“If Chicago doesn’t fix the horrible “carnage” going on, 228 shootings in 2017 with 42 killings (up 24% from 2016), I will send in the Feds!” @realDonaldTrump tweeted on January 24, 2017.


Can predictive policing save lives in Chicago?


Los Angeles implemented a predictive policing program in 2008 using analytical techniques to determine locations or hot spots that were most at risk for criminal activity. A 2015 UCLA study showed that these predictions, utilized to deploy its officers to the right places, prevented twice as much crime as trained analysts. These location centered predictions have been used effectively in many large cities, but Chicago’s crime is less geographically centralized.

Andrew Papachristos, a sociologist at Yale and Chicago native, found that Chicago’s homicides were concentrated in a relatively small number of social networks. In a 2014 study of a high-crime neighborhood on Chicago’s West Side, Papachristos found that 41% of all gun homicide victims in the community of 82,000 belonged to a network of people who had been arrested together and who were a mere 4% of the population. He came to the conclusion that much can be learned about crime by examining the company people keep.

With a $2 million grant from the National institute of Justice a professor at the Illinois Institute of Technology, Miles Wernick, created an algorithm of 11 weighted variables tied to a person’s past behavior, particularly arrests and convictions. Sample source questions included, Have you been shot before? Is your trend line for crimes increasing or decreasing? Do you have an arrest for weapons?

The Chicago police also emphasized that Dr. Wernick’s model intentionally avoided using variables like race, gender, ethnicity and geography that could discriminate in some way.

Police Department Superintendent, Eddie Johnson, reported that in a city of 2.7 million people, about 1400 are responsible for much of the violence and all of them are on the Strategic Subject List (SSL) created by Dr. Wernick’s model. So far in 2017, 70% of the people who have been shot in Chicago were on the SSL, as were 80% of those arrested in connection with these shootings. In one recent report on homicides and shootings over a two-day stretch, nearly everyone involved was on that list.

A small consulting firm in Chicago, Clarity Partners, built a software interface for the police department that can pull a photo of a shooting victim, with a spider web radiating to known associates and yet more ties extending to their associates, mapping an entire social network in a glance. These risk assessments of individuals have been astonishingly accurate.

So,then why isn’t the crime rate in Chicago going down?


A RAND study of the utilization of the SSL by the Chicago police found no lives saved through this program. Identifying the people with the highest probability of committing a crime or being the victim of a crime was done well, but this program’s pre-crime feature of police visiting people on the list before a crime was committed wasn’t having the positive impact that was expected. The police were using the predictions to make quicker/better arrests but the homicides were still occurring.

“Just creating a data-driven ‘most-wanted’ list misses the value of big data prediction,” said Andrew Ferguson, law professor and predictive policing expert at the University of DC. “The ability to identify and proactively intervene in the lives of at risk youth is positive, but you have to commit to the intervention piece. Just directing police toward those individuals for traditional policing is not enough.”

Interested in this predictive analytics problem or others?  Contact Data Science Executive Recruiters at Smith Hanley Associates.  Midwest/South:  Nihar Parikh, [email protected]  East/West:  Paul Chatlos, [email protected]

Leave a Reply

Your email address will not be published. Required fields are marked *