Over the years, Hollywood has made a living out of predicting what life in the future would consist of, including crime fighting. Sometimes those predictions come true, at least to a degree, and it looks like a 2002 movie about predicting crime before it happens is on the verge of being one of those times.
In 2002, Tom Cruise starred in the movie “Minority Report” as the chief of the DC Pre-Crime Unit. The basic premise of the movie was combating the historically high murder rates in the nation’s capital by using three psychics of sorts to predict future murders. Then the pre-crime officers utilize the “visions” supplied by the psychics to identify the suspect and victims and then arrest the suspect for murder PRIOR to the crime being committed. Although we have not progressed to the point of using psychics to predict crime or arresting suspects for crimes they might commit but never do, we are one step closer to predicting future crime so we can then take the steps necessary to stop it.
In August at the 2015 Joint Statistical Meetings in Seattle, Michael D. Porter presented the results of his research project “A Statistical Approach to Crime Linkage” and suggested that statistics could be used in a manner similar to the “Minority Report” psychics to predict future criminal activity, which would then allow commanders to redeploy assets as necessary to prevent that crime from occurring.
The idea of predicting crime based upon an analysis of past criminal activity is not a new one. Every time a commander beefs up patrols due to an increase in pickpockets or deploys decoys after a series of rapes, they are performing basic criminal analysis and acting upon it for the purpose of preventing criminal activity. The problem with this approach is it is very time consuming and is often a best guess type of response. But Porter’s suggestion is to utilize a statistical model, available as an easy to use software program for commanders, to sift through the piles of reports, eyewitness statements, office observations and historical data to reduce the man hours required to conduct the analysis portion and then provide more than a best guess response – it provides a statistically valid prediction of where then next similar crime will occur.
But will it work?
To test his theories, Porter teamed up with fellow researcher Brian J. Reich, associate professor of statistics at North Carolina State University, and used the software to predict future criminal activity in a real world setting. Porter and Reich selected Baltimore County Maryland, which between 2001 & 2006 had experienced 10,670 burglaries. In the end, the software was able to not only able to identify potential links in different burglaries, but also predict where the suspects in those linked crimes were most likely to strike in the future. Based upon their findings, the software was successful in detecting linked crimes in 85% of cases with only a 5% error rate; future activity was successfully predicted between 74% and 89% of cases and researchers determined that follow up crimes of this nature typically occurred within an 8.3 square mile area.
Of course, the software does not provide images of victims, identities of suspects or the address to be burglarized, but it does allow commanders to narrow their area of concern and, if a suspect is apprehended, potentially link previous crimes to that perp. Right now that is not enough to remove good old fashioned police work from the equation; officers patrolling the identified areas are still the last line of defense and best hope of actually catching the burglars. But, what it does do is allow police officers to move one step closer to living up to the motto “Protect and Serve” through crime prevention rather than crime response.
Disclaimer: The content in this article is the opinion of the writer and does not necessarily reflect the policies or opinions of US Patriot Tactical.