Using athlete monitoring systems to understand performance in Australian academy soccer players

  • Michael G. Sydney

    Student thesis: Doctoral Thesis


    The objective of athlete monitoring is to capture information to support evidencebased decisions underpinning the management of athlete performance. Commonly, metrics relating to external load, internal load, and player wellbeing form the foundation of athlete monitoring systems. Research in athlete monitoring is valuable as it provides frameworks and proposed methodologies to implement within real-world practice. However, the development, scope and application of athlete monitoring systems should be sport and context specific. Despite the ubiquity of athlete monitoring in youth soccer academies and development programs, the majority of research to date has investigated elite senior male player cohorts. Whilst this provides some knowledge to inform applied practice and decision making for coaches and practitioners, the internal (e.g., age, sex, body composition) and external (e.g., rules, equipment, and environment) variables that contribute to performance in male and female youth soccer player cohorts differ to elite senior male player cohorts. As such, current research on existing adult male centric cohorts may not be translatable to youth cohorts, particularly youth female cohorts. The coaching staff of the Football Federation Australia Centre of Excellence (FFA CoE) approached the University of Canberra Research Institute for Sport and Exercise (UCRISE), seeking to explore identified contextual variables that may influence the performance of players within their national academy program. The initial empirical studies (chapters three, four and five) were designed to investigate these ‘coach-derived, real-world’ inquiries. The empirical studies in chapters six and seven sought to address the low number of studies in youth female soccer players. To bridge the gap between applied practice and research in youth female soccer, the coaching staff of the participant group for chapters six and seven were approached by the author of the thesis, building upon the previously established relationships and memorandum of understanding. The scope of research undertaken in chapter six and seven extended beyond data collection. Education sessions for athletes, parents, staff, and coaches were provided as a service with the athlete monitoring systems established for data collection in chapters six and seven being the first to be employed within the youth teams of Canberra United. Hence, the coaching staff of the participant groups in this thesis were integral to the research process and overarching thesis. The current thesis is formatted as individual chapters, containing stand-alone studies aimed at investigating the use of athlete monitoring systems to understand performance in youth male and female academy soccer players based on ‘coach-derived, real-world’ inquiries. Following the introduction in chapter one, chapter two contains a comprehensive narrative literature review. The purpose of this narrative literature review is to: 1) detail the anthropometric and characteristic physical attributes of elite youth male and female soccer players; 2) examine and critique the methodologies used to quantify the external and internal load of competition and training and analyse contextual variables that influence performance; 3) critically analyse measures of player wellbeing and their relationship to player performance; and 4) explore the complex phenomena of player injury and illness risk, incidence and burden in relation to aforementioned variables. Collectively, the review of the literature identified the complexity of athlete monitoring practices to leverage evidence-based decisions to manage performance. The first empirical study, chapter three, investigated the influence of competition standard and playing position on the physical demands of Australian elite youth male soccer players. The novelty of this study was the examination of a single elite youth male player cohort competing across a duel, segmented competition structure consisting of the National Youth League (NYL) and National Premier League (NPL) during a single calendar year. In addition, this study characterised the external load of competition of elite youth male soccer players within an Australian context. The results of this study determined that competition standard influenced the external load of players, but this effect varied according to position. Specifically, central defenders (3.1%), external attackers (4.2%) and central attackers (3.8%) performed more total distance per minute in NYL (professional development level) compared to NPL (semi-professional) match-play. Central defenders (24.2%) and central attackers (17.0%) completed more high-speed running (> 5.0 m·s-1) per minute in NYL (professional development level) compared to NPL (semi-professional) match-play. Since the publication of chapter three, the method of quantifying external load using relative (i.e., distance divided by duration) measures have been questioned within the literature. As such, both relative and peak measures were utilised to examine the external load of elite youth soccer players detailed in chapters four, five and six. Having established the influence of competition standard on the external load according to playing position during competition matches in chapter three, chapter four sought to assess the physical impact and internal response of substitute players, compared to starting and full-match players. Generally, studies that investigate soccer players tend to exclude players who complete less than 60-minutes during competition match play to reduce variance attributed to substitution status in their results. Whilst this may seem appropriate from a methodological perspective, inherently, substitutes are a vital component of a coach’s strategy to influence match outcome, countering game-induced fatigue. As such, excluding substitutes from analysis limits the understanding of player performance during competition matches. An improved understanding of the external load and internal response across substitution status and playing position is warranted. Within the context of youth soccer players, such data are important to coaches and practitioners as the decisions to manage the performance of players according to substitution status is different in the lead up to and post competition during critical stages of career development. The results of chapter four determined that substitutes displayed lower peak total distance and high-speed running distance compared to starting and full-match players. However, substitutes recorded greater relative high-speed running compared to full-match players. The rate of perceived exertion was significantly lower in substitute players compared to full-match and starting (e.g., players replaced by substitutes) players. This study not only highlighted the important role of substitutes but also the need to analyse external load according to relative and peak metrics in conjunction with internal responses when seeking to determine a player’s physical impact as a substitute. Having assessed the demands of competition in consideration of contextual factors such as competition standard, playing position and substitution status, there was a need to investigate and understand the external load of elite youth male soccer players during training environments. Various-sided games (VSGs) have been an area of considerable interest within the literature, and a common training modality employed by coaching staff in the field to simultaneously address technical, tactical and player conditioning training objectives (e.g., prepare players for the demands of competition). Within the context of the FFA CoE, the design and implementation of conditioning-focused VSGs was unique, underpinned by specific knowledge and designed to support the tactical philosophy of the coaching staff. A key strength of this study was the ecological validity and incorporation of ‘coach-derived, real-world’ inquiries. The results of this study indicated that when using relative measures, VSGs were found to supersede the demands of competition. However, no VSG type was found to supersede the absolute measures of peak total and high-speed running distance of match-play. This study recommended that VSGs should be supplemented with high-speed running training drills to prepare players for the demands of competition. As with the initial aspects of this thesis, research in elite youth soccer players has tended to focus on male cohorts. Unfortunately, this has resulted in a distinct lack of research relating to the performance of youth female soccer players. This gap in research has in turn facilitated the use of athlete monitoring frameworks and evidence generated from male participant groups being applied to female sporting environments, leading to erroneous practices in applied settings. Yet, the popularity and professionalism of female soccer is rapidly growing. As such, the next aspect of this thesis is to examine the use of athlete monitoring practices to manage performance of academy female soccer players. Chapter six provided a practical investigation, aimed to determine the association between pre-competition perceived player wellbeing measures and subsequent relative and peak running performance of developmental academy youth female soccer players in the first 15-minutes of match-play. Notably, menstrual cycle was included as a random factor in the analyses. We hypothesised that player wellbeing measures and menstrual cycle phase would influence peak and relative running demands. However, results indicated that menstrual cycle phases explained little variance in the peak and relative running outputs, further highlighting the necessity for sex specific study designs. Results did indicate that changes in stress, fatigue and lower-body muscle soreness influenced external load. This thesis focussed on using athlete monitoring systems to manage performance in youth male and female soccer players. Although managing performance is a clearly defined objective for which the development, scope, and application of athlete monitoring systems can be used, management of performance and mitigation of injury and illness are intrinsically linked. The inability to train and compete as a result of injury and illness negatively influences performance and, in this sense, the two conceptual models of managing performance and reducing risk of injury and illness are difficult to separate. The risk factors that contribute to the occurrence of injury and illness in youth soccer players are complex, multivariate and currently, not well understood. To date, research and applied practice has commonly assessed the influence of one variable in relation to another (e.g., external load and injury and/or illness occurrence). However, such research and applied practice is conducted on the assumption of linearity and fails to account for the time-varying dependencies between variables. As such, shown in chapters three, four, five and six, reductionist statistical methods are frequently utilised to analyse performance, injury and illness in youth male and female soccer players using data collected from athlete monitoring systems. To address the complex, non-linear interactions between the multiple variables of performance, injury, and illness, chapter seven aimed to model injury and illness occurrence as a complex system. A Dynamic Bayesian Network (DBN) was used as probabilistic frameworks offer statistical advantages in the analysis of soccer athlete monitoring data, allowing the time-varying dependencies between variables to be modelled. Specifically, the temporal relationships between self-reported perceived player wellbeing, internal load, external load, acute:chronic workload ratio, menstrual cycle phase and injury and illness occurrence were modelled. The resultant DBN suggested that injury and illness incidences were not associated with the examined athlete monitoring variables. However, the results of this study did illustrate the temporal relationship between player wellbeing measures towards lower-body muscle soreness and mood. The findings of this study provide an example of how researchers and applied practice can employ a novel, multifactorial, network approach to explore the complex relationships between athlete monitoring variables. In chapter eight, the discussion chapter, results of each study are reviewed, outlining the knowledge that was developed throughout this thesis. Future directions of research are also explored. Finally, a framework of practical recommendations to maximise translation of the results outlined in chapters three to seven into applied practice for coaches and practitioners in the field is examined.
    Date of Award2023
    Original languageEnglish
    SupervisorJocelyn Mara (Supervisor), Nick Ball (Supervisor), Dale Chapman (Supervisor) & Martin R. Wollin (Supervisor)

    Cite this