How much attention should criminology pay to very rare forms of violence

How much attention should criminology pay to very rare forms of violence (e.g., serial killers, mass shootings) when designing theories? When developing policies? Why?