I suppose this is possible using just one data set for each attack type. But I want to find the average number of candies for hit 1, hit 2, hit 3, etc., for each method.

Statistics: For each method, create parallel dot plots, histograms, and/or boxplots for the total number of hits after 10 candies. Discuss shape, center, spread. I’m particularly interested in seeing whether some of my predictions (e.g., that the “over the top” method will generate the greatest spread because it’s the one that is hardest to hit in the “sweet spot” and the one I missed most often) are correct.

Does that help at all?