Fair Bracket Generation Through a Modified Fisher–Yates Shuffle: Implementation and Initial Evaluation in the GOHit Platform

Abstract

The fairness and transparency of tournament brackets play a critical role in digital competition platforms, yet many systems still rely on manual or deterministic ordering that can introduce bias. This study implements and evaluates a weighted adaptation of the Fisher–Yates Shuffle algorithm in the GOHit web-based competition management system to address these limitations. The weighted approach divides participants into ranking-based groups prior to shuffling to achieve a balance between randomness and competitive fairness. The algorithm was integrated into a PHP CodeIgniter 4 environment and tested using three ideal bracket sizes (8, 16, and 32 participants) across 30 independent trials each. Results show that the algorithm consistently produced consistent and unique bracket permutations, with no duplication or recurring patterns. Visual and comparative analysis further demonstrate substantial displacement of participant positions when compared to manual sequencing, reducing predictability while maintaining structural fairness. These findings indicate that the weighted Fisher–Yates Shuffle offers an effective, lightweight, and reproducible solution for fair bracket generation in small- to medium-scale digital competitions. The approach also provides a practical foundation for future enhancements in fairness-sensitive scheduling systems.