Interunit Reliability of Firstbeat Sport Sensors as Accelerometer-Based Tracking Devices in Basketball

Purpose: The aim of this study was to assess the interunit reliability of the movement load (ML) derived from the Firstbeat Sport sensors during basketball training sessions. Methods: Eight professional male basketball players (age 25.5 [4.7] y, stature 199 [8] cm, body mass 96.6 [7.8] kg, body fat 10.0% [2.2%]) were recruited to participate in this study. A total of 50 individual training sessions were recorded on one or multiple occasions (range: 1-21 individual sessions for each player) while players were wearing 2 Firstbeat Sport sensors firmly affixed to their chest roughly at the base of the sternum via textile straps. ML was calculated via Firstbeat Sports software for the entire basketball training sessions and for the following basketball-specific activity types: 5-on-5 full court (5v5, n = 30), 5-on-5 half-court plus 2 courts back and forth (5v5 1/2 + 2, n = 26), and 5-on-5 half-court (5v5 1/2, n = 26). Results: Linear mixed models revealed no statistical differences in ML data recorded using different sensors during the entire training sessions and the selected basketball-specific activity types (all P > .05, trivial). The coefficient of variation calculated as percentage and intraclass correlation coefficient calculated for the entire training sessions and basketball-specific activity types ranged from 2.51% to 5.97% and from .98 to 1.00, respectively. Conclusions: The ML derived from the Firstbeat Sport sensor showed acceptable interunit reliability when considering the full training sessions and basketball-specific activities in professional basketball. Overall, basketball practitioners can use Firstbeat Sport sensors interchangeably to monitor the external load of professional basketball players during training sessions.