Now consider we have multiple data points of police committing murder and getting away with it. How can we argue that it's rare, given that we have several instances of police getting away with it and not being considered part of said recorded statistics?