> Statistically that's a pretty sensible assumption.
Interesting, is there a source or some data you’re aware of that suggests that it’s a statistically safe assumption?
[0] https://www.nas.org/academic-questions/36/3/did-american-pol...
[1] https://time.com/4779112/police-history-origins/
[2] https://www.britannica.com/biography/Rodney-King
[3] https://en.wikipedia.org/wiki/George_Floyd
A very different bar. A minority of cops can be white supremacists and because of the power they wield it's still sensible to treat them like every interaction is with a a white supremacist. As an example, a cop can legally kill you in many cases (or deny you freedom or seize your assets). If you had, say, a 20% chance of encountering a cop who was a white supremacist it would be sensible to treat every interaction as if that were the case.
Consider how unevenly weighted the outcomes depending on whether you assume a cop is racist when factoring how sensible it is to assume they are.
This is part of why we have juries. The letter of the law must be nullified sometimes in the interest of justice.