Journal of E - Technology Zhou Yongming 14 4 2023 https://doi.org/10.6025/jet/2023/14/4/97-104 https://www.dline.info/jet/fulltext/v14n4/jetv14n4_2.pdf This article aims to conduct a sampling survey of basketball sports through data indicator mining methods and provide a reference for the selection and training of basketball players through training-weighted analysis. Firstly, this article collects and analyzes data on basketball players’ physical fitness, technical skills, tactical awareness, and other aspects to identify important data indicators. Next, this article uses a sampling survey method to analyse many basketball player data statistically and understand their performance in various indicators. On this basis, this article adopts the technique of training weighted analysis to determine the weights of different indicators, providing a more scientific basis for the selection and training of athletes.