Strategies optimization for rugby team using Genetic Algorithms (GA)

Rugby union is a complex sport where reliance on coaching intuition often leads to predictable gameplay and inconsistent results. To address these inefficiencies, this study develops a strategy optimization prototype utilizing a Genetic Algorithm (GA) to facilitate data-driven tactical decision-making. The system methodology involves pre-processing where irrelevant attributes have been removed, and player datasets are processed to simulate evolutionary phases, including selection, crossover, and mutation. The proposed model categorizes gameplay into Basic, Tactical, and Contingency plays, mapping specific attributes such as player weight and match statistics to optimized roles. Experimental results indicate that the system successfully identifies high-fitness player combinations for specific strategies, such as "Pick and Go," with performance validation achieved through convergence graph analysis. By automating the selection of optimal strategies, this prototype provides coaches with a systematic tool to enhance team adaptability and competitive advantage in high stakes matches.