• We should now be fully online following an overnight outage. Apologies for any inconvenience, we do not expect there to be any further issues.

Could a black person get away with this?

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

zinfamous

No Lifer
Jul 12, 2006
111,864
31,359
146
HumblePie thinks it's very very important to only look at incidents/interaciton with police, which is very different between blacks and whites, at an absurdly unbalanced rate controlled by demographics, and yet also thinks that it is irrelevant/indeterminable to ask why the rate of interactions is so vastly different.

it's interesting that he takes this stance....but it really isn't when Occam's Razor explains that Humble Pie is either just lazy, shit at statistics, or really only interested in specifically creating observable insignificant windows around data points in order to defend a pre-determined conclusion.
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
HumblePie thinks it's very very important to only look at incidents/interaciton with police, which is very different between blacks and whites, at an absurdly unbalanced rate controlled by demographics, and yet also thinks that it is irrelevant/indeterminable to ask why the rate of interactions is so vastly different.

it's interesting that he takes this stance....but it really isn't when Occam's Razor explains that Humble Pie is either just lazy, shit at statistics, or really only interested in specifically creating observable insignificant windows around data points in order to defend a pre-determined conclusion.

You can not have police killings without police interactions. Period. It is a much closer data stepping stone to finding out better data than per capita. The fact you don't understand this makes you literally mentally down.
 

ivwshane

Lifer
May 15, 2000
33,527
17,034
136
HumblePie thinks it's very very important to only look at incidents/interaciton with police, which is very different between blacks and whites, at an absurdly unbalanced rate controlled by demographics, and yet also thinks that it is irrelevant/indeterminable to ask why the rate of interactions is so vastly different.

it's interesting that he takes this stance....but it really isn't when Occam's Razor explains that Humble Pie is either just lazy, shit at statistics, or really only interested in specifically creating observable insignificant windows around data points in order to defend a pre-determined conclusion.


It’s all of the above.
 

ivwshane

Lifer
May 15, 2000
33,527
17,034
136
You can not have police killings without police interactions. Period. It is a much closer data stepping stone to finding out better data than per capita. The fact you don't understand this makes you literally mentally down.

And yet you some how think that posting the total number of incidents without any context is superior. Lol

Btw, the links I supplied, that you clearly didn’t read, also included context, so there goes that talking point.

Keep posting though, I enjoy watching you make a fool of yourself!
 
  • Like
Reactions: DarthKyrie

zinfamous

No Lifer
Jul 12, 2006
111,864
31,359
146
You can not have police killings without police interactions. Period. It is a much closer data stepping stone to finding out better data than per capita. The fact you don't understand this makes you literally mentally down.

yet again, you just stepped right over it. lol. It's sad that you keep trying, when those that do this kind of stuff for actual, you know, work, understand how terrible you are at it.
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
And yet you some how think that posting the total number of incidents without any context is superior. Lol

Btw, the links I supplied, that you clearly didn’t read, also included context, so there goes that talking point.

Keep posting though, I enjoy watching you make a fool of yourself!

It is superior than per capita. Per capita has ZERO context. It is just per capita. Per interaction has the context of per interaction. I am not stating you don't drill down further. I am saying that per capita is worthless, and the moment you start drilling down, the opposite picture gets painted very quickly.
 
Feb 4, 2009
35,862
17,403
136
It is superior than per capita. Per capita has ZERO context. It is just per capita. Per interaction has the context of per interaction. I am not stating you don't drill down further. I am saying that per capita is worthless, and the moment you start drilling down, the opposite picture gets painted very quickly.

How? Please explain ideally with data.
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
yet again, you just stepped right over it. lol. It's sad that you keep trying, when those that do this kind of stuff for actual, you know, work, understand how terrible you are at it.

LOL, you dont this shit for work, but I certainly do. If you actually think a per capita statistic is a better stat to use in this instance of measuring how often police kill black versus white people than per interaction then you are horrible at your job. Unless of course your job is in the media peddling lies and this shit.
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
How? Please explain ideally with data.

Okay lets break it down from the beginning. Start with definitions.


Per capita literally means per the average person. Per capita has a lot of value when applied correctly. If want to figure out the rate of a population groups average chance of an event occurring that has no outside vector that can control for it then it is a good starting point for population statistics. For example, let us say we want to know how likely a person will be struck by lightning. Since everyone has roughly equal chance of being on this planet when a storm will appear, it is a good baseline example. Since everyone in the population group has roughly the same base chance of being exposed you can do a per capita statistic. You get the count of how many people are struck by lightning in a year. Divide it by the population segment you want. Then you can see the value in that to a degree. Per capita only really has merit though when you can measure it for against another group with a different control value. So people in North America will have a different value than say people South America. You can go with even finer granularity.

But when you do a per capita based on an outside control vector it falls apart. Let's change it from struck by lighting to fall in a volcano. Chances of anyone not living near a volcano to fall in one is zero. Period. So while people in Hawaii may have a chance to fall into one, people in Kansas don't. Doing a per capita chance of people falling into a volcano for the entire population of the United States is just a dumb fucking statistic. It is meaningless. If you want to see how likely a person in Hawaii around a volcano may fall into a volcano versus another area with a volcano then that now becomes a statistic that has some value.

Now to go further with this thought process. Per capita deaths rates for say road vehicle crashes in America is a good number. The vast majority of the population rides in a road vehicle at some point during the year. Most of us do so every day, or least until the current virus crisis. So per capita comparisons of regions, demographics, vehicles used, and all sorts of other variables can be used.

Which leads me into the next point. Like volcanoes, deaths by police can only occur with police interactions. Police interactions with the AVERAGE person is rare. Many people can go their whole lives without interacting with a police officer. So a per capita comparison of police killings across the whole population segment is no longer a valid statistic as it isn't the average of that population. It becomes a useless, dumb statistic.

Like volcanoes, you have to do a per capita for the control variable. In this case it would be using the control variable of police interactions. Just like with volcanoes you would have to use the control variable of people actually around volcanoes. From that baseline point you can further refine your variables to bring out more valuable information. If you literally can't follow along with this post, there is no getting to you. This is as basic as it gets.
 

HomerJS

Lifer
Feb 6, 2002
39,380
32,883
136
LOL, you dont this shit for work, but I certainly do. If you actually think a per capita statistic is a better stat to use in this instance of measuring how often police kill black versus white people than per interaction then you are horrible at your job. Unless of course your job is in the media peddling lies and this shit.
When you are being profiled interaction numbers will be up.
 
  • Like
Reactions: DarthKyrie

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
When you are being profiled interaction numbers will be up.

Only for non violent crimes. Blacks disproportionately account for a large portion of violent crime responses by police to investigate. The vast majority of police involved fatalities occur when investigating a violent crime response. Over 90%. As I said, only about 9 to 10 unarmed black people have been typically killed each year by police.

You do have a point that profiling does increase police contacts. But increased police contacts that aren't for violent crimes don't increase police killings. They have no correlation. Increased violent criminal contacts do have a strong correlation with increased police killings. Despite that, blacks have the highest amount of police contacts for violent crimes than any other demographic in this country. Yet white people are still killed more often in that comparison.

Again, you are right that when racial profiling is used, it increases contacts with the police with that demographic. No one is denying that. We can argue the pros/cons over such policing methods, but suffice it to say it has no statistically significant bearing on police killings.
 
Feb 4, 2009
35,862
17,403
136
Okay lets break it down from the beginning. Start with definitions.


Per capita literally means per the average person. Per capita has a lot of value when applied correctly. If want to figure out the rate of a population groups average chance of an event occurring that has no outside vector that can control for it then it is a good starting point for population statistics. For example, let us say we want to know how likely a person will be struck by lightning. Since everyone has roughly equal chance of being on this planet when a storm will appear, it is a good baseline example. Since everyone in the population group has roughly the same base chance of being exposed you can do a per capita statistic. You get the count of how many people are struck by lightning in a year. Divide it by the population segment you want. Then you can see the value in that to a degree. Per capita only really has merit though when you can measure it for against another group with a different control value. So people in North America will have a different value than say people South America. You can go with even finer granularity.

But when you do a per capita based on an outside control vector it falls apart. Let's change it from struck by lighting to fall in a volcano. Chances of anyone not living near a volcano to fall in one is zero. Period. So while people in Hawaii may have a chance to fall into one, people in Kansas don't. Doing a per capita chance of people falling into a volcano for the entire population of the United States is just a dumb fucking statistic. It is meaningless. If you want to see how likely a person in Hawaii around a volcano may fall into a volcano versus another area with a volcano then that now becomes a statistic that has some value.

Now to go further with this thought process. Per capita deaths rates for say road vehicle crashes in America is a good number. The vast majority of the population rides in a road vehicle at some point during the year. Most of us do so every day, or least until the current virus crisis. So per capita comparisons of regions, demographics, vehicles used, and all sorts of other variables can be used.

Which leads me into the next point. Like volcanoes, deaths by police can only occur with police interactions. Police interactions with the AVERAGE person is rare. Many people can go their whole lives without interacting with a police officer. So a per capita comparison of police killings across the whole population segment is no longer a valid statistic as it isn't the average of that population. It becomes a useless, dumb statistic.

Like volcanoes, you have to do a per capita for the control variable. In this case it would be using the control variable of police interactions. Just like with volcanoes you would have to use the control variable of people actually around volcanoes. From that baseline point you can further refine your variables to bring out more valuable information. If you literally can't follow along with this post, there is no getting to you. This is as basic as it gets.

So exactly what does this mean in regard to Police shootings?
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
So exactly what does this mean in regard to Police shootings?

That when using the correct control variable with police interactions when doing a per capita comparison that white people are far more likely to be killed by police than black people. Even with removing the numbers for non-violent offenses, that some like HomerJS points out may be inflated due to racial profiling and he has a point there, it doesn't matter. Black people in this country are over represented when it comes to criminal cases of murder, rape, aggravated assault, and robbery. Going against any of those police interactions versus deaths per capita easily shows how more likely a white person is likely to be killed than a black person. That hasn't always been the case in our history. His has been a trend that started where white people were more than likely than black people since the 80's.

I already gave you my answer as to the strong correlation factors of why black people commit more violent offenses in the country previously.
 
  • Haha
Reactions: ch33zw1z

Sunburn74

Diamond Member
Oct 5, 2009
5,076
2,635
136
You do have a point that profiling does increase police contacts. But increased police contacts that aren't for violent crimes don't increase police killings. They have no correlation.
have you read the news at all in the last 5 years? Ever heard of Alton brown, Eric garner, philando castille, tamir rice? Or the white guy in phoenix recently who was shot and killed opening his own door and following police commands? The white Australian lady in Minnesota who called the cops to investigate a noise in a nearby alley and is shot and killed by them as she approaches their car as they arrive? Or the south Carolina black guy who is pulled over for not wearing a seat belt and is shot multiple times getting the drivers license he was asked to get whilst telling the cop sure let me get that license you want to see? Or the blacklady in Dallas who was playing video games and looked out her window only to be shot and killed by a police officer who was snooping around her back yard?

How can you say with a straight face that increased police contact for non violent calls doesnt increase police killings? They grossly increase police killings because if there's any sign of non compliance by the citizen police start shooting. This is why people are pissed and are protesting. If your neighbor calls the cops to check out your place, you shouldn't be killed opening the door.
 
Feb 4, 2009
35,862
17,403
136
That when using the correct control variable with police interactions when doing a per capita comparison that white people are far more likely to be killed by police than black people. Even with removing the numbers for non-violent offenses, that some like HomerJS points out may be inflated due to racial profiling and he has a point there, it doesn't matter. Black people in this country are over represented when it comes to criminal cases of murder, rape, aggravated assault, and robbery. Going against any of those police interactions versus deaths per capita easily shows how more likely a white person is likely to be killed than a black person. That hasn't always been the case in our history. His has been a trend that started where white people were more than likely than black people since the 80's.

I already gave you my answer as to the strong correlation factors of why black people commit more violent offenses in the country previously.

In regards to the thread who is more likely to be shot for wrestling with the Police

This guy:
AFFB92C7-320E-4DE4-BB98-0085D5DB9133.jpeg

or this guy

0B82571D-637B-4826-AF8C-66FDC7D61F2D.jpeg
 

zinfamous

No Lifer
Jul 12, 2006
111,864
31,359
146
Okay lets break it down from the beginning. Start with definitions.


Per capita literally means per the average person. Per capita has a lot of value when applied correctly. If want to figure out the rate of a population groups average chance of an event occurring that has no outside vector that can control for it then it is a good starting point for population statistics. For example, let us say we want to know how likely a person will be struck by lightning. Since everyone has roughly equal chance of being on this planet when a storm will appear, it is a good baseline example. Since everyone in the population group has roughly the same base chance of being exposed you can do a per capita statistic. You get the count of how many people are struck by lightning in a year. Divide it by the population segment you want. Then you can see the value in that to a degree. Per capita only really has merit though when you can measure it for against another group with a different control value. So people in North America will have a different value than say people South America. You can go with even finer granularity.

But when you do a per capita based on an outside control vector it falls apart. Let's change it from struck by lighting to fall in a volcano. Chances of anyone not living near a volcano to fall in one is zero. Period. So while people in Hawaii may have a chance to fall into one, people in Kansas don't. Doing a per capita chance of people falling into a volcano for the entire population of the United States is just a dumb fucking statistic. It is meaningless. If you want to see how likely a person in Hawaii around a volcano may fall into a volcano versus another area with a volcano then that now becomes a statistic that has some value.

Now to go further with this thought process. Per capita deaths rates for say road vehicle crashes in America is a good number. The vast majority of the population rides in a road vehicle at some point during the year. Most of us do so every day, or least until the current virus crisis. So per capita comparisons of regions, demographics, vehicles used, and all sorts of other variables can be used.

Which leads me into the next point. Like volcanoes, deaths by police can only occur with police interactions. Police interactions with the AVERAGE person is rare. Many people can go their whole lives without interacting with a police officer. So a per capita comparison of police killings across the whole population segment is no longer a valid statistic as it isn't the average of that population. It becomes a useless, dumb statistic.

Like volcanoes, you have to do a per capita for the control variable. In this case it would be using the control variable of police interactions. Just like with volcanoes you would have to use the control variable of people actually around volcanoes. From that baseline point you can further refine your variables to bring out more valuable information. If you literally can't follow along with this post, there is no getting to you. This is as basic as it gets.

Everyone understands this, you fucking idiot.

what you are walking over, again ignoring because it is very inconvenient for you, is why the interactions-with-cop between blacks and whites is so vastly different, considering that it doesn't track with what a per-capita measure would suggest.

it is inherently disproportionate, which is exactly why you are using it, and refusing to acknowledge that your data set is already biased against reasonable significance. The only justification you really have to ignore that, despite what we know of blacks and whites committing crimes at similar rates, as would be predicted per-capita, but prosecuted at vastly disproportionate rates, which rejects both per-capita and "Crime committed" rates, and again, this "police interaction rate" which, again, is inherently disproportionate, per capita, is that you must assume that there is a logical need--perhaps some law of inequality or some crap one has to make up--to justify the disproportionate encounters between police, across demographics.

yes, there is a disproportionate encounter rate, and that is exactly significant. You ignore that it has meaning, walk right past it, and continue onto your windowed set of data that is already inherently biased beyond any significance you can give to it. A responsible statistician would ask why blacks are disproportionately targeted, despite no significance commission of crime beyond their population rate that would support such a factor.

volcanoes aren't going out "searching for a demographic" that has no other reason to be near them more than the population that normally lives next to them. Using these thoroughly controlled variables in your argument exposes your pedestrian intellect for what we all know it is. If volcanoes and lightning storms had some method of seeking out demographics that don't normally live near them, beyond simple random chance that nature subscribes to them, then you could use this argument to defend why the data set you want to focus on is inherently biased, and maybe even ignore that bias as without meaning.

But you can't.

Now, please continue to make up some other bullshit, roll around in it, then declare yourself the king of shit hill, again.
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
Everyone understands this, you fucking idiot.

what you are walking over, again ignoring because it is very inconvenient for you, is why the interactions-with-cop between blacks and whites is so vastly different, considering that it doesn't track with what a per-capita measure would suggest.

it is inherently disproportionate, which is exactly why you are using it, and refusing to acknowledge that your data set is already biased against reasonable significance. The only justification you really have to ignore that, despite what we know of blacks and whites committing crimes at similar rates, as would be predicted per-capita, but prosecuted at vastly disproportionate rates, which rejects both per-capita and "Crime committed" rates, and again, this "police interaction rate" which, again, is inherently disproportionate, per capita, is that you must assume that there is a logical need--perhaps some law of inequality or some crap one has to make up--to justify the disproportionate encounters between police, across demographics.

yes, there is a disproportionate encounter rate, and that is exactly significant. You ignore that it has meaning, walk right past it, and continue onto your windowed set of data that is already inherently biased beyond any significance you can give to it. A responsible statistician would ask why blacks are disproportionately targeted, despite no significance commission of crime beyond their population rate that would support such a factor.

volcanoes aren't going out "searching for a demographic" that has no other reason to be near them more than the population that normally lives next to them. Using these thoroughly controlled variables in your argument exposes your pedestrian intellect for what we all know it is. If volcanoes and lightning storms had some method of seeking out demographics that don't normally live near them, beyond simple random chance that nature subscribes to them, then you could use this argument to defend why the data set you want to focus on is inherently biased, and maybe even ignore that bias as without meaning.

But you can't.

Now, please continue to make up some other bullshit, roll around in it, then declare yourself the king of shit hill, again.

Literally NO you dipshit. You are literally missing the point of my argument in the first place. The point of my ENTIRE GODDAMN ARGUMENT is that a general fucking per capita comparison across the entire population segment to prove that cops are more likely to kill black people is FLAT OUT FUCKING WRONG. That the moment you start to control for ANYTHING like encounter rates the fucking comparison flips and stays flipped. I don't give a SHIT how many control variables you are trying to control for at that point. It doesn't fucking matter.

I am not arguing why blacks commit more violent crimes you literal inbred fucking dipshit. I am arguing that it is absolutely stupidly asinine in-fucking-sane to use a general per capita comparison as an argument to prove racism in police involved killings. If you are trying to fucking strawman me any fucking more on this after I have made my self overtly clear on this point you are a fucking retarded buffoon.
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
have you read the news at all in the last 5 years? Ever heard of Alton brown, Eric garner, philando castille, tamir rice? Or the white guy in phoenix recently who was shot and killed opening his own door and following police commands? The white Australian lady in Minnesota who called the cops to investigate a noise in a nearby alley and is shot and killed by them as she approaches their car as they arrive? Or the south Carolina black guy who is pulled over for not wearing a seat belt and is shot multiple times getting the drivers license he was asked to get whilst telling the cop sure let me get that license you want to see? Or the blacklady in Dallas who was playing video games and looked out her window only to be shot and killed by a police officer who was snooping around her back yard?

How can you say with a straight face that increased police contact for non violent calls doesnt increase police killings? They grossly increase police killings because if there's any sign of non compliance by the citizen police start shooting. This is why people are pissed and are protesting. If your neighbor calls the cops to check out your place, you shouldn't be killed opening the door.

Wow anecdote stories..... don't prove any fucking thing. Those names you mentioned are numbers in the statistics I am referring to. Unless you are trying to prove a point that in the last 5 years the media has been over sensationalizing the bad deaths in police involved killings. If that is the point you are trying to prove then I can agree with that. The media certainly doesn't do it the same way with white people. That is the whole point of the argument I have with the OP. He is using an anecdote that doesn't at all match the statistics to sensationalism a narrative to make people like you believe the exact opposite of what actually happens.
 

JEDIYoda

Lifer
Jul 13, 2005
33,986
3,321
126
Zero sources have debunked anything I stated. What debunks that deaths per encounter rate is not the better statistic than deaths per capita? Nothing you or anyone else here has provided has shown that. Grow up.
at least your participation in the thread adds some comedic value.....haha
 

HumblePie

Lifer
Oct 30, 2000
14,665
440
126
In regards to the thread who is more likely to be shot for wrestling with the Police

This guy:


or this guy

Resorting straight to strawman are we? Don't know who those people are and don't care. Based on statistics, a white person is more likely to be shot for wrestling with a police officer than a black person. A white person is more likely to be the victim of police brutality than a black person. Especially when in comparison against opposite race. A white civilian is more than likely to be the victim of incorrect use of force by police by a black officer than the opposite. Of course a black officer is more than likely to use excessive force on a black civilian than a white one, as well as a white officer is more than likely to use excessive force on a white civilian than a black one.

Which leads to one of the underlying control vectors of the per capita based on encounter rates that zinfamous was trying to allude to. There are far more white cops than black cops. White cops are far more likely to kill white people, and black cops are far more likely to kill black people. That is a bit of the correlation of why white people are more likely to be killed than black people by police. But it is a much weaker correlation than violent crime encounters or when the person is armed while encountering the police. Also geographics plays a large correlation role as well.
 

JEDIYoda

Lifer
Jul 13, 2005
33,986
3,321
126
Literally NO you dipshit. You are literally missing the point of my argument in the first place. The point of my ENTIRE GODDAMN ARGUMENT is that a general fucking per capita comparison across the entire population segment to prove that cops are more likely to kill black people is FLAT OUT FUCKING WRONG. That the moment you start to control for ANYTHING like encounter rates the fucking comparison flips and stays flipped. I don't give a SHIT how many control variables you are trying to control for at that point. It doesn't fucking matter.

I am not arguing why blacks commit more violent crimes you literal inbred fucking dipshit. I am arguing that it is absolutely stupidly asinine in-fucking-sane to use a general per capita comparison as an argument to prove racism in police involved killings. If you are trying to fucking strawman me any fucking more on this after I have made my self overtly clear on this point you are a fucking retarded buffoon.
But you are spreading your uniformed manure really thick!!
 
  • Like
Reactions: DarthKyrie

cytg111

Lifer
Mar 17, 2008
26,221
15,629
136
Literally NO you dipshit. You are literally missing the point of my argument in the first place. The point of my ENTIRE GODDAMN ARGUMENT is that a general fucking per capita comparison across the entire population segment to prove that cops are more likely to kill black people is FLAT OUT FUCKING WRONG. That the moment you start to control for ANYTHING like encounter rates the fucking comparison flips and stays flipped. I don't give a SHIT how many control variables you are trying to control for at that point. It doesn't fucking matter.

I am not arguing why blacks commit more violent crimes you literal inbred fucking dipshit. I am arguing that it is absolutely stupidly asinine in-fucking-sane to use a general per capita comparison as an argument to prove racism in police involved killings. If you are trying to fucking strawman me any fucking more on this after I have made my self overtly clear on this point you are a fucking retarded buffoon.

You are your own worst counter argument.

"That the moment you start to control for ANYTHING like encounter rates"
- I would hope so? That there was a build-up to this black person being shot, the alternative would be straight up assassination? Maybe cops would do a drive by?

In all your glorious cognitive might, you have deduced that there is a buildup to a person losing his/her life at the hands of police and that that is proof of something... Something.
Would you like a participation trophy? Bless you.