Analysis of Footwork Performance and Heartrate Data for Choking Recognition in College Tennis

In many sports, proper movement and precise footwork signify key components of high quality performance. Conditioning specialists and coaches regularly work with athletes on footwork improvements to increase their chances to reach desired performance levels. However, choking under pressure sensation can make it difficult for athletes to achieve such goals as they may underperform in pressure situations. Proper footwork performance has been deemed essential for success in many sports and choking under pressure can impede performance; an observation can be made that footwork performance and choking under pressure may be related. Additionally, specific heartrate patterns in pressure situations have been established as stress indicators, presenting a potential relationship between heartrate and choking. Therefore, this paper presents a model for choking detection in college tennis athletes based upon footwork performance information and heartrate measurements. A data collection network is designed to capture footwork performance and heartrate measurements using two wearable sensors. There were three distinct footwork performance parameters and heartrate measurements captured from the collected data. Data were collected with five participants completing a total of 13 tennis sessions, eight matches and five practices. Choking occurrences were identified during the tennis matches by six observing tennis coaches. The observations served as measurements for the choking detection and recognition. The captured data reveal that choking moments can be associated with certain footwork performance and heartrate patterns. The data also reveal that choking patterns strongly vary individually. However, results reached with the collected data and the proposed model demonstrate that choking recognition cannot be successfully achieved solely based on the proposed footwork performance and heartrate variables. It is possible that the proposed recognition model and introduced variables may aid with detection of choking moments when combined with other sport specific information. The future plan for the study is to collect data with additional participants and to combine the measured data with other tennis specific variables such as swing speed.