HALM: A Cybersecurity-Themed Serious Game with Hybrid Learning Analytics for Adaptive Programming Education at Scale

The digital revolution has fundamentally reshaped the educational landscape, giving rise to a generation of "digital natives" who interact with technology intuitively. While mobile learning and serious games have gained traction, scaling these interventions to provide personalized, adaptive feedback in large-scale scenarios remains a challenge. Existing frameworks often lack a rigorous connection between in-game metrics and formal pedagogical models like TPACK and SAMR. This paper introduces the Hybrid Adaptive Learning Model (HALM), a comprehensive mathematical and pedagogical framework designed to predict and enhance student performance through real-time gameplay analytics. The model is applied within "Hackathon," a cybersecurity-themed serious game that challenges students to "hack" educational content to learn procedural programming. The study analyzes data from a large-scale deployment (N = 4, 338) of the Hackathon game, which functioned as a national competition and international educational program. The diverse participant profile included primary and secondary students spanning from the 9th to 16th grade levels. HALM integrates a hybrid vector of individual performance indicators (accuracy, response time, in-game score) and social engagement metrics to calculate a dynamic Learning Progress Gauge (LPG). Crucially, the model addresses the high computational complexity inherent in large-scale educational MMORPGs, reducing algorithmic complexity from O(n × z) to O(ϕ) through a vectorized optimization strategy. We cross-referenced students' in-game completion rates with their midterm scores in a formal "Procedural Programming" course. The analysis reveals a statistically significant positive correlation between game progression and academic achievement. Specifically, we identified a critical "mastery threshold" at 70% game completion, where the probability of achieving top-tier academic scores increases dramatically. Conversely, 35.6% of failing students had completed less than half of the game activities. The findings demonstrate that HALM effectively bridges the gap between the "digital world" of serious games and the "real impact" of classroom performance. By leveraging hybrid analytics, educators can identify at-risk students early and provide scalable, adaptive interventions.