Pick a year, note how many police shootings are listed there for that year, then generate 20 random numbers in the range [1, N] where N is the number of said shootings, and look at the corresponding killings. That should give some reasonable insights.
If any of the people in this thread wish to have a go at that, try here: http://www.killedbypolice.net/kbp2018
Pick a year, note how many police shootings are listed there for that year, then generate 20 random numbers in the range [1, N] where N is the number of said shootings, and look at the corresponding killings. That should give some reasonable insights.