FROM TOKYO TO LOS ANGELES: MODELING OLYMPIC OUTCOMES IN WOMEN'S ARTISTIC GYMNASTICS

Performance in women's artistic gymnastics is evaluated through difficulty (D) and execution (E) scores across four apparatuses: Vault, Uneven Bars, Balance Beam, and Floor Exercise. Understanding how these scores predict overall success can guide athlete development and Olympic preparation. This study aimed to identify key predictors of all-around performance and project future trends by analyzing official Olympic data. Final scores from the all-around competitions at the Tokyo 2020 (n = 80) and Paris 2024 (n = 58) Olympic Games were used to construct multiple linear regression models. Apparatus-specific D and E scores served as independent variables to predict total scores. Separate models were developed for each Olympic cycle and for the combined dataset. All models demonstrated excellent predictive strength (R2 > 0.998). Among the predictors, E scores-particularly on Uneven Bars and Balance Beam- showed the strongest contributions to total scores (beta > 0.25, p < .001). The combined model exhibited strong internal consistency and minimal multicollinearity, supporting its robustness and generalizability. Trend-based analysis indicated a continued increase in average total scores between 2020 and 2024, suggesting further gains toward the 2028 Los Angeles Games. These findings underscore the critical role of execution quality, especially on technical apparatuses, and support the strategic use of longitudinal performance modeling in elite gymnastics training. Coaches and sport scientists may benefit from focusing on apparatus-specific E score optimization to enhance Olympic performance outcomes.