Models We Can Trust: Toward a Systematic Discipline of (Agent-Based) Model Interpretation and Validation.