MODELLING PLAYER PERFORMANCE DATA FOR ORGANISATIONAL DECISION SUPPORT IN PROFESSIONAL AUSTRALIAN RULES FOOTBALL Samuel McIntosh B. App. Sci. (Hons), B. Ex & Sp. Sci, B. Hlth & Rhb. Sci. This thesis is submitted in partial fulfilment of the requirements for the award of DOCTORATE OF PHILOSOPHY Supervisor: Prof. Sam Robertson Co-supervisor: Dr. Stephanie Kovalchik Institute for Health and Sport Victoria University, Melbourne, Australia Western Bulldogs Football Club 2020 i ABSTRACT Within contemporary professional team sport organisations, operational decisions are increasingly becoming informed by objective data. Within the elite competition of Australian Rules football, the Australian Football League (AFL), an abundance of player and team performance data is collected and reported. However, the extent to which this data has been used in the team sport notational literature to inform organisational decision-making is limited. This thesis utilises a particular algorithmic player rating system, the ‘AFL Player Ratings’, and the subcategories used to construct this metric. Each study of this thesis models various applications of player performance data and presents it in a format for the purpose of providing organisational decision support to AFL clubs. The first study of this thesis establishes the validity of the AFL Player Ratings system. The second study identifies how performance profiles created from the proportion of rating points in each AFL Player Rating subcategory can be used to classify players into a priori determined player role categories. Additionally, it determines a level of similarity between the playing styles of each individual player competing within the AFL. The third study developed two separate models to objectively benchmark player performance, and to identify stages of peak performance and specific breakpoints longitudinally. The final study of the thesis investigated the relationship between subjective ratings of performance and basic player performance indicators, in order to gain an understanding of the extent to which human decisions are related to measurable aspects of a player’s performance. It also looked to compare subjective and objective ratings of player performance. Each of these studies address a different use of the data operationally, and provide a framework for clubs competing in the AFL. It outlines how objective player performance data can be modelled to inform various aspects of team and player individuality, value and potential, with a specific focus on supporting team selection, player drafting and recruitment. ii STUDENT’S DECLARATION Doctor of Philosophy by Publication Declaration “I, Samuel McIntosh declare that the PhD thesis by Publication entitled Modelling player performance data for organisational decision support in professional Australian Rules football is no more than 100,000 words in length including quotes and exclusive of tables, figures, appendices, bibliography, references and footnotes. This thesis contains no material that has been submitted previously, in whole or in part, for the award of any other academic degree or diploma. Except where otherwise indicated, this thesis is my own work”. Signed: Date: Name: Samuel McIntosh ID Number: s4524829 iii iv ACKNOWLEDGEMENTS Without the support of some extremely influential and intelligent people, I would not have been able to complete this thesis. As such, there are many people I would like to acknowledge for the motivation and mentorship they have given me though the past three and a half years. First and foremost, I would like to thank my principal supervisor Professor Sam Robertson for your expertise and insight. Your guidance throughout my journey as a research student has not only made it possible for me to submit this thesis, but has helped me to gain a world of experience and many valuable skills. I am forever grateful for your mentorship, and am very proud of all that we have achieved together. To my co-supervisor Dr Stephanie Kovalchik, thank you so much for the guidance you have given me and your outlook on all things regarding statistical modelling and R. I have learnt so much from your input, and am truly appreciative of all your hard work. Thank you to everyone a part of the Institute for Health and Sport at Victoria University, and the Western Bulldogs Football Club for providing me with the opportunity to conduct this research. Specifically, to Jason Smith, Chris Caporaso, Sam Power, Jason McCartney and all others who’ve been a part of the performance analysis, recruitment and list management teams at the Western Bulldogs Football Club during the past five years. I am so thankful for the opportunities you have each given me to apply myself and my research within the football club. Each of these opportunities has been for the betterment of myself as a practitioner, as well as the research conducted as part of this thesis. v To my fellow research students, in particular those of W202 at the Whitten Oval campus, thank you all for the laughs and support. For those of you still on your research journey, I look forward to seeing you all reach this final hurdle. To all my friends and family that have been on this ride with me, thank you all for the encouragement you have given me throughout the past three and a half years. To my parents, thank you for always believing in me, and encouraging me to continue to strive for more. Without your support and encouragement I would not have made it this far. I cannot thank you both enough for the life that you have given me and for supporting me throughout my studies. I hope you both enjoying reading this! Lastly, a very special thank you goes to my wife Jade. I cannot thank you enough for the love, kindness and patience you have given me, not only these past three and a half years, but also the many prior to that which has led me to here. Your consistent encouragement and support has empowered me achieve something I thought I would never accomplish. I could not have got this far without you. I look forward to starting a new chapter with you , and sharing our future successes together. vi PUBLICATIONS ARISING DURING CANDIDATURE The following work has been published in peer-reviewed journals in support of this thesis: 1. McIntosh, S., Kovalchik, S., & Robertson, S. (2018). Validation of the Australian Football League Player Ratings. International Journal of Sports Science & Coaching, 13(6), pp. 1064–1071, doi:10.1177/1747954118758000. (Chapter Three). 2. McIntosh, S., Kovalchik, S., & Robertson, S. (2018). Examination of player role in the Australian Football League using match performance data. International Journal of Performance Analysis in Sport, 18(3), pp. 451-462, doi:10.1080/24748668.2018.1486116. (Chapter Four). 3. McIntosh, S., Kovalchik, S., & Robertson, S. (2019). Multifactorial benchmarking of longitudinal player performance in the Australian Football League. Frontiers of Psychology, 10(1283), doi:10.3389/fpsyg.2019.01283. (Chapter Five). 4. McIntosh, S., Kovalchik, S., & Robertson, S. (2019). Comparing subjective and objective evaluations of player performance in Australian Rules football. PLOS ONE, 14(8), e0220901. doi:10.1371/journal.pone.0220901. (Chapter Six). vii PRESENTATIONS ARISING DURING CANDIDATURE The following work has been presented at scientific meetings in support of this thesis: 1. McIntosh, S., Kovalchik, S., & Robertson, S. (2018). Multifactorial benchmarking of longitudinal player performance in the Australian Football League. Presented at World Congress of Performance Analysis of Sport XII. (Preliminary work relating to Chapter Five). 2. McIntosh, S., Kovalchik, S., & Robertson, S. (2019). Comparing subjective and objective evaluations of player performance in Australian Rules football. Presented at 9th World Congress on Science and Football. (Preliminary work relating to Chapter Six). AWARDS ARISING DURING CANDIDATURE Victoria University College Centenary Award (2016-2019) – Awarded to a graduate research student who has demonstrated academic excellence, and is conducting research that is strategically aligned with their college’s research focus. Routledge Young Researcher Award (2018) – Awarded to best presentations by researchers under 35 at World Congress of Performance Analysis of Sport XII. viii TABLE OF CONTENTS ABSTRACT ............................................................................................................................... i STUDENT’S DECLARATION .............................................................................................. ii ACKNOWLEDGEMENTS ................................................................................................... iv PUBLICATIONS ARISING DURING CANDIDATURE .................................................. vi PRESENTATIONS ARISING DURING CANDIDATURE ............................................. vii AWARDS ARISING DURING CANDIDATURE ............................................................. vii TABLE OF CONTENTS .................................................................................................... viii LIST OF SYMBOLS AND ABBREVIATIONS ............................................................... xiii LIST OF TABLES ................................................................................................................. xv LIST OF FIGURES ............................................................................................................. xvii LIST OF APPENDICES ....................................................................................................