Real-life Minority Report: software for predicting murderers

Obviously the best way to prevent recidivism among violent criminals would be to simply imprison them for life following their first offense (we'd call it the "one strike and you're out" law), but thanks to that antiquated document known as the Constitution, we're forced to coddle convicts by ensuring that they don't endure "cruel and unusual punishment" to pay for their crimes. Well if we can't lock up all the muggers, jaywalkers, or tax cheats and just throw away the key, the next best option would seem to be predicting criminal behavior before it happens -- and though it sounds like straight-up science fiction, that's exactly what University of Pennsylvania criminologist Richard Berk intends to do. Using a data set consisting of 30 to 40 variables derived from anonymous Philadelphia probation department cases over a two-year period, Berk and his colleagues were able to craft an algorithm that supposedly separates those folks likely to re-offend from ex-cons not deemed to be as dangerous. The point of this program is not to make an end run around sentence limitations or to arrest people before they've committed a crime, but rather, to help probation units decide how to best allocate their resources among hundreds of potential re-offenders. To wit, each subject's variables are plugged into the software in order to create a so-called "lethality score" -- and although the aggregated data is still relatively small, Berk points out that childhood exposure to violence has already floated to the top as the single most likely predictor of murder. Another, less well-known, but equally accurate predictor: jailhouse tattoo across an inmate's chest that reads "Die, ___, Die."

[Via Slashdot]

[Josh @ Dec 4th 2006 8:36PM]

Wonderful picture and quotation. It's the little things that really count. I don't read this site for news.

[james @ Dec 4th 2006 8:48PM]

uhm u have a tear drop tattooed by your eye....guilty. Or how bout the gang MS-13 . they tattoo this right on their foreheads. And if they see a person of another gang...theymust fight til theother is dead or they are. Any if ANY OF YOU are reading this....come get my badge motherf@*&^$!

[Foof @ Dec 4th 2006 9:07PM]

Why would the tattoo matter? That just means "The ____, the." in German.

[Stuart @ Dec 4th 2006 9:09PM]

Well anyone who speaks German can't be that bad.

[Foof @ Dec 5th 2006 2:04AM]

That's what I'm saying, yo.

[Covert @ Dec 4th 2006 9:16PM]

No one who speaks German could be an evil man.

[Aaron @ Dec 4th 2006 9:16PM]

When does the on-line quiz come out?

[Brian @ Dec 4th 2006 9:38PM]

Question #1
Have you been thinking of murder?
a) yes
b) no
c) you'll never catch me coppers!

[Oddmanout @ Dec 4th 2006 10:23PM]

Well, don't criminal profilers already try to predict future crimes, as a way of preventing them, based on past activities and psychological information ? I suppose it's not so far fetched to have a machine do it.

But yeah, I think it's best used as a guide not a rule so to speak. I'd hate to go to prison just because I happen to have watched 'A Clockwork Orange' 327 times between the ages of 13 and 18 and that sent up a flag on some gizmo...

For example I mean...

[krusher_00 @ Dec 4th 2006 10:24PM]

Question 2:
1) Do you read engadget?
2) Regularly?
3) Did you really think we wouldn't know..?

[Ryan @ Dec 4th 2006 11:40PM]

It's not your credit score, it's your crime score...

[xVariable @ Dec 5th 2006 12:25AM]

So it's cruel and unusual punishment now to jail someone literally for the rest of their natural life? Is that what the courts have said, or is that editorializing.

Here all this time I thought that the least punishment for taking away someone else's life is to take away the perpetrator's freedom forever. You can rest assured that the victim has FAR less even than that, but I bet you don't think about THAT a lot, do you? Does anyone believe that, if there were a better way of determining someone's guilt with much greater certainty, that we wouldn't be putting A LOT more people to death for murder?

Until we can resequence the DNA of of murderers, we owe it to the victims and society at large to put them away FOREVER.

[sugarkang @ Dec 5th 2006 1:24AM]

xVariable:

i absolutely agree. but you're presupposing that we know who the guilty people are. think of all the innocent black people that got jailed during the civil rights movement in the 60s.

but that's the past you say...
that shit is still going on. the legal system we have is imperfect. we catch some bad guys, but mistakenly jail good guys too. and in some circumstances, we sentence good guys to death. for that reason alone, there shouldn't be a death penalty. yeah the bad guys might deserve death, but the good guys don't. and since we have to assume that our system is imperfect, the death penalty is a just a vehicle for killing innocent lives.

[Andrew C @ Dec 5th 2006 1:22AM]

Hmm, this is similar to something one of my professors worked on. I'm taking this modeling of systems class here at Pratt School of Engineering at Duke and on our first day of class our teacher showed us some of her projects.

One was this software that you would plug all of these random variables into such as whether a knife, gun, poison, etc was used, whether the body was moved after the crime, whether the weapon was brought to the scene etc. It was a list of around 30 different variables that detectives could easily determine.

Anyways, they ended up determining what these variables were for a bunch of actual murder cases in the past and put them into a matrix. From this they found certain patterns.

Then they put the system up to the test. They would give it murder cases that were already solved (such that they knew who the killer was and all of his characteristics) and would put all of the background variables of the murder into the system. It would then spit out stuff like male, age 27-30, married, tall, white, etc...For all of the cases that were thrown at it, the algorithms had high success rates with good confidence intervals. I thought it was impressive at least.

http://ieeexplore.ieee.org/iel5/10559/33412/01582571.pdf?arnumber=1582571

There is a link to one of the papers that she published.

apc

[xVariable @ Dec 5th 2006 2:05AM]

sugarkang: If you read my post a little more carefully, you'll see that exactly what I said: our system isn't perfect but, were we a lot more accurate at determining guilt, we'd be sentencing a lot more people to death for murder.

The point, as you've said, is that people that are truly guilty of murder should forfeit their lives. Instead, we have Engadget saying that imprisoning murders for the rest of their natural lives in cruel and unusual. So why aren't you decrying Evan and Engadget for being soft or murder?

They didn't say that because we sometimes convict innocent people (something I agree is wrong and never said otherwise) that it's cruel and unusual to put them in jail for the rest of their lives. No, Engadget said it's cruel and unusual to to put MURDERERS in jail for live. Not the wrongly accused, but those that are actually guilty of killing someone.

The fact that such a statement doesn't cause visceral revulsion in people speaks to the sickness of our society in my opinion. "Hey, they snuffed-out someone's life, that person is gone forever but, hey, to remove them forever from the population, as punishment and for everyone else's protection, is cruel and unusual. The victim probably had it coming anyway, right? I mean, if somebody's willing to commit the ultimate transgression they must have had a good reason. Right?"

[xVariable @ Dec 5th 2006 2:24AM]

For those that this sort of thing couldn't possibly be accurate: there was a story, from years ago, about a university that had developed software that could predict a crime in a given area BEFORE it happened.

That's right. THey compiled stats from a specific area, and were literally able to predict WHEN and WHERE the next crime would happen, and WELL in advance. This isn't bullshit. If anyone seriously thinks I'm making this up I suppose I could be compelled to Google it but, regardless, nothing about how I described their system is hyperbole.

I hate to tell you, but people aren't special. They are profileable and predictable.

[I? @ Dec 5th 2006 7:28AM]

And I hate to tell you, but statistics, although useful, don't work well to evaluate changing situations. The profile of today's murderer is probably very similar to 1500's heroes'. You just can't tell somebody is a reinciding criminal by fitting him to a statistical profile created based on past populations.

The world is changing quickly and ever quicklier pal, my grandma still has trouble understanding homossexuals aren't crazy. People start supporting some dumb scholar like Mr. Berk there, and rehabilitation will be stricken right off the dictionary.

[Covert @ Dec 5th 2006 8:22AM]

For what it's worth, it seems to me that Engadget was saying that it would be cruel and unusual punishment to lock up "muggers, jaywalkers, or tax cheats" for life after the first offense - there's no mention of murderers, xVariable. I'm sure you realize that the punishment wouldn't fit the crime to lock up a jaywalker for life after his first offense, so it's cruel and unusual.

[Charles Navarro @ Dec 5th 2006 8:54AM]

While I was doing my undergrad in CS, there was a grad student who wrote software that said if a student had copied another student's work. Too bad two of us had similar coding styles and thought processes, so it always flagged our assignments, even though we didn't code together, and it was obvious just by looking at the code. I wouldn't have much faith in this software either. There is something to be said for human intuition.