When it comes to data-ink ratio, optimize rather than maximize

Summary

Minimalist data visualizations aren’t necessarily ideal: The Story of Goldilocks and the Three ChartsAbstract¶Aspiring to minimalism is common when you’re designing a data visualization. Minimalism is often treated as ideal for understanding a graph quickly, but sometimes ‘wasting ink’ makes sense to add emotion and engagement. This article argues for a “Goldilocks” balance between data and decoration, showing example cases where more ‘chartjunk’ improve clarity, comprehension, and attention depending on context. Effective visualizations should calibrate design choices to audience and purpose. But, empirical research on the ideal data-ink ratio of visualizations is are.Figure 1:The balance between data-ink and chartjunk in data visualizationsSources: Darkhorse Analytics Blog - Data Looks Better Naked, Visual Business Intelligence - The Chartjunk DebateBest practices in data visualization often suggest minimizing decoration on your data viz and maximizing the data itself. But a recent review of literature on data visualization suggested that striking an optimal balance between ‘data and decoration’ on your chart can improve the focus and clarity of chart/graph design McGurgan et al. (2021).A high data-ink ratio, that is, a high proportion of ink or pixel encoding data or information, will make your charts and figures really stand out. However, this comes at the cost of losing some of the visual cues and signals that will help draw attention to it. Hence it is important to find the right balance for your purpose and audience.Chartjunk is stuff on your graph that’s not communicating actual data¶Figure 2:Data ink vs. ChartjunkData-ink is the total ink on a graph that represents data. The non-erasable core of the graphic, the non-redundant ink representing or encoding data information." Everything else is Chartjunk Tufte (2001).Data-ink ratio is like the signal-to-noise ratio of the graph:¶Data ink ratio = data-ink / total ink used to print the graphic.You can also think...