Abstract for: Teaching Ship Lifecycle Dynamics and the Commons Dilemma Through a Multiplayer Fishing Game Pre-Calibrated with Rule-Based Agents

System dynamics simulation games like Fishbanks effectively teach resource management concepts, but calibrating game parameters before classroom deployment remains a challenge. Instructors risk wasting limited class time on poorly balanced games. Additionally, most existing games do not model the full asset lifecycle (ordering, design, construction, operations, retirement) that characterizes real maritime fleet investment decisions. We developed a multiplayer fishing game with two ship types and construction pipeline delays, then pre-calibrated its parameters using seven rule-based agent archetypes across nine synthetic scenarios and 81 parameter combinations. We evaluated each configuration on pedagogical criteria including strategy differentiation, player solvency, and resource crisis visibility. The calibrated game was deployed in a university classroom with 20 students over 25 rounds. The classroom session produced a classic commons-dilemma arc: an initial ordering surge, fleet overshoot to 73% above maximum sustainable yield, and fish stock collapse to 2% of carrying capacity. End-game balances ranged from 30.55 to -2.14 million CNY despite identical starting conditions. The real session tracked synthetic predictions within 10% at key checkpoints, providing initial support for the pre-calibration approach. Full information transparency did not prevent resource depletion, suggesting competitive pressure may outweigh information availability. The square-root density function creates a design tension between realism and pedagogy by delaying depletion signals. Identical instructions produced substantially different outcomes, and anticipated but unpredictable regulatory intervention may have distorted investment behavior. These observations remain hypotheses requiring controlled experiments with structured post-session data collection for validation. Improve writing and visualization; brainstorming on creating rule-based agents