Examining how data becomes information for an upcoming opponent in football

Abstract

As the sport industry witnesses a surge in the type and volume of data-driven decisions, the general question of the process of information development remains: how is data used to develop meaningful information? And does the presence of novel quantitative data sources lend greater objectivity to match analysis? Study 1 examines how 12 football analysts use the same qualitative (video) and quantitative (event and position) data to develop information constituting a typical opponent report for an upcoming match, while Study 2 investigates the agreement between grade evaluations of these opponent reports by numerous professional coaches. Findings of Study 1 through independent-samples t-tests (t(18) = 3.922, p = 0.001) indicate a clear dominance of qualitative video data over quantitative event and position data in all opponent reports. Despite the presence of quantitative data sources, analysts tend to prefer annotated video data. Possible relations to previous experience and familiarity with data, coach–analyst preferences and biases are discussed. Results from Study 2 show extremely weak intra-class correlations (ICC) (r = 0.147; p = 0.011) between different grades awarded to the same video, depicting a clear lack of agreement in what coaches consider a good opponent report. Furthermore, coaches most valued the comprehensibility and relevance of the report. No significant associations were found between use of either data type and better grades. The subjectivity of the coaching process highlighting preferences regarding data validity and negotiations of adopting new key performance indicators (KPIs) is discussed, alongside limitations of the sample as well as the level of coach–analysts involved.