Hacker News Reader

Comment by vessenes | Hacker News Reader

Analyses like these are always helpful, and often raise lots of questions, maybe even more questions than they answer. As the parent of a black child, I have a personal interest in this data -- the statistic that some counties have 20x higher risks for unarmed blacks is pretty terrible.

In my mind, I tend to assume that criminals, active or former are more likely to be shot at than non-criminals, whether or not they are armed. I'd really like to see the data normalized against prior convictions or in-process-of-a-crime stats; that would help me understand:

1. Is the effect magnified or dampened by some sort of differences in black and white criminality in these areas?

2. Are these shootings happening while people mostly commission crimes, or are they, a-la Minnesota this week, something that appears to be just wholesale adrenaline-based killings by police officers?

View on HN

I looked for (but did not find) the researcher's definition of "armed." E.g. is that any weapon, such as a knife, club, tire iron, etc? Or just a firearm?

One thing to understand is that police are trained to take defensive action against an aggressive individual who is within a certain distance. Within that circle, an unarmed individual can reach and overpower an officer before he can draw his weapon. So even if you are unarmed, if are acting aggressively and you approach an officer you are likely to get shot or at least tased/pepper sprayed. If you ignore an order to stop where you are and put your hands up, and you continue to approach you are likely to get shot.

Now, at least on the surface this does not appear to be exactly the situation in the recent Minnesota case. But we only know what's in the media, and the media likes to sensationalize and report half the story.

That said, if a cop has his weapon drawn on you, do not move. Do not twitch. Keep both hands in view. Do not do anything that you are not explicitly asked to do.

This might be good advice, but it's terrible that we live in a country where this is good advice.

In what country would it be good advice to move aggressively towards a police officer with his weapon drawn?

In many European countries officers don't draw guns in the first place. They're trained to de-escalate situations, not make them more dangerous.

I've seen this in action in Amsterdam and was amazed. Rather than charging into the situation guns drawn and yelling police in Amsterdam, seeing that the situation was not dire, first simply made their presence known. They slowly and calmly moved in talking to the perps the whole time. In the end, the situation was peaceably defused.

In America, we are so used to violent interaction with authority that most cannot imagine it working any other way.

One of the first steps we should adopt in the US is the mandatory changing of police uniform color. Green or light blue uniforms are less menacing than the typical US all black uniform. I believe this simple change would reduce police violence.

I was once stopped by a cop in San Francisco for accidentally running a red light on my bicycle. I don't know if he was intentionally intimidating, but I nearly shat my pants.

Police uniforms are navy, not black.

Have you tried pulling a knife on an armed officer in Europe? I've lived in Europe and I would not feel safe doing that.

I haven't and I wouldn't. I don't imagine it would be much fun from me. But I would most likely survive the encounter.

If you google "uk police subdue knife", you can find many examples. Here's one: http://www.huffingtonpost.com/entry/london-police-knife-wiel...

Quote from the article: "In San Francisco, the suspect died in a hail of bullets. In London, the suspect was subdued."

Different situations, different people, I'm sure they're not directly comparable. But in general people are not expected to die from police action in Europe unless they're actively shooting at police officers or hostages or something similar.

What about running away? I would think that in most cases, you should probably assume that if you are fleeing, you won't be shot in the back. But in reality, well..

> This might be good advice, but it's terrible that we live in a country where this is good advice.

Actually, it would be terrible to live in an environment where it was not good advice.

After all, its good advice if and only if you have some confidence that deadly force will not be used if you cooperate, otherwise, the advice to cooperate when a gun is drawn is bad advice. Otherwise, you should only superficially cooperate while actively looking for an opportunity to overpower or escape (neither of which has a good probability if the attacker has a gun already drawn, but if you don't expect cooperation to be effective protection, you've got no good options.)

Now, it might be terrible live in a place where it is not less likely than it is to be useful advice -- because where we live now, the circumstances in which people might need the advice come up more often than they should. But it would be good advice (with the caveat above) even if it were less frequently needed.

But what do you do when you have two cops with weapons drawn and one wants you to hold still and the other wants you to drop to the ground?

Hold your arms still, drop to your knees, and hope someone is filming.

From the abstract:

> There is no relationship between county-level racial bias in police shootings and crime rates (even race-specific crime rates), meaning that the racial bias observed in police shootings in this data set is not explainable as a response to local-level crime rates.

Do they break that down by type of crime? "Crime rate" is a pretty broad term.

I missed that sentence -- I'm not 100% sure, but I think that's not quite what I was talking about.

That statement says that areas with higher-than-normal black or white crime rates vis-a-vis the other, that there is still anti-black racial bias in terms of shootings.

I'm saying I'm wondering if people stopped during commission of a crime are 10x more likely to be white per capita, and 90% of police shootings are unarmed blacks and shootings happen in 10% of crime stops, and the population is 90/10 black and white (call it 10000 people, with 1% involved in crime stops and once per person), then we would have the following stats: 90 white people stopped, 10 black people, 10 shootings, 9 of them black.

What is the 'race bias' here? Do we normalize to population, or criminal population, or stopped population? Those stats look like: no bias: (9 / 9000 blacks shot, 1 / 1000 whites shot), incredible bias( 9/10 black criminals vs 1 / 90 white criminals), and we don't know because we don't know if there's any stop bias.

I'm not clear we get an answer out of the abstract or the study for this question. But, it's an important one to me for a variety of reasons, not least which is that I'd like to know if risky / criminal behavior statistically carries higher penalties for some race or socioeconomic groups in our country.

> As the parent of a black child, I have a personal interest...

As a human, I have a personal interest in this data.

I applaud that. For myself as a white man, I became much more personally interested in statistics like this when my family became multi-colored.

Reply on news.ycombinator.com