Schwabish (2014) describes and outlines a basic limitation of donut charts as, Donut charts—in which the centre of the pie is punched out—just exacerbate the problem: the empty centre makes the reader estimate the angle and arrive at other qualitative part-to-whole judgments without being able to see the center where the edges meet. Surprisingly, donut plots have their own followers and we do observe them in business reports or in news media as well. One of the places where donut charts can be useful is where the number of variables to be displayed is small, such as data related to yes versus no or male versus female.
