1. The raw count shows our perception of police murders with respect to race because we typically see counts not population adjusted. If that was more on display to the white communities, maybe they (we, I’m part of that) would be more aware of the magnitude of the problem.

2. I did use appropriate ratios based on the number of white and black people per age bracket, I just didn’t plot it.

3. The largest variables missing are : how many stops, how many arrests, for what “crimes”, probability of conviction, and police density per population. I would prefer to have a per zip code breakdown, but the trouble is that even police murders was not a reliable statistic available until 6 years ago. Police unions would largely block such stats from getting out.