Water resources are crucial for national economic growth, but industrial development has led to severe pollution, sparking conflicts between government and enterprises over their responsibilities. Addressing these water pollution conflicts is urgent. The Graph Model for Conflict Resolution (GMCR) offers a simple and logical approach to analyzing such conflicts with minimal data. However, current models overlook decision makers’ (DMs) heterogeneity, bounded rationality, and the theoretical foundation for preference ranking. To overcome these limitations, this study integrates hybrid information to account for DMs’ diverse knowledge and experience. Prospect theory is employed to capture DMs’ bounded rationality, enhancing the accuracy of decision outcomes. The PROMETHEE method is introduced for preference ranking during GMCR modeling, guided by multi-attribute decision-making (MADM) theory. Additionally, the four basic stability concepts are redefined and proven, expanding GMCR’s theoretical applications. Finally, a case study of the Huangshui River water conflict and the comparative analysis demonstrate the model’s validity and advantages