چكيده به لاتين
In this research, the results of badminton doubles matches are predicted using mixed effect models to help improve player performance and increase win probability for future matches. First, the variables which affect the win probability of a doubles match were identified. Then, data was gathered from youth girls badminton matches with verification and supervision by the national team coach. Considering the conditions and the variables the mixed effect model was chosen for this research. Using this model, the variable coefficients were extracted to estimate win probability in addition to identifying effective variables. According to the results, the amount of time the players spend training each week had the most positive impact and the body mass index (BMI) had the most negative impact on win probability amongst the constant variables. The coach and the players can increase training time and implement specific training routines to control BMI in order to improve win probability. Amongst the random variables, game intelligence had the most positive impact and nervousness had the most negative impact on win probability. For game intelligence, the coach can use techniques during training and properly analyze games in order to help the players and increase win probability. But in regards to nervousness, there are both controllable and uncontrollable factors involved and the players should try and take control of these factors themselves.
After extracting variable coefficients, win and loss probability of matches were predicted and the model had an accuracy of 90% compared to the real results of matches.
