Barriers to coach decision-making during Australian football matches and how it can be supported by artificial intelligence

Abstract

Elite sport coaches are often expected to make high-stakes decisions in time-constrained environments, such as during matches. The quality or efficiency of these decisions is bound by context-specific constraints on their decision-making. Artificial intelligence-based tools could alleviate some of these constraints. In Australian football, it is not yet known what these context-specific constraints are, nor is there any understanding of the thoughts and perception of elite Australian football coaches towards the use of artificial intelligence to support their decision-making. This study is the first to explore these issues in the context of elite Australian football and aims to inform the effective design and integration of tools to support the in-match decision-making of coaches in a manner that is welcomed by coaches. Using semi-structured interviews with elite Australian football coaches (n = 6) and reflexive thematic analysis, two domains were explored: ‘barriers to decision-making’ and ‘thoughts and perceptions of a decision support system (DSS)’. Coaches discussed seven ‘cognitive’ and ‘environmental’ barriers, some of which may be addressed using artificial intelligence. Meanwhile, four themes were generated for the second domain, namely ‘willingness to adopt’, ‘expectation’, ‘concern and criticism’, and ‘DSS role’. The findings enable practitioners to design tools that are intended to address some of these barriers, while also providing insights into important considerations for the implementation and design of these tools that may increase the likelihood of achieving the endorsement of elite coaches and their support staff.