Association between match-related physical activity profiles and playing positions in different tasks: A data driven...

Area of Science:

• Sports Science
• Data Analytics
• Football (Soccer) Performance Analysis

Background:

• Optimizing training intensity in soccer is crucial for player development and performance.
• Understanding the neuromuscular and metabolic demands of different training drills is essential for effective periodization.
• Existing methods may not fully capture the nuances of player-specific intensity requirements across various drills.

Purpose of the Study:

• To propose and validate a data analytics framework for assessing soccer training intensity.
• To analyze the neuromuscular and metabolic characteristics of soccer drills (matches, small-sided games, match-based exercises).
• To compare the intensity profiles of training drills with match demands relative to player positions.

Main Methods:

• Utilized Global Positioning System (GPS) data from 28 semi-professional soccer players over one season.
• Employed a supervised machine-learning approach to identify positional differences in sport-specific drills.
• Applied a non-supervised machine-learning model to profile match neuromuscular and metabolic characteristics.

Main Results:

• Player positions exhibited distinct metabolic and neuromuscular characteristics during matches and match-based exercises, linked to tactical demands.
• These positional differences were not evident in small-sided games.
• The framework successfully evaluated the alignment between training drill stimuli and specific playing position match demands.

Conclusions:

• The proposed data analytics framework effectively differentiates training intensity based on playing positions.
• Match-based exercises better reflect positional tactical demands compared to small-sided games.
• Practitioners can use this framework to ensure training drills provide appropriate stimuli mirroring match requirements for individual players.