Guidelines
Suggest edit

Avoid dark yellowish-green colors when optimizing categorical palette preference

For aesthetic evaluation of categorical palettes, avoid dark yellowish-green hues on categorical color encodings to maximize aesthetics and mitigate disliked color-region choices for viewers judging palette appeal.

  • purpose:refine
  • basis:empirical
  • data:categorical
  • quality:aesthetics
  • lever:encoding
  • polish:palette
  • aesthetic:color:avoid
  • channel:color-hue:avoid

advice

Filter the disliked yellow-green region

Remove dark yellowish-green colors when average viewer preference matters. For example, avoid hues around 85°–114° with lightness around 35–75, and penalize nearby yellow-green colors that border this region.

reason

Why this region hurts preference

A preference-driven palette can drift into disliked yellow-green colors when cool blues and strong separation are both rewarded. Filtering that region keeps the palette from solving discriminability by introducing colors that people tend to dislike.

Mechanism: Preference weighting pulls toward cool blues, and discrimination weighting can then push their opposites toward darker yellows; filtering the dark yellowish-green region blocks that undesirable path.

Evidence: The paper excludes a dark yellowish-green region because those colors are generally disliked on average, and reports that removing this region still preserved enough color space for discriminable pairings while improving typical palette preference (https://doi.org/10.1109/TVCG.2016.2598918">Gramazio et al., 2017).

Notes: The paper also adds a penalty for nearby yellow-green candidates so border cases are sampled less often.

context

Use when optimizing for average viewer preference

  • User Goal: Raise average palette appeal.
  • Task: Generate or revise a categorical palette that must balance preference with discrimination.
  • Data: Categorical groups encoded by color.
  • Chart Setting: A palette that includes blues or other colors that could otherwise drive selection toward dark yellow opposites.
  • Audience: A broad audience whose average preference matters more than individual taste.
  • Success Criterion: Fewer disliked yellow-green colors without losing usable category separation.

exceptions

Do not use when a known audience specifically wants those colors

Break it when: The design is for a known observer or audience that specifically prefers these yellow-green colors. Why: The paper notes individual differences and that some observers may like them even though they are generally disliked on average.

costs

Filtering the region reduces some available contrasts

Sacrifice: You lose part of the yellow-green color space. Risk: If applied blindly, you may remove some otherwise discriminable oppositions against blues. Mitigation: Keep using other yellow regions outside the filtered dark zone.

mistakes

Common yellow-green failure

Mistake: Letting discriminability and coolness push the palette toward dark yellow opposites of selected blues. Why it fails: That interaction produces colors the paper identifies as generally disliked.

check

Inspect the hue-lightness region directly

Failure Sign: The palette contains muddy dark yellow-greens that feel noticeably less appealing than the other colors. Quick Check: Check whether any palette color falls in the hue range 85°–114° and the lightness range 35–75. Stronger Test: Compare the current palette against a version with those colors replaced and collect preference ratings.

fix

Replace or penalize the filtered region

  • Replace colors inside the filtered region with nearby colors outside it.
  • Penalize neighboring yellow-green candidates so the palette generator samples them less often.
  • Regenerate the palette and keep the version with the higher minimum pair preference.

References

Gramazio, C. C., Laidlaw, D. H., & Schloss, K. B. (2017). Colorgorical: Creating discriminable and preferable color palettes for information visualization. IEEE Transactions on Visualization and Computer Graphics, 23(1), 521–530. https://doi.org/10.1109/TVCG.2016.2598918