Video-based perceptual training as a method to improve reactive agility performance in rugby union players

Abstract

This study investigated the effectiveness of rugby-specific video-based perceptual training on the speed and agility of club-level rugby union players. Twenty-six players were randomly divided into either a video-based, field-based or conventional training group. For six weeks, the video-based and field-based groups performed agility training twice a week in addition to their regular rugby training, while the control group was restricted to regular rugby training. Speed, reactive agility and change of direction speed were tested pre- and post-intervention, and after a six-week retention period. Results revealed that both video- and field-based training were almost certainly more beneficial (13% and 17%, respectively) to improve reactive agility compared to conventional rugby training alone. In addition, both intervention groups maintained their reactive agility improvements following the retention period. Consequently, video-based training, likely because of a perceptual component, could be used as an alternative training method to improve reactive agility.

Keywords

Change of direction speed decision making rugby football

Introduction

The multifaceted sport of rugby union is classified as an invasion game that involves frequent bouts of high intensity activity.^1,2 Teams that can maintain possession, are able to advance beyond the advantage line and defend effectively are more successful. In competitive match-play, there is a fair amount of unpredictability and multiple interactions between team mates and their opponents. Successful play requires the players to correctly interpret and respond to these changing situations as quickly and accurately as possible. Accordingly, perceiving the performance-relevant information or cues in this ever-changing context not only influences decision making, but also reaction time. It has been suggested that these responses to various stimuli are a component of agility.^3–6

Evasive agility manoeuvres lead to 72% of successful tackle breaks in rugby union.⁷ It has been shown that players who receive the ball at high speeds and who used evasive side-stepping manoeuvres were more likely to be successful at tackle breaks and advancing beyond the advantage line. Furthermore, these types of play often lead to positive phase outcomes, i.e. scoring a try.⁸ Agility is an open motor skill and does not rely purely on change of direction speed (CODS). Reactive agility (RA) is initiated by a certain stimulus and therefore, which is relevant to the context, and is therefore influenced by perceptual and decision-making skills. Perceptual skills are the ability to detect and interpret relevant cues or information, and as a result produce the appropriate responses to sport-specific situations. Thus, RA is defined as a rapid whole-body change of direction or velocity in reaction to a stimulus.⁴

Various researchers reported that measures of RA can distinguish between higher-skilled and lesser-skilled players, including Australian Rules football,⁹ rugby league,³ netball¹⁰ and soccer.¹¹ These findings attributed skillfulness of highly skilled players to their advanced level of perceptual and decision-making abilities.^3,12 Highly skilled players do not necessarily demonstrate better visual skills, but rather their ability to select the most relevant visual cues are more developed.¹³ If players improve the perceptual component of RA in a rugby-specific manner, they may be able to recognize the advanced visual cues associated with the evasive manoeuvres and improve their defences, as well as improve their ability to out manoeuvre their opponents during attack. Successful tackle breaks are strongly associated with top-performing rugby union teams, while evasive manoeuvres have been identified as a key element in leading to successful tackle outcomes.^7,8

According to the models of Young, James and Montgomery¹⁴ and Sheppard and Young,⁴ agility performance may be enhanced if the player has specific knowledge of the given situation, has the ability to recognize patterns of play and is able to spot advanced visual cues. Abernethy¹⁵ also stated that success in open-skilled sport, such as rugby union, is determined by effective decision-making strategies. Furthermore, Young and Wiley⁹ and Farrow et al.¹⁰ showed that the superior decision-making skills of highly skilled players correlate with RA performance. Gabbett and Abernethy¹⁶ found that rugby league players with more playing experience (number of matches and years’ experience) had superior anticipatory skills, suggesting that rugby-specific training may promote perceptual skill development.

RA can be trained through a variety of methods, including video-based training (VT). Previously Serpell et al.¹⁷ found that VT improved RA in rugby league players. Previous studies reported on the effectiveness of perceptual VT in a number of sports including soccer goalkeeping,¹⁸ tennis,¹⁹ basketball²⁰ and soccer.¹¹ The current study is the first to examine VT in rugby union. This type of training includes perception–action coupling so that the players learn to produce the most appropriate action relative to how they perceive a situation.²¹ With sport-specific training, players gain appropriate knowledge based on the constant changing environment presented to them and which simulate game-specific situations in an open-skilled sport such as rugby union. Thus, sport-specific training enables players to produce more effective responses, which may transfer to match-play.^22,23

Therefore, the study set out to develop a video-based RA training programme specifically for rugby union, and to compare this intervention with a field-based RA training programme. Furthermore, a secondary objective was to assess the durability of the changes in RA after a six-week retention period.

Method

Participants

A total of 42 male club-level rugby union players, aged 19–23 years, volunteered to participate in the study. Of these, 26 players completed the entire study. Ten players dropped out due to illness or injury and a further five players failed to complete the minimum number of required training sessions. The players were first divided into forwards and backs and then randomly assigned to the VT group (n = 10), the FT group (n = 9) and the control group (C; n = 7). There was an equal distribution of forwards and backs in each group. The study was approved by the Institutional Ethical Review Committee (Humanities) of Stellenbosch University and written informed consent was obtained from each player prior to study enrolment.

Experimental design

The study interventions formed part of the players’ pre-season conditioning programme. All three groups attended four club practice sessions per week in the afternoon, and each session lasted approximately 90 minutes. Additionally, players in the VT and FT groups completed two RA training sessions per week over a six-week period. The players in the VT group completed their sessions individually and indoors, while the players in the FT group trained as a group on an outdoors rugby field. The sessions were scheduled in the morning and separated by at least 48 hours. After the intervention period, all players participated in rugby training and matches prior to the retention tests six weeks later.

Tests and measurements

For descriptive purposes, anthropometrical assessments were performed on the players, which included bio-electrical impedance analysis (Bodystat® Quadscan 4000, UK) to assess percentage body fat and lean body mass (%) together with stature (cm) and body mass (kg). To be included in the study, players needed to obtain at least 60% of the calculated normative data used by the Western Province Rugby Union in a multistage shuttle run and repeated sprint test. All the players were able to achieve higher than the minimum score required. Table 1 summarizes the specific inclusion for each test respectively.

Table 1.

Inclusion criteria for the various playing positions.

Position	Multistage shuttle run test	Repeated sprint test (m)
Backline	Level 9.4	480–490
Scrumhalf	Level 9.4	490–500
Loose forwards	Level 9.1	460–470
Locks and hookers	Level 8.7	440–450
Props	Level 7.8	420–430

Speed, agility and RA

The Oliver and Meyers²⁴ test battery includes three tests, namely a 10-m linear sprint, a 10-m planned change of direction agility sprint, with a single change of direction to the left and right, as well as a 10-m RA sprint to the left and right. The total time taken to complete these tests was recorded and used as the dependent variable. Green, Blake and Caulfield²⁵ described this test as a reliable field test to assess speed, CODS and RA in rugby union players (ICC = 0.88). The Fusion Sport photoelectric timing gates (Smartspeed®, Fusion Sport, Brisbane, Australia) were used to record sprint times. D’Auria et al.²⁶ reported that the typical error of measurement of the Smartspeed® system is ∼ 0.03s, with a coefficient of variance of 1.7% for 10-m assessments. Unlike most commercially available timing systems, this system’s reliability may be attributed to the error correction processing and on-board timing functions as explained by Earp and Newton.²⁷

Three timing gates were placed in a straight line at 0 m, 5 m and 10 m for the linear sprint test. The fourth and fifth timing gates were placed to the left and the right of the course in a Y-formation; these were used for the CODS and RA sprint tests. All tests were conducted indoors. The best effort was noted for the linear sprint and CODS, and the mean values for all the trials were calculated for the RA test, similar to Gabbett et al.,²⁸ considering that this test includes cognitive skills. Players started with their leading foot 30 cm behind the first timing gate for all the sprints. The players first completed two trials of the linear sprint, after which they performed a planned change of direction agility sprint initially to the left and then to the right, where after they performed the RA test. When the players crossed the middle timing gates during the RA test, lights were activated either on the left or right timing gates, which indicated the direction in which the player had to sprint. The Smartspeed® system controls the reactive element during the pre-selected protocol (i.e. 1-1-2 formation), allowing objective evaluation of RA. A 40- to 45-millisecond lag time exists between breaking the middle timing gates and the activation of the lights. Players performed repeated trials of the RA test until they completed two sprints to either side or a total of eight trials, whichever came first. A rest period was incorporated by not letting a player complete all his trials at once, but rather allowing players to go one after the other.

Training procedures

Both training interventions involved RA drills that encouraged decision making and pattern recognition.

Video-based training. Video clips of various rugby-attacking scenarios were recorded from frontal view in line with the movement, using eight experienced rugby union players. The clips varied from a single player up to six players per clip, and these clips were played in a randomized order. The eight players acted out the following moves; the side-step, cross-over step, fake pass, scissor movements, goose step, double dodge, run around an opponent and straight running. A total of 189 video clips were recorded of which 175 were used; from these, 12 video training sessions were created that were used for all the players (Table 2). Each session comprised 50 video clips randomly selected from the 175 total clips, and lasted approximately 10 minutes. The inclusion of video clips with more than one player and various running lines added ecological validity to the training programme. The clips allowed the player to react to the left, the right or straightforward.

Table 2.

Number of different scenarios presented in each VT session.

	S1	S2	S3	S4	S5	S6	S7	S8	S9	S10	S11	S12	Total
Side-step	10	7	8	7	9	10	8	7	7	9	10	9	101
Cross-over	2	4	0	0	3	2	1	1	1	1	0	2	17
Fake pass	3	1	3	2	3	2	2	2	3	2	2	3	28
Scissor movement	9	12	13	14	12	13	13	13	12	13	13	13	138
Goose step	1	0	2	1	1	0	2	1	1	0	2	1	12
Single dodge	3	4	3	4	4	4	5	4	5	3	3	3	45
Double dodge	4	5	3	4	3	3	3	3	3	4	4	3	42
Run around	10	11	13	12	12	11	13	18	15	16	12	13	156
Straight running	6	5	5	4	3	3	2	1	2	1	2	2	36
Kick	2	1	0	2	0	2	1	1	1	1	2	1	14

VT: video-based training.

Each player completed the video training sessions indoors and individually to ensure that his reaction was purely based on what was happening on the video clip and not due to the reaction of other players or environmental factors. The video clips were projected on a white screen (2.0 m × 1.5 m) and players were positioned 8 m in front of the screen on a marked line. They had to watch the video clip and react according to the movement of the player(s) on the screen. Players were instructed to start moving forward as soon as the clip (2–4 s) started and to defend the last player with the ball. The players immediately knew the outcome as the entire movement was displayed. There was an 8-second gap between each clip during which they had to return to the starting line. Eight cones were placed on the floor to indicate the possible direction of their response. Each session comprised 50 video clips and lasted approximately 10 minutes. All the moves mentioned previously occurred in each session in random order and varying in numbers. During these sessions, players’ decision making relied purely on the visual cues from the screen.

Field-based training. Training took place on an outdoors rugby field and were all conducted by the main researcher. Players participated as a group and were instructed to wear their rugby boots during training. Drills varied from coach command drills, pairs working together, mirror drills, evasion drills and small-sided games. Unlike the VT, these sessions encouraged players to react to auditory cues (i.e. the sound of the whistle, the bounce of the ball, sound of a teammate running around a player or coming up to a player’s side), as well as visual cues (i.e. the flight of the ball or opponent running towards a player). The sessions consisted of three to four drills and lasted for 20–30 minutes (Table 3).

Table 3.

FT sessions.

	Sessions
Week 1	Session 1 • Coach command drill • Triangle-mirror drill • Rugby, speed pass	Session 2 • Knee tag game • Ball reaction drill • Partner lane drills • React to situation 3 versus 2
Week 2	Session 3 • Tail catch • Rugby 1 versus 1 • 2 versus 2	Session 4 • Running and pass • Six-cone mirror drill • Line drill with auditory cues • 2 versus 1
Week 3	Session 5 • Footspeed attack/defence • 15-m box tag • Kick tennis	Session 6 • Round and round • Continues attacking • 5 versus 5 touch rugby
Week 4	Session 7 • Defence cone drop game • Ball over head throw (three players) • React, run and defend	Session 8 • Rugby 1 vs. 1 • Change reaction- progression • One-on-one touch shadow rugby
Week 5	Session 9 • Coach command drill • 3 versus 1 • Rugby, defeating the opponent	Session 10 • Two-tail catch • 6-cone mirror drill • 2 on 1 evasion • Rugby scattered defence 6 versus 4
Week 6	Session 11 • Defence cone drop grame • Colour T-drill • Continues break • Touchies	Session 12 • Knee tag game • Decision making 2 versus 1 • Decision making attack 3 versus 2 • Kick tennis

FT: field-based training.

Statistical analysis

Data analysis was performed using Microsoft Office Excel® and StatSoft® statistical software (Statistica, StatSoft Inc., Tulsa, OK, USA). The data in the tables are reported as mean ( $\bar{x}$ ) ± standard deviation (SD), unless otherwise specified. A single-factor analysis of variance (ANOVA) was used to analyse the descriptive data between the three groups. A multifactorial ANOVA was used to assess the possible differences between the groups and over time, followed by Fisher’s least significant difference post-hoc test at individual time points. The level of significance for all analyses was set at p ≤ 0.05.

Cohen’s effect sizes (ESs) were calculated to determine the within/between-group magnitude of differences in performance variables and practical significance of these differences in response to the intervention and retention periods. Threshold values for Cohen’s d were > 0.2 (small^S),>0.6 (moderate^M),>0.8 (Large^L) and>1.2 (very large^VL).²⁹

The main reason for focusing on magnitude-based differences is that the absolute changes in sprint and agility performances are usually very small; however, these small improvements may have large practical significance in the field. The smallest worthwhile change (SWC) for between-group comparisons was calculated as 0.2 multiplied by the between-subject standard deviation.³⁰ The SWC is presented in the illustrations as the grey-shaded area. This was performed to determine if one training programme resulted in (i) a practically beneficial, (ii) unclear/trivial or (iii) no practically beneficial effect on performance.²⁹ Chances of beneficial/trivial/no practically beneficial effects were qualitatively assigned as follows: <1%, almost certainly not; <5%, very unlikely; <25%, unlikely/probably not; <50%, trivial/unclear; >75%, likely/probably; >95%, very likely; and >99%, almost certainly.^31–33 For the training interventions to be practically meaningful, or show a substantial performance difference, the standardized mean difference (SMD) had to be 1.75 times the SWC to be considered practically relevant.^33–35 The SMD expresses the size of the intervention effect relative to the variability in the outcome measurement. The SMD is also known as Cohen’s d. SMD = (change in outcome with intervention 1 – change in outcome with intervention 2)/pooled standard deviation.

The illustrations (Figures 1 and 2) always report whether the first intervention mention is more beneficial compared to the second. In other words, if the triangle (the standardized mean difference) is to the right of the graph, the first listed intervention is more beneficial than the second listed intervention.

Figure 1.

Relative changes and qualitative outcomes for the training effect (pre to post) in (a) speed, (b) change of direction speed, and (c) reactive agility.

Figure 2.

Relative changes and qualitative outcomes for the retention period (post to retention) in (a) speed, (b) change of direction speed, and (c) reactive agility.

Results

The physical characteristics of the 26 participants are summarized in Table 4. The players were healthy male rugby union players, with a minimum of 10 years rugby-playing experience. There were no statistically significant differences in any of the descriptive characteristics among the three groups (p > 0.05).

Table 4.

Physical characteristics of the players (n = 26).

Variables	Control (n = 7)		VT (n = 10)		FT (n = 9)
Variables	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD	Range
Age (years)	22 ± 1	20–23	21 ± 1	19–23	20 ± 1	19–22
Height (cm)	184 ± 6	174–191	184 ± 10	171–206	185 ± 7	178–199
Body mass (kg)	89 ± 11	74–103	92 ± 12	77–109	94 ± 12	72–114
Body fat (%)	16 ± 4	12–24	15 ± 4	9–21	15 ± 5	9–24
Lean body mass (%)	84 ± 4	76–88	85 ± 4	80–91	85 ± 5	76–92
Repeated sprint (m)	700 ± 37	645–755	727 ± 15	715–760	694 ± 47	610–770
VO_2max (ml.min⁻¹.kg⁻¹)^a	52 ± 3	47–57	53 ± 4	47–59	53 ± 8	41– 68

VO_2max: estimated maximum aerobic capacity.

Table 5 shows the absolute results and ESs for the performance tests of the three groups before and after the intervention period, as well as after the retention period. No statistically significant differences existed between the groups for speed, CODS or RA time during the pre-testing (p > 0.05). Between-group comparisons in response to the intervention and retention periods are depicted in Table 5. The multifactorial ANOVA revealed no significant interaction effects for any of the variables except for the RA outcome variable, with both VT and FT significantly faster than C from pre to post (p < 0.001), while C significantly improved RA from post to retention (p = 0.001).

Table 5.

The sprint and agility performances of the rugby players after the intervention periods.

Time (s)		C Mean ± SD	VT Mean ± SD	FT Mean ± SD
Speed	Pre	1.91 ± 0.09	1.86 ± 0.08	1.89 ± 0.07
CODS		2.04 ± 0.14	1.97 ± 0.08	2.02 ± 0.14
RA		2.75 ± 0.16	2.79 ± 0.18	2.88 ± 0.13

Speed	Post	1.96 ± 0.10^PreS	1.85 ± 0.12	1.93 ± 0.10^PreS
CODS		2.05 ± 0.12	1.88 ± 0.09^PreL	2.03 ± 0.14
RA		3.00 ± 0.12^PreVL	2.69 ± 0.15^PreS	2.70 ± 0.15^PreVL

Speed	Retention	1.89 ± 0.15^PostS	1.85 ± 0.08	1.87 ± 0.05^PostM,PreS
CODS		1.99 ± 0.16^PostS,PreS	1.94 ± 0.12^PostS,PreS	1.96 ± 0.06^PostM,PreS
RA		2.69 ± 0.10^PostVL,PreS	2.70 ± 0.19^PreS	2.71 ± 0.13^PreVL

The magnitude of the within-group differences S, M, L, and VL indicate a small, moderate, large or very large difference (e.g. PreS: difference vs. Pre with a small standardized difference).

C: control group; VT: video training group; FT: field training group; CODS: change of direction speed.

Speed

There were no practically meaningful improvement in absolute speed in the VT group from pre- to post test (ES < 0.20), while the FT and C groups had a small practically meaningful deterioration in sprint speed for pre- to post test (ES = 0.44 and 0.49, respectively). Figure 1 illustrates that VT was probably more beneficial to maintain sprint speed from pre- to post testing than FT (2.7 ± 5.0% faster) and C (3.6 ± 5.3% faster). Field-based training (FT) had no advantage over rugby training alone to improve sprint speed. The VT group also retained the faster sprint speed after the retention period, while the FT group showed a small practically meaningful improvement from pre-testing to after the retention period (ES = 0.31) (Table 5).

Change of direction speed

VT resulted in a large practically meaningful improvement in CODS from pre- to post test (ES = 1.01), while the performances of the FT and C groups were unchanged (ES < 0.20). Figure 1 (pre–post) indicates that VT was very likely more beneficial to improve CODS than FT (4.7 ± 3.6% faster), as well as C (5.3 ± 4.8% faster) and that there was a negligible small difference in the training effects of FT and C (0.3 ± 4.5% difference). The VT group was unable to maintain its improvements in CODS from post to retention, but they were still practically meaningfully faster than at baseline (ES = 0.28). Both FT and C improved their CODS times from post to retention and they were both faster from pre to retention (ES = 0.51 and 0.31, respectively). However, FT was not more successful in improving CODS compared to C from pre- to post testing (only a 1.8 ± 4.6% greater improvement in FT) (Figure 1).

Reactive agility

The RA performance for both VT and FT groups were practically meaningfully better from pre- to post testing (ES = 0.58 and 1.22, respectively), and these improvements were maintained after the retention period (post to retention). Both VT and FT were almost certainly more beneficial to improve RA compared to rugby training alone from pre- to post test, with differences in the training effect of (13.4 ± 5.6%) and (16.7 ± 4.4%), respectively (Figure 1). Furthermore, FT was probably more beneficial than VT to improve RA (3.0 ± 4.4%). The C group performed worst form pre- to post test (ES = − 1.67) and only managed practically meaningful improvements in RA after the retention period (9.7 ± 3.1%; ES = 2.61) (Figure 2). The performance in VT and FT from post to retention practically stayed the same (ES < 0.20).

Discussion

The current study is the first to examine the effectiveness of a video-based RA programme for rugby union players as a possible alternative method to FT. The advantages of a VT programme is that it is performed at low to moderate intensities, is short in duration and can be used indoors. The main findings were that both VT and FT improved RA performance in club-level rugby union players and that the training effect after 12 sessions is meaningfully greater than with rugby training alone (C). However, outdoors field-based agility training is probably more beneficial than indoors VT to improve RA in rugby players. This study also showed that the combination of perceptual components with movement during training twice a week for six weeks was adequate to facilitate long-term motor learning, as the improvements in RA were maintained after a six-week period without receiving the training stimulus.

Speed

A 10-metre distance was chosen for this study because the ability to quickly accelerate is seen as a more important requirement of agility performance than maximal velocity and also because rugby union players seldom sprint distances further than 10 metres.^32,36 The fact that sprint speed time in the VT players remained constant throughout training and retention may be a reflection of the training intervention. The programme did not include sprint training and focussed specifically on developing the perceptual aspects of agility. FT and C initially dropped their performance after the training period, but improved after the retention period. One would have expected the FT group to improve their speed during the intervention period, due to the running-based activities and mini-games included in their training. This finding is therefore inconsistent with the findings of Gabbett,³⁷ who reported that skill-based conditioning games significantly improved 10-metre speed in rugby league players. However, in this study the running exercises mostly involved change of direction manoeuvres and were seldom straight running. Considering that the changes in FT and C were similar, it is assumed that the changes for both groups are due to regular rugby training and not as a result of the training intervention.

Change of direction speed

The VT group was the only group to meaningfully improve their times in the 10-metre CODS test from pre- to post test (4.6 + 5.1%). Surprisingly, field-based agility training provided no clear benefit to improve CODS compared to C from pre- to post test. During the VT sessions, players were only required to react to the player(s) on the screen with a resultant movement either to the right, the left or forward. This is very similar to the movement in the 10-metre CODS test. The FT group, on the other hand, was exposed to a greater variety of reaction drills and different movement patterns. Thus, the greater improvement in CODS in the VT group may be explained by the specific movement patterns developed during training and how it directly related to the test.

The improvement seen in CODS in the VT group is in contrast with the results of Serpell et al.,¹⁷ who found a non-significant change in CODS performance following a video-based RA training programme. This contradictory finding may be reflective of the shorter intervention used by Serpell et al.¹⁷ as their programme only lasted three weeks, compared to the six weeks of this study. Although Gabbett³⁷ employed a longer programme of nine weeks, their improvement of 0.5% in mean CODS in rugby league players following a skill-based conditioning games training programme was not statistically significant. Neitzke et al.³⁸ on the other hand found significant improvements in agility of male adolescents following six weeks of RA drills. This may suggest that improvements in CODS are only facilitated by specific RA drills which include a significant perceptual component.

The improvement in CODS in the VT group was not sustained after the retention period; however, it was still better than at baseline. However, the FT group improved after the retention period, which may suggest a delayed learning effect, as learning through game-based training takes longer than planned technique practice.⁵ The results further indicate that CODS and RA are two unique skills that need to be trained in different and specific manners.

Reactive agility

VT and FT were almost certainly more beneficial in improving RA than rugby training alone from pre to post; indicating that even short-term exposure to RA (12 sessions) may indeed improve RA performance. Rugby training alone (C) initially did worse following the intervention, and this may indicate that the type of rugby training performed at their club during the intervention period may not have provided players with any perceptual or decision-making components and that the improvements seen in the VT and FT could be due to the extra training they received.

Both intervention groups had meaningful improvements in their RA over time and when compared to the C group from pre to post testing, while both interventions also failed to produce meaningful improvements in CODS. Similar highly skilled Australian football players were able to outperform lesser-skilled players when asked to react to evasive agility manoeuvres, but performed similar in planned CODS, suggesting that the perceptual aspect is the differential component.^12,39 Previous investigations also found that athletes with superior decision-making ability in a sport-specific context transfer to enhanced speed in RA performance.^10,12 Important for this study is that the VT was also effective in improving RA performance. While decision-making time was not directly assessed, the preceding results showed that the VT group’s CODS improvement was not sustained after the retention period, albeit RA improvements were maintain in the VT group during retention. Consequently, one could assume that it was the perceptual-cognitive skills that improved since previous researchers have found that response time and decision-making time correlated mostly with RA time,⁴⁰ also as mentioned before training with perceptual-cognitive components most often result in delayed improvements only noted during retention or transfer tests.⁴⁰

A direct comparison of the two training methods in this study revealed that FT resulted in a slightly greater improvement (3.0 ± 4.4%) in RA than VT (p > 0.05). Although VT is also rugby specific and the players had to react to game-like situations, FT may allow for a more physical involvement and the use of more senses. FT provided a multi-stimulus environment upon which the players had to react, whereas VT involved a less demanding environment. Thus, FT training may have required more cognitive involvement than VT. Gabbett³⁷ also found that skill-based conditioning games significantly improved rugby league players’ attacking abilities, which may be due to increased ability to read patterns of play. The latter would thus support the findings of this study.

There were no practically meaningful declines in RA for both VT and FT after the retention period and both groups’ performances were still better than at baseline. Only minor changes were observed for both experimental groups (VT: 2.69 ± 0.15 seconds to 2.70 ± 0.19 seconds and FT: 2.70 ± 0.15 seconds to 2.71 ± 0.13 seconds) from post to retention and the magnitude of change was the same for the two interventions. The inclusion and benefits of the perceptual component in agility training is also highlighted by this study. This is in accordance with Serpell et al.¹⁷ who also found an improvement in RA performance in rugby league players when targeting the perceptual and decision-making components through VT.

There was a very large practically meaningful improvement in RA performance in the C group after the retention period (ES = 2.61), and this performance was also better than at baseline (ES = 0.42). The retention period coincided with the start of the rugby competition phase; therefore, one may speculate that the exposure to competition-like decision-making situations and the requirement for RA manoeuvres during games may be the reason for these improvements. If this is indeed the case, one can further speculate that the reason why the VT and FT groups did not further increase their RA performance is because these training programmes are more effective in improving RA than exposure to competition only. The sustained better RA performance in both the training groups indicates that perceptual skills have been learned, and that VT and FT are effective methods of motor learning.

Even though the testing methods in this study used a generic stimulus, the training programmes were rugby specific. Thus, the findings point to the importance of including sport-specific perceptual components as well as decision-making strategies when developing RA programmes and assessment methods. By improving perceptual-motor skills, the player may anticipate better and indirectly resulting in improved reaction time or in the case of this study, RA.

Practical applications

Rugby players are exposed to both planned and unplanned movement situations, but often coaches and trainers focus only on predictable agility produced in an unchanging environment. This study shows that unpredictable agility drills (i.e. field and video based) improve RA performance in experienced club-level rugby union players. This is an essential aspect when developing agility skills in rugby union players, since effective anticipation strategies allow for faster movement responses without the player having to wait for the entire movement to be displayed.⁴¹ Evasive agility manoeuvres can therefore only be effectively trained if a perceptual component is included.

Furthermore, it is also important that coaches and trainers consider the sport-specific experience of the players when designing training drills for reactive agility. Players should first master basic rugby skills and techniques before drills are introduced to improve their decision-making skills and reactions to game-specific stimuli.

Although the changes in VT was no different than C from pre to retention, it does not mean that VT is worthless. VT may be an effective alternative method or add-on to develop RA. It may be considered ideal for busy and/or intense training weeks and bad weather days as the training is indoors, short in duration and at low to moderate intensities.

Future studies and limitations

The findings of this study should be considered as a point of departure for further research in training methods to improve RA in rugby union players.

One of the limitations of this study was that RA was assessed with a randomized light-stimulus system and not a sport-specific stimulus system as was used by Gabbett and Abernethy¹⁶ and Henry et al.³⁹ By responding to different coloured lights, this study may have removed the specific cues that could have highlighted the beneficial aspects of the VT programme. Due to this limitation, it cannot be claimed that decision-making skills have improved, as one would need to isolate movement time from decision time. It is recommended that future studies include high-speed video footage to record decision time and response accuracy. This would allow researchers to understand which specific components of RA performance, i.e. information processing, decision making, movement time and/or anticipation lead to the overall improvement in performance. Furthermore, if response accuracy is included in future investigations, it may also elucidate the speed-accuracy trade-off in players. In other words, a quicker decision time (i.e. faster decision-making ability) may result in inaccurate responses.

In light of the aforementioned critique, one may also consider that the coloured lights may actually have acted as performance-relevant cues during the RA test, even though it was not sport-specific cues. Considering that rugby is a dynamic sport where the environment is constantly changing, the relevant cue in practice may not be the same as found during an actual match. A relevant cue becomes context or task specific. In this instance, the relevant cue which the player needed to detect and interpret and then produce an appropriate response for the given task was the randomized light. The task as well as appropriate response (going left or right) is therefore dictated by the light i.e. the relevant cue. If this is the case, then the chosen RA test could give a true reflection of the players’ perceptual skills after all.

Even though this study shows that RA was retained in both FT and VT six weeks after the intervention, it is still uncertain if these improvements would transfer to actual playing performance. Consequently, the inclusion of performance transfer tests in future studies is warranted.

Conclusion

Twelve VT sessions of 10 minutes each over six weeks caused practically meaningful improvements in RA performance of club-level rugby union players. These improvements in RA were comparable to FT, which included game-based training.

The most important finding of this study is that benefits are available from additional training that includes a perceptual component when developing or testing RA and that it is best when performed in a sport-specific manner. Finally, it can be concluded that speed and CODS may not be improved by RA training, as limited transfer exists between the skills. Training programmes should therefore be designed to developed speed and CODS as separate skills.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Duthie

Pyne

Hooper

. Applied physiology and game analysis of rugby union. Sports Med 2003; 33: 973–991.

Hughes

Bartlett

. The use of performance indicators in performance analysis. J Sports Sci 2001; 20: 739–754.

Gabbett

Benton

. Reactive agility of rugby league players. J Sci Med Sport 2009; 12: 212–214.

Sheppard

Young

. Agility literature review: classifications, training and testing. J Sports Sci 2006; 24: 919–932.

Young

McDowell

Scarlett

. Specificity of sprint and agility training methods. J Strength Cond Res 2001; 15: 315–319.

Chelladurai

. Manifestations of agility. J Canadian Assoc Health Phys Education Recreat 1976; 42: 36–41.

Wheeler

Askew

Sayers

. Effective attacking strategies in rugby union. Eur J Sport Sci 2010; 10: 237–242.

Sayers

MGL

Washington-King

. Characteristics of effective ball carries in super 12 rugby. Int J Perform Anal Sport 2005; 5: 92–106.

Young

Wiley

. Analysis of a reactive agility field test. J Sci Med Sport 2010; 13: 376–378.

10.

Farrow

Young

Bruce

. The development of a test of reactive agility for netball: a new methodology. J Sci Med Sport 2005; 8: 52–60.

11.

Helsen

Pauwels

The relationship between expertise and visual information processing in sport. In: Starkes

Allard

(eds). Cognitive issues in motor expertise. Advances in psychology 1993; Vol. 102, Amsterdam: North-Holland/Elsevier Science Publishers, pp. 109–134.

12.

Sheppard

Young

Doyle

. An evaluation of a new test of reactive agility and its relationship to sprint speed and change of direction speed. J Sci Med Sport 2006; 9: 342–349.

13.

Abernethy

Russell

. Expert-novice differences in an applied selective attention task. J Sports Sci 1987; 9: 326–345.

14.

Young

James

Montgomery

. Is muscle power related to running speed with changes of direction? J Sports Med Phys Fit 2002; 42: 282–288.

15.

Abernethy

. Visual search strategies and decision-making in sport. Int J Sports Psychol 1991; 22: 189–210.

16.

Gabbett

Abernethy

. Expert-Novice differences in the anticipatory skill of rugby league players. Sport Exercise Perform Psychol 2013; 2: 138–155.

17.

Serpell

Young

Ford

. Are the perceptual and decision-making components of agility trainable? A preliminary investigation. J Strength Cond Res 2011; 25: 1240–1248.

18.

Savelsbergh

GJP

Van der Kamp

Williams

. Anticipation and visual search behaviour in expert soccer goalkeepers. Ergonomics 2005; 48: 1686–1697.

19.

Farrow

Abernethy

. Can anticipatory skills be learned through implicit video based perceptual training? J Sports Sci 2002; 20: 471–485.

20.

Starkes

Lindley

. Can we hasten expertise by video simulations? Quest 1994; 46: 211–222.

21.

Passos

Araujo

Davids

. Manipulating constraints to train decision making in rugby union. Int J Sports Sci Coach 2008; 3: 125–140.

22.

Berry

Abernethy

Côté

. The contribution of structured activity and deliberate play to the development of expert perceptual and decision-making skill. J Sport Exercise Psychol 2008; 30: 685–708.

23.

Salonikidis

Zafeiridis

. The effects of plyometric, tennis-drills, and combined training on reaction, lateral and linear speed, power, and strength in novice tennis players. J Strength Cond Res 2008; 22: 182–191.

24.

Oliver

Meyers

. Reliability and generality of measures of acceleration, planned agility, and reactive agility. Int J Sports Physiol Perform 2009; 4: 345–354.

25.

Green

Blake

Caulfield

. A valid field test protocol of linear speed and agility in rugby union. J Strength Cond Res 2011; 25: 1256–1262.

26.

D’Auria S, Tanner R, Sheppard J, et al. Evaluation of various methodologies used to assess sprint performance. Paper presented at the Australian Institute of Sport Applied Physiology Conference, Canberra, Australia, 2006.

27.

Earp

Newton

. Advances in electronic timing systems: considerations for selecting an appropriate timing system. J Strength Cond Res 2012; 26: 1245–1248.

28.

Gabbett

Kelly

Sheppard

. Speed, change of direction speed, and reactive agility of rugby league players. J Strength Cond Res 2008; 22: 174–181.

29.

Hopkins

Marshall

Batterham

. Progressive statistics for studies in sports medicine and exercise science. Med Sci Sports Exerc 2009; 41: 3–13.

30.

Hopkins

. Measures of reliability in sports medicine and science. Sports Med 2000; 30: 1–15.

31.

Pyne

Gardner

Sheehan

. Fitness testing and career progression in AFL football. J Sci Med Sport 2005; 8: 321–332.

32.

Duthie

Pyne

Marsh

. Sprint patterns in rugby union players during competition. J Strength Cond Res 2006; 20: 208.

33.

Hopkins WG. Calculating likely (confidence) limits and likelihoods for true values (Excel spreadsheet), http://www.sportsci.org/resource/stats/xcl.xls (2002, accessed 16 September 2014).

34.

Hopkins WG. Probabilities of clinical or practical significance, Sportscience 6, sportsci.org.jour/0201/wghprob.htm (2002, accessed 16 September 2014).

35.

Batterham

Hopkins

. Making meaningful inferences about magnitudes. Int J Sports Physiol Perform 2006; 1: 50–57.

36.

Sayers

. Running techniques for running rugby. New Zealand Coach 1999; 20: 23–31.

37.

Gabbett

. Skilled-based conditioning games as an alternative to traditional conditioning for rugby league players. J Strength Cond Res 2006; 20: 309–315.

38.

Neitzke

Miller

Cheatham

. Preplanned and reactive agility training influence on agility test performance in male adolescents. J Strength Cond Res 2010; 24: 1.

39.

Henry

Dawson

Lay

. Validity of a reactive agility test for Australian football. Int J Sports Physiol Perform 2011; 6: 534–545.

40.

Scanlan

Humphries

Tucker

. The influence of physical and cognitive factors on reactive agility performance in men basketball players. J Sports Sci 2013; 31: 367–374.

41.

Wheeler

Sayers

MGL

. Modification of agility running technique in reaction to a defender in rugby union. J Sports Med 2010; 9: 445–451.