Distribution of the distribution of distances between spending centroids and median voter (take the weighted median precinct by registered voters). Plot should be a density chart that overlays this for uncontested election cycles, contested election cycles that the incumbent handidly won, and election cycles that ended in a runoff.
Similar plot as above, but instead of distance to median voter, do HHI.
Things needed: