# Profiling based off of high per capita incidence of violent crime (gender, race and religion) **URL:** \ **Category:** Great Debates\ **Created:** [September 29, 2016, 3:35pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209 "2016-09-29T15:35:08Z")\ **Posts on this page:** 14\ **Page:** 2
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [September 30, 2016, 4:06pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/21 "2016-09-30T16:06:55Z")
All of which comes back to Bayes’ theorem. If the prior is high enough you can get good results keying off it. The prior will only get high enough when you have enough parameters and they \*aren’t \*strongly correlated. In the example above, it’s an error to think that [Arab plus Muslim] gives much better results than just [Arab] or just [Muslim] when (WAG) 98% of Arabs are Muslim and (WAG) 50% of Muslims are Arab. Adding the extra factor \*looks \*like it’s improving specificity. It may well be “improving” the mistaken impression of specificity a lot while improving actual mathematical specificity almost zero. ---
**Author:** ![Hector\_St\_Clare](https://avatars.discourse-cdn.com/v4/letter/h/b5ac83/32.png) [@Hector\_St\_Clare](https://boards.straightdope.com/u/Hector_St_Clare)\ **Post date:** [September 30, 2016, 4:23pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/22 "2016-09-30T16:23:20Z")
> [@Kimstu](#): > > There are two important issues here: > > 1. What is the actual prevalence of offenders in the given population? > > 2. What rights are being infringed or curtailed by such profiling in the interests of security? > For issue (1), [it appears that the incidence of committing sexual assault among men in general is quite a bit higher](http://fivethirtyeight.com/features/what-if-most-campus-rapes-arent-committed-by-serial-rapists/)—as in, orders of magnitude higher—than for any of the other crimes/terrorist acts in the populations we’re discussing: > > Emphasis added. If there was similar evidence indicating that over one in ten or even one in twenty Muslim-Americans was involved in terrorism, you can bet your boots that most of the people who now oppose ethnic profiling in security screenings would take a very different view of it. > > But of course, that incidence is more like one in tens or hundreds of thousands at the most, not anywhere near one in ten. Which is what makes ethnic profiling pretty much useless as an anti-terrorism security measure. > > For issue (2), as **doreen** pointed out, it’s a blatantly false equivalence to present _women being cautious around men in social circumstances_ as analogous to _security agents ethnically profiling people_. > > If a woman you’re trying to pick up in a bar won’t smile at you or talk to you because she’s concerned you might be a rapist, **your rights are not being infringed in any way**. If a Muslim-American is being ethnically profiled for security reasons, on the other hand, that at the very least puts significant burdens on their freedom of travel, not to mention the risks of being put on a watchlist or otherwise flagged as a potentially untrustworthy person. > > Now, when it comes to men’s exercise of their actual rights, I definitely agree they should **not** be subjected to restrictive gender profiling. No fair, e.g., excluding men from elevators or subway cars or medical consultations or anywhere else that they might happen to be alone with a woman who might fear for her safety. And of course, nobody here except **ZPG Zealot** , whose views on the subject are very extreme, is in favor of any such thing. > **To sum up:** Women “gender-profiling” men for safety reasons in the social behavior of their private personal lives, on the grounds that the prevalence of sexual assaulters among men is alarmingly high, is not unfair to men. However, “gender-profiling” men so as to impede their expected freedoms and actions as co-workers, passengers, parents, customers, etc., **would** be unfair to men and should not be practiced. > > _A fortiori_, ethnically profiling certain populations for security reasons so as to impede their freedoms and actions, on the grounds that the prevalence of terrorists or other criminals among them is slightly elevated compared to the general population but still fucking **tiny** , is extremely unfair to those populations and should not be practiced. What about for crimes where quite a large amount of the ethnic group in question commits the crime? As an example, at post #42 here, **Blake** makes the argument that _the majority of children_ in certain Aboriginal communities in northern Australia were victims of abuse by their family, and therefore “simply being Aboriginal was deemed to be sufficient grounds for declaring a child at risk.” If he’s right about the numbers (and I have no reason to doubt him) then that seems like an example of a situation where ethnic profiling would be legitimate. > **[What do you know about Australia and it's Aborigines? (+ opinions on what you...](https://boards.straightdope.com/sdmb/showthread.php?t=555597)** > > For the non-Aussies: What do you know about the relationship between Aboriginal Australians and non-Aboriginal Australians? What do you know about our attitudes towards each other? How our cultures mix? How intergrated we are in every day life? ... ---
**Author:** ![AK84](https://avatars.discourse-cdn.com/v4/letter/a/85e7bf/32.png) [@AK84](https://boards.straightdope.com/u/AK84)\ **Post date:** [September 30, 2016, 4:33pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/23 "2016-09-30T16:33:15Z")
> [@LSLGuy](#): > > In the example above, it’s an error to think that [Arab plus Muslim] gives much better results than just [Arab] or just [Muslim] when (WAG) 98% of Arabs are Muslim and (WAG) 50% of Muslims are Arab. The actual numbers are sometjing like 10-15% Arabs are non-muslim and something like 15% muslims are Arabs. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [September 30, 2016, 4:43pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/24 "2016-09-30T16:43:12Z")
Thank you. ---
**Author:** ![Little\_Nemo](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/little_nemo/32/3120_2.png) [@Little\_Nemo](https://boards.straightdope.com/u/Little_Nemo)\ **Post date:** [September 30, 2016, 4:44pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/25 "2016-09-30T16:44:24Z")
> [@LSLGuy](#): > > The problem with this whole line of thought is widespread ignorance of Bayesian probabilities. > > This is too deep and rigorous an intro to the idea, but it’s a start: [Bayes' theorem - Wikipedia](https://en.wikipedia.org/wiki/Bayes%27_theorem) > > The probability that any particular person is a bad actor is the probability of bad actors in general times the probability that this person is part of the known group. > > e.g. If 1 in 1,000 men is a raper-of-women and 100% of rapers-of-women are men, the odds that any given woman should be scared of any given man are not 100%, but 0.1% The amount she should be scared of another woman is zero percent. So for this fact pattern the sound choice is 0% scared for other women and 0.1% scared for other men. Not 0% and 100%. > > Folks forget (or never knew) these basic facts. They don’t know how to validly reason about statistical risks. > > To the degree we permit public policy to be driven by irrational bunk mislabeled as “rational thinking”, we’re no better than the folks in Salem who believed dunking tortures correctly identified witches. > > IOW … 100% of the 19 hijackers on 9/11 were Saudi. That same day another 31 million Saudis did not hijack any airliners. 19/31,000,000 is a very small number. Even if we filter the Saudi population to just young and male we’re looking at 19/10,000,000. > > That’s called looking for a needle in a haystack. Trying to argue that we ought to treat all 10 million young male Saudis as high threats is logical garbage. But it does sell well to emotional non-rational people. Exactly. A lot of people don’t understand the difference between “the majority of rapists are men” and “the majority of men are rapists”. When they see evidence to support one of those statements, they treat it as if it proves both statements. ---
**Author:** ![Richard\_Parker](https://avatars.discourse-cdn.com/v4/letter/r/35a633/32.png) [@Richard\_Parker](https://boards.straightdope.com/u/Richard_Parker)\ **Post date:** [September 30, 2016, 4:47pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/26 "2016-09-30T16:47:02Z")
Don’t forget that for most crimes we only know the demographics of the perpetrators as a result of where we deploy law enforcement resources. IOW, if you profile black drug dealers, you catch more black drug dealers, justifying black profiling. If we put the full weight of the FBI on catching white nationalist terrorists (who are roughly as common as Muslim terrorists in the US), we would see a lot more foiled plots from them. Conversely, of course, the profiling might suppress the true rate. ---
**Author:** ![Damuri\_Ajashi](https://avatars.discourse-cdn.com/v4/letter/d/898d66/32.png) [@Damuri\_Ajashi](https://boards.straightdope.com/u/Damuri_Ajashi)\ **Post date:** [September 30, 2016, 4:57pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/27 "2016-09-30T16:57:04Z")
> [@LSLGuy](#): > > The problem with this whole line of thought is widespread ignorance of Bayesian probabilities. > > This is too deep and rigorous an intro to the idea, but it’s a start: [Bayes' theorem - Wikipedia](https://en.wikipedia.org/wiki/Bayes%27_theorem) > > The probability that any particular person is a bad actor is the probability of bad actors in general times the probability that this person is part of the known group. > > e.g. If 1 in 1,000 men is a raper-of-women and 100% of rapers-of-women are men, the odds that any given woman should be scared of any given man are not 100%, but 0.1% The amount she should be scared of another woman is zero percent. So for this fact pattern the sound choice is 0% scared for other women and 0.1% scared for other men. Not 0% and 100%. No matter how remote a risk, if the risk is large enough, it is not irrational to be risk averse. For example, I have a chance to invest 1% of my money in an investment that has a 10% chance of returning 20 times my money and a 90% chance of becoming worthless. It might make sense for me to take that risk because I can absorb the loss. I also have the opportunity to invest all (and no less than all) of my money in an investment that has a 90% chance of tripling my money but a 10% chance of becoming worthless. Even though the second investment has a higher expected return, a rational person could make the first investment while rejecting the second one. How well are you able to absorb the risk of being raped? ---
**Author:** ![monstro](https://avatars.discourse-cdn.com/v4/letter/m/ba9def/32.png) [@monstro](https://boards.straightdope.com/u/monstro)\ **Post date:** [September 30, 2016, 5:09pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/28 "2016-09-30T17:09:58Z")
> [@Velocity](#): > > I mean mainly individual behavior and inner thoughts, although other aspects of profiling can be discussed as well. > > President Obama once commented about how, as a black man, he used to encounter women clutching their purses tightly and holding their breath nervously when with him on an elevator, about how people would lock their car doors when he approached, etc. He was praised for these comments, and rightfully so - black men are often perceived as threats. > > But I wonder if, the word “black” was omitted and someone simply commented about how women take precautions around men and lock car doors when men approach and act uneasy and nervous around men, that such remarks would be rebutted with, “Yes, women \*do \*face a threat from men and just doing what they have to do.” > > I am a minority man living in a very liberal city, and I find the inconsistent logic jarring. It’s like society is saying that it’s wrong to consider me a potential threat because of my race, but it’s OK to consider me a potential threat because of my gender. I take precautions against strange men for two reasons: One, they are bigger and stronger than me, on average. Two, they are more likely to show _physical_ aggression. These two things are generalizations, of course. There are men who are smaller and weaker than me, and there are men who are more timid and passive than I am. But on the whole, I’m going to be at a big disadvantage when it comes to going toe-to-toe against a man who is up to no good. So of course, I’m going to be more afraid if I’m walking down a dark alley and a dude steps into my path versus a dudette. Maybe I’m dead either way in this day and age of guns, but at least with the average dudette, I’ve got a fighting chance. A few years ago there was a thread about whether it’s wrong for women to prefer sitting next to women passengers when riding the bus. I don’t think it is. Chances are if I sit next to a random guy, nothing is going to happen. But the cost of discriminatory seat selection is considerably less than being plopping down next to any Tom, Dick, and Harry and regretting it later. To be conscientious in this context is simply to appreciate the importance of sexuality in human behavior. Sitting next to a man (when you are a woman) may be perceived–rightly or wrongly–as being friendly towards certain advances. If you aren’t interested in receiving advances, avoiding male-adjacent seats is totally acceptable behavior that doesn’t harm anyone. If men don’t like this practice, then they should stop hitting on women in public. Scolding women for being discriminatory ignores the fact that they are simply responding to men’s discriminatory behavior. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [September 30, 2016, 5:27pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/29 "2016-09-30T17:27:05Z")
> [@Damuri\_Ajashi](#): > > No matter how remote a risk, if the risk is large enough, it is not irrational to be risk averse. > … > How well are you able to absorb the risk of being raped? You raise a valid point. The severity of the outcome properly factors into the decision on how risk-averse to be. But choosing to be very risk averse over something that is not in fact indicative of risk, but just incorrectly \*looks \*like it’s indicative of risk, merely magnifies the severity of your error. If you believe that a shadow touching you would be fatal, are you really better off running screaming from the room when you see one versus just having a cold sweat? You’re right that running screaming from the room is an appropriate level of response to a life-threatening situation. But you’ve goofed by mistakenly believing that seeing a shadow is life-threatening. Most profiling behavior by people and by bureaucracies falls into the [being afraid of shadows] realm. ---
**Author:** ![Kimstu](https://avatars.discourse-cdn.com/v4/letter/k/ecd19e/32.png) [@Kimstu](https://boards.straightdope.com/u/Kimstu)\ **Post date:** [September 30, 2016, 6:13pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/30 "2016-09-30T18:13:58Z")
> [@Velocity](#): > > Yes, but again, when Obama used to encounter women clutching their purses tightly and holding their breath nervously when alone with him on an elevator, or people would lock their car doors when he approached, Obama’s rights were not being infringed in any way - but was it not racist and offensive to him? I think you may have missed my distinction between “profiling” in _social interactions of one’s private personal life_, especially in response to a comparatively **large** risk probability, and “profiling” in the _ordinary exercise of one’s “expected freedoms and actions as co-workers, passengers, parents, customers, etc.”_, especially in response to a comparatively **small** risk probability. So yes, it is indeed unfair to “profile” non-white men _minding their own business in public places_ by avoiding them as probable criminals. But that’s not analogous to a woman ignoring a man’s _unsolicited social advances_ because she’s “gender-profiling” him as a possible rapist. To apply this to your Obama example: Yes, it was racist and offensive for white women to react fearfully to Obama minding his own business on a public street or in a public elevator, and I disapprove of such behavior. If, on the other hand, Obama approached a strange woman in a public bar and said “Excuse me ma’am, may I buy you a drink?” and she shut him down out of fear of a potential risk to her safety, her behavior would **not** be offensive and I don’t see any reason to disapprove of it. ---
**Author:** ![Kendall69](https://avatars.discourse-cdn.com/v4/letter/k/51bf81/32.png) [@Kendall69](https://boards.straightdope.com/u/Kendall69)\ **Post date:** [October 1, 2016, 4:35am UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/31 "2016-10-01T04:35:06Z")
> [@LSLGuy](#): > > The problem with this whole line of thought is widespread ignorance of Bayesian probabilities. > > This is too deep and rigorous an intro to the idea, but it’s a start: [Bayes' theorem - Wikipedia](https://en.wikipedia.org/wiki/Bayes%27_theorem) > > The probability that any particular person is a bad actor is the probability of bad actors in general times the probability that this person is part of the known group. > > e.g. If 1 in 1,000 men is a raper-of-women and 100% of rapers-of-women are men, the odds that any given woman should be scared of any given man are not 100%, but 0.1% The amount she should be scared of another woman is zero percent. So for this fact pattern the sound choice is 0% scared for other women and 0.1% scared for other men. Not 0% and 100%. > > Folks forget (or never knew) these basic facts. They don’t know how to validly reason about statistical risks. > > To the degree we permit public policy to be driven by irrational bunk mislabeled as “rational thinking”, we’re no better than the folks in Salem who believed dunking tortures correctly identified witches. > > IOW … 100% of the 19 hijackers on 9/11 were Saudi. That same day another 31 million Saudis did not hijack any airliners. 19/31,000,000 is a very small number. Even if we filter the Saudi population to just young and male we’re looking at 19/10,000,000. > > That’s called looking for a needle in a haystack. Trying to argue that we ought to treat all 10 million young male Saudis as high threats is logical garbage. But it does sell well to emotional non-rational people. Bad math on your part. When you look at more realistic numbers and add in common sense like how many Saudi are baseball game wearing a heavy coat and carrying a back pack then you narrow you hay stack down to a bucket full of hay, and the needle actually pricks you in the hand as you look for it. If you use that same common sense, at all large events, malls, theme parks, you will stop and they have stopped a lot of attacks you and I will never know about. Facts are profiling works, and the proof is in Israel, where the have more attacks than any place on the planet and stop a great number by profiling. ---
**Author:** ![octopus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/octopus/32/3716_2.png) [@octopus](https://boards.straightdope.com/u/octopus)\ **Post date:** [October 1, 2016, 7:20am UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/32 "2016-10-01T07:20:15Z")
> [@AK84](#): > > There is a big difference between understanding why someone acts some way and being upset by it. > > I can intellectually understand why a women might be afraid or at least wary of me as an unknown male. I do not have to like being thought of as a potential violent criminal. > > ETA: Ghetto in drag? Yeah, very very mature. I bet the woman doesn’t like being afraid. I’d rather her feel safe than worry about offending my delicate sensibilities by being careful. I have a daughter and I don’t care about her offending people if she acts in her self interest. Are ghetto and drag now trigger words? The point is, which you are trying to obfuscate, if you don’t profile in some cases it can be fatal. Even some state course recognize that. [Mass. High Court Says Black Men May Have Legitimate Reason To Flee Police | WBUR News](http://www.wbur.org/news/2016/09/20/mass-high-court-black-men-may-have-legitimate-reason-to-flee-police) ---
**Author:** ![octopus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/octopus/32/3716_2.png) [@octopus](https://boards.straightdope.com/u/octopus)\ **Post date:** [October 1, 2016, 7:26am UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/33 "2016-10-01T07:26:07Z")
> [@LSLGuy](#): > > You raise a valid point. The severity of the outcome properly factors into the decision on how risk-averse to be. > > But choosing to be very risk averse over something that is not in fact indicative of risk, but just incorrectly \*looks \*like it’s indicative of risk, merely magnifies the severity of your error. > > If you believe that a shadow touching you would be fatal, are you really better off running screaming from the room when you see one versus just having a cold sweat? You’re right that running screaming from the room is an appropriate level of response to a life-threatening situation. But you’ve goofed by mistakenly believing that seeing a shadow is life-threatening. > > Most profiling behavior by people and by bureaucracies falls into the [being afraid of shadows] realm. Well the numbers of people raped, robbed, or murdered by shadows is probably less than the number of toes I have. People on the other hand have and do act monstrously. Your house have locks? How about your car? ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [October 1, 2016, 2:55pm UTC](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209/34 "2016-10-01T14:55:00Z")
> [@octopus](#): > > Well the numbers of people raped, robbed, or murdered by shadows is probably less than the number of toes I have. People on the other hand have and do act monstrously. Your house have locks? How about your car? And as I have said all along, it’s perfectly reasonable, rational, and logically correct to make preparations and be wary, cautious, fearful, or whatever word you prefer. _In proportion to the actual incidence and severity of threat._ What I have been arguing all along is that most people have a vastly overblown and fact-free opinion about the incidence and severity of the threat. As such their responses are vastly overblown also. e.g. Yes, I have door locks. And I use them when I’m asleep or away. Some neighbors in my condo building have double doors with 2 separate locks each, sleep with a gun under their pillow and tell me they wake in fear every night in reaction to violent noises they hear. I hear no such noises although I do often wake to the sound of thunder. One of us is vastly misinformed about the actual incidence and severity of crime in our neighborhood and in our building. That individual is certainly free to react to their delusions. But I certainly don’t want to be policed by a police force operating under similarly deluded fantasies. Nor governed by a legislature in thrall to those fantasies. Get actual numbers. Make an actual risk calculation. And be guided accordingly. That is my whole and entire point. [Previous page](https://boards.straightdope.com/t/profiling-based-off-of-high-per-capita-incidence-of-violent-crime-gender-race-and-religion/767209.md?page=1)