Implementation of Stable Pairing Algorithms for Optimizing Educational Games: A Computational and Pedagogical Perspective

Abstract

The Gale-Shapley algorithm solves the problem of stable pair formation across various fields including economics, labor markets, biology, computer science, and physics. This study modifies the algorithm to use a single list of participants and calculates compatibility scores using Jaccard similarity coefficients from students' proficiency tests and academic performance. We compared the effectiveness of this modified algorithm by evaluating two groups of students engaged in digital educational games: an experimental group matched by the modified algorithm and a randomly matched control group. The results show that the modified algorithm forms pairs with superior compatibility, consistent performance, and balanced competition. These findings suggest integrating the Gale-Shapley algorithm into educational technologies can enhance learning environments. The results significantly impact educational practices indicating that systematic peer training can improve collaboration, competition, and student engagement.