The statistical analysis of athletics data: an investigation of the current scoring system for the decathlon

Woolf, A. (2008) The statistical analysis of athletics data: an investigation of the current scoring system for the decathlon. (MSc(R) thesis), Kingston University.

Full text not available from this archive.

Abstract

The current scoring system for the decathlon was introduced in 1985. This project has used personal best data from elite athletes to investigate the role of the scoring system in the changes in performance that have been observed over a forty year period. In the first part of the investigation statistical tests were used to determine how far one of the basic principles of the current scoring system, that all the events should be of equal merit and rewarded equally, was being upheld. In the second part multivariate techniques were applied to decathlon personal best point score data to establish that the scoring system had produced an underlying structure representative of the ten decathlon events. Changes in this structure over two twenty year periods were investigated. The solution from a principal components analysis for each period was used to provide a two-dimensional configuration of the point score data in terms of the ten events and to suggest possible numbers of clusters for use in the application of a cluster analysis technique. Changes observed in the content of the clusters of the ten events obtained from each of the two periods were compared with operational changes made to the decathlon in the same period. It was found that although the sprinter/ jumping events scored higher points than the throwing events, the point scores for the first three track events had significantly smaller variability than that for any of the field events. There was insufficient evidence to suggest that either track or field events predominate in providing an advantage to the decathlete. The investigation also found that for the personal best decathlon data from the past twenty years the point score achieved by an individual in the 1500m race did not make a statistically significant contribution to his total point score. The principal components solutions confirmed that the throwing events and the 1500m race were associated with lower point scores than the sprinter/jumping events for both of the twenty year periods. The only observable change in the content of the clusters between the two periods was linked to a change made to the specification of the javelin in 1986.

Item Type: Thesis (MSc(R))
Physical Location: This item is held in stock at Kingston University Library.
Research Area: Applied mathematics
Sports-related studies
Faculty, School or Research Centre: Faculty of Science (until 2011)
Depositing User: Katrina Clifford
Date Deposited: 17 Apr 2012 15:27
Last Modified: 16 Sep 2014 12:47
URI: http://eprints.kingston.ac.uk/id/eprint/21725

Actions (Repository Editors)

Item Control Page Item Control Page