Software Used To Predict Who Might Kill

eldavojohn writes "Richard Berk, a University of Pennsylvania criminologist, has worked with authorities to develop a software tool that predicts who will commit homicide. I could not find any papers published on this topic by Berk, nor any site stating what specific Bayesian / decision tree algorithm / neural net is being implemented." From the article: "The tool works by plugging 30 to 40 variables into a computerized checklist, which in turn produces a score associated with future lethality. 'You can imagine the indicators that might incline someone toward violence: youth; having committed a serious crime at an early age; being a man rather than a woman, and so on. Each, by itself, probably isn't going to make a person pull the trigger. But put them all together and you've got a perfect storm of forces for violence,' Berk said. Asked which, if any, indicators stood out as reliable predicators of homicide, Berk pointed to one in particular: youthful exposure to violence." The software is to enter clinical trials next spring in the Philadelphia probation department. Its intent is to serve as a kind of triage: to let probation caseworkers concentrate most of their effort on the former offenders most likely to be most dangerous.

popcorn

Hey I've seen that movie! Tom Cruise survives and gets to have the cute girl!

I believe the answer was...

I could not find any papers published on this topic by Berk, nor any site stating what specific Bayesian / decision tree algorithm / neural net is being implemented.

Yeah, I think if you ask for it to answer that question, the algorithm responds "I'm sorry dave, I'm afraid I can't do that."

Reference

This paper was published in the June 2006 issue of "The Journal of Quantitative Criminology".
Here are the pertinent details:
Title: Forecasting Dangerous Inmate Misconduct: An Application of Ensemble Statistical Procedures
Journal: Journal of Quantitative Criminology
Issue: Volume 22, Number 2 / June, 2006
Pages: 131-145

Abstract:
In this paper, we attempt to forecast which prison inmates are likely to engage in very serious misconduct while incarcerated. Such misconduct would usually be a major felony if committed outside of prison: drug trafficking, assault, rape, attempted murder and other crimes. The binary response variable is problematic because it is highly unbalanced. Using data from nearly 10,000 inmates held in facilities operated by the California Department of Corrections, we show that several popular classification procedures do no better than the marginal distribution unless the data are weighted in a fashion that compensates for the lack of balance. Then, random forests performs reasonably well, and better than CART or logistic regression. Although less than 3% of the inmates studied over 24 months were reported for very serious misconduct, we are able to correctly forecast such behavior about half the time.

Unfortunately, you've got to pay $30 to get this paper. Maybe some slashdotter with a school/corp subscription to Springer will put up the text? ;-)

Re:Reference

http://download.yousendit.com/CFD415EB623F072C [yousendit.com]

Re:Reference

Thanks! However, reading the paper, it seems that this paper is about the California Department of Corrections, and is not actually about who will commit homicide "on the outside." It's about which prisoners are "likely to engage in serious misconduct while incarcerated" (from the abstract). I don't know if this is the right paper. In fact, I'm going to guess that Berk hasn't published a paper on his new method. This paper may be a similar method, but there's no way to know that.

I also wonder in yousendit.com can handle a slashdotting. I guess we'll know soon!

Re:Reference

Interesting stuff. Here's a link to the full text:

http://130.58.240.179:8080/~erek/minorityreport.pd f [130.58.240.179]

Re:Reference

Well since I was making Hurricane forecasts by July of 2006 that said the season was essentially over for the USA... (Not bragging just I did) The forecasting of events really isn't a hard or difficult thing. In the case of prisoners, you would do better to give a small reward system for accuracy and do a survey system similar to Iowa Prediction Markets. (do your own lookup) It is often a reality that we can predict who is going to steal the most or who is going to kill and quite accurately.

Having worked in a prison as RN, I know pretty well what is going on with the crime scene. It isn't a mystery. The domestic ignorance of what is causing crime, and how to deal with it is mostly the problem. People really just don't like the factor set being told. So I will just to get some really low moderation (by telling the truth) tell approximately what is the profile of violent criminals.

A violent criminal usually falls either below 85 IQ or above 185 IQ. The frequency of this below IQ 85 is about 85% of the population of such persons with about 13% above IQ 185. Only a tiny fraction falls in the middle. Essentially a person who is violent is one who cannot adapt to their world due to low mental state and who rashly reacts to situations they are unable to handle. The very bright criminals of this type are in fact vicious cunning predators. The unique and common link of both groups is their unwillingness to defer gratification of desires for extended periods of time. The want something now and they demand it now and they get it now. They brush anything out of their way on the way including other human beings. If this profile matches to the behavior of some other groups of people we all know and love (CEO's) It isn't an accident. They fall out of this group as well. Essentially we have a party who is willing to force the system rather than work with it. This profile does have racial components. Certain (Nameless deliberately -- I am not suicidal) racial groups tend to do this more than others by wide margins. You could just as well determine these people by their credit rating. It would be just as accurate or more so than the networks.

Actually the most uniform behavior seen in prison is that the persons are ones who "flunked out of kindergarten." The reality here is that successful anti-crime programs generally teach people to defer gratification and to do things like saying those 3 magic words, "Please" and "Thank you." I know this sound simple. It really is. The Church of Scientology (I am not a member and don't intend to be one) has a very successful program that teaches this sort of stuff. It empties prisons when tried. The reality is that when people are taught how to actually deal with their desires and how to communicate with others and how to handle situations, most of them actually do so. This is a damning statement against our modern public schools who think that such teaching is not their duty. Frankly it is their only duty.

Re:Reference

A violent criminal usually falls either below 85 IQ or above 185 IQ. The frequency of this below IQ 85 is about 85% of the population of such persons with about 13% above IQ 185. Only a tiny fraction falls in the middle.

I defy you to find me a single study which supports this ridiculous claim. According to the IQ bellcurve [geocities.com] only one in 1,000,000 people will have an IQ of 174-200. So what you claim means that there are 300 people in the United States committing 13% of the violent crimes? Nonsense.

Re:Reference

The technique used in the paper breaks prisoners into two groups: about 20% of them are tagged as likely to engage in misconduct, and about 80% of them are tagged as unlikely. Then about 20 people from each group (representing 10% of the misconduct group, but only 2.5% of the unlikely group) wind up doing something especially bad-guy-like in the next two years. So it's much better than flipping a coin, which would put 500 people in each group, and would be more inaccurate and more expensive. (Because higher security for 500 people is more expensive than higher security for 200 people.)

Of course, it kind of sucks for the 180 people who aren't going to do something bad-guy-like who are stuck in the misconduct pool. But that number gets winnowed as technique gets better, which is what this research is about.

So, this system flags someone

Their probation officer pays more attention to them, and they feel trapped in the system. They can't move on and contribute positively, and lash out violently.

Thanks, that helps.

Pretty much.

Their probation officer pays more attention to them, and they feel trapped in the system. They can't move on and contribute positively, and lash out violently.

Alternately, their probation officer ignores them, and they get dumped out on the street, where they're unable to find a job and contribute positively, and turn to crime instead.

It's a real win/win.

Re:Utter BS

1) Convicted criminals are the only ones that concern probation officers.

2) Convicted criminals are the only ones they are likely to have the data to fill most of the fields for.

3) Probation officers have a job to do that does not involve tracking random citizens.

Thus, it seems unlikely it could be used for anything *but* the intended purpose without a fairly serious rework.

A bit uneasy?

This sounds like a really BAD idea to me. Either it works really well and then people will start asking why it isn't being used on the general population or it wont work and we'll be focusing our attention on the wrong people. What's the indicator of success? A reduction in homicide rates among people singled out? Our justice system is based on dealing with people AFTER they break the law, everyone, even people at "high risk" to commit crime have to actually do something wrong before you can take action. It may just be used to focus rehabilitation and surveillance efforts on high risk people, but the profiling potential for this must be obvious to the people who designed it, then all it takes is for a little public exposure of how this system could have saved some children if it had been used more aggressively. I'm a bit uneasy about any technology or system that seeks to punish people retroactively. The way the article describes it as working seems harmless now, but the potential of abuse is there. Definitely something to keep an eye on.

Guys we have a problem

Excerpt from the test:

...

21. Ever killed or tortured small animals?

22. If yes, did you often think they enjoyed it and wanted more?

23. Are you a minority?

24. Do you read Slashdot?

25. Regularly?

26. Would you punch a guy with glasses in the face?

27. Would you punch a clown in the balls?

...

Re:Guys we have a problem

[quote]26. Would you punch a guy with glasses in the face?[/quote]
I don't know, having glasses in his face would be pretty painful already.

Oh, stop it.

many places in europe (with virtually no legal gun ownership) are in fact much safer than the usa

http://en.wikipedia.org/wiki/Correlation_implies_c ausation [wikipedia.org]

I find it amusing that Europeans love to bemoan Americans for thinking, particularly when they travel, that Europe should be just like America; however, whenever a European or Euro-phile analyzes crime in the U.S., the only difference that ever gets brought up between the two places in question is the difference in gun control. Really ... so that's the most significant difference between "many places" in Europe and the U.S.? You don't think there are, just perhaps, some more significant social, economic, and cultural contributors to the difference in crime?

Europe and the U.S. are not the same place, and you'd have to control for a whole lot more variables than "gun control" in order to start comparing something as high-level as per-capita murder rates.

Re:Oh, stop it.

There are many things seriously wrong with the U.S., but none of them are easily solvable, nor are they trivial issues. There are thousands of possible inputs which can all have "crime" as one possible output, ranging from the legacies of slavery and discrimination (urban collapse, "white flight"), to drug policy, to basic taboos (sex "education" leading to ridiculously high teen pregnancy rates), to culture (glorification of violence, acceptability of violence in mainstream media). I could literally go on all day. Each one is an incredibly complex issue, in many cases rooted in generations of conflict and bad feelings and issues that people prefer not to discuss. A whole lot of very smart people have worked hard to solve them, and where we are today is the best compromise found so far.

In short, given the existence of fairly high crime rates here anyway, coupled with a well-justified sense of distrust of government and authority, and the extreme symbolic importance of the firearm, it would make little sense and cause great harm to intentionally disarm law-abiding people and remove the means with which they might defend themselves. This is particularly true since there's no convincing evidence showing that disarming law abiding citizens would reduce crime; rather, logically we'd expect to see it increase.

What people in other countries do may well be fine solutions for their needs (although I would probably disagree on fundamental philosophical grounds), but it's foolish to make sweeping cross-cultural comparisons and then blame the resulting difference on a single factor.

Does not apply outside the prison system?

This study was done on incarcerated criminals. Even attempting to apply the findings to people outside prisons would be a HUGE mistake. Now if they conducted a similar set of questions on a few thousand randomly selected members of the public and were able to show the same high correlations, that would be a different story entirely.

Be careful....

"Do you make up these questions, Mr. Holden? Or do they write 'em down for you?"

"The tortoise lays on its back, its belly baking in the hot sun, beating its legs trying to turn itself over but it can't. Not without your help. But you're not helping."

"WHAT DO YOU MEAN, I'M NOT HELPING?"

"I mean you're not helping! Why is that, Leon?"

And yet again

...the attempt to solve with technology what can't be solved by technology.

How about having social workers that deserve that job title? Do we soon replace all judgment on humans and human interaction with computers'?

It is this very dehumanization that causes violence among humans in the first place. How long until someone is flagged by this and it becomes a self-fulfilling prophecy because he feels trapped?

This whole anti-social project shouldn't even have started. What a waste.

This already happens....

actually there are many tools like this already in existence... modern probation work has been scientificalised and statisticalised to the extent that you can't do anything with an offender until you know what their various scores are. In the UK the risk of general reconviction is calculated statistically in the OGRS programme based on age, conviction, prison sentences etc. (http://www.homeoffice.gov.uk/rds/probation1.html [homeoffice.gov.uk]) . This also produces a level of risk that that person will commit a violent offence. There are other specialist tools for domestic violence - the Spousal Assault Risk Assessment which is a 20 item checklist. Also, for sex offenders their risk of reconviction is assessed by using the Thornton Risk Matrix 2000. Every offender who comes into the probation system also has an OASys assessment completed on them - which asks the assessor to score factors from 14 different areas such as accommodation, lifestyle, substance misuse etc. (http://www.probation.homeoffice.gov.uk/files/pdf/ Info%20for%20sentencers%203.pdf [homeoffice.gov.uk]).

Probation: People are missing something here...

"This will help stratify our caseload and target our resources to the most dangerous people," probation department director of research Ellen Kurtz said

Emphasis added.

This is being used by people who have already been tried, convicted, and sentenced and are being monitored and required to check in anyways. The model, further, was derived from the probation system (not from those already in jail):

"Using probation department cases entered into the system between 2002 and 2004, Berk and his colleagues performed a two-year follow-up study - enough time, they theorized, for a person to reoffend if he was going to."

This is just being used to help parole officers decide how to allocate their caseload. That's a Good Thing(TM). No one seems to be talking about applying it to society in a minority report fashion, and while such a harebrained scheme may eventually be table, it needs to be evaluated independently of whether it is a good idea for parole officers deciding how to allocate limited resources.

Something similar

There is a similar thing on this site [peltarion.com] showing how to predict which police cadets would become good cops and which would become bad cops. It's some form of neural net tutorial, but the conclusions are (at least to me) remarkable:

We can with 96.55% confidence say that a graduating cadet will be failing at his job in five years and we can with 99.17% confidence say that a graduating cadet will be performing an adequate job five years in the future.

I'm actually a probation officer

and tried a couple of similar package before. They're all snakeoil.
Nothing can replace years of professional practice and the ability to analyze the bumps on a perps skull.

In other news lately...

Scotland Yard agrees:

It looks like Scotland Yard [timesonline.co.uk] is also looking for scary new tactics in fighting crime. The latest idea of Laura Richards, head of analysis of the Metropolitan Police's Homicide Prevention Unit, sounds like a strangely familiar concept to those who have seen Minority Report. She aims to create a database of people who could supposedly commit a crime in the future, based on their psychological profile.

Even though preventing crimes is a noble motivation, this idea raises serious privacy issues.

As a sidemark it should be mentioned that Laura Richard also seems to be part of the team that "revealed" Jack the Ripper's face some time ago.

Arlo Guthrie QOTD

And I went up there, I said, "Shrink, I want to kill."

"I mean, I wanna, I wanna kill. Kill. I wanna, I wanna see, I wanna see blood and gore and guts and veins in my teeth. Eat dead burnt bodies. I mean kill, Kill, KILL, KILL!" And I started jumpin up and down yelling, "KILL! KILL!" and he started jumpin up and down with me and we was both jumping up and down yelling, "KILL! KILL!" And the sergeant came over, pinned a medal on me, sent me down the hall and said, "You're our boy."

--
BMO

This is Silly

I think ANYONE would kill under the right circumstances. Look at war for instance... All soldiers have to kill at one point or another during a war. Whether it's hand to hand combat, pointing a gun and firing or dropping a bomb. It's all killing. You also have crimes of passion where someone loses control and goes over the line. A parent who witnesses something horrific happening to a child will likely lash out in a rage which would certainly cause death under the right cirumstances. The same for a spouse. You can't predict who will or won't kill if you don't know the situation the person is in.

and in a couple of years time ...

this gets applied generally ... to work out who is going to comit some crime, why not round them up before they do it and save us all a lot of bother.

So: what if you know that you have all the contra indicators: black male youth, poor background, divorced parents, ...

Why bother to do anything: you can't get credit (you are going to be a criminal - right ?), you can't get to be an apprentice or into a good college (you are going to be a criminal - right ?), ...

I can see this happening. Be scared, real scared!

If only they'd run this on George Bush...

...before he was elected. More than 750,000 people might still be alive.

Re:Games

Sure. Once they start plugging in the stats from Halo 2, that will make it obvious as to who is willing to kill or suck the big one.

Edit: Bad Idea.

I thought Phillip K. Dick already explained why this was a bad idea...

There, I edited that for you buddy.
Let's just leave it at that's what you really intended, because otherwise I'll destroy all of my karma in spewing forth a slur of obscenities about how...
well, let's just leave it at that.

If kids couldn't tell the difference...

If kids couldn't tell the difference between pretend and real, we would have never gotten to Pac-Man. Have you ever looked at what kids used to play? They wouldn't look at any graphics on the screen. They would chase down real people tie them to a tree, and physically pretend to cut their scalp off. It is a game that you might have heard of, "Cowboys and Indians". They would pretend to kill each other in cold blood with guns. They would physically act out violent crimes when they would play "Cops and Robbers". If exposure to pretend violence were have any real effect on kids, we wouldn't have made it this far.

Re:If kids couldn't tell the difference...

Indeed. One of my earliest memories is of happily running around the playground and field at school, lobbing imaginary grenades at my friend, shooting them with my fingers and stabbing them with my empty fist. We called it "war", and we divided up into two teams to blow each other away again and again and again.

Amazingly, I've not grown up to be a mass murderer. (In fact, I've never even so much as had a real fight in my life)

Re:Maybe this will be as good as other psychologic

3) Do you ever fantasize about murder?

      I don't get it. This would include most married people...