By Mark Esposito, Weekend Contributor
Lucky or just good? That’s what police in Madison, Wisconsin are wondering after crime analyst, Caleb Klebig, successfully predicted the date and time of Scottie T. Patterson’s, 28, latest and last bank heist. Using data from other similar robberies, Klebig estimated that the then unknown Patterson would hit his next bank on a Wednesday or Thursday between 2 and 7 p.m. He narrowed the field of potential targets to five banks in greater Madison. Police staked out the banks and, sure enough, Patterson arrived right on cue at 2:40 p.m. on Wednesday. Confronted by the seemingly omniscient detectives while exiting the bank with the loot, Patterson made a break for it but was captured behind a nearby shopping center.
The prediction was made possible by advanced computer software and some sophisticated matrix analysis of factors by Klebig, who compared scores of factors from various known bank robberies before playing Karnac for the police. Patterson pled guilty to the robbery and three others saying he stole the money to buy heroin.
The story channels the sci-fi thriller from Steven Spielberg, Minority Report, about a dystopian world where crimes are predicted and punished before they occur. Chicago and LA police are already hot on this futuristic trail. Chicago uses complicated algorithms to track and then predict crimes. A list of most likely suspects is then developed from the interplay of crime scene factors and known methods of operation of career criminals. According to CPD Commander Jonathan Lewin it’s all for the public good:
“This [program] will become a national best practice. This will inform police departments around the country and around the world on how best to utilize predictive policing to solve problems. This is about saving lives.”
The program sounds Orwellian as not only offenders but their network of associates are also tracked using social networking theory and targets are then developed. Sometimes the targets are warned to prevent crime.
The program was founded with a grant from the National Institute of Justice. “These are persons who the model has determined are those most likely to be involved in a shooting or homicide, with probabilities that are hundreds of times that of an ordinary citizen,” NIJ representative Joan LaRocca said.
“Who the model has determined”? That should strike fear into most anyone who has ever turned on a computer. The practice raises several civil liberties issues not the least of which is the role of a Big Brother government concerned with whom its citizens associate. Not only that, the computer has no conscience and even reformed offenders stay in the database still suspected of crimes.
L.A. cops have taken the computer predictions a step further — taking orders from the machines to prevent burglaries. The software outperforms human analysts according to Sean Malinowski, a police captain in the Foothill division. In that area of L.A., computer models predicting theft crimes accounted for a 25 percent drop in reported burglaries, an anomaly compared to neighboring areas. “We are seeing a tipping point—they are out there preventing the crime. The suspect is showing up in the area where he likes to go. They see black-and-white [police cruisers] talking to citizens—and that’s enough to disrupt the activity,” says Malinowski.
The system in L.A. works to provide each patrol car with a printed map highlighted with red boxes,” 500 feet on each side, suggesting where property crimes—specifically, burglaries and car break-ins and thefts—are statistically more likely to happen.” Cops the use the data to set up patrols in likely areas.
The question not answered in Chicago or L.A. is whether the software will be used to predict just crime or criminals, too. Is it ethical to assume criminality before it’s even accomplished?
Source: MIT Technology Review; The Daily Caller; Channel 3000
Now what do you think? Good Police work or crime fighting gone too far:
~Mark Esposito, Weekend Blogger
28 thoughts on “Predicting Crime and Criminals — “Minority Report” Or Good Policing?”
Mark Esposito: ” Chicago and LA police are already hot on this futuristic trail.”
Two of the most corrupt police departments in the country are already doing this. Makes ME feel safer!
Very interesting article, Mark. Thank you.
If he did legitimate profiling…. I have not a problem…. But if it involved info supplied by the NSA etc…. Yeah I have a problem unless a valid warrant was issued…..
Logic dictates that this is not an intrusion into the lives of American citizens but an analysis of events and probabilities. The article states that the target banks and times of the projected robbery were zeroed in on, not the individual. If, as in the movies, people are scrutinized to this extent, then that would obviously open up the issue of our rights of privacy. However, to some degree the closer an individual’s activities gets him or her to a crime, the more they become a suspect. This has been police practice for hundreds of years, since the era of Louis Vidocq, the father of all detectives.
Given the choice in the article, what I think is good sophisticated police work.
As another here mentioned, as long as the data collection comports with true 4th Amendment constraints, then by all means, go for the pre-crime analysis.
I agree with Darren. Actually, there is nothing new under the sun, since law enforcement and security agencies have been using similar techniques for centuries, literally. The techniques are different. Instead of colored maptacks on a board, they now use electrons.
The model began to change with the security apparatchik set up by George W. Bush and his minions when 9-11 gave them the excuse they had been looking for. Of course, they had wanted an excuse since the days of Ronald Reagan, and even before.
It is a pipe dream to think the inventors of this technology won’t try to use the huge information vacuum cleaner that is the NSA, CIA, ATF and other such agencies. The creation of Homeland Security as an all-powerful agency with Cabinet status was a horrible idea, because it concentrated all this power in the hands of a select few whose motives are unknown, and who seem to be accountable to no one. The President and Congress are their bosses in name only.
Those who have the information have the power. Keeping the rest of us, including the President and Congress, in the dark enables them to keep that power.
Obviously, it is ethical to predict criminality before it is predicted if there is ever to be an effective way to actually prevent criminal acts.
Locking the barn door only after the horse thieves have stolen the horses, slaughtered the horses, and ground them up into beef stock flavored horsemeat, and burned the barn to the ground, is not an efficient way to keep said horses alive and well.
Without a way to recognize patterns accurately, accurate pattern recognition is as though made of some form of unobtainium.
Without accurate pattern recognition criminal activities may mistakenly be thought to be randomly chaotic and not, as my sense of biology suggests to me, largely the result of an interesting and sadly deterministic mechanism.
You took the words right out of my mouth. The specific instances in this story seem fine to me. The bank robber wasn’t entrapped and he was not arrested until after the crime had been committed (unlike Minority Report). The posting of patrol cars in predicted burglary areas also seems like a fairly passive move. No one is arrested, the crime is deterred by their presence.
Of course, it’s law enforcement, so if there’s a tool, they’ll find a way to abuse it. But what I read in this particular story seemed like a terrific use of science and technology.
LTMG, Thanks for the info on Markov. I’ll read up on it. And yes, we are becoming more and more illiterate as a culture and math is probably the best/worst example. I’m 61. When I was in my late 40’s I went back to school to get my teaching certification. I had to take standardized tests prior to getting into the UW education program. I went on a Saturday morning and was standing in line, conversing w/ other folks taking the test. They were almost all in their early 20’s. There were several people there taking only the math section. One was taking it the 3rd time, 2nd time, etc. A woman sheepishly said this was her 5th attempt @ passing. Now, I was an average math student. The only A’s I ever got were in geometry, I just got it. So, I walk into the test a bit shaken. I get to the math section, start answering the questions, saying to myself, “This isn’t hard!” Then I start to think, “am I’m kidding myself?” But, I have common sense, review a few of the sorta challenging questions and then say to myself, “Damn, this isn’t hard.” When I got my results I was in the 92 percentile. Well, I was not in the 92% when I was in college back in the 70’s and standardized tests were MUCH more difficult then. WTF happened the last 40 years, LTMF??
to not TI
@spinelli 10:58 pm. I suspect that the police are using Markov analysis and other stochastic process analytical techniques to suggest where and when TI interdict criminal intentions. Same also for many years in professional sports.
In the sports world, such analysis is not cheating. In the law enforcement world, it is not a violation of civil rights. It is applied mathematics. No magic, and nothing nefarious. Of course, try explaining the method in a country that is slipping backwards towards mathematics illiteracy.
rollotomasi, I loved your “role” is Hollywood Confidential. I’m sure the surveillance comes up dry most of the time. Surveillance, no matter how well researched and conducted, is not productive most of the time. You have NO CONTROL over what the subject will do. You merely use all your resources to try and be in a position if something does happen. Patience is a rare quality in our culture. It is critical to conducting surveillance. I did it for 30 years and still do a little now.
Sounds more like Numb3rs than Minority Report. 🙂
Is this like the horseplayer who tells us only about hitting the big one? How many misses were there, how many large stakeouts came up dry?
I don’t have a problem with this system as long as it uses arrest / crime reports for its database and not social media and other intrusive factors. I also don’t have a problem with it using crime patterns by known criminal individuals and their habits. It is doing more accurately what an intuitive local officer or agency does regularly, knowing where to patrol based upon certain conditions and based upon MOs the types of suspect likely to be involved.
Very interesting post, raff.
They’ve been using programs like this in MLB. Tendencies, history, etc. to position your infielders/outfielders. It helps predict what kind of pitch a hitter will swing at, or not swing. Direct marketing companies use data to predict what you might want to buy based on information they have on you. It’s typical of the mindset it took law enforcement this long to develop a system like this. I see no problem w/ this based on the info provided. There is no entrapment. Hell, cops use informants to entrap criminals to commit crimes and they’re set up to bust them when it goes down. In my thinking, this should help lessen that practice. It will never stop it.
What, prevent crime as if it were something as unproductive as fires and crashes??? What will happen to the criminal lawyers when they are found to be the ones planning their slippery clients m.o.’s so they can get everything dismissed, whitewashed, and then get paid with the proceeds of those crimes thanks to the Courthouse Gang’s warped thinking?????
Technology, wow. And wow, it happened in Madison. I can see that it needs be used wisely.
long as they’re not getting any help from nsa intercepts.
but how can you tell?
Great article Mark. I am concerned how the police will use this latest tool.
Comments are closed.