A DevOps team wants to improve how their application responds to unexpected outages. They decide to regularly simulate failure events in a controlled environment to test their runbooks and recovery procedures before real incidents occur. Which AWS Well-Architected Framework design principle does this practice BEST represent?