No Cover Image

Journal article 99 views 16 downloads

Performance indicators associated with match outcome within the United Rugby Championship

Georgia Scott, Neil Bezodis Orcid Logo, Mark Waldron Orcid Logo, Mark Bennett, Simon Church, Liam Kilduff Orcid Logo, Rowan Brown Orcid Logo

Journal of Science and Medicine in Sport, Volume: 26, Issue: 1, Pages: 63 - 68

Swansea University Authors: Georgia Scott, Neil Bezodis Orcid Logo, Mark Waldron Orcid Logo, Mark Bennett, Liam Kilduff Orcid Logo, Rowan Brown Orcid Logo

  • 62102.pdf

    PDF | Version of Record

    © 2022 The Authors. This is an open access article under the CC BY license

    Download (601.12KB)

Abstract

ObjectivesThe aims of this study were to: i) identify performance indicators (PIs) associated with match outcomes in the United Rugby Championship to; ii) compare efficacy of isolated data and data relative to opposition in predicting match outcome; and iii) investigate whether reduced PI statistica...

Full description

Published in: Journal of Science and Medicine in Sport
ISSN: 1440-2440
Published: Elsevier BV 2023
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa62102
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract: ObjectivesThe aims of this study were to: i) identify performance indicators (PIs) associated with match outcomes in the United Rugby Championship to; ii) compare efficacy of isolated data and data relative to opposition in predicting match outcome; and iii) investigate whether reduced PI statistical models can reproduce predictive accuracy.MethodsTwenty-seven PIs were selected from 96 matches (2020-21 United Rugby Championship). Random forest classification (RFC) was completed on isolated and relative datasets, using a binary match outcome (win/lose). Maximum relevance and minimum redundancy PI selection was utilised to reduce models. In addition, models were tested on 53 matches from the 2021-22 season to ascertain prediction accuracy. ResultsWithin the 2020-21 datasets, the full models correctly classified 83% (CI 77%-88%) of match performances for the relative dataset and 64% (CI 56%-70%) for isolated data. When models were reduced, these values were 85% (CI 79%-90%) and 66% (CI 58%-72%). In prediction on the 21-22 season, the reduced relative model successfully classified 90% of match performances (CI 82%-95%). Within the reduced relative model, five PIs were significant for match outcome: kicks from hand, metres made, clean breaks, turnovers conceded and scrum penalties. ConclusionsRelative PIs were more effective in predicting match outcomes than isolated data. Reducing features used in random forest classification did not degrade prediction accuracy, whilst also simplifying interpretation for practitioners. Increased kicks from hand, metres made, and clean breaks compared to the opposition, as well as fewer scrum penalties and turnovers conceded were all indicators of winning match outcomes within the United Rugby Championship.
Keywords: Game Statistics, Decision Modelling, Multivariate Analysis, Sports Performance, Team Sports.
College: Faculty of Science and Engineering
Funders: This work was supported by the ESPRC DTP (EP/EGF1069/; EP/T517987/1) and Ospreys Rugby.
Issue: 1
Start Page: 63
End Page: 68