CyberYuruGu: An Interpretative Gameplay Model Grounded in African Epistemology for Explainable Cybersecurity Reasoning and Human–AI Sense-Making

Abstract

Cybersecurity education and serious games have long been dominated by Western‑centric paradigms, leaving limited space for indigenous epistemologies that emphasise interpretative reasoning. To address this gap, CyberYuruGu is developed in this study as a cybersecurity game inspired by the Dogon Fox tradition. Unlike abductive reasoning or Bayesian inference, the Dogon Fox system encodes uncertainty through symbolic spatial traces, requiring interpretative reconstruction rather than probabilistic prediction. This resonates with Bachelard’s notion of epistemological rupture, where disruption and discontinuity are not obstacles but productive sources of new knowledge. Embedded within a socio‑technical framework informed by constructivist pedagogy and ethno‑informatics, the game enables participants and a large language model (LLM) to collaboratively interpret gameplay data. This design allows them to reconstruct multi‑layered attack scenarios under conditions of uncertainty. Developed in Python using Pygame, OpenGL, and WebGL, CyberYuruGu records gameplay traces as participants place attack and defence tokens, infer adversarial intent, and construct defensive narratives. Results revealed a cognitive bias toward protective reasoning, as defence paths achieved higher cumulative scores (8) than attack paths (5). While participants occasionally formed strong offensive sequences, defensive reasoning remained more consistent and complete. Thus, the Dogon Fox model contributes a relational and cosmological logic of trace interpretation that complements but is not captured by conventional probabilistic or abductive models.