The bolded above is your basic problem. The press takes great pains, especially as of late, to imply that race was a factor without providing anything other than a racial difference between the officer and the "victim". Of course, when the racial differences are reversed, this rarely occurs. That difference in reporting alone skews that "internal score" so much as to render it meaningless.
When a small racial subset (14%) of the population is responsible for about half of the nation's homicides, one (using that "internal score"") is not permitted to use that fact to infer that blacks, in general, are more likely murders (since that would be racist), but if less than 1% of police officers kill any (much less black) folks then it is somehow fair game (using your sacred, personal "internal score"?) to label police, in general, as more likely violent, racist killers.