalar-field view, map low-to-high values with one ordered scale instead of a rainbow palette.

Ordered color scales keep visual differences aligned with data differences. Rainbow palettes change unevenly across hue, create artificial bands, and make smoothly varying data look grouped into named colors.

Mechanism: A sequential scale preserves relative magnitude, so nearby values look nearby and larger value changes look larger. This reduces false grouping and makes low-to-high structure easier to read, including for some readers who cannot distinguish all rainbow hues.

Evidence: The paper reports that rainbow colormaps distort perceived value relationships, create false divisions in smoothly varying data, and become less accessible for colorblind readers; it recommends sequential colormaps for ordered or continuous data without a meaningful midpoint (Szafir, 2018).

• User Goal: Compare magnitude, locate highs and lows, or read smooth variation.
• Task: Compare ordered values from low to high.
• Data: Ordered or continuous values without a natural center value that readers must compare against.
• Chart Setting: A chart uses color as the main value encoding.
• Audience: Readers must interpret value differences at a glance, including some with color-vision deficiency.
• Success Criterion: Equal-looking color steps correspond more closely to equal data steps, without false bands.

• Break it when: The data has a meaningful midpoint such as a baseline or natural zero. Why: The source recommends a diverging colormap for direct comparison to that middle point.
• Break it when: The data is categorical. Why: The source notes that hue-based grouping can be useful for categories rather than ordered magnitudes.

Sacrifice: You give up strong emphasis on above-versus-below a central reference value. Risk: If a meaningful midpoint does exist, a purely sequential scale makes comparison to that midpoint less direct. Mitigation: Switch to a diverging scale when the midpoint is part of the task.

Mistake: Using a rainbow palette for smoothly varying values. Why it fails: Readers see artificial blue, green, yellow, and red bands, and equal numeric differences no longer look equal.

Failure Sign: Smooth data looks split into named-color regions or abrupt boundaries. Quick Check: Scan the scale and ask whether some hue transitions look much stronger than others for similar value steps. Stronger Test: Compare the same chart with a sequential scale; if apparent clusters or boundaries disappear, the original palette was distorting the data.

• Replace the rainbow palette with a sequential colormap.
• Make the color progression visually ordered from low to high.
• Reserve categorical hue sets for categorical data.
• Switch to a diverging colormap if the task depends on a meaningful middle value.