Developing an Artificial Intelligence Model to Predict Team Efficacy Based on Visual Cohesion Indicators: An Applied Study on Iran’s U21 Men’s Volleyball World Champion Team

Paper ID : 1258-SPORTCONGRESS

Authors

MUNTADHAR SAHEB ALNWYNI * KARBALA UNIVERSITY

Abstract

Abstract This study develops an artificial intelligence (AI) model to predict team efficacy among players of Iran’s U21 men’s volleyball world champion team, using visual cohesion indicators extracted from match video analysis. Visual cohesion—the team’s shared ability to perceive positions and movements synchronously without verbal communication—reflects collective intelligence, coordination, and overall performance efficiency under competitive pressure. A quantitative–visual approach supported by AI was employed. Data were extracted using Computer Vision and Social Network Analysis (SNA) techniques to measure (1) gaze direction and visual focus, (2) positional and movement patterns, (3) group visuomotor response speed, and (4) visual cooperation during offensive and defensive transitions. These indicators were processed through Machine Learning models (Random Forest and Neural Networks) to estimate their predictive power on collective efficacy, which was evaluated by certified experts and FIVB coaches. Results revealed a strong positive correlation between visual cohesion and team efficacy, with the AI model achieving over 90% predictive accuracy in high-pressure scenarios. Algorithms such as OpenPose and MediaPipe demonstrated precise temporal and perceptual identification of visual indicators, validating their reliability compared to traditional observational methods. This research contributes to sport psychology by introducing objective, AI-driven measures of team performance and perceptual coordination. It establishes a foundation for integrating computer vision into psychological and behavioral evaluation of team sports. Key Recommendations: (1) Employ visual analysis as a standard methodology for measuring collective perception and communication in sports; (2) Extend research to other team-based games such as football and basketball; (3) Utilize AI and neural networks to deepen understanding of perceptual–social interactions and optimize training programs in elite-level competitions.

Keywords

: Visual Cohesion Team Efficacy Artificial Intelligence (AI) Computer Vision Social Network Analysis (SNA) Machine Learning Neural Networks Collective Performance Sport Psychology Volleyball Iran U21 Men’s National Team

Status: Abstract Accepted (Poster Presentation)