Punch Trackers: Correct Recognition Depends on Punch Type and Training Experience.

Area of Science:

• Sports Science
• Biomechanics
• Wearable Technology

Background:

• Punch tracking technology is increasingly used in combat sports for performance analysis.
• Understanding the accuracy of different punch trackers (PT) is crucial for reliable data collection.
• Trained (TR) and untrained (UNTR) individuals may exhibit different punching mechanics, potentially affecting tracker accuracy.

Purpose of the Study:

• To evaluate the accuracy of three distinct punch trackers (Corner PT, Everlast PT, Hykso PT) in recognizing specific punch types.
• To compare the performance of these PTs between trained and untrained punchers.
• To determine if punch sequence or type influences recognition accuracy.

Main Methods:

• Subjects (n=10 TR, n=11 UNTR) performed standardized punch combinations.
• Data from Corner (CPT), Everlast (EPT), and Hykso (HPT) were compared against video analysis.
• Descriptive statistics and multilevel modeling were employed for data analysis.

Main Results:

• All tested PTs (CPT, EPT, HPT) demonstrated higher accuracy in trained individuals compared to untrained (p = 0.007).
• Straight punches were recognized more accurately than hooks and uppercuts across all devices (p < 0.001).
• CPT and HPT showed recognition accuracy dependent on punch order, with earlier punches identified better; EPT data export limitations prevented this analysis.

Conclusions:

• Corner PT and Hykso PT are viable options for tracking punch count and type in both trained and untrained individuals.
• Punch tracker accuracy is influenced by the puncher's training status and the specific punch type.
• Further research is needed to fully assess the EPT's capabilities due to data export restrictions.