Wait a second, why is the chance of getting shot and killed by the police a per capita statistic.

The denominator should be the number of police-initiated encounters. I could only find 2018 data, but it's likely in the same ballpark. [1]

Thus, the chance of getting shot and killed by police (2020):

993 in 28,880,900 = 3.44 x 10^-5

For African Americans (2020):

241 in 3,393,800 = 7.10 x 10^-5

So it's an order of magnitude more likely that you'll be shot and killed by the police than getting hit by lightning. It's in the same category as being killed by cataclysmic weather in the US.

As an aside, I also believe that the way you've argued your point here speaks to a lack of contextualization. Murder is more abhorrent than many other causes of death simply because things like lightning strikes or car accidents are done without prejudice or intent. Murdering someone with a gun requires intent at every step. In some cases, it's preventable, and action should be taken to ensure that mistakes do not happen.

[1] https://bjs.ojp.gov/content/pub/pdf/cbpp18st.pdf

Or even better, the number of citizen-initiated encounters, since the formulation is often "I wouldn't call the police for help because I don't want to end up getting shot by them when they show up and misread the situation."

"Wait a second, why is the chance of getting shot and killed by the police a per capita statistic."

To normalize data collected from jurisdictions with different populations.

"The denominator should be the number of police-initiated encounters."

It depends what you care about. If you're a black man in America, perhaps you care about the overall chance of dying this way.

Perhaps, conditional on having been stopped by police, the chance of being shot and killed is the same for white men and black men. But this would be irrelevant if black men are 10X more likely to be stopped in the first place.

Overall, I agree with what you've written, though on the disparity in shooting incidents regarding race Roland Fryer's paper[1] came to different conclusions (it's different data but a more extensive data set and approach):

> In stark contrast to non-lethal uses of force, we find that, conditional on a police interaction, there are no racial differences in officer-involved shootings on either the extensive or intensive margins. Using data from Houston, Texas – where we have both officer-involved shootings and a randomly chosen set of potential interactions with police where lethal force may have been justified – we find, after controlling for suspect demographics, officer demographics, encounter characteristics, suspect weapon and year fixed effects, that blacks are 27.4 percent less likely to be shot at by police relative to non-black, non-Hispanics. This coefficient is measured with considerable error and not statistically significant. This result is remarkably robust across alternative empirical specifications and subsets of the data. Partitioning the data in myriad ways, we find no evidence of racial discrimination in officer-involved shootings. Investigating the intensive margin – the timing of shootings or how many bullets were discharged in the endeavor – there are no detectable racial differences.

If looking at any kind of force though:

> Blacks are almost eighteen percent more likely to incur any use of force in an interaction, accounting for all variables we can in the data. Hispanics are roughly twelve percent more likely. Both are statistically significant. Asians are slightly less likely, though not distinguishable from whites.

He does give caveats and goes into an extensive discussion on the limitations of the data available (one being the lack of data, which I would say should be the first order of the day for anyone looking for solutions to this) and possible interpretations. As an aside, I'm now unsurprised when I encounter progressives (an irony that might be amusing if they weren't so insidious and invidious) who've not heard of Fryer, nor how quickly they move to dismiss his work or even smear him without ever bothering to check his work in the slightest.

I didn't find anything about lightning, sadly (that would've been a welcome funny moment in amongst these awful stats), although earlier in the pandemic I checked the chances of being hit by lightning (for use as a comparison to some other stats) and was surprised to find out how high they are. Just last month I read this article[2] on the BBC site and decided never to go out in a thunderstorm ever again. If you read to the end it gets worse:

> In 2016, Bangladesh declared lightning strikes a natural disaster when more than 200 people died in the month of May alone, including 82 people on a single day.

Shocking, literally and figuratively.

[1] https://www.nber.org/papers/w22399

[2] https://www.bbc.com/news/world-asia-58094680

Both those denominators are waaaay off. Lightning doesn’t have the equal probability to strike everywhere and anywhere and I imagine there are relatively few people in the path of a lightning strike (grounded or near a ground) when there is a high probability for lightning… so if you’re going to use people as a denominator, it’s probably closer to 1 million.

Also, very few people stick around when the cops are coming after someone and the denominator should be people the cops are going after (arrests + stops + deaths from cops). In general, cops don’t just randomly start shooting people in their line of site…

> And yet, you are about as likely to get hit and killed by lightning as you are to die from being shot by police.

And?

You’re also more likely to die of a lightning strike than you are in a school shooting [0]. Does that mean we should judge the horror of school shootings any differently?

- [0] https://www.statista.com/statistics/971473/number-k-12-schoo...

You're asking if we should judge the horror more differently?

Certainly. One (school shootings) is obviously far more horrible, for obvious reasons.

But, if you're commenting on how common something is (as the original poster did), then having an actual understanding of the frequency is important. For instance, I don't think we need armed police and metal detectors on every single campus. That would be an over reaction but one that you might make if you thought the problem was bigger and more widespread than it actually is.

I am somewhat surprised that I am having to explain why keeping perspective is important on HN. Sign of the times, I guess.

> then having an actual understanding of the frequency is important

In which case it’s important to add some more context to the numbers you cited.

Lightning strikes kill about 50 people/year. Police kill about 1,000 (US).

You’re significantly more likely do die at the hands of police than you are from lightning.

Does that make either occurrence particularly likely? That’s where things get messy. Lightning doesn’t discriminate. Cops do. A person of color has no disadvantage against lightning, but they do when police are involved.

Bottom line: this comparison is problematic and misleading. This was the point of my first response.

Understanding that on the whole it’s extremely unlikely to be struck by lightning doesn’t mean someone should go stand in a field during a lightning storm holding an umbrella.

If context is important then how many people go out in thunderstorms and how many of those shootings were justified?

> I am somewhat surprised that I am having to explain why keeping perspective is important on HN. Sign of the times, I guess.

Even on HN, people have an allergic reaction to even a suggestion that a report arguing for the conclusion they already believe in could be based on flawed data or flawed methodology.

(A sure-fire way to have this conversation on an arbitrary topic is to point out base rate fallacy - so many reports and arguments omit the base rate, either accidentally or on purpose, which often leads to making a sensation out of a non-story.)

Still, bad reasoning has to be pointed out, or else we fall into darkness. The big problems ahead of us are of the kind where intuition alone won't help us. We've had thousands of years honing our intuitions and refining them into saws, cultural norms, religious beliefs, etc. The problems we find hard today are precisely the ones where you need rational, correct reasoning.

> You are about as likely to get hit and killed by lightning as you are to die from being shot by me.

Chance of getting hit by lightning: 1 in 500,00 = 2.0 * 10^-6 [1]

Population of earth ~7674mil, so I can pop out have a little bit of a rampage, settle some scores and kill ~15,000 and you wouldnt consider that I had 'done much wrong'.