Search

There is a dearth of up-to-date literature scrutinising the experiences of soccer coaches, players and analysts involved in the collection and dissemination of findings from Performance Analysis (Henceforth PA). Given the rapid expanse of PA within soccer in recent years, a comprehensive and contemporary account of how PA is delivered and utilised by coaches appears warranted. Therefore, the purpose of this research was to investigate the challenges and opportunities experienced by coaches and analysts in embedding PA within applied soccer environments. Using an interpretivist theoretical framework and qualitative research method, this thesis included three different stages of data collection and analysis. In stage one, data were collected from eight full-time performance analysts working in the English Football League and English National League, exploring the use of PA across the phases of match-play (pre-match, match-day, and post-match). Data collected from semi-structured interviews and reflective thematic analysis identified two higher-order categories (Flying solo / coach control) and four themes classified as lower-order (identifying indicators / PA pragmatism / analysis ingredients / autocratic presentations). Consequently, this study highlighted the types of PA completed by performance analysts is context-specific and dependent on multiple factors (e.g., coach and analyst relationship) which can restrict the value of the PA produced. Consequently, PA appears to be undertaken on a continuum from coach-controlled to analyst-controlled processes, pending what aspects of analysis are used, when in the relationship between analyst and coach the process is developed and when in the matchday process they are carried out. Stage two of the research investigated PA "in-situ" through a case-study approach. This allowed a contextualised process for the use of PA at half-time to be developed and studied within a soccer academy. Reflective thematic analysis of interviews with the lead coach, assistant coach, head of coach development and six players, both before and after PA was available, highlighted two higher-order themes (Opportunity knocks / Conflict of interest) and four lower-order themes (Supporting current practice / Race to the tech / Prioritising performance / The best before the rest). Meanwhile, reflective thematic analysis of six half-time team teams before and after PA was introduced, generated two higher-order themes (Coach-controlled feedback and evidence informed evaluation) and four lower-order themes (Scouting report / effort and attitude / tactical deconstruction / inspiring individuals). This stage of the research highlighted that within the Elite Player Performance Plan (Henceforth EPPP), the role of the coach, and application of PA, appears complicated and multifaceted despite an enthusiasm expressed by all stakeholders to access PA. (NB the EPPP is the talent identification and development programme initiated by the Premier League in consultation with the Football Association and English Football League [Henceforth EFL] to identify and develop male footballers in England). In the third stage of the research, twelve coaches were interviewed to identify the ways in which they engage with PA to underpin their coaching practice and utilise the findings to aid player and team development. The data analysis identified two higher-order themes (external auditors and weaponisation of PA) and five lower-order themes (tick in a box, prescription PA, continuous professional denial, collateral damage and exit strategy). The results demonstrate that talent development programmes are complex environments, and several problems were revealed in the manner coaches are prepared for, and subsequently used, PA to support their practice. Despite an enthusiasm from coaches about PA, both a lack of education regarding how best to use PA and the influence of senior coaches within an academy impacted how PA was used. Furthermore, the desire of an academy to win matches, as well as to develop the players perceived to be the 'best' and achieve a specific EPPP category status influenced the use of PA. The empirical findings of this thesis have revealed the complexity experienced when utilising PA within applied coaches' practice. Despite coaches appearing enthusiastic at the prospect of implementing PA, the use of analysis is highly contextualised and reliant on several cultural and ethical issues including the coach-analyst relationship, coach education regarding the effective context-specific use of PA and the desire of a club to be seen to engage in PA to improve EPPP category status. Consequently, there appears to be several inter-related factors that contribute to PA being used sporadically by coaches within applied soccer environments. Therefore, as individuals and clubs embark upon using PA, an understanding of both the opportunities and challenges of PA should inform the development of the role of the analyst and the way in which PA is used to support coaching practice.

The aim of this thesis was to examine the physiological and biomechanical responses to multi-directional running in male and female team sport athletes. Chapter 4 compared measures of energy expenditure derived from indirect calorimetry and microtechnology, as well as high power and high-speed activity, during linear and multi-directional running. Measured energy expenditure was higher during the multidirectional trial (9.0 ± 2.0 cf. 5.9 ± 1.4 kcal.min-1), whereas estimated energy expenditure was higher during the linear trial (8.7 ± 2.1 cf. 6.5 ± 1.5 kcal.min-1). Whilst measures of energy expenditure were strongly related (r > 0.89, p < 0.001), metabolic power underestimated energy expenditure by 52% (95% LoA: 20-93%) and 34% (95% LoA: 12-59%) during the multi-directional and linear trial, respectively. Time at high power was 41% (95% LoA: 4-92%) greater than time at high speed during the multidirectional trial, whereas time at high power was 5% (95% LoA: -17-9%) lower than time at high speed during the linear trial. Chapter 5 explored the internal and external responses to linear and multi-directional running, specifically examining if measures of high speed and high power reflect changes in internal load. High speed distance (p < 0.001) was higher during the linear trial, whereas time at high power (p = 0.046) and accelerations performed (p < 0.001) were higher during the multi-directional trial. Summated HR (-0.8; ±0.5, p = 0.003), B[La] (-0.9; ±0.6, p = 0.002) and RPE (-0.7; ±0.6, p = 0.024) were higher during the multi-directional trial. There was a large difference in the ratio of high speed:summated HR (1.5; ±0.5, p = 0.001) and high speed:total V̇O2 (2.6; ±1.2, p < 0.001) between linear and multi-directional running, whilst high power:summated HR (0.3; ±0.5, p = 0.246) and high power:total V̇O2 (0.1;±0.8, p = 0.727) were similar. A small decrement in knee flexor torque was observed after the multi-directional (0.4; ±0.4, p = 0.017) and linear (0.2; ±0.3, p = 0.077) trials, respectively. Collectively, Chapters 4 and 5 reveal that more directional changes induce a greater internal response, despite reducing the high-speed distance someone is likely to cover. High power better reflects internal responses to multidirectional running than high speed, but microtechnology cannot be used to determine the absolute energy cost of multi-directional running. Chapters 6 and 7 explored alterations in side cut biomechanics in males and females immediately (Chapter 6) and 48 h (Chapter 7) after multi-directional running. In Chapter 6, 20 m sprint time was higher (ES: 0.65 - 1.17, p < 0.001) after multidirectional running, indicating the presence of fatigue. Males and females displayed trivial to moderate changes in trunk flexion (0.16 - 0.28, p = 0.082), peak hip internal rotation (0.46 - 0.54, p = 0.090), and knee flexion (0.17 - 0.41, p = 0.055) and higher knee abduction (0.40 - 0.51, p = 0.045) and internal rotation (0.59 - 0.81, p = 0.038) angular velocities, during the weight acceptance phase of side cuts after multidirectional running. Peak hip extensor (0.19 - 0.29, p = 0.055) and knee internal rotation moment (0.22 - 0.34, p = 0.052) displayed trivial to small increases after multidirectional running, whereas peak hip external rotation (0.44 - 0.57, p = 0.011), knee extensor (0.33 - 0.45, p = 0.003) moment and knee to hip extensor ratio (0.15 - 0.45, p = 0.005) were lower. In addition, IGRF displayed trivial to moderate changes (0.04 - 0.79, p = 0.066) and lateral GRF was lower (0.29 - 0.85, p = 0.002) after multidirectional running. In Chapter 7, CK concentration (2.4 - 4.94, p = 0.009), perceived muscle soreness (4.2 - 4.8, p < 0.001) and 20 m sprint time (0.6 - 0.9, p < 0.001) were higher 48 h after multi-directional running, indicating the presence of EIMD. Males and females displayed trivial to moderate changes in peak torso flexion (0.13 - 0.35, p = 0.055), hip internal rotation angular velocity (0.43 - 0.64, p = 0.073) and more knee internal rotation (0.31 - 0.5, p = 0.009) 48 h after multi-directional running. A tendency for an interaction between sex and time was noted for peak knee flexion (p = 0.068) and internal rotation angular velocity (p =0.057), with males only displaying a moderate increase. Males and females also displayed a lower peak knee extensor moment (0.43 - 0.56, p = 0.001) and a small increase in extensor moment (0.21 -0.46, p = 0.066) and knee external rotation moment (0.34 - 0.78, p = 0.062). An interaction between sex and time was noted for IGRF (p = 0.037); there was a large increase in IGRF at 48 h in females (1.4) but not males (0.08). For the first time, these data highlight multi-directional running which elicits fatigue and EIMD causes alterations in side cut biomechanics which can persist for at least 48 h. Specifically, both males and females performed side cuts in a more extended position, with higher peak angular velocities, and peak knee external rotation moments and less knee extensor moments both immediately and 48 h after multi-directional running.

The research described in this thesis used a standardised battery of tests called the 'Rugby League Athlete Profiling (RLAP)' battery for assessing the anthropometric and physical characteristics of UK-based rugby league players. The overall purpose of this research was to determine the utility of the RLAP battery, which involved establishing the use of RLAP across numerous professional clubs over a three-year period, determining the measurement properties of the tests included and investigating the factors associated with a change in the characteristics. An early version of the RLAP battery existed [called SPARQ] and was provided by the Rugby Football League with scope to alter this as part of this programme of research. Before determining if an alteration to the battery was required, it was essential to understand the tests that are currently used in rugby league for assessing the anthropometric and physical characteristics of players. As such, the systematic review initially sought to determine the volume of performance tests used in rugby league along with their measurement properties. Based on the results, it was evident that a shorter sprint distance (< 20 m) ought to be included in the battery. It was also clear that only one field-based method for measuring muscle strength was available, though had received minimal research. Furthermore, the review highlighted that no rugbyspecific intermittent running test had previously been used and that RLAP was the first battery to include such a test. Therefore, based on these results, the battery was rebranded to RLAP, which included a stature, body mass, a 10 m and 20 m sprint test, a rugby-specific intermittent test, a change of direction test, measures of lower- and whole-body power. With the RLAP battery confirmed, it was then used and the reliability (Chapter 4) and discriminant validity (Chapter 5) of its elements determined. Results indicated that the RLAP battery is reliable and does not require habituation. Furthermore, the calculation of the required change, which includes the worthwhile change and random error of each test, provides researchers and practitioners with a single value that can be used as an analytical goal to evaluate a true change in characteristics with confidence. All components of the RLAP battery (except 10 m sprint time) possessed adequate discriminant validity between youth, academy and senior rugby league players, suggesting this battery can accurately distinguish between playing standards. As noted in above, the review highlighted a rugby-specific intermittent test has yet to be established in the literature before its inclusion in the RLAP battery. Whilst it appeared to be suitable and, based on Chapters 3 and 4, is reliable and possesses discriminant validity, the test itself had received no previous attention. Given the novelty of this test, it was unknown if this test was better associated with the responses to rugby league match performance and what the physiological responses were to this test. As such, Chapter 5 sought to determine the concurrent validity of this test and compare it against the traditional Yo-Yo Intermittent Recovery test level 1 (Yo-Yo IR1). The results indicated the association between prone Yo-Yo IR1 distance and the external, internal and perceptual responses to simulated match-play was improved when compared to the Yo-Yo IR1. Chapter 6 demonstrated that starting each 40 m shuttle in a prone position increases the internal, external and perceptual loads whilst reducing the total distance achieved. The degree of shared covariance between the prone Yo-Yo IR1 and Yo-Yo IR1 suggest the rugby-specific test provided insight into additional characteristics associated with rugby league performance. In studies that have reported on the anthropometric and physical characteristics, few have considered the multiple factors that might influence these with no studies conducted in rugby league. Chapter 7 sought to determine the complex interaction between anthropometric and physical characteristics that requires careful consideration by those involved in developing youth and academy athletes. The results also revealed a number of contextual factors such as season phase, league ranking, playing age and playing position that can influenced the change in characteristics over the course of a competitive season. The findings of this study highlight how some characteristics are impaired towards the end of the season, thus providing a rationale for considering in-season training loads and the application of short training interventions to off-set these negative changes. Based on negative changes in some anthropometric and physical characteristics towards the end of the year, Chapter 8 reported on the efficacy of two in-season sprint interval interventions for enhancing the physical characteristics of rugby league players. Furthermore, the study provided insight into the sensitivity of the RLAP battery for detecting changes in the characteristics of rugby league players. The results highlighted that two weeks of rugby-specific and running-based sprint interval training appeared affective for promoting the physical characteristics of rugby league players with minimal deleterious effects on wellness and neuromuscular function. Using the reliability statistics from Chapter 1, the mean change for prone Yo-Yo IR1 in the rugbyspecific group met the required change whilst changes approached this value for the running-based group despite contrasting loads. In all, this study demonstrated that sprint interval training that includes sport-specific actions is a suitable and effective training modality that can be used in-season. In addition, the result demonstrated how the prone Yo-Yo IR1 was sensitive to change across the intervention period whilst others were not sensitive to sprint interval training due to the lack of specificity. This thesis provides a thorough evaluation of the RLAP battery that can be used by researcher and practitioners to assess the anthropometric and physical characteristics of rugby league players. The battery is reliable and possess discriminant validity, while the prone Yo-Yo IR1 has concurrent validity and is sensitive to change during a lowvolume in-season training intervention. Overall, this thesis provides justification for the tests included and comprehensively examines the utility of this battery for assessing the anthropometric and physical characteristics of rugby league players. Practically, this battery of tests can be used by researcher and applied practitioners in rugby league with an understanding of the reliability, validity and sensitivity of the tests along with some factors that might influence the characteristics of players across a season.

High speed running and repeated sprinting are component parts of training and match play among academy football players. Despite players having to self-pace running speed and the intervening recovery periods during match play, the way these qualities are trained and tested are often externally regulated with specific work-to-rest ratios and prescribed intensities. The aims of this thesis were to investigate high speed running separated by externally regulated and self-selected recovery periods under conditions that replicate training and testing practices analogous with football. Under controlled conditions replicating training practices common amongst academy players, Chapter 4 showed that high speed running and repeated sprinting separated by externally regulated recovery periods resulted in running speeds that differed by a smaller magnitude than those used in their prescription. These data question the fidelity of this approach and the ability of players to replicate prescribed running speeds in the field. Data from Chapter 4 also demonstrated that neuromuscular function was likely reduced 14 hours after high speed running (-5.6%; ES -0.44 ± 0.32; P = 0.01) and combination running (-6.8%; ES -0.53 ± 0.47; P = 0.07) . During 10 x 30 m repeated sprints there was a most likely higher percentage decrement (65%; 0.36 ± 0.21; P = 0.12) and most likely increased physiological load evidenced by between sprint heart rate recovery (-58.9%; ES -1.10 ± 0.72; P = 0.05) when sprints were interspersed by self-selected compared to externally regulated recovery periods (Chapter 5). Performance decrements were, however, attenuated in more mature players (Chapter 6). When considering biological maturity, prePHV players displayed a lower percentage decrement (2.1 ± 1.1%) than post-PHV (3.2 ± 2.1%) players across all sprints when recovery periods were externally regulated (37%; ES 0.41 ± 0.51; P = 0.03). When self-selected recovery periods were used, percentage decrement was lower in the post-PHV group. In Chapter 7, ratings of perceived exertion were used to guide 4 running speed and recovery distribution during a high speed running test performed to volitional exhaustion. Peak running speed in the self-paced (21.8 ± 1.4 km·h-1 ) was likely (4.1%: ES 0.63 ± 0.43; P = 0.03) higher than in the externally regulated YYIRT1 (20.9 ± 1.1 km·h-1); however, average running speed in the self-paced (13.5 ± 1.2 km·h-1) was likely (6.5%; ES 0.67 ± 0.51; P = 0.05) slower (12.7 ± 1.6 km·h-1). There was a moderate difference in total between shuttle recovery periods (13.3%; ES 0.58 ± 0.81; P = 0.16) in the self-paced (552 ± 132 s) compared to externally regulated versions (634 ± 125 s) of the YYIRT1. When exposed to running drills separated by self-selected and externally regulated recovery periods, academy footballers allocate insufficient recovery to preserve running performance and are unable to differentiate between sprinting and high speed running when prescribed according to specific speeds (Chapter 4) and subjective ratings of exertion (Chapter 7). Prescribing self-paced high intensity running interspersed with self-selected recovery periods results in higher physiological loads when compared to externally regulated recovery intermissions and therefore should be considered during training programmes that target adaptations in aerobic capacity. Despite this, where coaches are using high speed running programmes to improve speed and/or speed endurance, externally regulated recoveries are likely to result in the preservation of performance across the repetition range and, as such, are more beneficial to the intended adaptation.

The aim of the current thesis was to develop knowledge of the 'loads' associated with rugby league match-play, with a particular focus on the effects of altered mental loads before and during exercise indicative of a rugby league match. Chapter 3 examined the test-retest reliability of movement, physiological and perceptual measures during and after a novel rugby match simulation, where movement commands were more random than those typical of match simulations. The most reliable measure of external load during bouts of the simulation was relative distance (typical error [TE] and coefficient of variation [CV%] = 1.5-1.6 m.min-1 and 1.4-1.5%, respectively), with all other movement characteristics possessing a CV% < 5%. The most reliable measure of internal load, neuromuscular function and perceptual measures were for %HRmax during bout 1 (TE and CV% = 1.4-1.7% and 1-4-2.1%, respectively), MVC before (TE and CV% = 10.8-14.8 N·m and 3.8-4.6%, respectively), and average RPE (TE and CV% = 0.5-0.8 AU and 3.6-5.5%, respectively). The conclusion of this chapter was that randomisation of the movements during simulated activity to better reflect intermittent team sports has no detrimental effect on its reliability. Studies can therefore confidently examine alterations in several perceptual, neuromuscular, physiological and movement load measures related to rugby activity using stochastic movements. Chapter 4 examined the responses to a simulated rugby league protocol that was designed to include more random commands, and therefore require greater vigilance, than traditional team sport simulation protocols. The randomised simulation (RDM) was matched for the number and types of activity performed every 5.45 min in a control trial (CON), but included no repeated cycles of activity. The RDM trial was more mentally demanding than CON (Effect size (ES) = 0.56; ±0.57). Self-paced mean sprint performance increased in RDM (22.5 ± 1.4 vs. 21.6 ± 1.6 km∙h-1; ES = 0.50; ±0.45), which was accompanied by a higher RPE (14.3 ± 1.0 vs. 13.0 ± 1.4; ES = 0.87; ±0.54) and a greater number of errors in the Stroop Test (10.3 ± 2.5 vs. 9.3 ± 1.4 errors; ES = 0.65; ±0.67). MVC peak torque (CON = -48.4 ± 31.6 N.m, RDM = -39.6 ± 36.6 N.m) and voluntary activation (CON = -8.3 ± 4.8%, RDM = -6.0 ± 4.1%) was similarly reduced in both trials. Providing more random commands, requiring greater vigilance, can therefore alter performance and associated physiological, perceptual and cognitive responses to team sport simulations. Chapter 5 describes the subjective task load of elite rugby league match play using the NASA-TLX and examines their association with several contextual match factors, technical ii performance and external movement demands. Linear mixed modelling revealed that various combinations of contextual factors, technical performance and movement demands were associated with subjective task load (NASA-TLX). Greater number of tackles (η2 = 0.18), errors (η2 = 0.15) decelerations (η2 = 0.12), increased sprint distance (η2 = 0.13), losing matches (η2 = 0.36) and increased perception of effort (η2 = 0.27) lead to most likely - very likely increases in subjective total workload. These data provide a greater understanding of the internal load and their association with several contextual factors, technical performance and external movement demands during rugby league competition. The purpose of the final empirical chapter (Chapter 6) was to describe the effects of mental fatigue on simulated rugby league performance and to determine the effects of caffeine supplementation on simulated rugby league performance in the presence of mental fatigue. Completing a mentally demanding task increases participants' subjective rating of mental fatigue (pre = 29 ± 25 AU; post = 55 ± 20 AU) immediately before completing a simulation protocol. Impairments in sprint speed (ES = -0.18; ±0.19), sprint to contact speed (ES = -0.20; ±0.27), high-intensity running (ES = -0.30; ±0.24), high metabolic power > 20 W·kg-1 (ES =-0.50; ±0.51) and time to complete a passing accuracy task (ES = 0.54; ±0.63) were observed after mental fatigue. Caffeine supplementation (5 mg.kg-1) attenuated several adverse effects of mental fatigue before exercise replicating the demands of rugby league match play, with increased sprint speed (ES = 0.40; ±0.18), high-intensity running (ES = 0.50; ±0.53), high metabolic power > 20 W·kg-1 (ES = 0.33; ±0.38) and decreased time to complete a passing accuracy test (ES =-0.70; ±0.45). Mental fatigue affected internal loads, external loads and skill performance during simulated rugby league match play that appear to be centrally regulated by a decreased motivation and increased perception of effort. However, a single dose of caffeine taken 60 min before performance can attenuate several of these negative effects. In summary, the current thesis highlights several interactions between the physical and mental loads associated with actual and simulated rugby league performance.

The aim of this thesis was to investigate the influence of physical collisions on internal (physiological and perceptual) and external (locomotive and accelerometer) load during simulated rugby league performance and fatigue responses in the days after. Chapter 4 examined the influence of physical contact type on internal and external load using a traditional soft tackle bag and custom-built tackle sled. Using a traditional tackle bag to simulate physical collisions resulted in likely faster sprint to contact speed (16.1 ± 1.5 c.f. 14.8 ± 1.1 km.h -1 ) but possibly lower overall high-speed running distance (27.7 ± 2.4 c.f. 28.4 ± 2.6 m.min-1 ). Also, the heavier tackle sled likely increased time at 91-100% HRpeak (12:58 ± 13:21 c.f. 6:44 ± 8:06 min:s) and resulted in greater lower limb fatigue reflected by the likely larger decrease in countermovement jump (CMJ) performance (5.9 ± 4.9 c.f. 2.6 ± 5.4%). Also of note was the variation in number of tackles detected using the automatic tackle detection feature compared to the actual number in the match simulation. During the Bag and Sled simulations ~53 and ~59 tackles were detected compared to 48 performed. The purpose of Chapter 5 was to investigate the influence of sprint to contact speed and contact type on automatic tackle detection using microtechnology. Repetitions were divided into three speed categories; walking, jogging and striding (1, 2.5 and 4 m.s -1 ) and four conditions: i) no contact standing upright (NCST), ii) no contact dropping to the ground in a prone position (NCGR), iii) contact with the tackle bag and remaining upright (CST), iv) contact with the tackle bag and going to ground (CGR). Similar tackle detection accuracy was observed between NCGR and CST conditions with one tackle observed in 41 and 43% of trials, respectively. While CGR resulted in the greatest frequency of correct tackle detection (62%), during 16% of trials two tackles were detected. During NCST, there were no tackles detected and 100% accuracy. The PlayerLoadTM results demonstrated that the metric can detect differences in movement speed, the inclusion of physical contact and changes in orientation during short periods of activity (8-10 s). In Chapter 6 the rugby league movement simulation protocol for interchange players (RLMSP-i) was modified to include a tackle shield collision to investigate the reliability of PlayerLoadTM metrics to quantify collision load. The coefficient of variation (%CV) for locomotive metrics ranged from 1.3 to 14.4%, with greatest variability observed for high-speed running distance (8.0 and 14.4% for Bouts 1 and 2, respectively). Accelerometer metrics CV% were 4.4 to 10.0%, while internal load markers were 4.8 to 13.7%. All variables presented a CV% less than the calculated moderate change during one or both bouts of the match simulation except from high-speed distance (m.min-1 ), %HRpeak and RPE (AU). The aim of Chapter 7 was to investigate the influence of contact type on external load metrics including PlayerLoadTM derivatives whilst controlling for total running distance. Participants were randomly assigned to one group to complete the match simulation with either a tackle shield (n = 10), tackle bag (n = 7) or no-contact (n = 10). Total PlayerLoadTM, PlayerLoadTM 2D (AU), PlayerLoadTM slow (AU) and PlayerLoadTM slow-ratio (%) were analysed from the accelerometer in addition to high- and low-speed running and sprint speed. Total PlayerLoadTM was likely lower for the Bag group compared to the Run group (498 c.f. 460 AU), with no clear differences between the other groups. 3 PlayerLoadTM slow for the Shield group (167 ± 26 AU) was very likely greater than both the Bag (133 ± 11 AU) and Run groups (128 ± 20 AU) but no clear difference was observed between the Bag and Run groups. No differences were observed in PlayerLoadTM 2D between any groups. High-speed running distance was likely lower in the Shield group (1056 ± 225 m) compared to the Bag group (1326 ± 245 m) and very likely lower compared to the Run group (1318 ± 175 m). Total PlayerLoadTM is not sensitive to contact type during simulated rugby league activity but does reflect greater high-speed running distance during a rugby league match simulation. However, PlayerLoadTM slow can detect the types of contact and might be preferred for quantifying match and training loads associated with physical contact. The purpose of the final empirical chapter (Chapter 8) was to determine the influence of contact type on in neuromuscular, perceptual and biochemical parameters associated with exercise-induced muscle damage. The participants were again assigned to one of three groups to complete the match simulation with a tackle shield (n = 6), tackle bag (n = 7) or no contact (n = 7). In addition to internal and external load measured during the match simulation, venous blood, muscle function and soreness measures were collected immediately (+0), +24 and +72 hours after the match simulation. Upper body neuromuscular performance and knee flexion torque likely decreased in the Shield group +0 and +72 hours after the simulation compared to the other groups while CMJ power likely decreased more in the Run group. All three groups demonstrated a very likely increase in IL-6 and IL-10 concentration immediately after the match simulation, but differences between the groups were unclear and values returned to baseline +24 hours after the simulation. In conclusion, current automatic tackle detection metric should be used with caution, particularly in training sessions where physical contact is replicated. Instead PlayerLoadTM and associated derivatives from the embedded accelerometer can provide a useful measure of contact-specific load during training and competitive matches. Physical contact type affected external load by modifying a participant's running strategy during simulated match performance, thereby influencing site-specific fatigue during and after a simulated rugby league match. However, regardless of contact type, large increases in cytokine and leukocyte concentration are apparent with a return to basal values 24 hours after. Therefore it is not recommended to use such biomarkers in applied settings to quantify the magnitude of muscle damage specifically associated with physical contact.

Little is known about the muscle function capabilities of trained middle-aged males and how they differ to younger counterparts. Accordingly, the overall aim of the research documented in this thesis was to compare the acute muscle function responses to resistance exercise in middle-aged and young resistance trained males. The first study (Chapter 3) examined the intra- and inter-day reliability of an ecologically valid device (FitroDyne rotary encoder) for measuring upper and lower-body muscle function during three popular multi-jointed resistance training exercises (bench press, squat, and bent-over-row), and confirmed that it was capable of detecting moderate changes in muscle function across a range of submaximal loads. In the second study (Chapter 4) the load-velocity and load-power relationships were investigated during the same exercises among 20 young (age 21.0 ± 1.6 y) and 20 middle-aged (age 42.6 ± 6.7 y) resistance trained males, and it emerged that, despite their regular training, the middle-aged males were unable to achieve velocities at low external loads and peak powers at all external loads as high as the young males across a range of external resistances. Study three (Chapter 5) proceeded to compare the internal (heart rate (HR), OMNI-ratings of perceived exertion (RPE) and sRPE) and external (peak velocity and power and volume load) loads experienced during high volume squatting exercise, and the fatigue responses among nine young (age 22.3 ± 1.7 years) and nine middle-aged (age 39.9 ± 6.2 years) resistance trained males. The findings highlighted that internal, but not certain markers of external (peak power and volume load), load responses can be monitored during exercise in a like manner between these age groups. Moreover, compared to young resistance trained males, middle-aged males can expect greater decrements in peak power after lower-limb resistance exercise. In the final study (Chapter 6), the time-course of recovery in nine trained young (age 22.3 ± 1.7 years) and nine trained (39.9 ± 6.2 years) and nine untrained (44.4 ± 6.3 years) middle-aged males after high volume lower-body resistance (muscle damaging) exercise was investigated. Of practical importance, it emerged that compared to the young males, the trained middle-aged males experienced greater symptoms of muscle damage and an impaired recovery profile, the implication of which is the need for trained middle-aged males to adopt strategies to enhance their recovery. Furthermore, both middle-aged groups experienced similar symptoms of muscle-damage, albeit the untrained group demonstrated greater losses in peak power at low and high external loads. For the first time, the current research has determined that middle-aged males, despite regular resistance training, are subject to losses in peak velocity and power output across a range external loads, compared to young males. When undergoing lower-body resistance training to ameliorate these decrements, applied practitioners can use internal load markers and peak velocity, but not peak power or volume load, to monitor trained young and middle-aged males alike. Furthermore, the muscle damage response (24 to 72 h), and losses in peak power (0 to 72 h), after lower-body resistance exercise are greater in trained middle-aged than young males. Consequently, future research should seek to corroborate these observations in upper-body exercise and determine the effectiveness of strategies (e.g. nutritional intake) to enhance recovery in middle-aged males.

This theses includes 4 case studies covering - (1)an examination of the anthropometric and physical characteristics of youth female team handball players (16.07 ± 1.30 y) in non-elite (n= 47), elite (n= 37) and top-elite players (n= 29); (2) a comprehensive analysis of team handball match play in youth English U18 Men’s National League players through the assessment of player movement demands, technical actions and heart rate during match play and secondly, the impact of team handball competition on fatigue during and after matches; (3) an investigation into neuromuscular fatigue and well-being of English handball players during a training camp and an international tournament; (4) the effect of two different interchange strategies on performance and pacing strategy during a simulated team-sports protocol.

It is well documented that engaging in resistance exercise can lead to further improvements in endurance performance. Whilst, not fully understood, it is speculated that increased motor unit recruitment, improved muscle coordination and enhanced utilisation of stored elastic energy after resistance-based exercise improves exercise economy. Nevertheless, while prolonged exposure to resistance training improves endurance performance in the long-term, a consequence of such training when unaccustomed is the appearance of exercise-induced muscle damage (EIMD). Exercise-induced muscle damage is well known to affect athletic performance requiring muscular strength and power; however, its effects on markers of endurance exercise are unclear. Therefore, the aim of this thesis was to investigate the effects of EIMD on endurance performance, with an emphasis on the physiological (oxygen uptake; , minute ventilation; ), metabolic (blood lactate; [La]), perceptual (rating of perceived exertion; RPE) and kinematic (stride length; SL, stride frequency; SF) responses during sub-maximal endurance exercise.

Over the last two decades, plyometric training has been extensively adopted by athletes, coaches and sport scientists with a primary aim to improve vertical jump height. The focus of these plyometric programmes has been to train the lower-extremity musculature in order to enhance jump performance. However, the lower-extremities are not the only contributing factor to vertical jump performance, as the use of an arm-swing during vertical jumping has also been shown to contribute to achieving maximum vertical jump height, yet training programmes for improving the arm-swing during the vertical jump are limited. Therefore, the primary aim of this thesis was to examine the full arm-swing mechanics during vertical jumping, and to then develop and assess the suitability of an upper-extremity plyometric programme for increasing both arm-swing kinematics and jump height. Firstly, a descriptive study was conducted to assess if an arm-swing countermovement was utilised during the vertical jump, which was deemed the prerequisite for using plyometric training to improve the arm-swing. Then an experimental study was conducted comparing vertical jumps performed with and without an arm-swing countermovement. The results showed that jumps performed with an arm-swing countermovement significantly increased mean peak shoulder angular velocity (ω) (+67.5 deg·s-1) and mean jump height (+ 6.2 cm) when compared to jumps performed using no arm-swing countermovement. During the final chapter of this thesis, a group of elite basketball players volunteered to participate in upper-extremity plyometric training aimed at increasing vertical jump height by training only the upper-extremities. Vertical jump height and full body kinematics were analysed using a 3 dimensional (3D) motion capture system, and key kinematic jump variables and various arm-swing performance measurements were collated both before and after a 4 week upper-extremity plyometric intervention. The use of upper-extremity plyometric training significantly increased the mean jump height (+ 7.2 cm), mean peak shoulder ω (+ 167.1 deg·s-1), mean peak frontal shoulder ω (+ 121 deg·s-1) and mean active range of motion at the shoulder joint (+ 5.3°), when compared to a control group. Furthermore, the use of a large active range of motion armswing during the arm-swing countermovement was shown to be the preferred arm-swing condition for increasing arm-swing kinematics. The increase in arm-swing kinematics and jump height after the 4 week upper-extremity plyometric programme was attributed to the participants’ improved ability to use the stretch-shortening cycle, elastic energy transfer system and stretch reflex system. Therefore, the use of upper-extremity plyometric exercises as part of a training regime for improving vertical jump performance should be advocated.

This study monitored a cohort of youth rugby league players from one professional club in England, across three competitive seasons (under-15 to under-17 age group). The aims were to establish which dimensions of growth and performance characterized players who were either coach-selected or unselected each season and to evaluate annual developments in growth and performance. It was also necessary to establish the credibility of various measurement techniques that are implicated in the talent identification process. In the assessment of sprint performance, GPS measurements systematically underestimated both distance and timing gate speed but can be used to reliably evaluate sprint performance, particularly for measurements of peak speed (95% Limits of Agreement (LoA) = 0.00 ± 0.8 km·h-1; CV = 0.78%). Using a larger sample of youth team sport players (n = 60), multiple linear regression analysis, incorporating mean and peak GPS speeds as predictors of timing gate speed, yielded a prediction model that was able to provide a valid alternative to timing gates in the assessment of sprint performance over 30 m. With regards to the reliability of assessments of sport-specific skill in youth rugby league players, no comparisons met the pre-determined analytical goal of ‘perfect agreement’, meaning that up to 56% of players’ skill could be misinterpreted. The credibility of such assessments should be questioned and alternative tests considered. In the period between the under-15 and under-16 group, there were large annual increments in speed (5.02 Δ%), force (13.82 Δ%) and power (19.85 Δ%) generated over 10 m sprint intervals and predicted vertical jumping power (13.02 Δ%), with concomitant developments in body mass (5.14 Δ%), lean body mass (3.20 Δ%) and predicted muscle of the quadriceps (10.12 Δ%). A discriminant function analysis also highlighted 30 m force and 10 m acceleration as significant predictors of selected players in the under-15 group and under-16 group, explaining 47.3% and 40.7% of the between-group variance, respectively – which was the case independent of age at peak maturity. However, there were 5 no differences between selected and unselected players in the under-17 group. During match time, there were differences between selected (57.1 ± 11.9 min) and unselected (44.1 ± 12.3 min) players for average playing interval in the under-16 group. In turn, selected players covered more total distance (5181.0 ± 1063.5 m c.f. 3942.6 ± 1108.6 m, respectively; P = 0.012) and high intensity distance (1808.8 ± 369.3 m c.f. 1380.5 ± 367.7 m, respectively; P = 0.011) than unselected players. When age at peak height velocity (PHV) was statistically controlled, only distance in zone 3 and summated-HR remained higher in the selected players of the under-16 group. Conversely, higher values amongst the unselected under-16 players for total and relative distance in zone 4, 5 and high intensity were revealed. There was a relationship in the under-15 group (R = 0.702, P < 0.001), under-16 group (R = 0.607, P < 0.001) and under-17 group (R = 0.671, P < 0.006) between the number of successful ball carries and 10 m sprinting force, thus supporting the use of 10 m sprinting force as a predictor of match performance. The relationship (r = 0.51, P = 0.044) between aerobic capacity and HIT·min-1 in the under-17 group also provides preliminary evidence of aerobic endurance as a potential predictor of match running intensity. It was concluded that players who are coach-selected are not characterized by match related performance variables but are offered greater match exposure during the under-16 age group, resulting in larger running distances. Unselected players are unrewarded for higher intensity running during matches when maturational age is statistically controlled and are also equally effective in regard to tackling and ball carrying outcomes. These results collectively indicate the inability of match performance measurements to contribute to talent identification processes in players of this type. The changes in growth and performance should be used to guide talent development practices of rugby league coaches. In particular, the assessment of force (i.e. the product of acceleration and body mass) should be considered as an important factor in differentiating between higher and lower ability players, as well as relating to match performance.

The aim of the present thesis was to examine the potential for acute carbohydrate-protein (CHO-P) ingestion to enhance performance and recovery from exercise designed to simulate the demands of multiple-sprint sports (MSSs). Chapter 3 of the thesis explored the inter- and intra-day reliability and concurrent validity of non-motorised treadmill ergometry (NMT) for the assessment of short-distance sprint performance [i.e. 10-30 m). There were no significant mean differences between NMT variables recorded on the same day or between days. Ratio limits of agreement indicated that the best agreement was in 20 [1.02 */-=- 1.09) and 30 m [1.02 */* 1.07) sprint times, peak [1.00 */T 1.06) and mean (0.99 */+ 1.07) running speed and step length (0.99 */-=- 1.09) and frequency (1.01 */+ 1.06). The poorest agreement was observed for time to peak running speed (1.10 */* 1.47). Significant differences were observed between NMT and over-ground sprint times across all distances, with times being lower (faster) by approximately 25-30% over-ground. The correlations between NMT and over-ground variables were generally modest (r5 = 0.44 - 0.67), and optimal for time to cover 30 m on Day 2 (rs = 0.8). Chapter 4 sought to examine the efficacy of CHO-P ingestion during 4 h of recovery from the Loughborough Intermittent Shuttle Test (LIST) when compared to CHO matched for energy (ISOEN) or CHO (ISOCHO) in a typical CHO beverage. There were significant increases over time in muscle soreness, and reductions in extensor and flexor peak torque (by approximately 9%, 9% and 8%, and 13 %, 13% and 11% at 60 deg-s-1) and jump performance (10%, 7% and 5%) with the ingestion of CHO-P, ISOEN and ISOCHO, respectively. Beverage type x time interactions were not significant for any of these variables, indicating that changes in each variable were similar for all groups. Decrements in sprint performance assessed on the NMT were typically small and not different between beverage types (<4%), although sprint times over 20 and 30 m remained elevated for 48 h post-exercise. Accordingly, Chapter 4 provided no clear evidence for a benefit of ingesting CHO-P in the hours after exercise to enhance recovery of muscle function and selected performance variables following MSS activity. Chapters 5 and 6 of the thesis aimed to examine the effect of CHO-P ingestion during simulated MSS exercise. In Chapter 5, it was observed that sprint times, HR and gut fullness increased over the course of the LIST, with no influence of consuming each of the different beverages. In contrast, there was a main effect of time (P < 0.001), and drink (P = 0.042) observed for RPE, which was lower (P < 0.001) during the LIST in the CHO-P condition (16.9 ± 1.4) than in either the ISOCHO (17.8 ± 1.1) or ISOEN (17.7 ± 1.3). However, time to exhaustion was not different (P = 0.29) between CHO-P (468.3 ± 268.5 s), ISOCHO (443.4 ± 286.3 s) and ISOEN (446.2 ± 282.08 s), although these times did equate to a non-significant mean improvement of 4% in the CHO-P trial. Chapter 6 demonstrated that during a modified version of the LIST with two self-regulated blocks of exercise intensity, participants had a higher average speed (8.1 ± 0.3 cf. 7.9 ± 0.5 knvlr1) during the final (self-regulated) 15 min block of the LIST in the CHO-P condition compared to CHO. Whilst the mechanisms for such an improvement are not certain, the attenuated rise in RPE observed in Chapter 5, and increased blood urea concentration observed in Chapter 6, with CHO-P ingestion may suggest altered central fatigue and/or increased protein oxidation enhances performance during MSSs.

The effectiveness of recovery interventions following prolonged multiple sprint team sports matches has rarely been studied despite the potential for exercise-induced muscle damage to adversely affect training in the days following games. The lack of research related to this topic is probably owing to the wide variability that exists in the movement demands of players between matches and the impact that this has on the subsequent rate and magnitude of recovery which makes it difficult to detect meaningful differences when conducting research with small sample sizes. Therefore, the purpose of this thesis was to develop a rugby league-specific match simulation protocol that replicates the movement demands, physiological responses and subsequent recovery from matches in order to study the effectiveness of recovery interventions. Hence, two time-motion analysis studies were conducted using a semi-automated image recognition system to inform the development of the rugby league match simulation protocol (RLMSP). Whilst mean total distance covered over the duration of the match was 8,503 m, ball in play and stoppage work-to-rest ratios were 1:6.9 and 1:87.4, respectively, for all players. Furthermore, a significant decline in high and very high intensity running locomotive rates were observed between the initial and final 20 min periods of the match. Thus a RLMSP was devised to replicate the overall movement demands, intra-match fatigue and recovery from a senior elite rugby league match. Not only was there a low level of variability in the movement demands during the RLMSP over consecutive trials, but with the exception of creatine kinase, the rate and magnitude of recovery following the RLMSP was similar to that that has been published following competitive matches. Therefore, the RLMSP devised in this thesis may be a more appropriate research tool for assessing the effectiveness of recovery interventions following match related exercise than following actual match play.