Introduction

Study 1 - Evaluation of the Pneumatically Controlled Predator Individual Scrummaging Machine

In order for the scrummaging machine to effectively assess scrummaging performance, the task of exerting force against the shoulder pad must demand the same specific technical skills that are required during scrummaging in a game. If this were not the case then the force values provided by the machine would not be specific to rugby scrummaging and would therefore offer reduced information on scrummaging performance. The evaluation of the Predator individual scrummaging machine would hopefully answer the following two questions:

1. Do the force values provided by the scrummaging machine reflect specifically scrummaging performance, or do they simply represent general strength?
2. Does the scrummaging machine demand the specific scrummaging technical skills, and therefore is it able to distinguish between good and poor scrummaging technique?

Evaluation Method: Twelve college/university level rugby forwards (age: 19 -22 years, body mass: 101.5 ± 14.6 kg) participated in the evaluation study. Following familiarisation with the Predator equipment, the subjects were required to apply maximal force using their most efficient and effective scrum technique against the pneumatic scrum machine. Scrummaging performances were videoed for technique assessment, including hip and knee angles, and the slope of the back. Coordinates of the subjects' joint centres were reconstructed from digitised video data using 2D Direct Linear Transformation (DLT). Subjects also performed a maximal squat jump from a static position, with an ideal scrummaging knee angle of 120°(O'Shea,1996), from a Kistler force platform. The purpose of the vertical squat jump was to obtain a measure of players explosive strength in an activity that demanded minimal skill. Maximal force data generated during the minimal skill demanding squat jump were correlated with the maximal force applied during scrummaging.

Evaluation Results and Discussion: Figure 2 displays the relationship between peak force exerted during the vertical squat jump and maximal force exerted whilst scrummaging on the Predator individual scrummaging machine.

Figure 2. Relationship between squat jump peak force and Predator scrum force.

Although both tests were measuring explosive leg strength, Pearson correlation analysis showed only a moderate relationship with a correlation of r = 0.55 (p = 0.06), which is perhaps not as strong as one would have expected. If scrummaging performance was solely due to leg strength then all subjects would be positioned along the line of best fit. This line of best fit given by the regression equation y = 0.0607x + 149.95 could therefore be taken as representing a players expected Predator scrum force based upon their squat jump peak force, if leg strength was the sole determinant of scrummaging force. Leg strength obviously is not the sole determinant of scrummaging force as the correlation coefficient is only r = 0.05, and other factors such as scrummaging technique contribute to the determination of scrummaging force. However this assumption is useful to help explain the selection of players for the technique groups detailed below.

Looking closely at Figure 3, Subject 1 (S1) exerted 53kg more than his expected value given by the line of best fit, and Subject 12 (S12) relatively achieved the lowest scrum force value, exerting 81 kg less than expected. It was hypothesised that those subjects obtaining scrum force values above the best fit line would display better scrummaging technique than those subjects located below the line of best fit. The three subjects that performed the best in relation to their squat jump peak force score made up the hypothesised 'Good Technique' group (refer to Figure 4). These three subjects technique were compared to the three subjects of the 'Poor Technique' group who performed the worst in relation to their squat jump value.

Scrummaging technique within the rugby literature recommend to maintain the back in a near horizontal position with shoulders slightly above the hips (Greenwood, 1992; O'Shea, 1996). Figure 5 illustrates that the 'Good Technique' group displayed the recommended technique with a back slope of around 5 degrees above the horizontal. The 'Poor Technique' group exhibited a steep back slope approaching 20 degrees.

The hip angles are displayed in figure 6. The 'Poor' group tended to have larger hip angles throughout the scrummaging performance. There was little difference in knee angles between the two groups although it appeared that the 'Good' group initiated knee extension earlier than the 'Poor' group (Figure 7).

Looking at both the hip and knee angles together provides a better indication of what is happening during the scrummaging performance. Figure 8 displays the hip and knee angles for the 'Poor' group. It is evident that the 'Poor' group relied on excessive dropping of the hips to initiate the movement, with there being a delayed knee extension. This dropping of the hips results in larger hip angles, which causes the steep slope of the back. Figures 9 and 10 show Subject 12 from the 'Poor' group clearly dropping his hips to initiate the movement.

Technique analysis of the 'Good' group indicated that there was simultaneous knee extension with the initial hip extension (Figure 11). This ability to co-ordinate the knee and hip extension, maintains a shallow slope of back, and results in a more powerful forward drive.

Evaluation Summary: The evaluation study was carried out to answer the following questions:

• Do the force values provided by the scrummaging machine reflect specifically scrummaging performance, or do they simply represent general strength?
• Does the scrummaging machine demand the specific scrummaging technical skills, and therefore is it able to distinguish between good and poor scrummaging technique?

Results from the study indicate that the scrummaging force values provided by the Predator individual scrummaging machine do specifically reflect scrummaging performance. The scrummaging force values are scrummaging specific and are dependent upon technical skills and the control of explosive strength. Subjects that obtained better scrum force values relative to their explosive strength ('Good' group) tended to display technique similar to that recommended in the rugby literature (Greenwood, 1992; O'Shea, 1996), and was distinctively different from subjects that obtained relatively low scrum force values ('Poor' group).

The disparity in performance between the two tests of maximal force is a consequence of differences in scrummaging technique. It can therefore be concluded that the pneumatically controlled Predator individual scrummaging machine is capable of distinguishing players that perform well during scrummaging, from those that have a good level of explosive leg strength but limited scrummaging technical skills.

The findings of the Evaluation study highlight that accurate assessment of physical performance requires sport specific tests, which replicate the demands of the game. As most rugby clubs do not have access to a scrummaging machine with force measurement capabilities, they tend to rely on general physical performance tests to get an indirect measure of a players scrummaging potential. The second study compares scrummaging machine performance scores with performance scores attained in frequently used standard field tests for leg power.

Study 2 - Comparison of Scrummaging Machine Scores with Standard Field Tests for Power in Rugby

In the game of rugby, the development of leg power is of particular importance for forwards in the scrum and maul. The forward is required to hold the push in the scrum and give an explosive push when the ball is put in. The most frequently used standard field tests for leg power according to Nicholas (1997), which are also recommended in numerous training manuals and by the England Rugby Football Union (Reddin, 1999), are the vertical jump and a timed short sprint test. The performance unit for the vertical jump is usually expressed in centimetres (cm), and for the timed sprints in seconds (s).

Recently a number of researchers have taken body mass into consideration for the timed sprints and calculate momentum obtained during the sprints (kg.m/s) (Quarrie et al., 1996; Deutsch & Sleivert, 2000). Rugby researchers do not appear to take body mass into account when expressing performance in the vertical jump, even though there is the Lewis formula (Wilkinson and Moore, 1995) which calculates the power from the vertical jump from subjects jump height and body mass. The Lewis formula equation is power (watts) = 21.72 x mass (kg) x square root of height jumped (m). An interesting finding for the vertical jump reported by Nicholas (1997) is that backs generally score higher than forwards, with the outside backs being the best performing player sub-group (Quarrie et al., 1996). Based upon these results for the vertical jump, which tests for explosive leg power, would you put the outside backs into the scrum?!!! The suitability of these indirect field tests, especially if body mass is not considered, and their applicability to rugby union is therefore questionable.

This study was designed to compare performance scores attained and the maximal force applied in these tests with the maximal force that could be applied against the pneumatically controlled Predator individual scrummaging machine. The comparisons would hopefully answer the following question:

How valid are performance scores from standard field tests of leg power in rugby for assessing maximal force application potential of rugby forwards?

Comparison Method: Ten college/university level rugby forwards (age: 19 -22 years, body mass: 101.5 ± 15.5 kg) were tested performing a maximal standard vertical jump from a Kistler force platform and a 15m sprint from a standing start through a photo-electric timing system which recorded times at 5m intervals. Following familiarisation with the Predator equipment, the subjects were then required to apply maximal force using their most efficient and effective scrum technique against the pneumatic scrum machine. The equipment allowed the subjects to move through a knee angle of 120° for greatest force application (Zatsiorsky, 1995) to maximum extension, while keeping the hips horizontal with the shoulder or slightly below, providing sufficient resistance to measure maximal force application in kilograms (kg).

Correlations were carried out between the maximal force applied during scrummaging against the Predator individual scrummaging machine and each of the following:

• height jumped in vertical jump (cm)
• maximal jumping force from the force platform (N)
• calculated power from subjects jump height and body mass using the Lewis formula (watts).
• fastest sprint times at 5, 10 and 15 metres (s)
• calculated momentum from subjects mass and sprint times (kg.m/s)
• calculated average force from the subjects' mass and sprint times (N).
  This average force was calculated using an original formula devised by the researchers, average force (N) = 2 x sprint distance (metres) x mass (kg) / time squared. (Average force = 2dm/t2)

Comparison Results and Discussion: Rugby studies, which have used the vertical jump and sprint tests, have tended to compare performance between players based on height jumped and movement time over specified distances. When comparing the Predator scrum force with the best vertical jump height, Pearson correlation analysis showed there was an insignificant relationship with a correlation coefficient of r = 0.38 (Figure 12). This finding is in agreement with Quarrie and Wilson (2000) who found an insignificant correlation of r = - 0.13. A much stronger relationship (r = 0.67, p<0.05) is demonstrated in figure 13 when the Predator scrum force is compared with maximal vertical force applied to the force platform during the vertical jump. However, how many rugby clubs possess a force platform?

When body mass was taken into account for the vertical jump and power was calculated using the Lewis formula a strong relationship was found with a correlation of r = 0.83 (p = 0.002), which is illustrated in figure 14. There was no relationship between scrum force and the sprint times at 5m (r = -0.26), 10m (r = -0.06) or 15m (r = 0.05), with figure 15 displaying the insignificant relationship with 15 metre sprint times.

When body mass was taken into consideration with the calculation of the momentum obtained during the sprints, high correlations of r = 0.74, r = 0.76, and r = 0.80 (all p values < 0.01) respectively were obtained. Figure 16 illustrates the relationship for the 15metre sprint. When average force during each of 5m, 10m and 15m sprints was calculated using the formula - average force = 2dm/t2 and compared to the maximal force applied during scrummaging against the Predator individual scrummaging machine, thereby comparing force with force, the strongest relationships, of all performance scores investigated, were found. Correlation coefficients of r = 0.88, r = 0.88, and r = 0.90 (Figure 17) respectively were obtained (all p values < 0.001).

Comparison Summary: The comparison study was carried out to answer the following question:

How valid are performance scores from standard field tests of leg power in rugby for assessing maximal force application potential of rugby forwards?

Results from the study indicate that the raw data of these recommended standard tests is of little use in evaluating the ability of forwards to apply force in a scrum. The raw data needs to be converted into appropriate units, which take into account the subject's mass and have a direct, accurate and pertinent relationship with the performance objectives of the playing position. Most training and testing manuals do not take this fact into consideration.

Conclusion

This article began by asking, "what is a valid method to accurately assess scrummaging performance?" The pneumatically controlled Predator individual scrummaging machine, one of a small number of scrummaging machines now available that possess force measurement devices, was evaluated in terms of its ability to effectively assess scrummaging performance. The evaluation study demonstrated that the Predator individual scrummaging machine demands specific scrummaging technical skills and is capable of distinguishing players that perform well during scrummaging, from those that have a good level of explosive leg strength but limited technical skills. The comparison study then compared maximal force applied against the Predator individual scrummaging machine with performance scores attained in recommended standard tests of leg power for rugby union. The highest relationship between Predator scrum force and the standard tests for leg power were obtained for sprinting when the time data was converted to average force. Without this additional calculation the raw data was shown to be quite meaningless.

The average force calculation provides all rugby clubs with a relatively simple test for rugby related leg power. However, with the new era of professional rugby and an increased scientific approach, rugby clubs should evaluate the validity of the physical tests they currently use and consider adopting rugby specific tests that measure scrummaging performance directly. This would increase confidence in the validity of the test data and result in the real weaknesses being addressed, thereby improving the preparation of players for competition.