This article was downloaded by: [University of Stellenbosch] On: 27 April 2013, At: 06:38 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Sports Sciences Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rjsp20
Analysis of the physical demands of international rugby union D.A. McLean
a
a
The Queen's College Glasgow, 1 Park Drive, Glasgow, G3 6LP, UK Version of record first published: 14 Nov 2007.
To cite this article: D.A. McLean (1992): Analysis of the physical demands of international rugby union, Journal of Sports Sciences, 10:3, 285-296 To link to this article: http://dx.doi.org/10.1080/02640419208729927
PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-andconditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Journal of Sports Sciences, 1992,10, 285-296
Analysis of the physical demands of international rugby union D.A. McLEAN The Queen's College Glasgow, 1 Park Drive, Glasgow G3 6LP, UK
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
Accepted 12 September 1991
Abstract
The aim of this study was to investigate the physical demands of international rugby union. Five games in the 1989-90 Five Nations Championship were analysed using video-recordings of live television transmissions. When the ball was in open play, the average running pace of players central to the action ranged from 5 to 8 m s - 1 . This together with scrum, lineout, ruck and maul was classified as high-intensity exercise. The density of work was measured by timing the work:rest ratios (W:RRs) throughout each game. The mean duration of the work periods was 19 s and the most frequent W:RRs were in the range of 1:1 to 1:1.9. On average, a scrum, lineout, ruck or maul occurred every 33 s. The ball was in play for an average of 29 min during a scheduled time of play of 80 min. To complement the time-motion analysis, blood samples were taken from six players throughout a first-class game. The highest measured blood lactate (BLa) concentrations for each individual ranged from 5.8 to 9.8 mM. Running speed, duration, BLa levels, physical confrontation and, most particularly, the density of work as illustrated by the W:RRs indicate that the game places greater demands on anaerobic glycolysis than previously reported. This has implications for the physical conditioning of rugby union players. Keywords: Rugby union, time-motion analysis, physical fitness.
Introduction
The desire to develop appropriate conditioning programmes for games players has necessitated the development of time-motion game analysis. Although such analyses were reported prior to 1976, it was recognized that the methods used were neither objective nor reliable (Reilly and Thomas, 1976). Since then, published studies have used the pivotal components of match analysis validated by Reilly and Thomas (1976). The components are categorization of player activity, measurement of distance covered for each activity and the frequency of each activity. Although measurement of distance covered for each activity is possible, the process is more involved than measuring duration. Mayhew and Wenger (1985) identified that the duration of each level of intensity of activity is as representative of the physical demands as distance covered. More recent publications have used duration rather than distance (Docherty et al., 1988; Mayhew and Wenger, 1985; Treadwell, 1988). Other modifications have been made to suit different sports and computer technology included to facilitate data storage, analysis and presentation (Ohashi et al, 1988; McKenna et al, 1988; Treadwell, 1988). 0264-0414/92 © 1992 E. & F.N. Spon
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
286
McLean
Much of the early work on rugby union was reported in the book Fitness and Training for Rugby (Rugby Football Union, 1978), which contained information on distances run and number of sprints undertaken. Unfortunately, the methods and sources were not disclosed. More recently, papers using rugby as the subject have clearly reported their methods and results of time-motion analysis of specific playing positions (Docherty et al., 1988; Treadwell, 1988). To identify how fit players should be to play the game at the highest level, the game itself should be analysed rather than an individual player. The physical demands of a team game result in part from its rules and structure, as well as the skill and tactical ability of all the players involved. This requires analysis of the game at the highest level, since these factors are constraints on individual performance. The part that any one individual plays in a game may also be constrained by his fitness. Thus the analysis of one individual may underestimate the physical demands of the game. It is accepted that team games like rugby are of an intermittent nature. The physiological demand of intermittent activity depends not only on running speed and duration but on the density of physical work. The pattern of work:rest ratios (W:RRs) throughout the game is just as important a determinant of the physical demands of the game as total time or distance covered at different intensities. Work:rest ratios have not previously been reported in studies of rugby (Docherty et al, 1988; Treadwell, 1988). The measurement of blood lactate (BLa) has been done as an adjunct to the time-motion analysis of team games. One such study of rugby showed BLa concentration to be 2.8 mil at the end of a game. The conclusion drawn by Docherty et al. (1988) was that alactic metabolism was implicated. There are, however, several factors which do not support this conclusion. To give a clearer view of the type of metabolism used during international rugby union, measurements of BLa should be undertaken throughout a competitive senior game of the highest possible standard. In response to the demands to improve physical fitness of elite rugby union players, this paper intends to identify: 1. The density of high-intensity exercise by measurement of W:RRs in full international matches. 2. The BLa concentrations throughout a first-class game in Scotland.
Methods Time-motion analysis 1. Categorization of activity. The following classification of categories was used (Docherty et al, 1988): Low-intensity 1. Standing 2. Walking 3. Jogging
High-intensity 4. Run with elongated stride 5. Sprint 6. Non-run intense activity
These descriptors were used during the initial observation of video-recordings of live
The physical demands of rugby union
287
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
television transmission of matches in the Five Nations Championships during the 1989-90 season. They showed that as soon as the ball went dead, players switched into a low-intensity mode. This was regarded as a recovery phase which included both active and passive recovery. At the restart, it appeared that high-intensity exercise began for players central to the action. This assumption has to be tested. 2. Identification of high-intensity work. The distances used to calculate average running speeds were measured using a system of visual cues similar to the system validated by Reilly and Thomas (1976). The method involved sampling passages of play where the television coverage included sufficient visual cues to identify the distance travelled by a player central to the action. The sample was also representative of the range of distances run (11-70 m) and of the different playing positions in a team. The slow frame advance and pause facilities were used to enable the path of the run to be plotted on a scaled plan of a rugby pitch and measured (n = 16). These runs were timed using a digital stopwatch. The average speed was calculated in metres per second (m s" 1 ). To identify within-observer reliability, this process was repeated 3 weeks later without reference to the first measures. The standard deviation (S.D.) of the difference between duplicate measures was 1.3 m and 0.09 s respectively.
3. Work:rest ratio (W:RR). Each phase of high- and low-intensity exercise was timed in order to identify the W:RRs. Play was judged to start immediately the ball was thrown in at the lineout or the scrum was formed. If, for example, a penalty was awarded and the option was taken to kick direct to touch and no purposive action followed, then this period of time was included in the previous recovery period. This provided data for each contiguous W:RR throughout each game. The accuracy was checked using two methods. All the times recorded for each period of work and rest for the entire match were summed. This should produce a total of approximately 80 min. It cannot be assumed to be exactly 80 min, since it is at the referee's discretion as to how much time is added for stoppages. However, assuming the real time played to be 80 min, of the five international matches analysed the average error was 4.3%. The second method used was as described in the previous section. Passages of play were retimed (n= 15; range of duration = 3-40 s). The S.D. of the differences between duplicate measures was 1.69 s.
Blood lactate analysis Two Scottish First Division clubs agreed that during penalty kicks or stoppages for injury, blood samples could be taken from six subjects. Three subjects were selected from each side to include front, second and back row, halfback and three-quarter. Five samples were taken from each subject (each subject gave his informed consent). Finger prick blood samples were collected and 3.5 p\ samples analysed using a semi-automated analyser (Analox LM3). Analysis was carried out between 3 and 5 min after the sample was taken. The machine was calibrated using an independently produced standard. The S.D. of the differences between duplicate measures was 0.16 mM (n = 30). Three days after the match, a maximum treadmill test (Schnabel and Kindermann, 1983) was used to identify peak BLa concentration. It involved running up an incline of 4° at 20 km
McLean
288
h 1 until exhaustion. Post-test blood samples were collected at 1 and 4 min. The players were asked to follow the same diet for 24 h before each event.
Results
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
Ratification of high-intensity work by measurement of running speed during international matches Over five games, the average running speeds of the player either in possession or in closest pursuit of the ball was 5-8 m s" 1 . This range of running speeds compares with a similar analysis of soccer (van Gool et al., 1988). For rugby union, this running pace does reflect high-intensity exercise. The optimal pace of training runs used by the current Scottish international squad to improve aerobic fitness is in the range 4.2-4.7 m s" 1 (6.4-5.7 min mile" 1 ). General descriptors of activity Table 1 outlines the general descriptors of activity during five international matches. The average playing time was 29 min. This included scrums and lineouts, so that the forwards' active involvement was greater than the backs. The average duration of passages of play was 19 s. The large standard deviation reflects the variability of their duration. Figure 1 shows that the distribution is positively skewed. Table 1. Playing time and distance covered at high intensity during five matches in the 1989-90 Rugby Five Nations Championship
Playing time (min) first half second half total
France
Ireland
Wales
England
vs
vs
vs
vs
vs
England
Scotland
Scotland
Wales
France
14.2 12.8 27.0
14.0 14.0 28.0
14.7 17.1 31.8
12.5 16.2 28.7
14.7 13.9 28.6
21.0 18.6 20.5
16.3 11.4 19.3
17.0
5.6
9.2
20.6 16.3
17.8 10.5
21.7 16.6 19.4 13.5
19.6 17.0 19.9 15.4 19.7 19.7
70.0
51.0
47.0
56.0
Average duration of work periods (s) first half 18.9 S.D. 14.5 second half 16.1 S.D. 11.6 17.4 total S.D. 13.1 Longest work period (s)
45.0
8.8
Scotland
Table 2 contains standard information on the number of scrums, lineouts, rucks and mauls. The fact that rucks and mauls outnumbered scrums by 56% and lineouts by 44% has implications for conditioning.
289
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
The physical demands of rugby union
21-25
11-15
31-35
41-45
51-55
Wai V Scot ScotVFra EngVWal FraVEng Ire V Scot
Duration in seconds Fig. 1. Duration of work periods in international matches, 1989-90.
Table 2. The average number of scrums, lineouts, rucks and mauls during five games in the 1989-90 Rugby Five Nations Championship S.D. Range Total First half Second half Scrum Lineout Ruck/maul
15 22 37
17 19 36
32 41 73
7.7 3.3 13.7
24-45 38^6 62-88
Work:rest ratios (JV.RRs)
Figure 2 illustrates the distribution of W:RRs in five games. The distribution is very similar in all games and follows a normal distribution apart from the first bar which shows W:RRs of 1: > 4. Stoppages for injury and kicks at goal were responsible for the prolonged rest periods. The work:rest ratios which occur most frequently are 1:1-1.9 (26.6%) and 1-1.9:1 (20.2%). Figure 3 shows that there are nearly twice as many W: RRs where the work period is less than the rest period. Figures 4a and 4b provide an example of patterns of W: RRs during the first half of a game. These figures show how the distribution of different W:RRs varies with time. Figures 5a and 5b allow comparison of two short passages of play which illustrate the extremes of intensity of work which can occur in rugby union. In Fig. 5a, the third to the eighth passages of play involve six consecutive passages of play where the duration of work exceeds rest (a total of 137 s of work to 71 s rest). In Fig. 5b, the opposite is the case, with work lasting 123 s and rest 228 s. Analysis of W:RRs identified that the pattern shown in Fig. 5a was an extreme and uncommon occurrence. In the five games analysed, the maximum
McLean
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
290
1:3-19
1:2-Z9
1:1-1.9
1-1.9:1
WalVScot ScotVFra Ire V Scot /ZFraVEng EngVWal
Work: Rest Ratios Fig. 2. Work'.rest ratios in the 1989-90 Five Nations Championship showing similarities in distribution.
Fig. 3. Proportion of work:rest ratios averaged over five internationals in 1989-90.
number of consecutive passages of play with work greater than rest was six and the average was four. The number of occasions when two or more such passages of play occurred in a game was eight. The range was 4-8. Included in all passages of play is the high-intensity muscular effort required during scrums, lineouts, rucks and mauls (Table 2).
291
The physical demands of rugby union (a)
20 18.
16. 14. w R
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
R
W:RR=1:20
12. 10. 8. 6. 4. 2.
•• •
1 1 1 JL
0
6 17 2 18 21 30 10 8 38 5 28 7 6 4 19 25 18 12 9
9 4 12 12 5 33 13
Duration of each work period in s e c o n d s
(b)
(nme 21 mm 24 sec)
30. 25. 20.
W R
15.
R
10_ 5.
• I in in
in
In • I
l.llll inn
26 37 31 7 20 14 12 8 7 11 14 4 38 21 37 39 16 51 13 19 15 19 5 15 8 13 6
cn™2imin248ec)
Duration of each work period in seconds
( H - M ™ 391*28.80
Fig. 4. Contiguous work:rest ratios during (a) the first quarter and (b) the second quarter of Wales us Scotland in 1989-90. Blood lactate concentrations during match play
Table 3 shows that the highest measured BLa concentrations for each individual were in the range 5.8-9.8 mM. The highest BLa recorded during the game was compared with the player's maximum BLa, which was identified using a maximal treadmill test. The highest
292 (a)
McLean 7 6. 5.
w
4.
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
I 3. 2. 1.
I 14
(start oi second haio
(b)
14
37
20
24
7
28
21
Duration of each work period in seconds
23
(r,me7nin4isec)
4.5
4.0. 3.5. 3.0. W
2.5.
R
2.0.
R
1.5. 1.0. 0.5. 0.0 21 (Stand second half)
15
44
8
18
19
Duration of each work period in seconds
I (Time 8 min 14 sec)
Fig. 5. Contiguous work:rest ratios during (a) Wales vs Scotland and (b) England vs Wales, 1989-90.
The physical demands of rugby union
293
Table 3. Blood lactate concentrations (HIM) of first division players during a game of rugby union Team A
TeamB
Prop
No. 8
Stand-off
Prop
No. 8
Centre
15 min 30 min Half time 65 min Full-time
3.9 5.1 5.7 5.2 5.8
6.9 7.0 6.9 6.4 5.6
7.5 6.2 4.1 7.3 4.6
5.1 5.9 5.5 5.4 5.9
3.6 5.9 6.5 7.7 9.8
6.1 6.6 5.2 5.1 5.9
Mean S.D.
5.1 0.68
6.6 0.52
5.9 1.4
5.6 0.3
6.7 2.0
5.8 0.63
% of peak BLa
85%
57%
65%
61%
75%
56%
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
Time elapsed
BLa concentration recorded during the game is expressed as a percentage of maximum BLa in Table 3.
Discussion
The major differences between this paper and any other on time-motion analysis of rugby union is that it focuses (1) on the game and not the individual, and (2) on work:rest ratios of high intensity, not summations of time spent on different intensities of exercise. Previous papers have operated on the assumption that the physical demand of activity is determined by intensity (i.e. running speed) and 'volume', which is either total distance run or total time spent at different running speeds. The component of intensity not included in previous studies is density of intermittent work. To illustrate the point, if a player runs at a speed of 7.5 m s~1 over 120 m and does this six times with 80 s rest between each run, the W:RR would be 1:5. If the recovery was reduced to 40 s, the overall physical demand would be far greater yet the running speed and the volume have remained the same. Holmyard et al. (1988) investigated the effect of two different rest periods on mean power output during 10 maximal 6-s sprints on a non-motorized treadmill. Mean power output was maintained with 60 s recovery but dropped significantly on the fifth to tenth sprint with 30 s recovery. Both W:RRs produced BLa concentrations of between 17 and 18 nw. These results support the need to analyse W:RRs to produce a better understanding of the physical demands of rugby. All but 20% of the W:RRs illustrated in Fig. 2 were more severe than those used by Holmyard and co-workers as well as those in a similar study by Tumilty et al. (1988). Figure 1 also shows that approximately 80% of the work periods were longer than those used in these studies. It is accepted that during the passages of play referred to in this study, the players were not sprinting at full pace at all times. However, the average running paces of 5-8 m s" 1 are significantly higher than steady-state running pace for these players and as such are characterized by an exponential rise in BLa concentration (Kindermann et al., 1979). All of this information points to the considerable demands placed by the game on anaerobic metabolism. The BLa analysis carried out in this study confirms that anaerobic metabolism is required
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
294
McLean
for playing rugby union. The highest figures recorded for each individual were in the range 5.8-9.8 mM and exceed both the average of 4 mM, where aerobic metabolism is thought to be the major source of energy (Jacobs, 1981), and the range of 3.05-5.5 mM for steady-state running (Heck et al., 1985). In comparison to peak measurements of 14 mM in soccer (Ekblom, 1986), the levels in the present study could not be regarded as high. The timing of collection of blood samples was dictated by stoppages and was not specific to passages of intense play. Since it has been suggested that in soccer and rugby the periods of low-intensity exercise allow BLa to be metabolized (Mazzeo et al., 1982), the figures will tend to underestimate the true levels of peak BLa. The findings in this paper are in conflict with the results and conclusions of the only other study to report BLa levels in rugby union. Docherty et al. (1988) reported levels of 2.8 mM and duration of work periods not greater than 8.6 s. The low BLa level could be due to two factors. A single blood sample was taken 5 min after the game and the standard of players used in the analysis was not made clear, so they may not have been elite players. The conclusion reached by Docherty and co-workers was that the alactic energy source predominates and should be trained. Certainly, short sprints are an integral part of the game but the concept of work being alactic in isolation is rather theoretical. During sprints of 6 s duration, the three energy systems are active concurrently (Boobis, 1987). The present study suggests that the tempo of the game at international level places a large demand on anaerobic metabolism. This is supported indirectly by the moderate levels of aerobic power recorded for rugby union players (Maud and Shultz, 1984). The mean VO2 max of the Scotland squad in January 1990 was 53.4 + 2.7 ml kg" 1 min" 1 (unpublished personal data). The lower the aerobic power the greater the likelihood that energy demands will be met by anaerobic glycolysis. This is especially the case in rugby union where training does not include extensive steady-state exercise, which enables athletes to operate at a high percentage of their VO2 max without BLa accumulation. The need to analyse the game at the highest level requires compromises to be made in the methodology of time-motion analysis and BLa measurement. Ideally, blood samples should have been taken during an international match; however, such access was not available. Since the intensity of physical effort increases with the skill levels of the players involved (Ekblom, 1986), the results reported will tend to underestimate BLa concentrations during an international match. There are similar restrictions in obtaining access to all international grounds to make dedicated video-recordings which are ideal for time-motion analysis. The alternative was to use recordings of live television transmissions. This did not detract from the identification of W:RRs. The major intrinsic error resulted from the judgement of when play started and stopped, which is common to all time-motion analyses. Provided that suitable passages of play were selected, it was also possible to verify that when the ball was in play the players central to the action were required to run at high intensity. For a minority of occasions, the camera angle or the television director's choice of picture prevented a sequence of play being analysed. However, a representative sample of play was available to measure average running speed. The provision of the groundsman's plan of the visual clues of pitch markings and patterns of grass cut allowed distance to be measured using methods similar to that validated by Reilly and Thomas (1976). Within-operator reliability was acceptable given the range of average running speeds (5-8 m s1) representative of high-intensity exercise. The measurement error of distances run throughout the whole game was poor compared to that of Reilly and Thomas (1976), who achieved less than a 1% error, which highlights
The physical demands of rugby union
295
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
a limitation of live transmission for analysis. Nevertheless, the approximate distance of 2000 m, run at high intensity, is sufficiently accurate to be of use to coaches to help structure their training programmes for rugby union. In conclusion, the results of this study relate to the physical demands of the game and not to any one individual playing position. It does, however, apply more closely to playing positions which are central to the action throughout the game, but will be an overestimate for others. The identification of intensity of physical work by measurement of running pace and W:RRs, supplemented by the measurement of BLa concentrations, suggests that the extent of anaerobic glycolysis in rugby union is greater than previously reported. The data on W:RRs are of practical importance in the preparation of conditioning programmes for rugby union players.
Acknowledgement I gratefully acknowledge the suggestions of Drs C.S. Williams and M.A. Nimmo in respect of this manuscript.
References
Boobis, L.H. (1987). Metabolic aspects of fatigue during sprinting. In Exercise: Benefits, Limits and Adaptations (edited by D.A.D. MacLeod, R. Maughan, M.A. Nimmo, T. Reilly and C. Williams), pp. 116-43. London: E. and F.N. Spon. Docherty, D., Wenger, H.A. and Neary, P. (1988). Time-motion analysis related to physiological demands of rugby. Journal of Human Movement Studies, 14, 269-77. Ekblom, B. (1986). Applied physiology of soccer. Sports Medicine, 3, 50-60. Heck, J.H., Mader, A., Hess, G., Mucke, S., Muller, R. and Hollman, W. (1985). Justification of the 4 mmol 1-1 lactate threshold. International Journal of Sports Medicine, 6, 115-30. Holmyard, D.J., Cheetman, M.E., Lakomy, H.K.A. and Williams, C. (1988). Effects of recovery duration on performance during multiple treadmill sprints. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 134-42. London: E. and F.N. Spon. Jacobs, I. (1981). Lactate, muscle glycogen and exercise performance in man. Acta Physiologica Scandinavica, 495 (suppl.). Kindermann, W., Simon, G. and Keul, J. (1979). The significance of the aerobic-anaerobic transition for the determination of work load intensities during endurance training. European Journal of Applied Physiology, 42, 25-34. McKenna, M.J., Patrick, J.D., Sandstrom, E.R. and Chennells, M.H.D. (1988). Computer-video analysis of activity patterns in Australian Rules Football. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 274-81. London: E. and F.N. Spon. Maud, P.J. and Shultz, B.B. (1984). The US national rugby team: A physiological and anthropometric assessment. The Physician and Sportsmedicine, 12, 86-99. Mayhew, S.R. and Wenger, H.A. (1985). Time-motion analysis of professional soccer. Journal of Human Movement Studies, 11, 49-52. Mazzeo, R.S., Brooks, G.A., Budinger, T.F. and Schoeller, D.A. (1982). Pulse injection 1 3 C tracer studies of lactate metabolism in humans during rest and two levels of exercise. Biomedical Mass Spectrometry, 9, 310-14. Ohashi, J., Togari, H., Isokawa, M. and Susuki, S. (1988). Measuring movement speeds and distances in soccer. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 329-33. London: E. and F.N. Spon.
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
296
McLean
Reilly, T. and Thomas, V. (1976). A motion analysis of work-rate in different positional roles in professional football match-play. Journal of Human Movement Studies, 2, 87-97. Rugby Football Union (1978). Fitness Training for Rugby. Twickenham: RFU. Schnabel, A. and Kindermann, W. (1983). Assessment of anaerobic capacity in runners. European Journal of Applied Physiology, 52, 42-6. Treadwell, P.J. (1988). Computer-aided match analysis of selected ball games (soccer and rugby union). In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 282-7. London: E. and F.N. Spon. Tumilty, D. McA., Hahn, A.G., Telford, R.D. and Smith R.A. (1988). Is 'lactic acid tolerance' an important component of fitness for soccer. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 81-6. London: E. and F.N. Spon. van Gool, D., van Gerven, D. and Boutmans, J. (1988). The physiological load imposed on soccer players during real match-play. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 51-9. London: E. and F.N. Spon.
Journal of Sports Sciences Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rjsp20
Analysis of the physical demands of international rugby union D.A. McLean
a
a
The Queen's College Glasgow, 1 Park Drive, Glasgow, G3 6LP, UK Version of record first published: 14 Nov 2007.
To cite this article: D.A. McLean (1992): Analysis of the physical demands of international rugby union, Journal of Sports Sciences, 10:3, 285-296 To link to this article: http://dx.doi.org/10.1080/02640419208729927
PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-andconditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Journal of Sports Sciences, 1992,10, 285-296
Analysis of the physical demands of international rugby union D.A. McLEAN The Queen's College Glasgow, 1 Park Drive, Glasgow G3 6LP, UK
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
Accepted 12 September 1991
Abstract
The aim of this study was to investigate the physical demands of international rugby union. Five games in the 1989-90 Five Nations Championship were analysed using video-recordings of live television transmissions. When the ball was in open play, the average running pace of players central to the action ranged from 5 to 8 m s - 1 . This together with scrum, lineout, ruck and maul was classified as high-intensity exercise. The density of work was measured by timing the work:rest ratios (W:RRs) throughout each game. The mean duration of the work periods was 19 s and the most frequent W:RRs were in the range of 1:1 to 1:1.9. On average, a scrum, lineout, ruck or maul occurred every 33 s. The ball was in play for an average of 29 min during a scheduled time of play of 80 min. To complement the time-motion analysis, blood samples were taken from six players throughout a first-class game. The highest measured blood lactate (BLa) concentrations for each individual ranged from 5.8 to 9.8 mM. Running speed, duration, BLa levels, physical confrontation and, most particularly, the density of work as illustrated by the W:RRs indicate that the game places greater demands on anaerobic glycolysis than previously reported. This has implications for the physical conditioning of rugby union players. Keywords: Rugby union, time-motion analysis, physical fitness.
Introduction
The desire to develop appropriate conditioning programmes for games players has necessitated the development of time-motion game analysis. Although such analyses were reported prior to 1976, it was recognized that the methods used were neither objective nor reliable (Reilly and Thomas, 1976). Since then, published studies have used the pivotal components of match analysis validated by Reilly and Thomas (1976). The components are categorization of player activity, measurement of distance covered for each activity and the frequency of each activity. Although measurement of distance covered for each activity is possible, the process is more involved than measuring duration. Mayhew and Wenger (1985) identified that the duration of each level of intensity of activity is as representative of the physical demands as distance covered. More recent publications have used duration rather than distance (Docherty et al., 1988; Mayhew and Wenger, 1985; Treadwell, 1988). Other modifications have been made to suit different sports and computer technology included to facilitate data storage, analysis and presentation (Ohashi et al, 1988; McKenna et al, 1988; Treadwell, 1988). 0264-0414/92 © 1992 E. & F.N. Spon
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
286
McLean
Much of the early work on rugby union was reported in the book Fitness and Training for Rugby (Rugby Football Union, 1978), which contained information on distances run and number of sprints undertaken. Unfortunately, the methods and sources were not disclosed. More recently, papers using rugby as the subject have clearly reported their methods and results of time-motion analysis of specific playing positions (Docherty et al., 1988; Treadwell, 1988). To identify how fit players should be to play the game at the highest level, the game itself should be analysed rather than an individual player. The physical demands of a team game result in part from its rules and structure, as well as the skill and tactical ability of all the players involved. This requires analysis of the game at the highest level, since these factors are constraints on individual performance. The part that any one individual plays in a game may also be constrained by his fitness. Thus the analysis of one individual may underestimate the physical demands of the game. It is accepted that team games like rugby are of an intermittent nature. The physiological demand of intermittent activity depends not only on running speed and duration but on the density of physical work. The pattern of work:rest ratios (W:RRs) throughout the game is just as important a determinant of the physical demands of the game as total time or distance covered at different intensities. Work:rest ratios have not previously been reported in studies of rugby (Docherty et al, 1988; Treadwell, 1988). The measurement of blood lactate (BLa) has been done as an adjunct to the time-motion analysis of team games. One such study of rugby showed BLa concentration to be 2.8 mil at the end of a game. The conclusion drawn by Docherty et al. (1988) was that alactic metabolism was implicated. There are, however, several factors which do not support this conclusion. To give a clearer view of the type of metabolism used during international rugby union, measurements of BLa should be undertaken throughout a competitive senior game of the highest possible standard. In response to the demands to improve physical fitness of elite rugby union players, this paper intends to identify: 1. The density of high-intensity exercise by measurement of W:RRs in full international matches. 2. The BLa concentrations throughout a first-class game in Scotland.
Methods Time-motion analysis 1. Categorization of activity. The following classification of categories was used (Docherty et al, 1988): Low-intensity 1. Standing 2. Walking 3. Jogging
High-intensity 4. Run with elongated stride 5. Sprint 6. Non-run intense activity
These descriptors were used during the initial observation of video-recordings of live
The physical demands of rugby union
287
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
television transmission of matches in the Five Nations Championships during the 1989-90 season. They showed that as soon as the ball went dead, players switched into a low-intensity mode. This was regarded as a recovery phase which included both active and passive recovery. At the restart, it appeared that high-intensity exercise began for players central to the action. This assumption has to be tested. 2. Identification of high-intensity work. The distances used to calculate average running speeds were measured using a system of visual cues similar to the system validated by Reilly and Thomas (1976). The method involved sampling passages of play where the television coverage included sufficient visual cues to identify the distance travelled by a player central to the action. The sample was also representative of the range of distances run (11-70 m) and of the different playing positions in a team. The slow frame advance and pause facilities were used to enable the path of the run to be plotted on a scaled plan of a rugby pitch and measured (n = 16). These runs were timed using a digital stopwatch. The average speed was calculated in metres per second (m s" 1 ). To identify within-observer reliability, this process was repeated 3 weeks later without reference to the first measures. The standard deviation (S.D.) of the difference between duplicate measures was 1.3 m and 0.09 s respectively.
3. Work:rest ratio (W:RR). Each phase of high- and low-intensity exercise was timed in order to identify the W:RRs. Play was judged to start immediately the ball was thrown in at the lineout or the scrum was formed. If, for example, a penalty was awarded and the option was taken to kick direct to touch and no purposive action followed, then this period of time was included in the previous recovery period. This provided data for each contiguous W:RR throughout each game. The accuracy was checked using two methods. All the times recorded for each period of work and rest for the entire match were summed. This should produce a total of approximately 80 min. It cannot be assumed to be exactly 80 min, since it is at the referee's discretion as to how much time is added for stoppages. However, assuming the real time played to be 80 min, of the five international matches analysed the average error was 4.3%. The second method used was as described in the previous section. Passages of play were retimed (n= 15; range of duration = 3-40 s). The S.D. of the differences between duplicate measures was 1.69 s.
Blood lactate analysis Two Scottish First Division clubs agreed that during penalty kicks or stoppages for injury, blood samples could be taken from six subjects. Three subjects were selected from each side to include front, second and back row, halfback and three-quarter. Five samples were taken from each subject (each subject gave his informed consent). Finger prick blood samples were collected and 3.5 p\ samples analysed using a semi-automated analyser (Analox LM3). Analysis was carried out between 3 and 5 min after the sample was taken. The machine was calibrated using an independently produced standard. The S.D. of the differences between duplicate measures was 0.16 mM (n = 30). Three days after the match, a maximum treadmill test (Schnabel and Kindermann, 1983) was used to identify peak BLa concentration. It involved running up an incline of 4° at 20 km
McLean
288
h 1 until exhaustion. Post-test blood samples were collected at 1 and 4 min. The players were asked to follow the same diet for 24 h before each event.
Results
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
Ratification of high-intensity work by measurement of running speed during international matches Over five games, the average running speeds of the player either in possession or in closest pursuit of the ball was 5-8 m s" 1 . This range of running speeds compares with a similar analysis of soccer (van Gool et al., 1988). For rugby union, this running pace does reflect high-intensity exercise. The optimal pace of training runs used by the current Scottish international squad to improve aerobic fitness is in the range 4.2-4.7 m s" 1 (6.4-5.7 min mile" 1 ). General descriptors of activity Table 1 outlines the general descriptors of activity during five international matches. The average playing time was 29 min. This included scrums and lineouts, so that the forwards' active involvement was greater than the backs. The average duration of passages of play was 19 s. The large standard deviation reflects the variability of their duration. Figure 1 shows that the distribution is positively skewed. Table 1. Playing time and distance covered at high intensity during five matches in the 1989-90 Rugby Five Nations Championship
Playing time (min) first half second half total
France
Ireland
Wales
England
vs
vs
vs
vs
vs
England
Scotland
Scotland
Wales
France
14.2 12.8 27.0
14.0 14.0 28.0
14.7 17.1 31.8
12.5 16.2 28.7
14.7 13.9 28.6
21.0 18.6 20.5
16.3 11.4 19.3
17.0
5.6
9.2
20.6 16.3
17.8 10.5
21.7 16.6 19.4 13.5
19.6 17.0 19.9 15.4 19.7 19.7
70.0
51.0
47.0
56.0
Average duration of work periods (s) first half 18.9 S.D. 14.5 second half 16.1 S.D. 11.6 17.4 total S.D. 13.1 Longest work period (s)
45.0
8.8
Scotland
Table 2 contains standard information on the number of scrums, lineouts, rucks and mauls. The fact that rucks and mauls outnumbered scrums by 56% and lineouts by 44% has implications for conditioning.
289
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
The physical demands of rugby union
21-25
11-15
31-35
41-45
51-55
Wai V Scot ScotVFra EngVWal FraVEng Ire V Scot
Duration in seconds Fig. 1. Duration of work periods in international matches, 1989-90.
Table 2. The average number of scrums, lineouts, rucks and mauls during five games in the 1989-90 Rugby Five Nations Championship S.D. Range Total First half Second half Scrum Lineout Ruck/maul
15 22 37
17 19 36
32 41 73
7.7 3.3 13.7
24-45 38^6 62-88
Work:rest ratios (JV.RRs)
Figure 2 illustrates the distribution of W:RRs in five games. The distribution is very similar in all games and follows a normal distribution apart from the first bar which shows W:RRs of 1: > 4. Stoppages for injury and kicks at goal were responsible for the prolonged rest periods. The work:rest ratios which occur most frequently are 1:1-1.9 (26.6%) and 1-1.9:1 (20.2%). Figure 3 shows that there are nearly twice as many W: RRs where the work period is less than the rest period. Figures 4a and 4b provide an example of patterns of W: RRs during the first half of a game. These figures show how the distribution of different W:RRs varies with time. Figures 5a and 5b allow comparison of two short passages of play which illustrate the extremes of intensity of work which can occur in rugby union. In Fig. 5a, the third to the eighth passages of play involve six consecutive passages of play where the duration of work exceeds rest (a total of 137 s of work to 71 s rest). In Fig. 5b, the opposite is the case, with work lasting 123 s and rest 228 s. Analysis of W:RRs identified that the pattern shown in Fig. 5a was an extreme and uncommon occurrence. In the five games analysed, the maximum
McLean
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
290
1:3-19
1:2-Z9
1:1-1.9
1-1.9:1
WalVScot ScotVFra Ire V Scot /ZFraVEng EngVWal
Work: Rest Ratios Fig. 2. Work'.rest ratios in the 1989-90 Five Nations Championship showing similarities in distribution.
Fig. 3. Proportion of work:rest ratios averaged over five internationals in 1989-90.
number of consecutive passages of play with work greater than rest was six and the average was four. The number of occasions when two or more such passages of play occurred in a game was eight. The range was 4-8. Included in all passages of play is the high-intensity muscular effort required during scrums, lineouts, rucks and mauls (Table 2).
291
The physical demands of rugby union (a)
20 18.
16. 14. w R
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
R
W:RR=1:20
12. 10. 8. 6. 4. 2.
•• •
1 1 1 JL
0
6 17 2 18 21 30 10 8 38 5 28 7 6 4 19 25 18 12 9
9 4 12 12 5 33 13
Duration of each work period in s e c o n d s
(b)
(nme 21 mm 24 sec)
30. 25. 20.
W R
15.
R
10_ 5.
• I in in
in
In • I
l.llll inn
26 37 31 7 20 14 12 8 7 11 14 4 38 21 37 39 16 51 13 19 15 19 5 15 8 13 6
cn™2imin248ec)
Duration of each work period in seconds
( H - M ™ 391*28.80
Fig. 4. Contiguous work:rest ratios during (a) the first quarter and (b) the second quarter of Wales us Scotland in 1989-90. Blood lactate concentrations during match play
Table 3 shows that the highest measured BLa concentrations for each individual were in the range 5.8-9.8 mM. The highest BLa recorded during the game was compared with the player's maximum BLa, which was identified using a maximal treadmill test. The highest
292 (a)
McLean 7 6. 5.
w
4.
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
I 3. 2. 1.
I 14
(start oi second haio
(b)
14
37
20
24
7
28
21
Duration of each work period in seconds
23
(r,me7nin4isec)
4.5
4.0. 3.5. 3.0. W
2.5.
R
2.0.
R
1.5. 1.0. 0.5. 0.0 21 (Stand second half)
15
44
8
18
19
Duration of each work period in seconds
I (Time 8 min 14 sec)
Fig. 5. Contiguous work:rest ratios during (a) Wales vs Scotland and (b) England vs Wales, 1989-90.
The physical demands of rugby union
293
Table 3. Blood lactate concentrations (HIM) of first division players during a game of rugby union Team A
TeamB
Prop
No. 8
Stand-off
Prop
No. 8
Centre
15 min 30 min Half time 65 min Full-time
3.9 5.1 5.7 5.2 5.8
6.9 7.0 6.9 6.4 5.6
7.5 6.2 4.1 7.3 4.6
5.1 5.9 5.5 5.4 5.9
3.6 5.9 6.5 7.7 9.8
6.1 6.6 5.2 5.1 5.9
Mean S.D.
5.1 0.68
6.6 0.52
5.9 1.4
5.6 0.3
6.7 2.0
5.8 0.63
% of peak BLa
85%
57%
65%
61%
75%
56%
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
Time elapsed
BLa concentration recorded during the game is expressed as a percentage of maximum BLa in Table 3.
Discussion
The major differences between this paper and any other on time-motion analysis of rugby union is that it focuses (1) on the game and not the individual, and (2) on work:rest ratios of high intensity, not summations of time spent on different intensities of exercise. Previous papers have operated on the assumption that the physical demand of activity is determined by intensity (i.e. running speed) and 'volume', which is either total distance run or total time spent at different running speeds. The component of intensity not included in previous studies is density of intermittent work. To illustrate the point, if a player runs at a speed of 7.5 m s~1 over 120 m and does this six times with 80 s rest between each run, the W:RR would be 1:5. If the recovery was reduced to 40 s, the overall physical demand would be far greater yet the running speed and the volume have remained the same. Holmyard et al. (1988) investigated the effect of two different rest periods on mean power output during 10 maximal 6-s sprints on a non-motorized treadmill. Mean power output was maintained with 60 s recovery but dropped significantly on the fifth to tenth sprint with 30 s recovery. Both W:RRs produced BLa concentrations of between 17 and 18 nw. These results support the need to analyse W:RRs to produce a better understanding of the physical demands of rugby. All but 20% of the W:RRs illustrated in Fig. 2 were more severe than those used by Holmyard and co-workers as well as those in a similar study by Tumilty et al. (1988). Figure 1 also shows that approximately 80% of the work periods were longer than those used in these studies. It is accepted that during the passages of play referred to in this study, the players were not sprinting at full pace at all times. However, the average running paces of 5-8 m s" 1 are significantly higher than steady-state running pace for these players and as such are characterized by an exponential rise in BLa concentration (Kindermann et al., 1979). All of this information points to the considerable demands placed by the game on anaerobic metabolism. The BLa analysis carried out in this study confirms that anaerobic metabolism is required
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
294
McLean
for playing rugby union. The highest figures recorded for each individual were in the range 5.8-9.8 mM and exceed both the average of 4 mM, where aerobic metabolism is thought to be the major source of energy (Jacobs, 1981), and the range of 3.05-5.5 mM for steady-state running (Heck et al., 1985). In comparison to peak measurements of 14 mM in soccer (Ekblom, 1986), the levels in the present study could not be regarded as high. The timing of collection of blood samples was dictated by stoppages and was not specific to passages of intense play. Since it has been suggested that in soccer and rugby the periods of low-intensity exercise allow BLa to be metabolized (Mazzeo et al., 1982), the figures will tend to underestimate the true levels of peak BLa. The findings in this paper are in conflict with the results and conclusions of the only other study to report BLa levels in rugby union. Docherty et al. (1988) reported levels of 2.8 mM and duration of work periods not greater than 8.6 s. The low BLa level could be due to two factors. A single blood sample was taken 5 min after the game and the standard of players used in the analysis was not made clear, so they may not have been elite players. The conclusion reached by Docherty and co-workers was that the alactic energy source predominates and should be trained. Certainly, short sprints are an integral part of the game but the concept of work being alactic in isolation is rather theoretical. During sprints of 6 s duration, the three energy systems are active concurrently (Boobis, 1987). The present study suggests that the tempo of the game at international level places a large demand on anaerobic metabolism. This is supported indirectly by the moderate levels of aerobic power recorded for rugby union players (Maud and Shultz, 1984). The mean VO2 max of the Scotland squad in January 1990 was 53.4 + 2.7 ml kg" 1 min" 1 (unpublished personal data). The lower the aerobic power the greater the likelihood that energy demands will be met by anaerobic glycolysis. This is especially the case in rugby union where training does not include extensive steady-state exercise, which enables athletes to operate at a high percentage of their VO2 max without BLa accumulation. The need to analyse the game at the highest level requires compromises to be made in the methodology of time-motion analysis and BLa measurement. Ideally, blood samples should have been taken during an international match; however, such access was not available. Since the intensity of physical effort increases with the skill levels of the players involved (Ekblom, 1986), the results reported will tend to underestimate BLa concentrations during an international match. There are similar restrictions in obtaining access to all international grounds to make dedicated video-recordings which are ideal for time-motion analysis. The alternative was to use recordings of live television transmissions. This did not detract from the identification of W:RRs. The major intrinsic error resulted from the judgement of when play started and stopped, which is common to all time-motion analyses. Provided that suitable passages of play were selected, it was also possible to verify that when the ball was in play the players central to the action were required to run at high intensity. For a minority of occasions, the camera angle or the television director's choice of picture prevented a sequence of play being analysed. However, a representative sample of play was available to measure average running speed. The provision of the groundsman's plan of the visual clues of pitch markings and patterns of grass cut allowed distance to be measured using methods similar to that validated by Reilly and Thomas (1976). Within-operator reliability was acceptable given the range of average running speeds (5-8 m s1) representative of high-intensity exercise. The measurement error of distances run throughout the whole game was poor compared to that of Reilly and Thomas (1976), who achieved less than a 1% error, which highlights
The physical demands of rugby union
295
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
a limitation of live transmission for analysis. Nevertheless, the approximate distance of 2000 m, run at high intensity, is sufficiently accurate to be of use to coaches to help structure their training programmes for rugby union. In conclusion, the results of this study relate to the physical demands of the game and not to any one individual playing position. It does, however, apply more closely to playing positions which are central to the action throughout the game, but will be an overestimate for others. The identification of intensity of physical work by measurement of running pace and W:RRs, supplemented by the measurement of BLa concentrations, suggests that the extent of anaerobic glycolysis in rugby union is greater than previously reported. The data on W:RRs are of practical importance in the preparation of conditioning programmes for rugby union players.
Acknowledgement I gratefully acknowledge the suggestions of Drs C.S. Williams and M.A. Nimmo in respect of this manuscript.
References
Boobis, L.H. (1987). Metabolic aspects of fatigue during sprinting. In Exercise: Benefits, Limits and Adaptations (edited by D.A.D. MacLeod, R. Maughan, M.A. Nimmo, T. Reilly and C. Williams), pp. 116-43. London: E. and F.N. Spon. Docherty, D., Wenger, H.A. and Neary, P. (1988). Time-motion analysis related to physiological demands of rugby. Journal of Human Movement Studies, 14, 269-77. Ekblom, B. (1986). Applied physiology of soccer. Sports Medicine, 3, 50-60. Heck, J.H., Mader, A., Hess, G., Mucke, S., Muller, R. and Hollman, W. (1985). Justification of the 4 mmol 1-1 lactate threshold. International Journal of Sports Medicine, 6, 115-30. Holmyard, D.J., Cheetman, M.E., Lakomy, H.K.A. and Williams, C. (1988). Effects of recovery duration on performance during multiple treadmill sprints. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 134-42. London: E. and F.N. Spon. Jacobs, I. (1981). Lactate, muscle glycogen and exercise performance in man. Acta Physiologica Scandinavica, 495 (suppl.). Kindermann, W., Simon, G. and Keul, J. (1979). The significance of the aerobic-anaerobic transition for the determination of work load intensities during endurance training. European Journal of Applied Physiology, 42, 25-34. McKenna, M.J., Patrick, J.D., Sandstrom, E.R. and Chennells, M.H.D. (1988). Computer-video analysis of activity patterns in Australian Rules Football. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 274-81. London: E. and F.N. Spon. Maud, P.J. and Shultz, B.B. (1984). The US national rugby team: A physiological and anthropometric assessment. The Physician and Sportsmedicine, 12, 86-99. Mayhew, S.R. and Wenger, H.A. (1985). Time-motion analysis of professional soccer. Journal of Human Movement Studies, 11, 49-52. Mazzeo, R.S., Brooks, G.A., Budinger, T.F. and Schoeller, D.A. (1982). Pulse injection 1 3 C tracer studies of lactate metabolism in humans during rest and two levels of exercise. Biomedical Mass Spectrometry, 9, 310-14. Ohashi, J., Togari, H., Isokawa, M. and Susuki, S. (1988). Measuring movement speeds and distances in soccer. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 329-33. London: E. and F.N. Spon.
Downloaded by [University of Stellenbosch] at 06:38 27 April 2013
296
McLean
Reilly, T. and Thomas, V. (1976). A motion analysis of work-rate in different positional roles in professional football match-play. Journal of Human Movement Studies, 2, 87-97. Rugby Football Union (1978). Fitness Training for Rugby. Twickenham: RFU. Schnabel, A. and Kindermann, W. (1983). Assessment of anaerobic capacity in runners. European Journal of Applied Physiology, 52, 42-6. Treadwell, P.J. (1988). Computer-aided match analysis of selected ball games (soccer and rugby union). In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 282-7. London: E. and F.N. Spon. Tumilty, D. McA., Hahn, A.G., Telford, R.D. and Smith R.A. (1988). Is 'lactic acid tolerance' an important component of fitness for soccer. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 81-6. London: E. and F.N. Spon. van Gool, D., van Gerven, D. and Boutmans, J. (1988). The physiological load imposed on soccer players during real match-play. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 51-9. London: E. and F.N. Spon.
