A NEW PERSPECTIVE ON ANALYSIS AND EVALUATION OF BASKETBALL TEAM PERFORMANCE AND PLAYERS USING CLUSTER ANALYSIS
Basketball is a game that demands a complete list of criteria to be enumerated in order to fully comprehend the game and assess strategy and decisions while limiting unpredictability. Its success heavily depends on identifying the players' playing styles both offensive and defensive, preferred and assembling the ideal team. This paper reviews a new perspective on analysis and evaluation of basketball team performance and players using both h-cluster and k-cluster techniques. Though there are known techniques for analysis and player performance evaluation which have been developed overtime, several other techniques have been proposed. However, cluster analysis hasn’t been much explored in player performance. This study used both the k-cluster and the h-cluster to assess and analyze the basketball team's and players' performance. There are 23 factors in the NBA player statistics from the 2017–2018 regular season that describe a team's overall offensive, defensive, offensive rebounding, and general capabilities. After standardizing the variables, two clustering techniques—hierarchical clustering and K-means clustering—were used to conduct the cluster analysis. K-means clustering produced superior outcomes than h-cluster. According to the study, participants in cluster 3 spend a lot more time than those in the other 3 clusters. The cluster 2 players spend the least amount of time. While players in cluster 2 scored very little throughout the 2017–2018 season, players in cluster 3 end up with a lot more points. For offensive rebound (OREB), players in cluster 4 had higher values compared to other clusters. While cluster 4 had the highest blocks meaning players in cluster 4 are mostly defensive players. An overall analysis showed in details which player performed better generally and where to improve for next season.
Authors : Ajinaja, M.O., Abiona, A.A. and Adewuyi, J.A.
Category : Open Access Volume (Issue) : 8(2) Date Uploaded : 28th January 2023