There are some biases inherent to the relevant data. For example, you might be right about this specific situation (police shooting unarmed black people), but what about other instances of violence or misconduct?
We need to consider where data about that comes from. Are we to trust the accuracy of police reports and their own statements as they pertain to misconduct? IA investigations that frequently find no wrongdoing even when it’s plainly obvious that the situation was handled improperly? Or at least could have been avoided? Should we rely upon charges filed against police and their conviction rate?
There’s no official national database for this stuff by the way, and localities almost never produce such metrics willingly, so it’s up to someone to comb through public records. And that, like the case of body cam videos, assumes that police will properly follow FOIA requests.
I’m not suggesting that we assume the absolute worst, but we need to recognize that whatever the data says, the situation is almost definitely worse than that due to these biases.