Inter-unit reliability of microtechnology devices

Subjects: Three trained male subjects.

Procedures: To evaluate the inter-unit reliability of GPS-based and LPS-based metrics, this study adapted the methodology of Delaney et al (1). Twenty devices were utilized: 10 Hz EVO devices from GPSports (Canberra, Australia, n=5), 10 Hz S5 devices from Catapult (Melbourne, Australia, n=5), 10 Hz T6 devices from Catapult (Melbourne, Australia, n=5), and 10 Hz Apex devices from StatSports (County Down, Northern Ireland, n=5). These devices were secured to a sprint sled with each device tilted at a 38° angle towards the sky to ensure equal exposure of the embedded antenna, simulating the average angle of a device worn on an athlete’s back. Each device was positioned at least 2cm apart on the sled, which was attached to a harness secured to the waist of the participants.

Four 8-minute team sport simulations, consisting of alternating periods of walking, jogging, running, and accelerations/decelerations, were conducted. Participants received audio cues throughout each simulation instructing them on which movement to perform.

Analysis: After data collection, all devices were downloaded and trimmed using proprietary software: GPSports (GPSports, Canberra, Australia), Openfield (Catapult, Melbourne, Australia), and StatSports Apex (StatSports, County Down, Northern Ireland). The following variables were analyzed: total distance (m), maximum speed (km/h), and distance within discrete velocity bands: Zone 1 (<5.4 km/h), Zone 2 (5.4 – 10.8 km/h), Zone 3 (10.8 – 14.4 km/h), and Zone 4 (> 14.4 km/h).

Inter-unit reliability was assessed by calculating the coefficient of variation (CV) to determine the agreement of measures between devices. The CV was categorized as good (CV <5%), moderate (CV = 5-10%), or poor (CV >10%).


As shown in Table 1, the reliability of all commonly utilized GPS- and LPS-based metrics was measured as good (CV <5%) across all devices.

View this research