How I Approach Gang Mapping Research in Chicago
Gang maps of Chicago have been around longer than most people realize. The early versions came from law enforcement agencies doing their own internal analysis, usually using GIS tools that most residents couldn't access. What you see online now is a patchwork of sources: some are legitimate research projects from universities, some are citizen journalists compiling public data, and some are just someone's opinion drawn on a map with no citations at all. I've spent more time than I care to admit vetting these things, and most of what passes for a "gang map" falls apart under basic scrutiny. Before you go anywhere near a gang map, you need to understand what data actually supports it. The Street Crime Incident Report (SCIR) database from the Chicago Police Department is publicly available through their open data portal. It contains records of street-level crimes with coordinates going back to around 2018. That's one source. The federal Bureau of Alcohol, Tobacco, Firearms and Explosives maintains a National Gang Threat Assessment that covers Chicago as part of its broader regional analysis. The Chicago Tribune and Sun-Times have published their own mapping projects over the years, though their coverage tends to peak during specific news cycles rather than maintaining continuous updates. University of Chicago's Crime Lab has done peer-reviewed work on gang territory patterns that you can reference if you're looking for academic grounding. I built my first working map back in 2019 by pulling SCIR data and cross-referencing it with neighborhood association boundary files. The immediate problem I ran into was that police incident reports don't actually label gang affiliation for the vast majority of entries. Only a small fraction of incidents include any gang notation, and those notations are inconsistent across districts. You end up with a map that shows where crimes happened, but telling someone that a cluster of dots in Englewood represents gang activity specifically is a leap that most map makers gloss over. I learned to present the data honestly: here are the incident clusters, and here's where publicly documented gang presence overlaps or doesn't overlap with those clusters. The gap between the two is where most misinformation lives.
The most useful tool I found was QGIS for actual map construction because it handles shapefiles and coordinate systems without costing anything. ArcGIS is better if your organization already has a license, but for personal research it's overkill. I also use Python with the geopandas library to clean and process the raw incident data before plotting it. A typical workflow involves downloading the CSV export from the city's data portal, filtering for relevant incident types, merging in geographic boundaries, and then doing spatial joins to see where known gang territories fall within hot zones. This usually takes me about forty-five minutes for a fresh analysis, though the first time through a neighborhood you haven't worked with before can stretch to two or three hours depending on data quality. Here's something most people writing about this get wrong: gang boundaries are not static. They shift based on law enforcement pressure, economic conditions in the area, leadership changes, and sometimes just rival moves. A map you find online that claims to show current territories is probably at least six months out of date, and in some cases a year or more. I've seen people cite maps from 2017 as if they reflect 2024 conditions. The only way to keep current is to feed fresh incident data into your model regularly and watch where the clusters move over time rather than relying on any single snapshot. There's also the issue of what actually counts as a gang map. Some maps show only sanctioned or police-recognized gangs. Others include sets, car clubs, and informal groups that operate in the same style but aren't on any official list. The gang terminology itself varies wildly between sources. What one database calls a "set" another might call a "crew," and some organizations rebrand or split without anyone updating their records. I kept a running spreadsheet of naming variations across the different databases I was pulling from, and it ended up being fifty pages long. Without that kind of tracking, you'll miscount membership and misattribute territory simply because the labels changed.
If you're trying to use this research for something practical, like understanding your neighborhood or writing a report, the biggest pitfall is confirmation bias. You find a map that matches what you already believed, and you stop digging. I caught myself doing this when I was first researching the South Side. The map I was using showed heavy concentration in certain areas, and I kept skipping over neighborhoods that didn't fit the pattern because it was easier to just accept the existing visualization. The workaround was to run the raw incident data through a different classification system and compare the results. When the two methods produced different maps, I had to figure out which one was actually closer to what was happening on the ground. That process took weeks and completely changed my initial interpretation. The second counter-intuitive thing I learned is that more data doesn't always mean a better map. When I first started, I thought pulling every single incident type would give me the clearest picture. It didn't. Violent crime incidents clutter the map and make patterns harder to see. I ended up focusing on a narrower set of indicators: gun-related incidents, assaults with known gang notation, and thefts in areas where gang activity was already documented. That filtered dataset produced a much cleaner and more accurate representation than the full raw feed ever did. What I can't recommend is using any gang map as a definitive source without checking the methodology. I've seen maps sold or shared that have no visible data sources, no dates, and no explanation of how territories were determined. Those are not research. They're opinion dressed up as cartography. If someone can't tell you where their boundary lines come from, you shouldn't treat their map as factual. The same goes for maps that show gang membership numbers without citing their source. That kind of data is almost always invented or wildly outdated.
Get the Full Details
For people who want to do this properly, the path is straightforward even if it's tedious. Download the city's open data. Learn basic SQL or Python for data cleaning. Use QGIS for visualization. Cross-reference with academic sources and news archives. Update your map quarterly at minimum. And always, always note the limitations and date range of whatever you're presenting. The goal isn't to produce something that looks authoritative. The goal is to produce something that's actually accurate, and accuracy requires admitting what you don't know as much as what you do. I keep a folder of bad maps I've encountered over the years. There are dozens of them. They share common flaws: stale data, missing methodology, overconfident territory lines, and the occasional outright fabrication. None of them survive a simple fact-check against public incident records. That's the baseline standard. If your map can't hold up against the raw crime data that the city already publishes, it's not worth sharing.
