We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
FORECASTING IN SPORT.
- Authors
Yiannakis, Andrew; Selby, Michaäl J. P.; Douvis, John; Joon Young Han
- Abstract
The rationalization of modern sport has made it possible for social scientists to predict the results of sports events with greater accuracy. In this study we applied multivariate time series analysis to determine the degree to which soccer results could be predicted with three teams in the English Premier League. Success was based on the model's ability to predict the outcome for each of three dependent binary variables; that is, to win, to lose or to draw in the last 10 games of the season. Multivariate ARIMA correctly predicted the outcome with a success rate of nine out of 10 for Winning, eight out of 10 for Losing and nine out of 10 for Drawing. A mix of both shared and new variables in different sets of interactions help predict Winning, Losing and Drawing. A theory of team empowerment is proposed to better explain the utility of the input variables in predicting game outcome. The authors also suggest that multivariate time series analysis may hold promise as an effective forecasting tool in the sociological analysis of sport.
- Subjects
SPORTSCASTERS; STATISTICAL analysis in sports; SPORTS; SPORTS journalism; SPORTS teams; ATHLETES; SOCCER; FORECASTING; MASS media &; sports
- Publication
International Review for the Sociology of Sport, 2006, Vol 41, Issue 1, p89
- ISSN
1012-6902
- Publication type
Article
- DOI
10.1177/1012690206063508