1. IntroductionRugby union match-play is characterised by high-intensity intermittent activity, that consists of sprints, collisions, accelerations, decelerations and high-speed running (HSR), interspersed with periods of lower intensity activity and recovery.1Wundersitz D.W.T. Josman C. Gupta R. et al.Classification of team sport activities using a single wearable tracking device.,2Cahill N. Lamb K. Worsfold P. et al.The movement characteristics of English Premiership rugby union players. When combined, the different types of running and collisions can be considered as the physical characteristics or external load of rugby union match-play.3Cunniffe B. Proctor W. Baker J.S. et al.An evaluation of the physiological demands of elite rugby union using Global Positioning System tracking software., 4Jones M.R. West D.J. Crewther B.T. et al.Quantifying positional and temporal movement patterns in professional rugby union using global positioning system., 5Roberts S.P. Trewartha G. Higgitt R.J. et al.The physical demands of elite English rugby union. Much of the published literature describing the physical characteristics of different levels of rugby competition has reported upon locomotor metrics derived from microsensor technology,4Jones M.R. West D.J. Crewther B.T. et al.Quantifying positional and temporal movement patterns in professional rugby union using global positioning system.,6Beard A. Chambers R. Millet G.P. et al.Comparison of game movement positional profiles between professional Club and senior international rugby union players.,7Cunningham D. Shearer D.A. Drawer S. et al.Movement demands of elite U20 International Rugby Union Players. whilst in contrast limited research has reported upon the collision characteristics of rugby union8Yamamoto H. Takemura M. Iguchi J. et al.In-match physical demands on elite Japanese rugby union players using a global positioning system. using microsensor technology derived metrics.Professional rugby union players can partake in up to 4 different competition levels during 1 season – this is evident in the Northern Hemisphere where players often compete in International, European, domestic and British & Irish Cup fixtures and competitions. The challenge is that the season fluctuates between these competition levels, and a player may partake in domestic, European and International fixtures, on consecutive weeks. The management of players through the various competitions is of high importance to practitioners. This type of comparison has been explored in rugby league, a different yet similar code to rugby union, where differences in physical characteristics across professional, semi-professional and junior-elite match-play have been described in the literature.9Performance analysis of professional, semi-professional, and junior elite rugby league match-play using global positioning systems. Distance covered, number of sprints and sprint distance were reported by McLellan and Lovell to be significantly different between professional rugby league match-play and junior-elite rugby league match-play.9Performance analysis of professional, semi-professional, and junior elite rugby league match-play using global positioning systems. These authors suggested that junior-elite rugby league match-play may not adequately prepare players to meet the physical characteristics of professional rugby league.9Performance analysis of professional, semi-professional, and junior elite rugby league match-play using global positioning systems. Gabbett has also compared the physical characteristics of rugby league players competing at three different competition levels.10Influence of playing standard on the physical demands of junior rugby league tournament match-play. A higher relative total distance in Division-1 and -2 competitions was identified, when compared to Division-3.10Influence of playing standard on the physical demands of junior rugby league tournament match-play. Gabbett also reported a lower number of relative collisions in Division-3, when compared to higher competition levels.10Influence of playing standard on the physical demands of junior rugby league tournament match-play. Gabbett suggests that a thorough understanding of the different physical characteristics across competition levels can be used to help plan training and prepare players for competition specific match-play, as well as to guide recovery strategies from different competitions.10Influence of playing standard on the physical demands of junior rugby league tournament match-play.There is limited published literature in rugby union exploring competition level differences. Some recent research has reported the differences in physical characteristics between senior-international match-play and under 20s match-play.11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. Cunningham and colleagues11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. reported that a number of microsensor technology derived metrics were higher in senior international backs, including – relative distance, high-metabolic load distance (HMLD), high metabolic load efforts (HMLEs), and heavy decelerations (3–4 m·s−2). When compared to U20 internationals, senior international forwards were reported to have a greater relative HMLD and number of severe decelerations (>4 m·s−2), but a lower high speed running (HSR) and sprint distance per minute.11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. Cunningham et al.11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. concluded that U20s rugby may act as a good “stepping stone” towards senior international rugby union match play – but a key consideration they discussed was the necessity for an increased body mass to successfully transition to senior international match-play. These authors did not however, describe the collision characteristics of senior-international match-play or under 20s match-play. Beard and colleagues explored differences in senior international rugby union and senior professional club rugby union match-play. The authors reported that a measure of repeated high intensity locomotive efforts (RHILEs) was able to distinguish the two levels of competition, with senior international match-play being characterised by a significantly higher number of RHILEs. All positional categories demonstrated a higher relative RHILE in international rugby match-play, compared with senior professional club match-play. The findings of Beard et al. also provide some information that could be used in the management of players in-season, as they can transition between international and senior club competitions. Knowledge of the potential differences in the physical characteristics of rugby union match-play associated with different competition levels has implications for the planning of training and recovery. As an example, research has indicated that an inverse relationship exists between running performance and collision activities in rugby union.12Tee J.C. Lambert M.I. Coopoo Y. Impact of fatigue on positional movements during professional rugby union match play. Furthermore, the association of post-match creatine kinase and collision activities in rugby union has also been investigated.13Roe G. Darral-Jones J. Till K. et al.The effect of physical contact on changes in fatigue markers following rugby union field-based training. Costello and colleagues14Costello N. Deighton K. Preston T. et al.Collision activity during training increases total energy expenditure measured via doubly labelled water. examined the energetic cost of collision activities during training in professional youth male rugby league players. Energy expenditure was significantly higher (mean difference: 1185 ± 132 kcal) with the inclusion of one collision session across a five day training period, compared to the same training schedule without the collisions.14Costello N. Deighton K. Preston T. et al.Collision activity during training increases total energy expenditure measured via doubly labelled water. The collision session involved 20 individual collision events, and the corresponding non-collision session replicated the kinematic and locomotor demand of the collision session exactly, without collisions (participants ran past each other without making contact).14Costello N. Deighton K. Preston T. et al.Collision activity during training increases total energy expenditure measured via doubly labelled water.The findings of both Cunningham et al.11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. and Beard et al.6Beard A. Chambers R. Millet G.P. et al.Comparison of game movement positional profiles between professional Club and senior international rugby union players. support the notion that the physical requirements, specifically metabolically demanding efforts, are greater in senior international rugby union match-play, when compared with U20s and senior professional club match-play. However, neither of these research studies included collision-related characteristics. Cunningham and colleagues reported the importance of body mass in senior international rugby union,11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. which will influence the metabolically demanding movements and collisions. Beard and colleagues refer to the need for future research to investigate RHILEs and the collision characteristics of different levels of rugby union competition, so that practitioners can optimally prepare players for the physical intensity of match-play. No research has explored or discussed the challenges faced by rugby union practitioners when players interchange between different competition levels within the same season.

Considering the limited number of published studies comparing and contrasting the physical characteristics of different professional rugby union match-play competition levels, the aim of this study was to evaluate whether differences in physical characteristics (running-related and collision-related) derived from microsensor technology exist between four different professional rugby union competition levels.

2. Methods

Elite male rugby union players from two senior professional clubs (n = 118; age = 24.7 ± 4.1 years; body mass = 101.6 ± 12.2 kg; height = 186.5 ± 7.0 cm) competing in the British and Irish Cup, Guinness Pro14, and European Rugby Champions Cup participated in this study. Participants from both clubs also included players who represented their country at senior international level. Participants were informed about the purpose of the study, the relevant study procedures and the potential applications of the study findings. Ethical approval was granted by the University Human Research Ethics Committee and all participants signed a consent form.

A total of 2869 data files from match-play were retrospectively included in the data set. The four different competition levels were: senior international rugby union (INT), European Rugby Champions Cup (ERCC), domestic international club competition involving clubs from Ireland, Wales, Scotland, Italy and South Africa (PRO14), and British and Irish Cup involving clubs from Ireland, England and Wales (B&I). This resulted in the following stratification of data files per competition; INT: n = 232, ERCC: n = 504, PRO14: n = 1545, B&I: n = 588.

Data were collected via microsensor technology – StatSports Apex Units (Apex Pro Series Pod, StatSports, Down, UK) (Firmware Version 2.45). The microsensor technology unit collects 10 Hz global positioning system (GPS) data (augmented to 18 Hz), accelerometer data at a rate of 600 Hz, magnetometer data at a rate of 10 Hz and gyroscope data at a rate of 400 Hz. This microsensor technology unit has demonstrated good inter-unit reliability for most metrics,15Thornton H.R. Nelson A.R. Delaney J.A. et al.Inter-unit reliability and effect of data processing methods of global positioning systems. and more recently has been identified to have very strong – nearly perfect relationships compared to video-coded events.16Tierney P. Blake C. Delahunt E. The relationship between collision metrics from microsensor technology and video-coded events in rugby union. All matches were played between September 2017 and May 2018. All players wore the same microsensor technology unit for the duration of the data collection period in an attempt to reduce inter-unit variation, and units were worn in a pocket stitched to the player's jersey. Units were turned on outdoors following manufacturer guidelines before the pre-match warm-up (at least 30 min before kick-off). All data traces from match-play were visually inspected. Any “velocity spikes” – including: a sharp increase in the velocity trace that exceeded the individual's maximum velocity; velocities greater than 12 m·s−1; and an unnatural rapid increase in velocity – were deleted from the participants' data, and remaining relevant data were recalculated. Data were processed via the proprietary StatSports Apex Rugby Software (Version 3.0.11171) and exported to Microsoft Excel for data storage. Statistical analyses and data visualisation, but not data processing, were completed in R (Version 3.6.1). The microsensor technology derived metrics collected in this study are described in Table 1, with a brief definition of each metric.

Table 1Microsensor derived variables and definitions.

Consistent with previously reported procedures11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. match data were timed so that microsensor technology unit output was extracted for total playing-time. Therefore, time off the field, including half time, injuries, sin-bins, and warm-ups were not included in the calculation of match intensity. Time-off during match play, for example injury time and stoppages during each half were included. All durations below 15 min were excluded from this analysis to remove outlier data that would demonstrate an extremely high intensity. This approach is consistent with a previously published study describing the physical demands of specific phases in rugby union.17Tierney P. Tobin D.P. Blake C. et al.Attacking 22 entries in rugby union: running demands and differences between successful and unsuccessful entries. Microsensor technology derived data for distances, counts and manufacturer metrics, as outlined in Table 1, were calculated to quantify ‘intensity’; all metrics were divided by minutes to produce ‘relative’ metrics.

One-way ANOVA was performed to identify if there were significant differences (p-value of <0.05) between any of the competition levels across the 12 microsensor technology derived physical characteristic dependent variables. Following the observation of a significant one-way ANOVA test result, Tukey Honest Significant Differences (HSD) tests were completed to perform multiple pairwise-comparisons within player positional groups (Prop, Hooker, Second Row, Back Row, Scrum Half, Out Half, Centre, Back Three) between competition levels (INT, ERCC, PRO14, B&I). p-values < 0.05 were determined to be statistically significant.

4. Discussion

The aim of our study was to evaluate whether differences in physical characteristics exist between 4 different professional rugby union competition levels. We believe that our study is the first to compare and contrast the physical characteristics of four different professional rugby union competition levels – the novelty lies in our inclusion of both running-related and collision-related metrics.

Collisions per minute and collision load per minute were significantly higher in INT match-play compared with ERCC, PRO14 and B&I match-play. Collision intensity was significantly lower in B&I match-play when compared with ERCC and PRO14 match-play. There was no significant difference in collision intensity between ERCC match-play and PRO14 match-play. The collisions per minute results compare favourably to those of rugby league match-play, whereby it has been reported that a higher frequency of collisions is experienced during higher levels of competition.10Influence of playing standard on the physical demands of junior rugby league tournament match-play. Anecdotally, a higher collision demand reflects what elite coaches and players often described when talking about an “increased level of intensity” at INT and ERCC competition levels. However, we cannot be certain as to whether the increase in collision load per minute is solely due to an increase in frequency (collisions per minute) or magnitude of the collisions. There is no published literature in rugby union that we can make direct comparisons with regard to the collision metrics that we describe in this study. However, the higher collision characteristics at INT competition levels are of particular importance to practitioners, who have to ensure the development of players' abilities to dominate the contact areas of the game whilst also overseeing their technical proficiency and running-related fitness. Our study develops upon the recommendations of Beard and colleagues6Beard A. Chambers R. Millet G.P. et al.Comparison of game movement positional profiles between professional Club and senior international rugby union players., who advised that further research is warranted, particularly related to repeated effort bouts (including tackling and collisions), in order to enhance practitioners' knowledge of the transitional gap between club and international competition levels. It is crucial that practitioners have an understanding of the differences in the collision characteristics between professional rugby union competition levels, especially when considering preparation phases for international tournaments.Both the INT and ERCC competitions were characterised by lower total distance- and HSR distance-per minute when compared to PRO14 and B&I match-play. The typical qualitative perceptive commentary that INT match-play is ‘more intense’ than club match-play (ERCC, PRO14, B&I) is not supported by the traditional locomotor metrics, namely total distance and HSR distance. Furthermore, our results challenge those of Cunningham et al.11Cunningham D.J. Shearer D.A. Drawer S. et al.Movement demands of elite under-20s and senior international rugby union players. who demonstrated a higher relative total distance in backs in INT, compared with U20 match-play, yet a lower relative distance in forwards. This is also consistent with research that has reported a decreasing trend for distance per minute and high-speed running distance per minute in older age groups for both forwards and backs in adolescent rugby union.18Read D.B. Jones B. Phibbs P.J. et al.Physical demands of representative match-play in adolescent rugby union. Therefore, our findings suggest that with an increase in competition level, there is a tendency for distance- and HSR distance-per minute to decrease, and collisions- and collision load-per minute to increase. This has implications for physical preparation, in particular locomotor intensity in training and preparation for the increase in collision intensity observed at the INT competition level.We observed that relative HMLD was significantly higher during PRO14 match-play compared with INT, ERCC and B&I match-play. However, relative HMLE was significantly higher during INT and ERCC match-play compared with PRO14 and B&I match-play. This provides an interesting area for discussion; it is plausible that less space to cover distance is available at INT and ERCC levels of match-play due to higher skill level, defensive structures, tournament tactics and style of play; this is represented when comparing HMLD and HMLEs. The number of RHILEs (not including collisions) has been reported to differ between senior international rugby union match-play and senior club rugby union match-play,6Beard A. Chambers R. Millet G.P. et al.Comparison of game movement positional profiles between professional Club and senior international rugby union players. which supports the premise that the physical characteristics of rugby union match-play are competition level dependant. Whilst the relative HMLE was higher during INT and ERCC match-play, the distance accumulated during these efforts was lower compared to PRO14 and B&I match-play, resulting in a lower relative HMLD (during INT and ERCC match-play). Seeing as the HMLE intensity was greater, but the HMLD intensity was lower, exploring the number of efforts relative to the distance covered (HMLEs divided by HMLD) would seem to be relevant.A substantial challenge for practitioners is the interchange of players between competition levels during the same season – as an example, it is possible that a player could be involved in two PRO14 matches, immediately followed by two ERCC matches and then an INT series of matches. During INT competitions, it is not uncommon for players to be “released” to participate in club training and PRO14 fixtures. This exposure of players to different levels of competition presents a challenge from a physical preparation planning and recovery perspective, especially considering that collision activities are associated with fatigue,13Roe G. Darral-Jones J. Till K. et al.The effect of physical contact on changes in fatigue markers following rugby union field-based training. and that INT competition can present a greater collision intensity compared to PRO14. Based upon the findings of Costello et al.14Costello N. Deighton K. Preston T. et al.Collision activity during training increases total energy expenditure measured via doubly labelled water. who reported an increase in energy expenditure with the inclusion of collision activities during a five day training week, it would seem rationale to acknowledge that the increased collision intensity that we report in INT competition would have substantial nutritional implications. The data in the current study may support load management decisions – for example, players that are “released” to play in the PRO14 that have been exposed to heavy collision-oriented match and training load in INT competition, and need to acutely prepare for a potentially higher high-speed running and distance per minute demand. The design of training programmes in conjunction with coaches would have significant impact on the type of physical loading that players would be exposed to, and inversely, the recovery strategies implemented acutely between competition levels could see an important advantage for player health. The collaboration of practitioners between competition levels and teams would seem of the utmost importance.

We did not observe any differences in accelerations >2 m·s−2 and decelerations >2 m·s−2 between competition levels; this would suggest that this lower threshold of accelerations and decelerations (2 m·s−2) is not sufficiently sensitive to detect differences in the physical characteristics across competition levels. We suggest that the count of accelerations and decelerations at this lower intensity holds little practical utility and should not be used for competition level comparisons, however acceleration and deceleration events of a higher magnitude (>3 m·s−2) may offer a differing view, given the overall trend for acceleration intensity (>3 m·s−2) to decrease as the competition level increases, and in particular for the back three positional group. However, practitioners should be cognisant of the magnitude of variability of acceleration and deceleration count intensities derived from microsensor technology, when interpreting their own data. Our data suggests that characteristics derived from inertial sensors (and in particular collision-related characteristics) may offer greater utility in differentiating the physical characteristics associated with different professional rugby union competition levels.

There are a number of limitations that warrant consideration. There is no data for the Scrum Half position at the INT level. This is due to the fact that a particular player from a team not included in this study played an overwhelmingly large percentage of INT fixtures during the study period. Our data represents the average physical characteristics of match-play for each competition level. We did not consider the peak physical characteristics of match-play for each competition level. It is probable that the peak physical characteristics of match-play are different across competition levels. We suggest that future research should compare and contrast the peak physical characteristics of different rugby union competition levels. Our data may be population specific, and could be influenced by coaching style at various competition levels. However, the inclusion of data from two separate professional teams and 2869 data files from match-play should give practitioners some level of confidence in the results described. Due to data sensitivity with competing teams within the dataset, we were unable to access potentially identifiable data in this dataset (player ID and date), and as such our analysis is limited at exploring the interaction of these on the data.