Research Article: What’s in a game? A systems approach to enhancing performance analysis in football

Date Published: February 17, 2017

Publisher: Public Library of Science

Author(s): Scott McLean, Paul M. Salmon, Adam D. Gorman, Gemma J. M. Read, Colin Solomon, Jaime Sampaio.


Performance analysis (PA) in football is considered to be an integral component of understanding the requirements for optimal performance. Despite vast amounts of research in this area key gaps remain, including what comprises PA in football, and methods to minimise research-practitioner gaps. The aim of this study was to develop a model of the football match system in order to better describe and understand the components of football performance. Such a model could inform the design of new PA methods.

Eight elite level football Subject Method Experts (SME’s) participated in two workshops to develop a systems model of the football match system. The model was developed using a first-of-its-kind application of Cognitive Work Analysis (CWA) in football. CWA has been used in many other non-sporting domains to analyse and understand complex systems.

Using CWA, a model of the football match ‘system’ was developed. The model enabled identification of several PA measures not currently utilised, including communication between team members, adaptability of teams, playing at the appropriate tempo, as well as attacking and defending related measures.

The results indicate that football is characteristic of a complex sociotechnical system, and revealed potential new and unique PA measures regarded as important by SME’s, yet not currently measured. Importantly, these results have identified a gap between the current PA research and the information that is meaningful to football coaches and practitioners.

Partial Text

Since the 1960s, football researchers have investigated the physiological, technical, and tactical components of football to determine the key performance indicators (KPIs) that predict successful performance [1, 2]. In more recent times, advances in computer and video aided match analysis systems, as well as increased global visibility and reach, has led to a substantial increase in football performance analysis (PA) literature and methods [1, 3, 4]. Despite more than five decades of research in this area, current football PA methods remain beset by various issues, including a lack of standardised operational definitions, a lack of match context, and the discrete measurement of isolated variables [1, 4]. Furthermore, previous PA research has had only a minimal impact on practice [5, 6], suggesting a lack of transferability of research outputs to practice [1, 5]. One reason for this is that football match performance has not yet been described in its entirety. Accordingly, there remains a substantial number of features that need to be defined and measured in football PA to ensure that the data are of benefit to practitioners [1, 5].

The aim of this article was to present a first-of-its-kind WDA of the ‘football match system’ in order to examine the state of the art in football PA. The WDA was developed based on two workshops involving highly experienced football SME’s. The study was designed to determine the composition of high performance football, and to then use this to identify key knowledge gaps within the PA literature.

The current study indicated that a different approach is needed to advance the current approaches used in PA for football. In particular, the novel measures identified in the current study require new measurement techniques, and the complexity engendered during football matches requires an integrated approach that considers multiple aspects of performance. The challenge for researchers is to develop and test these new measures to move PA research forward and to better align PA with the needs of coaches. In our opinion, to align research and practice more closely, the integration of sport scientists and football experts is required to fully understand PA in football. It appears that current PA measures are driven by researcher-based approaches that are largely impractical and unusable in practice. Through demonstrating that football is a complex system that requires new ideas and potentially more sophisticated, yet useable, measurement techniques, we hope that this article provides the impetus to bridge this research-practice gap.