A method to inform team sport training activity duration with change point analysis

Area of Science:

• Sports Science
• Performance Analysis
• Biomechanical Analysis

Background:

• Training duration is critical for enhancing athlete skills and physical qualities.
• Balancing skill development and physiological targets is a key challenge in sports training design.

Purpose of the Study:

• To demonstrate change point time-series analyses for informing training activity duration in Australian Football.
• To identify optimal durations for training activities based on player behavior metrics.

Main Methods:

• Utilized change point time-series analyses on five player behavior features: disposal frequency, efficiency, pressure, possession time, and movement velocity.
• Employed both univariate and multivariate analyses to identify specific and overall moments of change in player behavior.
• Evaluated six training activity iterations to find common change point locations.

Main Results:

• Identified specific moments of change in player behavior that can inform minimum or maximum activity durations.
• Univariate analysis pinpointed changes for individual metrics, while multivariate analysis identified a single overall change point.
• Common change point locations were found between 196 and 252 seconds across training activity iterations, indicating significant behavioral shifts.

Conclusions:

• Change point analysis provides valuable insights into athlete behavior during training activities.
• These methods can guide practitioners in making data-driven decisions about training activity durations.
• Optimizing training duration based on identified behavioral changes can enhance skill development and physiological outcomes.