| Abstract | Computerised notational analysis is now widely used in the sports industry. Most, if not all, professional sports teams or individuals use some form of video analysis through a computer and software packages. However there is little research into the efficiency of match analysis systems, furthermore, the areas of human-computer interaction and artificial intelligence in match analysis systems have also been neglected.
Human-computer interaction is the study of the relationship, which exists between human users and the computer systems they use in the performance of their various tasks (Faulkner 1998). Whilst there are many definitions of artificial intelligence, it is first important to understand intelligence. There are very many definitions of intelligence; Negnevitsky (2001) defined intelligence as, ‘The ability to learn and understand, to solve problems and make decisions. The definition offered by Lapham and Bartlett (1995) was, ‘The faculty of understanding’; ‘the action or process of understanding’. Negnevitsky considered artificial intelligence to be a science that aims, ‘To make machines do things that would require intelligence if done by humans’. Lapham and Bartlett expressed the concept as, ‘The capacity of a machine to simulate or surpass intelligent human behaviour’.
This paper therefore investigates and discusses the efficiency and ergonomics of selected squash match analysis systems. The three systems analysed were Focus X2 manual system using a mouse, Focus X2 Voice Interactive system and the SWEAT (Murray and Hughes, 2001) system using keyboard data entry. The three systems were analysed in real time match analysis and lapsed time analysis using winner and error analysis. 4 matches were analysised in lapsed time and real time for each analysis system. Whilst the analysis procedure was being completed, video recording of the data entry were recorded. The study investigated data inputs per minute and analysis time, in lapsed time analysis. Whereas in real time analysis the paper examined the analysis times, errors made, error corrections, error correction times and total analysis time. A training study was conducted prior to the efficiency analysis to enable the researcher to gain sufficient learning of systems and prevent bias. A percentage difference calculation stated by Hughes et al. (2004) was used to perform an intra-operator reliability investigation in real time analysis and lapsed time analysis, overall highest errors being 3.9%, which were deemed satisfactory (Howells, 2006).
From the results, it found that the Focus X2 manual system was the most efficient in both lapsed time and real time analysis. The efficiency profiles of the analysis highlight the strengths and weakness of the human computer interaction of the analysis systems. The findings of the study also concluded that from the three-match analysis systems there were significant ergonomic characteristics that affected the operator in the case of repetitive injuries. Furthermore, discussions into the development of technology to enhance the efficiency and ergonomic characteristics of match analysis software and computers. It was found from the study that more research into human computer interaction is required and also to research the use of artificial intelligence and the application to performance analysis. |
|---|