1. 26
    1. 2

      I’m not going to say stacked area graphs are always bad, I’m sure there are cases where they work well, but I too have found myself with difficulty interpreting them due to the baseline issue. Overlapping line charts can look messy, but messy and harder to misinterpret is usually better for a data visualisation than clean and easier to misinterpret.

      On the other hand, if you’re stacking proportions (so the graph height is always 1), I think they’re pretty great. That’s essentially a pie chart with a time axis, which is very useful.

      1. 1

        I think for plotting proportions a stacked bar or clusters of bars both work pretty well. If the proportions change much then I prefer clustered bars, otherwise it can become a bit hard to read, imo