What is it about?
When an esports team loses a match, understanding what went wrong is not always straightforward. Was the problem a poor strategy, an incorrect tactical decision, a breakdown in teamwork, or simply a mistake in executing an action? Our paper introduces the Esports Performance Screening (EPS) framework, a new approach designed to help coaches and players systematically analyse performance and identify the underlying causes of mistakes. Drawing on established concepts from traditional sports coaching and game research, the framework examines performance across five interconnected levels: strategy, tactics, tasks, actions and operations. It also explains how players and teams create advantages over their opponents, and how these advantages influence the outcome of competitive situations. The framework can be used both to review performances after a match and to guide decision-making during gameplay. Although illustrated through examples from League of Legends, it is designed to be applicable across different invasion-based esports, including games such as Counter-Strike and Rocket League.
Featured Image
Why is it important?
Despite the rapid growth of competitive esports, coaches and educators have relatively few structured resources to help them understand and develop player performance. Training often relies on repeated gameplay, reviewing match recordings and the personal experience of coaches and players. While these approaches can be valuable, they do not always help identify why particular mistakes occur or what players should practise to improve. A key contribution of the EPS framework is its ability to distinguish between errors in decision-making and errors in execution. This matters because different problems require different solutions. A player who makes the correct decision but fails to execute it needs different training from a player who executes an inappropriate decision perfectly. By providing a shared language for analysing performance, the EPS framework offers a way for coaches, players and educators to communicate more clearly, identify specific development needs and design more targeted training activities. Importantly, because the framework is not tied to a single game, it offers opportunities for developing coaching resources and performance analysis tools that can be applied across multiple esports titles. Future research will investigate how effectively the framework supports coaching, learning and performance in practice.
Perspectives
One of the challenges I have encountered throughout my research into esports performance is the gap between our understanding of what contributes to successful performance and the practical resources available to help coaches and players develop those capabilities. Traditional sports have spent decades developing coaching methods, performance analysis frameworks and approaches to teaching tactical decision-making. Yet many esports coaches are still expected to develop their own methods, often with limited formal education or access to evidence-informed resources. This paper represents an important step towards addressing that gap. By bringing together expertise from sport psychology, coaching science and human-computer interaction, we wanted to develop a framework that could help coaches and players better understand the complexity of esports performance. My hope is that EPS will provide a foundation for future research, practical coaching resources and educational tools that make it easier for players to understand their performance, regulate their behaviour and work more effectively with their teammates. Ultimately, I would like to see esports coaching develop beyond simply identifying mistakes towards helping players understand why those mistakes occur and how they can learn from them.
Michael Trotter
Halmstad University
Read the Original
This page is a summary of: Diagnosing Performance in Invasion-Based Esports: The Esports Performance Screening (EPS) Framework, Proceedings of the ACM on Human-Computer Interaction, November 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3831337.
You can read the full text:
Contributors
The following have contributed to this page







