Multi-factorial modelling of player physical development within the Australian Football League (AFL) participation pathway

HAYCRAFT Jade-thesis_nosignature.pdf - Submitted Version (17MB) | Preview

Haycraft, Jade Alexandra Ziems (2018) Multi-factorial modelling of player physical development within the Australian Football League (AFL) participation pathway. PhD thesis, Victoria University.


National sporting organisations such as the AFL invest significant resources into establishing talent identification (TID) and talent development models to ensure the future success of their sport. The process usually involves predicting player potential through subjective assessments of game performances in combination with objective inputs inclusive of physical fitness testing, movement ability assessments, and match activity profiles. However, elite sporting performance is the product of a player’s ability to overcome and master the dynamic interactions between organismic, task, and environmental constraints that may impede or facilitate physical fitness and skill development. Organismic constraints such as growth, maturity, and learning stages of individual players can all influence physical fitness, movement ability, and match activity profiles. Specific environmental constraints include differences in game play, skill level, game rules and policies, and field sizes. Task constraints are the limitations imposed by a set task, for example; the goals and rules of AFL, the sporting actions required, and the physical fitness qualities needed for high-level match performance. Of particular interest is the as relative age effect (RAE), a selection bias among players of differing skill and maturity levels caused when adolescent players are grouped into annual age-grouped teams. The extent to which these constraints contribute to variations in physical fitness and match activity between levels of junior AF competition is unclear. This thesis investigates the interactions between physical fitness characteristics, the RAE, and match activity profiles across multiple levels of the AFL participation pathway, and the subsequent implications for player TID and development. The first study was a systematic review examining physical fitness measures of players across the AFL participation pathway levels to quantify longitudinal changes observed in physical fitness characteristics. Only studies examining physical fitness tests were included, with 27 meeting the inclusion/exclusion criteria. Sprint time (20-m) was the most reported test across the AFL participation pathway, followed by vertical jump (VJ), AFL agility, and 20-m multi-stage fitness test (MSFT). The fastest times for 20-m sprint were for elite AFL players (range 2.94 – 3.13 s), with local level players the slowest (3.22 – 4.06 s). State Junior Under (U) 18s had higher jumps than senior players; with the lowest jumps reported for Local U10s (range 31- 66 cm). Interestingly, no elite-level data were reported for the AFL agility or 20-m MSFT, with AFL agility times only reported for talent pathway levels (8.17 – 9.12 s). However, 20-m MSFT scores were reported across the junior levels of the AFL participation pathway (6.1 – 13.5 shuttles). Talent squad players exhibit similar test scores between competition levels irrespective of the physical fitness test, with the exception of 20-m sprint and VJ. It was suggested that physical tests can discriminate between local participation level players, but are less useful within the AFL talent pathway. Study II examined the influence of age-policy changes on the RAE across the AFL talent pathway. The study design was a retrospective cross-sectional analysis of junior AFL players attending the annual National Draft, State, and State U16 combines between 1999 and 2016. Birth-date data was obtained for players attending the AFL State U16 (n = 663), State (n = 803), National (n = 1111) combines, with comparisons made against corresponding aged-matched Australian general population birth-rates. Specifically, Under 16s and State players had greater birth frequencies (2% to 4.9%, p < 0.05) for earlier months in the selection year. Age-policy changes at the National level reduced birth distribution bias for some months; however the RAE remained for other months (March 3.9%, June 6.1%, and July 4.3%, p ≤ 0.05). State U16s and National players had 2- 9% lower birth frequencies for the later months in the selection year compared general population. It seems that selection bias towards older players is instigated at the AFL’s State U16, and maintained through to the State and National levels, with age-policy changes only partially successful at addressing the RAE at the National level. Study III investigated the levels of association between physical fitness and match activity profiles of players within the AFL participation pathway. A total of 287 players across seven pathway levels were assessed on the 20-m sprint, AFL agility, VJ and running VJ, 20-m MSFT, and Athletic Abilities Assessment (AAA). Match activity profiles were obtained from global positioning system (GPS) measures; relative speed, maximal velocity, and relative high speed running (HSR). Correlational analyses revealed moderate relationships between sprint and jump test scores and match activity profiles in Local U12, Local U14, National U16 and National U18s (r = 0.32-0.78, p ≤ 0.05), but not jump tests in National U18s. AFL agility time was moderate-to-strongly associated with all match activity measures in Local U12, Local U14, Local U18, and National U16s (r = 0.37-0.87, p ≤ 0.05), and with relative speed in Local U18s (r = 0.84, p ≤ 0.05). Relative speed and HSR were moderate-to-strongly associated with 20-m MSFT in Local U14, Local U18, and National U18s, and AAA score in Local U12, and Local U18s (r = 0.41-0.95, p ≤ 0.05). Match activity profile demands increased between Local U12 and National U16s, then plateaued across the talent pathway levels. Physical fitness seemed to relate more strongly to match activity profiles in younger adolescent and National level players’, therefore recruiters should consider the dynamic changes physical changes between AFL participation pathway levels. The final study examined the utility of physical fitness and movement ability tests in differentiating and classifying players into specific AFL participation pathway levels. Players (n = 293) completed the same physical fitness tests battery as Study III; 5-m, 10-m and 20-m sprint, AFL agility, VJ, running VJ, 20-m MSFT, and AAA. A multivariate analysis of variance between AFL participation pathway levels for each test was conducted, and a non-linear analysis (classification tree) determined the extent players could be allocated to relevant levels. The magnitude of the difference between physical fitness and movement ability was age-level dependent, with the largest standardised effects between Local U12, Local U14s, and older levels for most physical fitness tests (Effect Size (ES): -4.24 to 4.65). The 20-m, 5-m, AFL agility, 20-m MSFT, overhead squat, and running VJ (right) all contributed substantially to the classification model, with over half of the players accurately classified into the appropriate AFL participation pathway levels (57%). The National U16 players were most accurately classified based these tests (87%); however, no National U18 players were classified. Talent selectors should consider differences in physical fitness and movement ability patterns between players when selecting players into the talent pathway; however other contextual factors (i.e., skills, psychological, and socio-cultural factors) are needed to establish a multi-component TID model in older levels of the talent pathway. Physical development differences between players, age-grouped competition levels, and talent levels reported in these studies should be considered in the planning, implementation, and review of AFL development programs. Establishing associations between common physical fitness and movement ability tests used for TID in AFL and match activity profiles allows coaches and talent selectors to make more informed player selection decisions. Also, non-linear modelling facilitates TID decisions as it can highlight under or over-performing players; flagging them to selectors for further investigation of other contextual factors influencing a player’s potential. Future studies should focus on including more areas that encompass TID and development (i.e., skills, psychological, and socio-economic attributes) to provide a comprehensive understanding of the interactions that determine a player’s success at an elite level. This research is not limited to AFL, with the methods used having the potential to provide more informed TID and development processes across other sports.

Item type Thesis (PhD thesis)
Subjects Current > FOR Classification > 1106 Human Movement and Sports Science
Current > Division/Research > Institute for Health and Sport
Current > Division/Research > College of Sports and Exercise Science
Keywords thesis by publication; talent identification; talent development; physical fitness; relative age effect; match activity profiles; AFL
Download/View statistics View download statistics for this item

Search Google Scholar

Repository staff login