“We predict you will commit a crime. Confess now and save yourself the trouble of actually committing the crime.”

Software used to predict who might kill

University of Pennsylvania criminologist Richard Berk, a trained statistician, never met a data set he didn’t like.

Now, using fresh data from the Philadelphia probation department, Berk and three colleagues have built an innovative model for predicting which troublemakers already in the system are most likely to kill or attempt a killing.

With the homicide rate in Philadelphia outpacing last year’s by at least 7 percent, a computer model for “forecasting murder” is in the works, Berk said, to be delivered to the probation department in the new year, with clinical trials of the new tool to begin in the spring.

Initial research suggests the software-based system can make it 40 times more likely for caseworkers to accurately predict future lethality than they can using current practices.

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.

“If we have 100 probationers I can accurately find the one murderer who statistically will be in that group if I devote resources to all 100 as if they are murderers. The problem is that for that one murderer who is a ‘true positive,’ I have 99 false positives. […] “Now suppose I can identify the 10 at highest risk. For that one true positive I now have nine false positives,” Berk said, “and that may be something we choose to live with.”

  1. Miguel Correia says:

    uau… Minority Report without the pre-cogs. 😉

  2. Named says:

    does computer game or movie violence count?

  3. Mike says:

    #2, I don’t know about computer violence, but given the recent growth in popularity of the “Saw” type slasher, gross-out flicks, it really says a lot about people that they flock to watch people being gruesomely tortured for entertainment.

  4. Chris Evans says:

    Do the words “self-fulfilling prophecy” mean anything to anyone?

    Point the finger and keep harassing someone who hasn’t done what you accuse him of ‘potentially doing’ and guess what’s most likely to happen…

    The mealy-mouthed “If it saves just one person” justification is fast becoming my hottest hot-button.

  5. V says:

    Great. There’s just one problem: it’s useless. Statistics analyze trends on a large scale, and the only results they will deliver are the trends on a large scale.

  6. Ballenger says:

    “Now suppose I can identify the 10 at highest risk. For that one true positive I now have nine false positives,”

    McDonald’s, Wendy’s and BK can produce better results now on forecasting how many people out of a thousand will order a cheeseburger with a marshmallow on it. Nine false positives in one hundred wouldn’t meet their standards.

    “But before that can begin in earnest, the public has to decide how many false positives it can afford in order to head off future killers, and how many false negatives (seemingly nonviolent people who nevertheless go on to kill) it is willing to risk to narrow the false positive pool.”

    Please, the thermometer hasn’t popped on this turkey. It needs more time in the lab, not a “public decision” on usability.

  7. tallwookie says:

    ya know I saw this on /. last night, and I *instantly* thought of minority report.

  8. Calin says:

    The only use this might have is give a fairly good % possiblity of future violent crime by people up for parole. Not that their parole should be based strictly on this number….but it could be a variable in the process.

  9. David Darwin says:

    Regarding Mr. Berk,

    There was a time when we used to say “better to let 10 guilty be free, than to imprison a single innocent”. Now that the criminalization of citizens has become lucrative big business, innocence and guilt are becoming non-issues. I think now it’s how many, how fast, and where can we put’ em.

    Innocent until proven guilty . . . Mr. Berk’s new “tool” merely provides a theoretical result and should be regarded as just that, a statistic and nothing more. Documenting possibilities of a person’s actions, marketing them for profit, and then a clouded perception of fluff for fact . . . how lurid and sociopathic! Mr. Berk should sell used cars and get out of the “people” business.

    Blah, blah, blah . . . and then he farts out “Now suppose I can identify the 10 at highest risk. For that one true positive I now have nine false positives, and that may be something we choose to live with.”

    F*#k you! I see a 1976 Ford Pinto in your future, Mr. Berk.

    It’s unfortunate but nonetheless true that crime must first occur, in reality, before we establish guilt. As I understand our system now, guilt is established through concrete and lawfully obtained evidence and processes. Welcome to the bungle . . .!

    So, Mr. Berk, that is the ONLY sad truth I am willing to live with and not a bean counter’s preconceived notion and assumtion of FUTURE human action and character.

  10. Arbo Cide says:

    This is something that happens already. Criminal sentencing, people look at factors like this. That’s why we have 3 strikes you’re out. And

  11. Smartalix says:


    Well said.

  12. Mr. Fusion says:

    #9, well said and I agree.

    #10, Wrong. Three strikes looks at past behavior and punishes on that basis. Only it punishes those who were caught, not those who committed a crime. Since minor crimes such as shoplifting may be considered felonies, it also isn’t really a good measure of a person’s true anti social behavior or propensity to further commit heinous crimes either.


Bad Behavior has blocked 6423 access attempts in the last 7 days.