Anomaly detection systems usually focus on spotting unusual behavior, but sometimes anomalies follow patterns too. AD-MERCS is a new method that looks for patterns in both normal and abnormal behavior to detect anomalies better. It finds different areas where patterns exist and figures out what makes instances deviate from these patterns. Tests show that this approach works well for many types of anomalies. Also, by pinpointing patterns and conditions in simpler spaces, the system can explain why something is considered an anomaly in a straightforward way.