AGIR: a theory-based framework to evaluate impacts of game-based interventions for sustainability transitions

Game-based interventions (GBI) increasingly embed serious games within broader transformation processes addressing sustainability challenges. Yet their evaluation remains largely session-centred and learning-oriented, and says little about how in-game dynamics translate into shifts in real-world practices. This gap matters most for GBI that aim to shape behaviours, coordination mechanisms and institutional trajectories over long time horizons. This article proposes the framework for Assessing Game-based Interventions in Real-world contexts (AGIR), a theory-based framework for assessing multi-level change processes in GBI. Drawing on programme evaluation theory, behavioural sciences and serious games research, AGIR organises evaluation into six levels (L0–L5), from game design and in-session dynamics (L0–L1) through learning (L2) and behavioural change (L3) to practice change (L4) and transformative systemic impacts (L5). The framework makes explicit the mechanisms, contextual factors and temporal pathways that link intervention design to sustainability outcomes. Its contribution is threefold. It strengthens theoretical foundations by linking design choices with evaluation concepts. It provides a common reference for comparing interventions across contexts via standardised levels and hypothesised mechanisms. It also functions as a formative tool supporting ex ante hypothesis formulation, in itinere adjustments and ex post impact analysis. By shifting evaluation from end-point outcome measurement toward a process-oriented understanding of contribution, AGIR answers long-standing calls for rigorous yet practical frameworks suited to complex, action-oriented interventions. Though resource-intensive, it offers practitioners and researchers a structured way to assess whether and how GBI contribute to sustainable change.