Yeah, still not a valid response, because you’re relying on flawed data to determine “where there is the most crime.”
If the geographical placement of police presence in a city is dictated by the past ten years of police activity, and police activity is influenced in part by racial profiling (this is true in NYC), then the data from that decade is skewed to implicate minority areas as having higher crime rates than they actually do. If a bot is policing based solely on that flawed dataset, it will spend more time in minority areas than it should.
This does not mean there is no crime in these areas, or even that there is not a disproportionate amount of crime than is present in other areas. It does mean that the police presence is higher than it should be.