I care very little about alleged dog whistles these days. You can only cry wolf so many times and all that.
> So it doesn't even matter if the statistics show the police kill extra in proportion to the neighborhoods the patrol and that in turn is proportional to the crime rate, because you haven't Baysianed deep enough to find the cycle.
Maybe you're right, but the conversation as far as I'm aware is still very much fixated on the former question. We are locked on this question and we can't meaningfully "baysian" our way to deeper questions without more authoritative data. If one actually cared about solving the problem, he ought to support this sort of data collection initiative.