Hacker News Reader

Comment by emaginniss | Hacker News Reader

That leaves 50% in the unknown category. Who knows whether those other 50% have the same distribution as the knowns? It's also possible that most of the unknowns are stranger murders since the offender might be harder to identify in those cases.

So, the ratio is somewhere between 10% stranger to 90% family and acquaintance and 60% stranger to 40% family and acquaintance.

View on HN

Here's a 1977 NYT story [0] during a time in which there were many more murders and the stranger-rate nationwide was estimated at 20%:

- 1,622 homicides

- 1,246 where (known vs. stranger) relationships were established

- 354 of that 1,246 attributed to strangers

Today's clearance rate is lower [1], but it's not really apples-to-apples comparison for a lot of reasons relating to recordkeeping conventions. But clearance rate for white homicide victims in NYC is still around ~85%. The past few years, the % of overall victims that were killed by strangers is around 5% [2].

So yes, there is still some uncertainty in the exact ratio of stranger/friend/unknown suspects. But still not rational to automatically assume that Lee was the victim of random crime.

[0] https://www.nytimes.com/1977/04/20/archives/murders-by-stran...

[1] https://www.cbsnews.com/newyork/news/crime-without-punishmen...

[2] https://www.nydailynews.com/new-york/nyc-crime/ny-stranger-o...

I assume gang related violence kinda skews the concept of the victims and perpetrators knowing each other.

Indeed. Homicide spreads like a disease among people who are in connected social graphs. https://www.chicagomag.com/city-life/december-2013/the-small...

This is fascinating. I live in Baltimore and follow a lot of street connected profiles as sort of an OSINT information gathering hobby. I’ve always seen on instagram where said population has RIP in their bios sometimes for 6-7 people. Really puts it in perspective that in a high homicide rate area, if you’re in the network, your risk is almost Warzone (or higher) on a per capita basis

The most fascinating thing to me is that the biggest indicator that you are going to be murdered is that you have been arrested in association with someone else who did get murdered. Police can address homicide rates by intervening in the lives of these future victims. When some low-level neighborhood criminal gets shot, go find all that guy's friends and offer them jobs in other cities.

They do this in Baltimore, it’s called GVRS. They find and assign risk scores (related to arrest but also some other stuff) and go seek out the person, offer them help or threaten to come down hard on them if they commit crimes

The pilot had a homicide reduction of 30% iirc but we will see over time as it expands to whole city how it plays out

That makes complete sense to me. Being involved in any sort of business where the underlying activity is criminal in nature means that contract disputes and even competition for market share have to be underwritten by violence and the willingness to resort to violence with no legal system to turn to for redress.

In the same way that being served with a lawsuit means likely to spending some time with lawyers or in court, a low-level criminal being shot indicates that his coworkers in the enterprise are likely to spend some time in shootouts.

Reply on news.ycombinator.com