There's value in trying your [dataviz] worst

In the past I’ve always asked students to create their best possible graphs in ggplot2 to practice creating clear, engaging data visualizations. Recently, I’ve realized value in adding a few early exercises that encourage students to make their worst. Why is it good to make ggplot2 graphs so (intentionally) bad? Here are four ways that creating a purposefully disgusting graph promotes learning and exploration in ggplot2: Making the worst graph encourages creativity while eliminating the stress of producing a perfect final visualization (unless, of course, it’s perfectly hideous) Making the worst graph requires extensive customization and code Making the worst graph means critically thinking about and identifying contributors to “bad” graphs Making the worst graph is surprisingly fun I was recently inspired to try my absolute dataviz worst by a #tidytuesday prompt using data from Sarah Leo’s efforts to improve imperfect graphs in The Economist.