"Since August 2015, the Boston Police Department has documented every reported crime in the city—creating a digital chronicle of extraordinary proportions."
The dataset before us represents one of the most comprehensive crime records in American municipal history. Nearly a million individual incidents, each timestamped, geolocated, and categorized, paint a picture of urban life that no survey or interview could capture.
By the Numbers
This isn't just data—it's a living record of the challenges Boston faces daily. From minor larcenies to violent assaults, from the historic streets of the North End to the neighborhoods of Roxbury and Dorchester, every incident tells a story.
The Annual Rhythm
Crime in Boston follows predictable patterns. The data reveals seasonal fluctuations, with incidents peaking during summer months when more people are outdoors and active. Winter brings a natural decline, though indoor crimes like domestic incidents may increase.
Crime rates dropped significantly during 2020 due to COVID-19 lockdowns, but rebounded in 2021-2022 as the city reopened. This natural experiment revealed how human movement patterns directly correlate with crime incidents.
District Distribution
Not all districts are created equal. Boston's 12 police districts show vastly different crime profiles, influenced by population density, commercial activity, nightlife presence, and socioeconomic factors.
| District | Total Incidents | Violent Crime % | Primary Type |
|---|---|---|---|
| B2 (Roxbury) | 123,863 | 24.8% | Assault |
| D4 (South End) | 112,407 | 18.2% | Larceny |
| C11 (Dorchester) | 110,094 | 22.1% | Assault |
| B3 (Mattapan) | 91,224 | 26.3% | Assault |
| A1 (Downtown) | 88,397 | 12.4% | Larceny |
District B2 (Roxbury) leads in total incident count, a pattern that has persisted throughout the decade. However, the nature of crime varies significantly—downtown District A1 sees primarily property crimes, while southern districts experience higher rates of violent incidents.
What Lies Ahead
This chapter establishes the foundation. In the chapters that follow, we'll dive deeper into the spatial patterns that reveal geography's role in crime, the temporal rhythms that show when danger peaks, the machine learning models that can predict violence, and the optimization algorithms that can improve patrol efficiency.
Crime data represents only reported incidents. Many crimes go unreported, particularly in communities with strained police relationships. The patterns we identify reflect reporting behavior as much as actual crime occurrence.