Christoph Trinn analyzes the possibilities and limitations of predicting the violent escalation of internal conflicts. Although these do follow certain simple empirical laws, the outbreak of violence is fundamentally unpredictable. The existence of escalation patterns and the impossibility of forecasting escalation arise from the same phenomenon: the power law as an expression of self-organized criticality. Conflict intensities stem from a unified causal mechanism that produces protests and assassinations as well as revolutions and civil wars. The author provides a scientifically grounded and empirically extensive investigation that points to ways for risk assessment and crisis management.