AUTHOR=Link Daniel , Weber Marcus , Linke Daniel , Lames Martin 

TITLE=Can Positioning Systems Replace Timing Gates for Measuring Sprint Time in Ice Hockey?

JOURNAL=Frontiers in Physiology

VOLUME=9

YEAR=2019

URL=https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2018.01882

DOI=10.3389/fphys.2018.01882

ISSN=1664-042X

ABSTRACT=<p>This study explores whether positioning systems are a viable alternative to timing gates when it comes to measuring sprint times in ice hockey. We compared the results of a single-beam timing gate (Brower Timing) with the results of the Iceberg optical positioning system (<italic>Optical</italic>) and two radio-based positioning systems provided by InMotio (<italic>Radio 1</italic>) and Kinexon (<italic>Radio 2</italic>). The testing protocol consisted of two 40 m linear sprints, where we measured sprint times for a 11 m subsection (<italic>Linear Sprint 11</italic>), and a shuttle run (<italic>Shuttle Total</italic>), including five 14 m sprints. The exercises were performed by six top-level U19 field players in regular ice hockey equipment on ice. We quantified the difference between measured sprint times e.g., by Mean Absolute Error (MAE) (s) and Intra Class Correlation (ICC). The usefulness of positioning systems was evaluated by using a Coefficient of Usefulness (CU), which was defined as the quotient of the Smallest Worthwhile Change (SWC) divided by the Typical Error (both in s). Results showed that radio-based systems had a higher accuracy compared to the optical system. This concerned Linear Sprint 11 (MAE<sub>Optical</sub> = 0.16, MAE<sub>Radio1</sub> = 0.01, MAE<sub>Radio2</sub> = 0.01, ICC<sub>Optical</sub> = 0.38, ICC<sub>Radio1</sub> = 0.98, ICC<sub>Radio2</sub> = 0.99) as well as Shuttle Total (MAE<sub>Optical</sub> = 0.07, MAE<sub>Radio1</sub> = 0.02, MAE<sub>Radio2</sub> = 0.02, ICC<sub>Optical</sub> = 0.99; ICC<sub>Radio1</sub> = 1.0, ICC<sub>Radio2</sub> = 1.0). In Shuttle Total, all systems were able to measure a SWC of 0.10 s with a probability of &gt;99% in a single trial (CU<sub>Optical</sub> = 4.6, CU<sub>Radio1</sub> = 6.5, CU<sub>Radio2</sub> = 5.1). In Linear Sprint 11 an SWC of 0.01 s might have been masked or erroneously detected where there were none due to measurement noise (CU<sub>Optical</sub> = 0.6, CU<sub>Radio1</sub> = 1.0, CU<sub>Radio2</sub> = 1.0). Similar results were found for the turning subsection of the shuttle run (CU<sub>Optical</sub> = 0.6, CU<sub>Radio1</sub> = 0.5, CU<sub>Radio2</sub> = 0.5). All systems were able to detect an SWC higher than 0.04 s with a probability of at least 75%. We conclude that the tested positioning systems may in fact offer a workable alternative to timing gates for measuring sprints times in ice hockey over long distances like shuttle runs. Limitations occur when testing changes/differences in performance over very short distances like an 11 m sprint, or when intermediate times are taken immediately after considerable changes of direction or speed.</p>