Guidelines
Suggest edit

Use a diverging palette when values deviate from a baseline

For baseline comparison in quantitative charts, use a diverging color gradient on the value encoding to improve insight and mitigate weak separation between below-baseline and above-baseline values for readers.

  • purpose:refine
  • basis:heuristic
  • data:quantitative
  • quality:insight
  • lever:encoding
  • operator:difference
  • channel:color-hue:use
  • aesthetic:color:use

advice

Diverging gradient

Use a diverging color gradient when the message is how values differ from a baseline. For example, map below-baseline and above-baseline values to clearly different hues and place a light gray center at the baseline instead of white.

reason

Why diverging palettes sharpen baseline reading

When the main question is which side of a baseline a value falls on, a single sequential scale hides that split.

Mechanism: Two distinct hue directions make below-baseline and above-baseline values immediately separable, and a neutral center marks the baseline itself.

Evidence: The post recommends diverging palettes when the goal is to emphasize how a variable diverts from a baseline, says both sides should use clearly distinguishable hues, and says the center color should ideally be light gray rather than white (Muth, 2018).

context

When to use a diverging palette

  • User Goal: See whether values are below or above a reference point.
  • Data: Quantitative values are compared against a meaningful baseline.
  • Chart Setting: Color carries the direction and amount of deviation.
  • Success Criterion: Readers can tell both magnitude and side of the baseline quickly.

exceptions

When not to use a diverging palette

Break it when: There is no meaningful baseline and the message is only low-to-high magnitude. Why: Diverging palettes are for deviation from a reference point.

costs

Tradeoffs of diverging palettes

Sacrifice: The scale becomes more complex than a single sequential gradient. Risk: An arbitrary midpoint can suggest a meaning the data does not support. Mitigation: Use a diverging palette only when the baseline is genuinely important to the message.

mistakes

Common diverging-palette mistake

Mistake: Using similar hues on both sides of the baseline or centering the palette on white. Why it fails: The split around the baseline becomes harder to read.

check

How to check a diverging palette

Failure Sign: The midpoint color does not clearly mark the baseline or the two sides look too similar. Quick Check: Verify that the legend midpoint matches the baseline value. Stronger Test: Check whether a reader can tell from color alone which values are above and below the baseline.

fix

How to fix a weak diverging palette

  • Switch from a sequential scale to a diverging scale.
  • Use clearly distinguishable hues on the two sides of the midpoint.
  • Replace a white midpoint with a light gray midpoint at the baseline.

References

Muth, L. C. (2018). What to consider when choosing colors for data visualization. https://www.datawrapper.de/blog/colors