Common mistakes include inconsistent notation usage, mixing abstraction levels within single models, unclear decision criteria, missing exception paths, undefined roles and responsibilities, overly complex diagrams, inadequate labeling, missing start/end points, incorrect sequence flows, and failing to validate models with process participants before finalization.