Menu

HomeHow it worksContact UsContact Us

Agreement and reliability of force-plate, smartphone-based, and wearable-sensor countermovement jump assessments for neuromuscular monitoring in male professional volleyball players

Bartosz Wilczyński, Tymon Środa, Mateusz Sikorski, Mariola Gepfert, Jakub Jarosz, Katarzyna Zorena · Frontiers in Physiology · 2026

Researchers had 20 professional male volleyball players perform two-leg and single-leg countermovement jumps measured at the same time by force plates (VALD), a smartphone app (MyJump Lab 3), and a wearable sensor (Baiobit), then repeated testing 12 days later. The smartphone app read jump height about 1.9-4.2 cm higher than the force plate's impulse-momentum method (a calculation difference, not a timing error), the wearable had bigger individual variability, and the force plate was most reliable for two-leg jumps (ICC 0.929, smallest detectable change ~2.8 cm) while single-leg jumps were less reliable overall.
Takeaway: Stick with one device and one jump-height calculation method for tracking, and only treat changes larger than that setup's error margin (e.g., ~2.8 cm on force plates) as real.
Abstract (source)

Background: Countermovement jump testing is widely used to infer lower-limb neuromuscular performance, readiness, fatigue, and adaptation in applied exercise physiology. However, commercially available force-plate, inertial-sensor, and smartphone-based systems use different signal sources and calculation

Methods: , which may affect the physiological interpretation of jump-performance changes.

Objective: To evaluate the agreement and test–retest reliability of bilateral and single-leg countermovement jump outcomes obtained from force plates, an inertial measurement unit, and a smartphone application in male volleyball players.

Methods Twenty professional male volleyball players were enrolled. Bilateral and left/right single-leg countermovement jump trials were recorded concurrently using VALD ForceDecks, MyJump Lab 3, and Baiobit. VALD impulse–momentum-derived jump height was used as the reference outcome. Agreement analyses compared MyJump Lab 3 and Baiobit with VALD using bias, limits of agreement, ICC(2,1), SEM, and MDC 95 . Test–retest reliability was assessed between two sessions separated by 12 days. Sensitivity analyses examined VALD flight-time-derived height, MyJump flight time, best-trial values, and strict quality-control exclusions.

Results: MyJump Lab 3 systematically overestimated VALD impulse–momentum jump height by +1.88 to +4.16 cm. This discrepancy was markedly attenuated when VALD flight-time-derived height was used as the reference, while MyJump flight-time bias was negligible, indicating a calculation-

Method: effect rather than temporal event-detection error. Baiobit showed smaller mean bias in some single-leg conditions but poorer individual-level agreement, with wider limits of agreement and less consistent absolute-agreement ICCs. VALD showed the strongest bilateral CMJ reliability (ICC = 0.929; MDC 95 = 2.79 cm), whereas single-leg reliability was generally lower.

Conclusion: Commercially available jump-testing systems can lead to

method-dependent differences in the interpretation of neuromuscular performance. Device-, task-, and metric-specific measurement error should be considered before using jump-height changes to infer readiness, fatigue, or adaptation in professional male volleyball players.

Primary studyOpen accessInjury Prevention & Rehab
Read the original