Data-driven analysis of key factors influencing the height of the thrust in artistic swimming

The Thrust is a core technical skill in Artistic Swimming, and its height is a key indicator for evaluating movement quality and difficulty. Traditional training often relies on coaches' subjective experience, lacking objective quantitative analysis. This study aims to employ a data-driven approach to quantitatively analyze the key factors influencing the height of the Thrust. By collecting video data of the Thrust actions and basic physical fitness test data from 12 elite artistic swimmers, key indicators such as thrust height, body angles, and movement timing were extracted using computer vision techniques (e.g., OpenPose), combined with physical data including upper limb power and core stability. Predictive models were constructed using Pearson correlation analysis, multiple linear regression, and Random Forest algorithms, with model performance evaluated through R² values and feature importance scores. The results indicated that upper limb power (contribution rate 42%) and body unfolding angular velocity (contribution rate 31%) are the two most critical factors affecting the height of the Thrust. This study provides an objective and quantitative diagnostic basis for Thrust training, demonstrating the application potential of data-driven methods in enhancing the scientific level of artistic swimming training.