Mafia Murders as a Complex Systems: Temporal Patterns and Statistical Modeling

The rise and fall of the mafia phenomenon in Sicily has profoundly influenced Italy's modern history. One of the main instruments historically employed by Cosa Nostra to impose its power has been targeted homicides. Rather than being random events, these acts are often later revealed through law enforcement operations and judicial investigations to be part of the complex political strategies of organized crime. In collaboration with the Prosecutor's Office of Palermo, we compiled a mostly comprehensive archive of homicides in Sicily from 1945 to the present day. The dataset includes names, dates, and locations of the events, covering key periods such as the so-called Mafia Wars. By analyzing the time series, we identified analogies with complex systems like earthquake dynamics, neural activity, and financial markets, particularly the emergence of scaling laws and non-trivial inter-event time distributions. We further modeled the data using a Hawkes process, revealing that the temporal structure of the homicide series provides insights into its underlying generative mechanisms. We will discuss if certain bursts of activity can be interpreted as cascades of triggered events, and if there is a characteristic timescale. This approach provides a quantitative description of the escalation dynamics of organized crime, contributing to the developing of data-driven approaches for historical and criminological analysis.