you are thinking of serial killers. But i can try. this stuff takes a long time to dig up though, so bear with me.
Some of my formulas were screwed up previously, here is the gist:
I didn't go to the spreadsheets, but I didn't see any immediate correlation between murder and either population density or education - so far, poverty is our best predictor.
I wonder if the trying to correlate the type of murder (murder 1, murder2, manslaughter, etc.) or "reason," for murder (crime of passion, financial gain, depression, etc.) would have a discernible demographic or geographic dimension... I'll try to find some stats along these lines to play with.
FBI expanded offense data for Homicides (all sorts of nifties, here): http://www.fbi.gov/about-us/cjis/ucr...d-offense-data
Serial killings are recurring murders that are unique events, as opposed to a single event.
If you kill ten people at a party, you are a mass murderer.
If you kill 1 person a day for 10 days, you are a serial killer.
Going over the data, the general trend is:
"If you have low poverty AND low urbanization, you are likely to have low murder rates."
individually, low poverty or low urbanization are not that strongly correlated, but when both are present, it almost always means low murder rate.
That seems to be a big part of it, anyway.
Thanks for posting this. I haven't had time to crunch the numbers, but doesn't it look as if the US has more murders just from people killing each other with their bare hands than many other countries in our peer group's overall murder rate?