Effective criminality forecasting enhances public safety by optimizing the knowledge of crime evolution, increasing the transparency of police actions at the condition to be at a the right decision making level. This kind of tools is dedicated for tactical and strategic issue in order to optimize the operational level. It is not built as an execution tool but needs to be used at the decision making level. This paper presents an adaptive time series analysis method that improves both the accuracy and interpretability of crime predictions. Utilizing a five-year dataset of interventions by the French Gendarmerie—covering approximately 80% of France’s territory—the model analyzes historical incidents to forecast future occurrences. It integrates two components: tendency, capturing short-term trends from recent incidents with higher weights assigned to more recent data, and seasonality, accounting for long-term patterns using data from the same periods over the past four years with decreasing weights for older data. Developed and validated using real-world data, the method produces accurate and explainable forecasts. This explainability satisfies legal requirements for accountability and aids law enforcement in decision-making processes. The approach effectively captures evolving crime patterns, offering valuable insights for proactive policing while balancing Precision and interpretability.