본문 바로가기
  • Home

The identification of optimal data range for the discrimination between won and lost

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2020, 25(7), pp.103-111
  • DOI : 10.9708/jksci.2020.25.07.103
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : July 13, 2020
  • Accepted : July 29, 2020
  • Published : July 31, 2020

Doryung Han 1 Hyongjun Choi 2

1경기대학교
2단국대학교

Accredited

ABSTRACT

Performance indicators have often investigated and developed in order to identify foundational elements and factors for an enhancement of performance in sports. In order to identify the valid performance indicators it is important that the indicators used within a performance analysis system discriminate between the winning and losing performances within a match (Hughes and Bartlett, 2002). However, the performance indicators proposed in research studies on basketball performance have not been used for real-time analysis and feedback within a coaching context. Such real-time support for the coach and players has been described within research on other sports (Choi et al., 2004; O'Donoghue, 2001; Palmer et al., 1997). Within the process of real-time feedback, the identification of relevant performance indicators that distinguish winning and losing performances should be the first stage of the development of a real-time analysis system. Therefore, this study investigated the differences between winning and losing teams in terms of a set of performance indicators gathered during the analysis of 10 English National Basketball League matches. Winning and losing teams were compared using whole match data (N=10) as well as individual quarters (N=40). A series of Wilcoxon Signed Ranks tests was used to identify the relevant performance indicators that discriminate between winning and losing performers within whole matches and individual quarters. The tests found that 3 point shots made (p<0.05) and Assists (p<0.05) were significantly different between winning and losing teams within matches. However, 2 point shots made (p<0.05), 2 point shots attempted (P<0.05), percentages of 2 point shots scored (p<0.05), 3 point shots made (p<0.05), Defensive Rebounds (p<0.05) and Assists (p<0.05) were significantly different between winning and losing performance within quarters. The analysis task should be based on relevant performance indicators which explain the current performances to performance analysts and coaches. Within a real-time analysis and feedback scenario, this will have the additional benefit of supporting a decision based on immediate performance within the most recent quarter. Consequently, the real-time analysis system would use performance indicators which have the property of construct validity to support the decisions of the coach

Citation status

* References for papers published after 2022 are currently being built.