@johnpowell I was saying the statistic they chose isn’t perfect. I was not saying I was going to try to devise the perfect formula, or even that there is a perfect formula.
However, if I were going to assess the dangerousness of a place compared to another place, I would want to take into account differences that won’t be reflected directly in the numbers, including yes making some effort to acknowledge evidence that crimes may be happening but unreported in some places, or more often reported and recorded in one place than another.
In fact, to assess dangerousness, I would even try to assess danger that is not resulting in harm, for example that exists but is usually avoided because the local people know it’s dangerous.
Taking my own experience of Chicago, for example, I attended university there, and one of the first things they told every new student was to NEVER go beyond certain streets, as well as other aspects of the extreme security they had available near campus. Locals also often gave warnings about not to go into certain (very very large) parts of the city – and in many cases they meant even on a bus or car or non-commuter train. There were large areas where there was probably a lot of danger and violence but it wasn’t getting reported. I saw people brutalize each other and no one was doing anything about it, even in “safe” parts of town. So there was both a lot of unreported violence, and enormous areas where it was considered very dangerous (at least for well-off-looking people) but the danger was mostly being avoided.
If I wanted a serious assessment of dangerousness, I’d want that kind of information more than the number of officially known crimes. Also, such crime stats make places where people report most disturbances and the police investigate and record them seem relatively more dangerous than they are compared to places where it’s less common to report or record crimes.