Rugby league is a high-speed, intermittent team sport, played over a period of 80 minutes, with players covering between 3,000 and 8,000 m, depending on their playing position (14,17,21,23,30,36). Although match running performance is intermittent in nature, the importance of high-speed running has become increasingly apparent for elite players (1,12,13,16,28,31,32,35). For example, a reduction in high-speed running of approximately 5–12% between the first and second halves of matches has been reported for elite rugby league players (18,20,28,31,35). Decrements in high-speed running seem to be a characteristic of elite rugby league matches, with subelite players showing no change between halves (28). Furthermore, elite National Rugby League (NRL, Australia) players have been reported to run at higher intensities (∼21% greater distance at high speed) than Super League players (United Kingdom) and, although reductions in high-speed running were apparent between halves of a match in both groups, NRL players were able to preserve their running speed by ∼23% more than that of Super League players (32). These findings were suggested to reflect the superior playing standard of NRL players (32). Over 50% of repeated high-speed efforts occur within 5 minutes of a try and in close proximity to other “critical” periods during elite rugby league matches (13). Collectively, the evidence suggests that the capacity to sustain high-speed running, particularly during critical match periods, is likely to facilitate the performance of elite rugby league players.
A reduction in high-speed running distance during other team sports, such as soccer, was originally reported in the 5-minute period that immediately followed the most intense (i.e., peak in high-speed movement) period of the match, suggesting that soccer players do not maintain high-speed running for sustained periods of time (3,5,24). Similar observations have since been made in rugby league matches and are thought to indicate “transient fatigue or pacing” (16,18,20,35). For example, Hulin et al. (16) reported reductions of approximately 50% (∼17 m·min−1 vs. ∼8 m·min−1) in high-speed running between the peak and subsequent periods of matches among rugby league players. Two techniques have been used to identify peak periods in team sports matches; one uses the highest running intensity recorded across consecutive 5-minute periods (3,4,16,24), whereas the other uses a rolling average (8,10,33). Although the former of these techniques seems to underestimate the latter, perhaps because of an overlap in peak running activity between consecutive 5-minute periods (33), both are able to detect transient fatigue during matches. From a metabolic perspective, this should be anticipated, given the limited capacity of nonoxidative energy systems, which are relied on to support high-speed exercise (29). Reductions in high-speed exercise during matches have been attributed to a variety of mechanisms that are both centrally and peripherally mediated (3–5,24,34,35). These temporary reductions in high-speed movement have many potential consequences in a match scenario. For example, the ability to perform demanding defensive duties for prolonged periods, yet retain the ability to execute subsequent high-speed offensive plays such as counterattacks, can shift the “momentum” of a match. However, the ability to perform these actions could be compromised if they were to follow an intense period of activity. Acute fatigue induced after a high-speed bout could also expose players to an increased risk of injury (11). Because of the potential deleterious effects of transient fatigue during matches in rugby league, specific high-speed training programs are often adopted to prepare players for a “worst-case scenario” during matches (1,2,17).
In a recent study, Kempton et al. (20) reported differences between the peak and subsequent 5-minute periods during elite rugby match performance to reaffirm the presence of transient fatigue. Replicating a previous technique applied in elite soccer (6), the authors also investigated the effect of “contextual” factors, such as ball-out-of play time. Notably, Kempton et al. (20) reported that ball-out-of play time was lower during the 5-minute peak periods compared with the subsequent 5-minute periods of running speed. Accordingly, it was suggested that the observed high-speed running reductions could be explained by the time that the ball was in play, rather than an indication of transient fatigue or pacing (20). A reduction in ball-in-play time might have been anticipated, given the known influence of external factors, such as team tactics and opponents activity on movement profiles in team sport (22,27). It is possible that ball-in-play time has a particularly strong influence on the movement parameters of rugby league players, as these periods could be used for recovery in preparation for the subsequent phase of play. Moreover, there is potentially little advantage to performing high-speed movements when the ball is not in play, because of the positional and territorial restrictions of the game, which draws players toward the gain line before a restart. Fundamental features of rugby league, such as the offside rule, also limit the space available for players to move into when the ball is not in play. These suggestions logically question the existence of transient fatigue.
The findings of previous studies could have implications for a large body of research in team sport that has attempted to validate the phenomenon of transient fatigue. However, no study has attempted to explore the relationships between ball-in-play time and high-speed running during peak and subsequent 5-minute periods in elite rugby league matches. Furthermore, no study has adopted the correct statistical procedure to directly evaluate of the effects of ball-in-play time on transient fatigue during team sports matches. For example, it is possible to statistically control for confounding variables, assuming that they have an influence on the dependent measures. Ball-in-play time is one of the suggested confounding variables and ought to be treated in this way. However, covariates must meet specific assumptions before they can be considered as such. One example of these assumptions is the presence of a relationship between the candidate covariate (ball-in-play time) and dependent variable (high-speed running), which is an analysis that can be simply performed using a statistical software package. Given the potential implications of previous findings, it is important that a correct statistical approach is adopted for corroboratory purposes. Therefore, the first aim of this study was to evaluate the differences in high-speed running between peak 5 minutes, subsequent 5 minutes, and mean 5 minutes of elite rugby league matches. The second aim of this study was to establish the relationship between high-speed running and ball-in-play time during 5-minute peak and 5-minute subsequent periods, thus establishing the role of ball-in-play time as a covariate. The findings of this study will be useful to coaches in the design of appropriate conditioning practices that more closely replicate the movement characteristics of match play.
Experimental Approach to the Problem
To determine the presence of transient fatigue, 5-minute periods of peak high-speed running were compared with the subsequent 5-minute period and the match mean 5-minute period for each player during the 73 performances. The same comparisons were performed on ball-in-play time to determine whether this was affected by match periods. Although it is known that between-match variability can be high (∼14% in rugby league matches) (19), this study followed a repeated-measures design and did not compare between matches, teams, or players. A post hoc power calculation (G*Power, Version 22.214.171.124; Universität Kiel, Kiel, Germany) showed that with an alpha level of p ≤ 0.05, the current sample (n = 73) was large enough to detect the smallest of significant effect sizes (ES = 0.5) with a power of 0.97. A correlational analysis was also performed to confirm the assumed relationship between ball-in-play time and high-speed running, as this is currently thought to confound the typical analyses of transient fatigue in team sports, such as rugby league.
After institutional ethics approval, 24 male rugby league players from the same professional Super League club participated in the study (mean ± SD: 18.9 ± 1.2 years, 1.83 ± 0.05 m, and 95.9 ± 9.7 kg). All players trained 3 times per week, consisting of technical/skill work and strength and conditioning. Data were collected from 5 different matches, providing a total of 73 match files for analysis. Of the 5 matches analyzed, 4 were won and 1 was lost, with a mean score deficit of 23 ± 18 points. Both the club, players, and their parents or guardians (where necessary) were fully informed of the procedures and provided written informed consent for their data to be included in the study. Among the squad, the players were divided into positional groups as follows: props (n = 5), back row (n = 5), outside backs (wingers, centers, and full back; n = 10), and pivots (stand-off, scrum-half, and hooker; n = 4). Players who did not participate for at least 5 minutes subsequent to their 5-minute peak high-speed period were excluded from the data analysis. The study was approved by the University of Chester Institutional Review Board.
Each participant was fitted with a 5-Hz nondifferential Global Positioning System (GPS, SPI-Pro; GPSports, Canberra, Australia), which was securely placed into a purpose-made sleeveless vest (GPSports) between the scapulae. The GPS data were downloaded using Team AMS (GPSports) software and exported to a purpose-designed spreadsheet for further analysis. High-speed running (>14.4 km·h−1) (27,31,35,36), expressed as meter per minutes, was recorded for each 5-minute segment of the match, with the highest locomotive rate being taken as the “peak period” (i.e., the highest predefined period), the 5-minute period that followed being recorded as the “subsequent period,” and the mean of all 5-minute periods taken as the “mean period” (24). The current demarcation of high-speed has been used previously to define match-related fatigue in rugby league matches (31,35). The GPS devices are commonly used to demarcate high- (>14.4 km·h−1) and low-speed zones (<14.4 km·h−1) (18,34,35) and are reliable enough to detect changes in high-speed running between match periods (26,37). The same GPS units were fitted to each player throughout all matches, thus controlling for any potential interunit variability (9). The amount of time (seconds) the ball was in play (ball-in-play time) for each of the 5-minute periods (peak, subsequent, and mean) was analyzed after the match using a video recording and a manual notational analysis procedure (Sportscode Gamebreaker, Version 8.2.1; Sportstec, New South Wales, Australia). The intrarater reliability of the analyst, expressed as a coefficient of variation, was 2%. The same experienced analyst was used throughout the study.
The data were initially checked for normality using a Kolmogorov–Smirnov test and homogeneity of variances. To consider ball-in-play time as a covariate, relationships between ball-in-play time and high-speed running at each level of the independent variable (peak, subsequent, and mean periods) were evaluated using Pearson's correlation coefficient (r). The strength of the relationships was considered as the following: <0.3 = weak, 0.3–0.5 = moderate, and >0.5 = strong (7). The effects of other factors, such as field position, were not included in the hypothesis of this study and were not deemed relevant to the current analysis because transient fatigue has not been shown to depend on this factor. A 1-way analysis of variance with repeated measures (ANOVA-RM) was used to compare high-speed running and ball-in-play time between peak, subsequent, and mean periods of all matches. Follow-up pairwise differences were identified using paired t tests to remove the bias of correction methods (25), and ESs were calculated and interpreted as the following: ≤0.2 = trivial, >0.2 = small, >0.6 = moderate, >1.2 = large, >2.0 = very large, and >4.0 = extremely large (15).
Secondary to traditional hypothesis testing, magnitude-based inferences (MBIs) were used to identify clinical differences in the dependent variables between peak, subsequent, and mean 5-minute periods. Raw data were log transformed to account for nonuniformity of effects. For the MBIs, threshold probabilities for a substantial effect based on the 90% confidence limits (15) were as follows: <0.5% = most unlikely, 0.5–5% = very unlikely, 5–25% = unlikely, 25–75% = possibly, 75–95% = likely, 95–99.5% = very likely, and 99.5% = most likely. Thresholds for the magnitude of the observed change in the dependent variables were determined as the within participant SD × 0.2 (small), 0.6 (moderate), and 1.2 (large) (15). Effects with confidence limits across a likely small positive or negative change were classified as unclear (15). The uncertainty of effects was based on 90% confidence limits for all variables. A custom spreadsheet was used to perform all calculations (http://www.sportsci.org/).
The homogeneity of regression slopes, an assumption of analysis of covariance with repeated measures (ANCOVA-RM), was analyzed using a customized model. Statistical significance was set at p ≤ 0.05, and data were reported as mean ± SD unless otherwise stated. All data were analyzed using Statistical Package for the Social Sciences (SPSS for Windows, version 18.0; SPP Inc., Chicago, IL, USA).
Relationships Between Ball-In-Play Time and High-Speed Running
There was no relationship between ball-in-play time and high-speed running for peak (r = −0.01, 95% confidence interval [CI] [0.24–0.22], p = 0.934), subsequent (r = 0.11, 95% CI [−0.12 to 0.30], p = 0.360), or mean 5-minute periods (r = −0.13, 95% CI = [0.35–0.10], p = 0.27) (Figures 1–3).
Comparison of 5-minute Periods
There was an effect of match period for high-speed running (F(1.6,117.4) = 291, p < 0.001), with post hoc tests revealing differences (p < 0.001; ES = 1.6, 95% CI [1.1–2.1]; most likely lower; mean difference = 14.6, 95% CI [12.6–16.5] m·min−1) between the peak and subsequent, peak, and mean (p < 0.001; ES = 1.7, 95% CI [1.4–2.0]; most likely lower; mean difference = 13.7, 95% CI [11.7–14.6] m·min−1) but not between the subsequent and mean periods (p = 0.201; ES = −0.3, 95% CI [−0.5 to −0.1]; unlikely higher; mean difference = −1.4, 95% CI [−3.0 to 0.1] m·min−1) (Figure 4). For ball-in-play time, there was an effect of match period (F(1.5,109.4) = 19.5, p < 0.001), with post hoc tests revealing differences between the peak and subsequent (p < 0.001; ES = 0.8, 95% CI [0.5–1.4]; most likely lower; mean difference = 30, 95% CI [18–43] seconds), peak, and mean periods (p = 0.036; ES = 0.5, 95% CI [0.3–0.7]; most likely lower; mean difference = 13, 95% CI [4–22] seconds) and the subsequent and mean periods (p = 0.002; ES = 0.6, 95% CI [0.4–0.8]; most likely higher; mean difference = −18, 95% CI [−27 to −9] seconds) (Figure 5).
The assumptions of repeated-measures ANCOVA were violated based on the nonsignificant relationship (p > 0.05) between ball-in-play time and high-speed running (Figures 1–3). Although homogeneity was reported between the regression slopes, signified by no interaction between the match period and ball-in-play time (p = 0.504), the covariate and the treatment effect were not independent of one another, denoted by the differences in ball-in play time between match periods (Figure 5).
Using the same statistical approach to others (18), there were differences (“most likely”) in high-speed running between peak 5-minute and subsequent 5-minute periods of rugby league matches (Figure 4). However, there were no differences in high-speed running between the subsequent 5-minute and the mean 5-minute periods (Figure 4). Similar conclusions have been reached previously among rugby league players (18). These findings show that high-speed running cannot be maintained for prolonged periods but that subsequent bouts of running do not necessarily need to be performed below the match mean. Although it is more common to observe differences in high-speed running between subsequent and mean 5-minute periods in team sports, it is possible that players reduced their running speed in response to afferent cues, such that the mean physical demands can be recommenced (34). Collectively, the current findings are largely as anticipated and indicate the occurrence of transient fluctuations in match running that are reflective of fatigue or the attempt to avoid a reduction in match running below the match mean.
Because of recent suggestions that ball-in-play time directly influences high-speed running fluctuations between 5-minute match periods (6,20), the analysis was extended to account for ball-in-play time as a covariate. However, the principal finding of this study is that ball-play-time has no influence on the fluctuations in high-speed running between peak 5-minute and subsequent 5-minute periods of rugby league matches. There was no relationship observed between ball-in-play time and high-speed running for any match period. This study is the first to recognize that transient fatigue can be analyzed in rugby league matches without accounting for ball-in-play time. The authors would encourage others to take similar statistical steps to establish the generalizability of our findings to a larger population. If similar results are found, ball-in-play time could be disregarded when evaluating transient fatigue during matches. Other approaches, such as multilevel modeling can also be used in a similar manner to account for the potential influence of other contextual factors on transient fatigue.
It is understandable, and entirely logical, to question the potential effects of ball-in-play time on high-speed running, given the potential opportunity for players to recover during this time. Consistent with previous reports (20), the current findings show that ball-in-play time is also different between peak, subsequent, and mean 5-minute periods. Although this might seem to show that ball-in-play time directly influences high-speed running, these assumptions are misleading and are not based on a suitable statistical approach. That is, despite corresponding systematic mean changes in both of these variables across match periods, there is no proportional relationship between them. In other words, the ball being in or out of play does not determine the high-speed running of a rugby league player.
Previous authors have recognized that reduced ball-in-play time in the 5-minute subsequent (after peak) period could be related to a conscious tactical decision made by players to kick the ball into touch, ostensibly to reduce the demands of the match (15,16). This could also be considered as part of a complex pacing strategy, used by players to avoid damaging physiological “failure,” thus guarding against a more pronounced physical expression of fatigue (2,4,34). Indeed, it is extremely likely that some rugby league players do perform lower speed running during this period and use this time to recover from a high-speed bout of running. However, with the current findings in mind, it is equally as likely that some players use the ball-out-of play time to reorganize their field positions for tactical purposes, resulting in higher speed movements rather than lower speed recovery. This might have been anticipated as it is extremely unlikely that all players recover at the same time, and it might be necessary for some players to provide cover for others who have been more involved in the match in the preceding period. Therefore, reductions in ball-in-play time do not necessarily rule out high-speed running. Future research is necessary to establish the effects of other “contextual factors,” such as defensive or offensive periods (20), while adopting a suitable statistical approach.
This study has reaffirmed the presence of transient fatigue during elite rugby league matches and questioned the influence of ball-in-play time as a confounding factor. Therefore, rugby league practitioners who currently adopt training practices, intending to extend bouts of high-speed running and facilitate transient recovery, should not be discouraged from their approach. For tactical reasons, rugby league players and coaches should also be aware that not all players are recovering when the ball is out of play and, by contrast, might be performing high-speed running. This has further implications for training practices that use ball-out-of play periods to permit recovery, as this is not reflective of every match scenario.
The current findings question the validity of training drills designed to mimic ball-out-of-play periods by transiently lowering running demands. Practitioners could use this information, alongside published data, to design training drills that distribute running demands of matches more accurately and encourage high-speed running periods when the ball is not active. For example, coaches might consider having players perform repeated effort sprints between bouts of game-specific activities rather than allowing players to walk or rest. This is potentially quite important as the intensity of training drills (such as phase plays and conditioned games) typically focus on ball-in-play scenarios and do not consider the intensity of running demands when the ball is “dead.” This often includes movement in different areas of the pitch and using different forms of locomotion. In addition, strength and conditioning practitioners who quantify transient fatigue during matches and use this as a performance indicator do not need to statistically control for ball in play time to do so.
1. Austin D, Gabbett TJ, Jenkins DJ. Repeated high-speed exercise in a professional rugby league. J Strength Cond Res 25: 1898–1904, 2011.
2. Black G, Gabbett TJ. Match speed and pacing strategies in rugby league: An examination of whole-game and interchange players, and winning and losing teams. J Strength Cond Res 28: 1507–1516, 2013.
3. Bradley PS, Di Mascio M, Peart D, Olsen P, Sheldon B. High-speed activity profiles of elite soccer players at different performance levels. J Strength Cond Res 24: 2343–2351, 2010.
4. Bradley PS, Noakes TD. Match running performance
fluctuations in elite soccer: Indicative of fatigue, pacing or situational influences? J Sports Sci 31: 1627–1638, 2013.
5. Bradley PS, Sheldon W, Wooster B, Olsen P, Boanas P, Krustrup P. High-speed running in English FA premier league soccer matches. J Sport Sci 27: 159–168, 2009.
6. Carling C, Dupont G. Are declines in physical performance associated with a reduction in skill-related performance during professional soccer match-play? J Sport Sci 21: 63–67, 2011.
7. Cohen J. Statistical Power Analysis for the Behavioral Sciences. (2nd ed.). Hillsdale, NJ: Erlbaum, 1988. pp. 19–74.
8. Delaney J, Tannath SJ, Thornton HR, Bennett KJ, Gay D, Duthie GM, Dascombe BJ. Establishing duration-specific running intensities from match-play analysis in rugby league. Int J Sport Physiol Perf 10: 725–731, 2015.
9. Duffield R, Reid M, Baker J, Spratford W. Accuracy and reliability of GPS devices for measurement of movement patterns in confined spaces for court-based sports. J Sci Med Sport 13: 523–525, 2010.
10. Furlan N, Waldron M, Shorter K, Gabbett TJ, Mitchell J, Fitzgerald E, Osborne MA, Gray AJ. Running-intensity fluctuations in elite rugby sevens performance. Int J Sport Physiol Perf 10: 802–807, 2015.
11. Gabbett TJ. Influence of fatigue on tackling technique in rugby league players. J Strength Cond Res 22: 625–632, 2008.
12. Gabbett TJ. Influence of playing standard on the physical demands of professional rugby league. J Sport Sci 31: 1125–1138, 2013.
13. Gabbett TJ, Gahan CW. Repeated high-speed effort activity in relation to tries scored and conceded during rugby league match-play. Int J Sport Physiol Perform 13: 530–534, 2016.
14. Gabbett TJ, King T, Jenkins D. Applied physiology of rugby league. Sport Med 38: 119–138, 2008.
15. Hopkins WG, Marshall SW, Batterham AM, Hanin J. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 41: 3–13, 2009.
16. Hulin BT, Gabbett TJ, Kearney S, Corvo A. Physical demands of match play in successful and less-successful elite rugby league teams. Int J Sport Physiol Perform 10: 703–710, 2015.
17. Johnston RD, Gabbett TJ, Jenkins DG. Applied sport science of rugby league. Sport Med 44: 1087–1100, 2014.
18. Kempton T, Sirotic A, Cameron M, Coutts A. Match-related fatigue reduces physical and technical performance during elite rugby league match-play: A case study. J Sport Sci 31: 1770–1780, 2013.
19. Kempton T, Sirotic AC, Coutts AJ. Between match variation in professional rugby league competition. J Sci Med Sport 17: 404–407, 2013.
20. Kempton T, Sirotic AC, Coutts AJ. An integrated analysis of match-related fatigue in professional rugby league. J Sport Sci 33: 39–47, 2015.
21. King T, Jenkins D, Gabbett TJ. A time-motion analysis of professional rugby league match-play. J Sport Sci 27: 213–219, 2009.
22. Lago C, Casais L, Dominguez E, Sampaio J. The effects of situational variables on distance covered at various speeds in elite soccer. Eur J Sport Sci 10: 103–109, 2010.
23. McLellan CP, Lovell DI, Gass GC. Performance analysis of elite rugby league match play using global positioning systems. J Strength Cond Res 25: 1703–1710, 2011.
24. Mohr M, Krustrup P, Bangsbo J. Match performance of high-standard soccer players with special reference to development of fatigue. J Sport Sci 21: 519–528, 2003.
25. Nakagawa S. A farewell to Bonferroni: The problems of low statistical power and publication bias. Behav Ecol 6: 1044–1045, 2004.
26. Petersen C, Pyne D, Portus M, Dawson B. Validity and reliability of GPS units to monitor cricket-specific movement patterns. Int J Sport Physiol Perform 4: 381–393, 2009.
27. Rampinini E, Coutts AJ, Castagna C, Sassi R, Impellizzeri FM. Variation in top level soccer match performance. Int J Sport Med 28: 1018–1024, 2007.
28. Sirotic AC, Coutts AJ, Knowles H, Catterick C. A comparison of match demands between elite and semi-elite rugby league competition. J Sport Sci 27: 203–211, 2009.
29. Spencer M, Bishop D, Dawson B, Goodman C. Physiological and metabolic responses of repeated-sprint activities specific to field-based team sports. Sport Med 35: 1025–1044, 2005.
30. Sykes D, Twist C, Hall S, Nicholas C, Lamb K. Semi-automated time-motion analysis of senior elite rugby league. Int J Perform Anal Sport 9: 47–59, 2009.
31. Sykes D, Twist C, Nicholas C, Lamb K. Changes in locomotive rates during senior elite rugby league matches. J Sport Sci 29: 1263–1271, 2011.
32. Twist C, Highton J, Waldron M, Edwards E, Austin D, Gabbett TJ. Movement demands of elite rugby league players during Australian National Rugby League and European Super League matches. Int J Sport Physiol Perform 9: 925–930, 2014.
33. Varley MC, Elias GP, Aughey RJ. Current match-analysis techniques' underestimation of intense periods of high-velocity running. Int J Sport Physiol Perf 7: 183–185, 2012.
34. Waldron M, Highton J. Fatigue and pacing in high-speed intermittent team sport: An update. Sport Med 44: 1645–1658, 2014.
35. Waldron M, Highton J, Daniels M, Twist C. Preliminary evidence of transient fatigue and pacing during interchanges in rugby league. Int J Sport Physiol Perform 8: 157–164, 2013.
36. Waldron M, Twist C, Highton J, Daniels M, Worsfold P. Movement and physiological match demands of elite rugby league using portable Global Positioning Systems. J Sports Sci 29: 1223–1230, 2011.
37. Waldron M, Worsfold P, Twist C, Lamb K. Concurrent validity and test–retest reliability of a global positioning system (GPS) and timing gates to assess sprint performance variables. J Sport Sci 29: 1613–1619, 2011.