This paper presents an evaluation system for the effectiveness of the forehand stroke in table tennis using an optimized deep learning (DL) model with cloud computing. The study aims to assess the performance of a bidirectional long short-term memory (BLSTM) model with an improved dragonfly algorithm (IDFOA) in analyzing sensory data from table tennis strokes. The data was collected from 16 players using a low-cost, non-intrusive inertial sensor (BNO055) attached to a table tennis racket. The sensor data includes acceleration, gyroscope, magnetometer, and Euler angle measurements. The data was processed and analyzed to evaluate the quality of the forehand strokes based on five criteria: racket grip angle, forward swing, follow-through, racket movement speed, and overall stroke quality. The BLSTM-IDFOA model was compared with traditional LSTM models, and the results showed that the BLSTM-IDFOA model provided more accurate predictions with lower error rates. The study also discusses the hardware setup, data acquisition, preprocessing, and the proposed methodology for evaluating the effectiveness of the forehand stroke. The results indicate that the BLSTM-IDFOA model is the most effective regression approach for this task. The study contributes to the field of sports analytics by providing a novel approach to evaluate the quality of table tennis strokes using DL and cloud computing. The findings have implications for improving the teaching and training of table tennis players, as well as for developing automated systems for sports performance evaluation.This paper presents an evaluation system for the effectiveness of the forehand stroke in table tennis using an optimized deep learning (DL) model with cloud computing. The study aims to assess the performance of a bidirectional long short-term memory (BLSTM) model with an improved dragonfly algorithm (IDFOA) in analyzing sensory data from table tennis strokes. The data was collected from 16 players using a low-cost, non-intrusive inertial sensor (BNO055) attached to a table tennis racket. The sensor data includes acceleration, gyroscope, magnetometer, and Euler angle measurements. The data was processed and analyzed to evaluate the quality of the forehand strokes based on five criteria: racket grip angle, forward swing, follow-through, racket movement speed, and overall stroke quality. The BLSTM-IDFOA model was compared with traditional LSTM models, and the results showed that the BLSTM-IDFOA model provided more accurate predictions with lower error rates. The study also discusses the hardware setup, data acquisition, preprocessing, and the proposed methodology for evaluating the effectiveness of the forehand stroke. The results indicate that the BLSTM-IDFOA model is the most effective regression approach for this task. The study contributes to the field of sports analytics by providing a novel approach to evaluate the quality of table tennis strokes using DL and cloud computing. The findings have implications for improving the teaching and training of table tennis players, as well as for developing automated systems for sports performance evaluation.