# Fairly simple math question (controversial subject) **URL:** \ **Category:** Factual Questions\ **Created:** [June 1, 2020, 5:58pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861 "2020-06-01T17:58:03Z")\ **Posts on this page:** 10\ **Page:** 1
**Author:** ![Zyada](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/zyada/32/3028_2.png) [@Zyada](https://boards.straightdope.com/u/Zyada)\ **Post date:** [June 1, 2020, 5:58pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/1 "2020-06-01T17:58:03Z")
So, I’m trying to figure out your chance of dying in a police encounter in the U.S. based on race. This website: [Our visualizations – Fatal Encounters](https://fatalencounters.org/our-visualizations/) shows that there have been 13,337 white people who have died in a police encounter, and 7617 black people. So somewhere around twice as many white people have died in police encounters as black people. However, black people only represent \<14% of the population: [https://www.census.gov/quickfacts/fact/table/US/POP010210](https://www.census.gov/quickfacts/fact/table/US/POP010210) whereas at least 60% of the population is white. So, just given these numbers, what is the actual likelihood of being killed by a police person if you are black vs. white? ---
**Author:** ![Snarky\_Kong](https://avatars.discourse-cdn.com/v4/letter/s/a183cd/32.png) [@Snarky\_Kong](https://boards.straightdope.com/u/Snarky_Kong)\ **Post date:** [June 1, 2020, 6:02pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/2 "2020-06-01T18:02:59Z")
You probably want: (# of X race killed by police in a year)/(# of X race that died in a year). ---
**Author:** ![Thudlow\_Boink](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/thudlow_boink/32/320_2.png) [@Thudlow\_Boink](https://boards.straightdope.com/u/Thudlow_Boink)\ **Post date:** [June 1, 2020, 6:24pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/3 "2020-06-01T18:24:48Z")
> [@Snarky\_Kong](#): > > You probably want: (# of X race killed by police in a year)/(# of X race that died in a year). I thought the OP was asking for (# of X race killed by police in a year)/(total # of X race living in the US). “What is your chance…” can be a very ill-defined kind of question. **Your** chance is going to depend on all sorts of factors. It might be clearer to ask “What percentage (or proportion, or fraction) of all X’s are Y’s?” If my calculations and the OP’s numbers are correct, approximately 0.0053% of white people are killed by police, and .0173% of black people are killed by police, which is about 3.26 times as high. ---
**Author:** ![Zyada](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/zyada/32/3028_2.png) [@Zyada](https://boards.straightdope.com/u/Zyada)\ **Post date:** [June 1, 2020, 6:45pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/4 "2020-06-01T18:45:06Z")
Thanks, **Thudlow**! ---
**Author:** ![Snarky\_Kong](https://avatars.discourse-cdn.com/v4/letter/s/a183cd/32.png) [@Snarky\_Kong](https://boards.straightdope.com/u/Snarky_Kong)\ **Post date:** [June 1, 2020, 7:10pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/5 "2020-06-01T19:10:21Z")
> [@Thudlow\_Boink](#): > > I thought the OP was asking for (# of X race killed by police in a year)/(total # of X race living in the US). > > “What is your chance…” can be a very ill-defined kind of question. **Your** chance is going to depend on all sorts of factors. It might be clearer to ask “What percentage (or proportion, or fraction) of all X’s are Y’s?” > > If my calculations and the OP’s numbers are correct, approximately 0.0053% of white people are killed by police, and .0173% of black people are killed by police, which is about 3.26 times as high. Your method doesn’t include people that may be killed by the police in the future. % of black people killed by police in a given year not % of black people killed by police. ---
**Author:** ![CookingWithGas](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/cookingwithgas/32/485_2.png) [@CookingWithGas](https://boards.straightdope.com/u/CookingWithGas)\ **Post date:** [June 2, 2020, 4:15pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/6 "2020-06-02T16:15:43Z")
> [@Zyada](#): > > So, I’m trying to figure out your chance of dying in a police encounter in the U.S. based on race. > […] > So, just given these numbers, what is the actual likelihood of being killed by a police person if you are black vs. white? This question is not well-formed. First I would like to know how you define “encounter.” But setting that aside, what do you mean: \*What are the chances that a white vs. non-white person living in the U.S. will die in a police encounter? \*This is simple to answer based on the numbers of each class of person living in the U.S. and the numbers of people who die in such encounters, such as in **Thudlow Boink’s** response. However, it may not answer the underlying social question you want to get to. \*If a person has a police encounter, what are the chances that a white vs. non-white person will die in that encounter?\*This is more complicated because you must know the number of police encounters by race to begin with. Police encounters includes more than just arrests, and are not even all documented. And the number of police encounters by race will probably show a bias against non-whites to begin with. And finally, police encounters are sometimes initiated by police and sometimes initiated by other people. An encounter that involves an armed robbery suspect shooting at police and being killed when the police return fire is a much different scenario that other recent events in the news. Using these statistics does not distinguish between those scenarios. So the math is quite easy but as is almost always the case in statistical analysis you have to ask the right question. ---
**Author:** ![puddleglum](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/puddleglum/32/137_2.png) [@puddleglum](https://boards.straightdope.com/u/puddleglum)\ **Post date:** [June 3, 2020, 4:55pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/7 "2020-06-03T16:55:28Z")
There is another site called [Mapping Police Violence](https://mappingpoliceviolence.org/) that has dramatically different numbers. There numbers are 39% fewer killings in 2019, than Fatal Encounters. The Washington Post’s Fatal Force site shows 44% fewer killings in 2019 than Fatal Encounters. ---
**Author:** ![Grey](https://avatars.discourse-cdn.com/v4/letter/g/b782af/32.png) [@Grey](https://boards.straightdope.com/u/Grey)\ **Post date:** [June 3, 2020, 6:30pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/8 "2020-06-03T18:30:46Z")
[What the data say about police shootings](https://www.nature.com/articles/d41586-019-02601-9) is an interesting view. ---
**Author:** ![Chronos](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/chronos/32/134_2.png) [@Chronos](https://boards.straightdope.com/u/Chronos)\ **Post date:** [June 3, 2020, 7:49pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/9 "2020-06-03T19:49:55Z")
How can there be fewer killings than fatal encounters? Do a significant number of fatal encounters result in multiple killings? ---
**Author:** ![Falchion](https://avatars.discourse-cdn.com/v4/letter/f/a88e57/32.png) [@Falchion](https://boards.straightdope.com/u/Falchion)\ **Post date:** [June 3, 2020, 8:26pm UTC](https://boards.straightdope.com/t/fairly-simple-math-question-controversial-subject/854861/10 "2020-06-03T20:26:04Z")
> [@puddleglum](#): > > There is another site called [Mapping Police Violence](https://mappingpoliceviolence.org/) that has dramatically different numbers. There numbers are 39% fewer killings in 2019, than Fatal Encounters. The Washington Post’s Fatal Force site shows 44% fewer killings in 2019 than Fatal Encounters. Looking at the dataset for Fatal Encounters (which you can download from the website) provides some hints, I think. For example, there are two entries for May 25, 2020 that state (names redacted): > [@](#): > > Police responded to a domestic violence call around 12.30am where they found a man who said he’d been in an argument with his wife. The man said his wife, [REDACTED], started shooting and chased him out of the house, barricading herself inside with her 9-year-old daughter and 12-year-old son. She released the boy from the home. Police entered the home at 6:30 a.m., and the woman barricaded herself in a closet with the girl. She shot two officers, who retreated. Officers entered again around 9 a.m. and found the woman and girl, [REDACTED], dead in the closet. The database classifies this as a suicide (Mom) and a murder (daughter). Less emotional (at least for me), also on May 25, 2020: > [@](#): > > At around 9 p.m., police in Houston were attempting to pull over [REDACTED], who allegedly was driving impaired. [REDACTED]took off down Veterans Memorial Drive, and police pursued. When [REDACTED] reached the 10000 block of Veterans Memorial Drive, he crashed into 63-year-old [REDACTED] who was riding a bicycle. Eventually, [REDACTED]'s vehicle stopped when he crashed head-on into a Buick driven by [REDACTED], 57. Both men were killed. It as the bicyclist and the Buick driver that were killed (not the suspect). I had trouble opening the dataset for Mapping Police Violence, but it’s methodology is described as counting: “A case where a person dies as a result of being shot, beaten, restrained, intentionally hit by a police vehicle, pepper sprayed, tasered, or otherwise harmed by police officers, whether on-duty or off-duty.” Neither of those events would seem to qualify, although I can’t confirm that they aren’t in the MPV database. I can’t tell what “Fatal Encounters” methodology is except that it describes itself as a database of “people killed during interactions with law enforcement”. Taking no position on whether these should “count”, it could explain the difference in the numbers.