Repeatability and Mixed-Model Predictors of Tennis Serve Velocity

Authors

• Zoe Weidley, University of LouisvilleFollow
• Kaleigh Boyd, University of LouisvilleFollow
• Chad Workman, University of LouisvilleFollow
• Tristen Cooper Schiafo, University of LouisvilleFollow
• Liz Dupree, University of LouisvilleFollow
• Ernest Rimer, University of Louisville HealthFollow
• Andreas Stamatis, University of LouisvilleFollow

Abstract

Serve velocity is a key performance indicator (KPI) in tennis. Single-test measurements can be noisy, especially in small samples, which can weaken conclusions about what relates to KPIs. Repeated trials allow us to quantify repeatability and model KPIs without treating trials as independent. PURPOSE: To estimate repeatability of key tests and examine predictors of serve velocity while accounting for repeated serves. METHODS: Eleven (5M, 6F) NCAA Division I collegiate players completed repeated trials of serve velocity (Pocket Radar Gun; 5 serves), counter movement jump (CMJ; 3 trials each for Output and Hawkin), range of motion tests (Output; 3–6 trials), and medicine-ball rotational scoop throws (Output; 5–7 trials), plus body mass and sex. Repeatability was estimated using intraclass correlation (ICC) from random-intercept models, for single trials (ICC₁) and for the mean of trials (ICC_mean). Serve-trial velocity (55 serves) was modeled with a mixed-effects model (random intercept for athlete) and fixed effects of CMJ Output height, medicine-ball rotational scoop throw peak velocity, sex, and body mass. Days between testing sessions (days-gap) was examined as a sensitivity covariate. RESULTS: Single-trial repeatability was high for serve velocity (ICC₁=0.93) and CMJ measures (Output ICC₁=0.97; Hawkin ICC₁=0.99), and moderate for torso rotation (ICC₁=0.81) and medicine-ball rotational scoop throw peak velocity (ICC₁=0.75). Using participant means improved precision (ICC_mean range: 0.94–0.99). In the mixed model, fixed predictors explained 0.68 of serve-trial variance (marginal R²=0.67), and most variance was between athletes (conditional R²=0.93). CMJ Output height showed the strongest positive association with serve velocity; adding days-gap did not meaningfully change results. CONCLUSION: Repeated trials produced highly repeatable serve and CMJ measures and supported mixed-model analysis of serve velocity. Power-related tests (CMJ and medicine-ball rotational scoop throw peak velocity) showed the most consistent relationships with serve velocity in this dataset.

Recommended Citation

Weidley, Zoe; Boyd, Kaleigh; Workman, Chad; Schiafo, Tristen Cooper; Dupree, Liz; Rimer, Ernest; and Stamatis, Andreas (2026) "Repeatability and Mixed-Model Predictors of Tennis Serve Velocity," International Journal of Exercise Science: Conference Proceedings: Vol. 15: Iss. 8, Article 27.
Available at: https://digitalcommons.wku.edu/ijesab/vol15/iss8/27