Action-Evaluator: A Visualization Approach for Player Action Evaluation in Soccer.

Area of Science:

• Sports Analytics
• Human-Computer Interaction
• Data Visualization

Background:

• Player action evaluation is crucial for soccer performance analysis and improving match outcomes.
• Existing methods lack integration of complex contextual information like team tactics and player locations.

Purpose of the Study:

• To address limitations in current soccer player action evaluation.
• To develop a system that supports comprehensive analysis by integrating action scores with match context.

Main Methods:

• Collaborated with soccer analysts and coaches to define domain problems.
• Designed a tailored visualization integrating action choice, tactics, and player locations.
• Developed the Action-Evaluator visual analytics system.

Main Results:

• The Action-Evaluator system facilitates efficient player analysis and investigation.
• Analysts can understand player performance across various match situations.
• The system provides insights for improving player action choices.

Conclusions:

• The Action-Evaluator system offers a comprehensive approach to player action evaluation in soccer.
• It effectively integrates complex contextual data for deeper performance insights.
• Demonstrated usefulness through case studies and expert interviews.