Journal article
A probability based approach for the allocation of player draft selections in Australian Rules Football
Journal of Sports Science and Medicine, Vol.5(4), pp.509-516
2006
Abstract
Australian Rules Football, governed by the Australian Football League (AFL) is the most popular winter sport played in Australia. Like North American team based leagues such as the NFL, NBA and NHL, the AFL uses a draft system for rookie players to join a team's list. The existing method of allocating draft selections in the AFL is simply based on the reverse order of each team's finishing position for that season, with teams winning less than or equal to 5 regular season matches obtaining an additional early round priority draft pick. Much criticism has been levelled at the existing system since it rewards losing teams and does not encourage poorly performing teams to win matches once their season is effectively over. We propose a probability-based system that allocates a score based on teams that win 'unimportant' matches (akin to Carl Morris' definition of importance). We base the calculation of 'unimportance' on the likelihood of a team making the final eight following each round of the season. We then investigate a variety of approaches based on the 'unimportance' measure to derive a score for 'unimportant' and unlikely wins. We explore derivatives of this system, compare past draft picks with those obtained under our system, and discuss the attractiveness of teams knowing the draft reward for winning each match in a season. ©Journal of Sports Science and Medicine (2006).
Details
- Title
- A probability based approach for the allocation of player draft selections in Australian Rules Football
- Authors
- Anthony Bedford (Author) - RMIT UniversityAdrian J Schembri (Author) - RMIT University
- Publication details
- Journal of Sports Science and Medicine, Vol.5(4), pp.509-516
- Publisher
- Journal of Sports Science and Medicine
- Date published
- 2006
- ISSN
- 1303-2968
- Organisation Unit
- University of the Sunshine Coast, Queensland; School of Science, Technology and Engineering; School of Health and Behavioural Sciences - Legacy
- Language
- English
- Record Identifier
- 99450776102621
- Output Type
- Journal article
Metrics
182 Record Views