The assessment of dynamic power is an important aspect of fine-tuning a high-level athlete's training program (5) and has also been shown to provide important information in regard to the ability of the athlete to excel at their chosen sport (3-5). A commonly reported method of assessing upper-body power is the ballistic bench throw (1,5,7,15,17,21), with studies ranging from cross-sectional assessment of the acute factors such as the loading/power spectrum (6,14) and force/velocity profile (9) to longitudinal training studies (18). In regard to the testing protocol, the movement commonly consists of a replication of the bench press, with the exception of the barbell being propelled from the hands at the termination of the movement. In this case, the countermovement is performed at the chest, with the barbell being ballistically lifted throughout the full concentric range of motion (ROM). Although the ballistic bench throw has been shown to result in a more intense exercise than the traditional bench press at an identical load (9), performing a countermovement test throughout the full ROM may not provide important information about the athlete's ability to produce high levels of force from a static start (13) or during concentric contractions, which are initiated in the midrange of the movement (10). These attributes are particularly important because they represent a variety of contraction types and functional ranges that may be utilized during sporting movements (8). In this context, the performance of a ballistic bench throw testing battery with the onset of the concentric contraction in both the full ROM and midrange of the movement and from both a countermovement and static start may provide valuable information.
Although implementing these different bench throw testing protocols may provide valuable information regarding athlete performance, time constraints and the requirement for specialized equipment could make this unfeasible. Knowledge of the correlations (relationships) between measurements of peak force and displacement produced during these selected tests of upper-body bench throw performance would ensure that the testing battery is time efficient, provides more global information about an athlete's performance, and does not contain redundant tests. Therefore, the aim of this study was to examine the relationship between the 4 different bench throw testing techniques-full ROM countermovement (FC), full ROM static start (FS), half ROM countermovement (HC), and half ROM static start (HS). The first component of this study was to assess the reliability of each of the techniques. We hypothesized (H1) that given the subjects' extensive training experience and familiarity with the movements, each of the testing protocols would provide reliable test-retest results. The second aspect of the study was to determine which of the 2 simplest testing protocols, FC and FS bench throws, provides the best representation of all-round upper-body bench throw performance. We hypothesized (H2) that FS bench throws would provide the best representation, as the force production requirements of this movement are the most likely to represent both the full- and mid-range exercises.
Experimental Approach to the Problem
The testing protocol was divided into 2 components, the test-retest reliability assessment and the single-session comparison of the different bench throw protocols. The testing sessions were performed in the athletes' preseason and were performed approximately 3 months after return from the off-season break.
The reliability component of the study consisted of a subset of the participants performing a series of FC, FS, HC, and HS bench throw tests in a Smith machine on 2 separate occasions at least 2 weeks apart. Measures of peak bench throw displacement (d) and peak force (f) for each of the trials were recorded from a position transducer and force plate, respectively. The test-retest reliability of the outcome measures was determined using intraclass correlation coefficient (ICC) analysis.
If these outcome measures were found to be reliable, they were to be used in the second component of the study-the relationship between each of the 8 outcome variables (FCd, FCf, FSd, FSf, HCd, HCf, HSd, and HSf). This was examined by repeating the experimental protocol used in the reliability assessment on all subjects. To assess the relationship between the outcome measures, a correlation matrix was created.
A pool of 26 male, semi-professional rugby league players (age: 23 ± 3 years, height: 180.8 ± 5.6 cm, mass: 95.7 ± 13.0 kg) with a minimum of 1 year of resistance training experience volunteered to participate in this series of experiments. All subjects in both components of the study had extensive ballistic training experience and had performed variations of each of the testing protocols as part of their preseason training regime. All 26 subjects completed the comparison study; however, only a subset of 11 (age: 23 ± 2 years, height: 181.3 ± 2.2 cm, mass: 92.7 ± 9.0 kg) were available for the 2 testing sessions required for the reliability study due to training and trial game commitments. Ethical approval was received from the Central Queensland University human research ethics committee for both of these segments of the experiment (approval number H05/09-105). All subjects completed an informed consent form outlining any risks involved in the study, along with a preactivity readiness questionnaire before participation.
All tests were performed using a force plate (Advanced Mechanical Technology Inc., Watertown, MA, USA) mounted weights bench. The bench was modified by replacing the cushioned board with a sheet of medium-density fiberboard cut to the same dimensions. The custom-made bench was designed with no cushioning to maximize the transfer of force from the subject to the force plate. This instrumented weights bench was positioned within a Smith Machine (Calgym, Robina, Australia) to restrict the movement of the barbell to a vertical, linear plane. Displacement was measured during the remaining tests using a linear position transducer (IDM Instruments, Hallam, Australia). Analogue to digital conversion of the data was performed using a combination of Labview PCI-6034E and PCI-6024E data acquisition cards (National Instruments, Austin, TX, USA). The data were synchronized and sampled at 1,000 Hz using a custom-written Labview software program (National Instrument). This combination of a linear position transducer for displacement measurement and a force plate for direct assessment of force allows for a valid assessment of the linear kinetic profiles of a single-dimension ballistic movement (12).
Warm-up and Lifting Standardization
The testing session began with a warm-up comprising 5 minutes of steady-state rowing at 150 watts on a rowing ergometer (Concept2, Morrisville, VT, USA) (9). After finishing the warm-up, the subject's standardized hand position during lifting was determined. This was performed to ensure repeatability of the tests, as small changes in hand positioning would potentially alter the ROM of the elbow and shoulder joints, thereby impacting adversely on the study results. Once in the lifting position the subjects were instructed to grip the barbell and lower it until their upper arm was parallel to the ground (measured using a goniometer mounted spirit level). To create a standardized lifting technique, the elbow angle at this position was 90°. If hand positioning was incorrect, the subjects were instructed to place the barbell on the lifting rack and adjust their hand position. This step was repeated until correct hand positioning was obtained, which was measured using 1 cm markings fixed onto the barbell and recorded on the subject's information sheet. Before the bench throws, the subject's bench press ROM was assessed with points at full, half, and terminal ROM subsequently determined using the linear transducer. This provided the position for the countermovement during the half ROM bench throws.
To determine an appropriate loading protocol for testing, the subject's 1 repetition maximum (1RM) was predicted based on the peak force levels recorded during an isometric maximal voluntary contraction (MVC), which replicated the bench press lift and was based on a technique reported previously (9). Two high-tensile adjustable cables were used to restrict ROM during the bench press exercise to approximately 2% of the distance from the chest to full elbow extension, in accordance with the position of peak force during a bench press exercise reported previously (16). The subject was required to perform a 3-second isometric MVC at this position, with a 2-second force ramp before maximal exertion. The peak force recorded during the test was assessed using the force plate mounted weight bench, with the testing position determined using the string potentiometer. Analysis of these data revealed a mean peak force of 1,285 ± 199 N, which corresponded to a predicted 1RM bench press strength of 130 ± 20 kg. This testing procedure was performed in a separate session before the bench throw tests. Based on the predicted 1RM results, all ballistic force tests were performed with an absolute load of 60 kg. This load was chosen because it provided a mass that would allow the subjects to be performing near their peak power output (23) while ensuring that it was not too excessive for the lower strength subjects.
The bench throw tests consisted of a total of 3 sets of 3 repetitions for each protocol, with a 1-minute rest between repetitions. The repetitions during each set were separated by a 15-second rest interval, both to reset the data collection software and to allow recuperation. Due to the submaximal load and the intermittent repetition scheme performed, this time frame was deemed suitable for subject recovery. Previous studies have used external control of the movement to ensure it was restricted to the upper-body pressing musculature, such as strapping the subject to the bench (14). We deemed this to be unnecessary due to the subject's extensive resistance and power training history, and believed that if anything this would have disrupted their technique and potentially produced unreliable results. The individual sets consisted of 4 different throwing conditions: full ROM countermovement (FC), full ROM static start (FS), half ROM countermovement (HC), and half ROM static start (HS).
Full ROM Countermovement Throws
During the FC set, the 60-kg barbell was unhooked from the Smith machine and lifted to the point where the subject's elbows were fully extended at the terminal ROM. On instruction by the investigator, the subject performed a ballistic bench throw. This movement required the subject to lower the barbell to the chest, perform the countermovement, and explosively throw the barbell into the air as high as possible. The subject was then required to catch the barbell during its descent, absorb the downward force, and then return it to the initial starting position.
Full ROM Static Start Throws
The starting position for the FS bench throws was at the full ROM position, with the barbell lightly touching the chest. These throws were concentric only, with no countermovement performed before the test. The Plyopower (Fitness Technologies, Skye, Australia) brake was adjusted to maximal setting throughout all static start tests, which provided a descent braking force of approximately 220 kg. This ensured that the descent of the barbell was controlled by the brake and not by the subject. This reduced the potential for elastic energy storage that would occur if the subject lowered the barbell to the starting position or was required to isometrically hold the barbell in position. Once the barbell reached the starting position, a 5-second rest interval was enforced before an audible signal to perform the next repetition.
Half ROM Countermovement Throws
The HC throws were performed using the same protocol as the full ROM countermovement throws; however, an audible signal was produced by the software when the barbell had descended to the midpoint of the ROM. At this point, the participant was required to perform the countermovement and throw the barbell for maximal height.
Half ROM Static Start Throws
For the HS bench throws, the Plyopower (Fitness Technologies, Skye, Australia) brake was maintained at a maximal setting, holding the barbell stationary at the subject's half ROM position. The subject was instructed to grip the barbell and perform a concentric-only throw upon an audible signal provided by the investigator.
Data Collection and Analysis
Both the force plate and the position transducer were set to initiate recording data in response to a keystroke performed by the investigator. This ensured that data for the entire movement were collected. Only the positive phase of the movement was analyzed, and therefore, it was necessary to determine the cutoff points for the concentric data. The onset of the positive phase of the movement was deemed to have occurred at the time of the first positive displacement data point during the concentric phase of the countermovement. The terminal cutoff point was defined as the point at which the barbell left the subject's hands, which was determined based on the terminal ROM position data recorded during the preceding ROM tests. An example of the displacement cutoff points and the corresponding force trace are provided in Figure 1.
Both the peak bench throw displacement (d), deemed the displacement of the barbell during the throw minus the bench press ROM displacement (measured using the position transducer), and the peak force (f) occurring during the countermovement (assessed using the aforementioned force plate system) were recorded and analyzed. The repetition with the highest throw displacement was used for analysis.
As a measure of relative reliability, the ICCs (22) were used, and these coefficients were calculated using a 2-way analysis of variance (ANOVA) based on absolute agreement. Absolute reliability was analyzed by the use of 2 indices. First, the SEM values were derived from the square root of the mean square error term from the respective repeated ANOVA (22). Conceptually, the SEM defines the range of plausible values that surrounds a single measurement where the true (population) value may lie. Second, the typical error (TE) of the change scores (also known as the minimum detectable change) was calculated for each test by multiplying the SEM value by the square root of 2. This TE value accounts for the error associated with repeated measurements (23).
A Pearson's correlation matrix was calculated for the 8 measures of bench throw performance, with each of the 4 different protocols (FC, FS, HC, and HS) yielding 2 different outcome measures (f, d). According to Cohen's criteria (11), the Pearson's correlations were interpreted as follows: trivial, <0.10; small, 0.10-0.29; moderate, 0.30-0.49; large, >0.50, with a significance level set at p < 0.05.
The results for the assessment of reliability are provided in Table 1. All testing protocols demonstrated ICC values that exceeded 0.80, which indicates acceptable test-retest reliability (22). The results of the comparison study are provided in Table 2, with the associated correlation matrix provided in Table 3. Both the FCd and FSd were strongly correlated with every measure of displacement for each testing protocol. The FCf measures were only moderately correlated with FSf and HCf throw peak force and the correlation with HSf was weak and nonsignificant. In contrast, the FSf was significantly correlated with all peak force results, with an r value in excess of 0.60 for both the midrange protocols.
The finding that each of the 4 testing protocols produced reliable results, which was consistent with previous studies (1,14) and supported our first hypothesis (H1), subsequently allowed us to perform the second component of this investigation-a comparison of the 2 full ROM testing protocols to determine which one provided the optimal overall measure of upper-body pressing power. The results of this comparison suggested that both full ROM displacement measures were strongly correlated with every measure of displacement. However, of the measures of peak force, only the FS test provided information about the athlete's ability to produce force in the midrange of the movement. This finding supported our second hypothesis (H2) and provides evidence that if time efficiency is important, the FS bench throw is the best overall measure of upper-body pressing power. When the difference in r values was formally examined (20) FSf, when compared with FCf, provides a significantly (p < 0.05) superior representation of the athlete's ability to produce force during a throw initiated in the midrange of the movement without a previous countermovement. This finding further supports the advantage of performing FS bench throws during athlete testing. From a practical perspective, the moderate to strong correlations of FSd with all other measures (except FCf) suggest that, in the absence of an instrumented weight bench such as the one used in the present study, the use of displacement-only measures provides a reasonable representation of upper-body bench throw performance. This is important considering that the cost and portability of linear position transducers, in comparison with a force plate, may make them more readily accessible to training staff.
Although the mechanisms underlying this finding were not examined as part of this study, it seems logical to assume that the different force/velocity profiles of the movements may explain this difference. As observed in previous studies (13), elastic energy contribution associated with the stretch-shortening cycle has a dramatic effect on peak force production, as evidenced by the ≈20% lower values recorded during the FS trials. Furthermore, the position of peak force production in the ROM is also different, with ex post facto analysis of the time to peak force suggesting that during the FC bench throws, peak force is produced during the countermovement (data not shown), which corresponds to a phase within the previously recognized sticking region of the bench press (16). In contrast, during the FS bench throws, peak force is not produced until 481 ± 218 milliseconds after the start of the concentric phase of the movement. This shifts the position of peak force production after the sticking region and well into the midrange of the movement, which supports the finding that this testing protocol provides the best representation of midrange force production of the 2 full ROM bench throw tests.
Of further interest is the weak relationship between peak force in the 2 full ROM tests, which may be due to the force profile of each test. Based on the relationship between peak force during each full ROM test and peak force during the 2 midrange tests, it would seem that the differences in athlete's technique during the countermovement test has a major effect on the force produced during the movement. During an FC bench throw, it is difficult to directly control the athlete's technique. Factors such as the uncontrolled descent velocity and barbell impact with the chest would alter the peak force recorded when a system implementing a force plate is used. This would result in a peak force reading with only a limited relationship to functional performance, which is evidenced by the low relationship between peak force and bench throw displacement during the FC bench throw. It would seem that despite this lack of a strong relationship between peak force and functional performance, based on our findings the peak force measure is reliable. This is likely due to the trained status of the subjects' involved in this study. Based on these findings, we conclude that although peak bench throw force recorded during an FC throw via a force plate is reliable, its convergent validity with functional performance is questionable. Interestingly, this conclusion does not appear to hold true for the HC throw. Peak force recorded during this form of the test had a strong relationship with every measure of full ROM performance, which would suggest that the midrange countermovement may be less reliant on the technique of the movement.
As with any study, a number of limitations were present. First, the use of an absolute load based on the group's mean strength levels would not provide the optimal loading protocol for each individual athlete. Despite this, the use of an absolute testing load is common both in previous bench throw (2) and in upper-body bench pressing strength (19) studies, and, based on the investigators' personal observations, the assessment of a team athlete's performance in a gym-based setting. One of the primary motives for this absolute load and a driving factor in the creation of this study is the need to ensure that a testing protocol is time efficient. Although the loading protocol used in the present study achieved this aim, further studies assessing relative loading are required. Of particular interest would be the relationship between these testing protocols, and the effect on reliability, of midrange loading protocols based on midrange strength levels, which have been shown to be dramatically higher than full ROM strength (10). Second, although the 2 full ROM testing protocols were deemed to be easier to perform in a gym-based setting, the FS throw incorporated the Plyopower (Fitness Technologies) braking mechanism to reduce elastic energy contribution to the movement. It is unlikely that many gym-based testing settings would possess this type of equipment; however, this can be replicated in a number of ways such as incorporating the stoppers on a Smith machine or having training partners hold the barbell in the full ROM position.
If time efficiency is an important component of an athlete's testing program, we recommend assessing throw height during full ROM static start bench throws. This is a simple test that provides the best overall measure of upper-body throwing performance while requiring minimal equipment and can therefore be implemented in almost any training or testing setting.
The authors would like to thank ASICS Oceania for their support.
1. Alemany, JA, Pandorf, CE, Montain, SJ, Castellani, JW, Tuckow, AP, and Nindl, BC. Reliability assessment of ballistic
jump squats and bench throws. J Strength Cond Res
19: 33-38, 2005.
2. Baker, D. A series of studies on the training of high-intensity muscle power
in rugby league football players. J Strength Cond Res
15: 198-209, 2001.
3. Baker, D. Comparison of upper-body strength and power
between professional and college-aged rugby league players. J Strength Cond Res
15: 30-35, 2001.
4. Baker, D. Differences in strength and power
among junior-high, senior-high, college-aged, and elite professional rugby league players. J Strength Cond Res
16: 581-585, 2002.
5. Baker, D and Nance, S. The relation between strength and power
in professional rugby league players. J Strength Cond Res
13: 224-229, 1999.
6. Baker, D, Nance, S, and Moore, M. The load that maximizes the average mechanical power
output during explosive bench press throws in highly trained athletes. J Strength Cond Res
15: 20-24, 2001.
7. Baker, D and Newton, RU. Methods to increase the effectiveness of maximal power
training for the upper body. Strength Cond J
27: 24-32, 2005.
8. Bloomfield, J, Ackland, TR, and Elliott, BC. Applied Anatomy and Biomechanics in Sport
. Melbourne, Australia: Blackwell Scientific Publications, 1994.
9. Clark, RA, Humphries, B, and Bryant, AL. A comparison of force curve profiles between the bench press and ballistic
bench throws. J Strength Cond Res
22: 1755-1759, 2008.
10. Clark, RA, Humphries, B, and Bryant, AL. An examination of strength and concentric work ratios during variable range of motion training. J Strength Cond Res
22: 1716-1719, 2008.
11. Cohen, J. Statistical Power Analysis for the Behavioral Sciences
. Hillsdale, NJ: Lawrence Erlbaum, 1988.
12. Cormie, P, McBride, JM, and McCaulley, GO. Validation of power
measurement techniques in dynamic lower body resistance exercises. J Appl Biomech
23: 103-118, 2007.
13. Cronin, JB, McNair, PJ, and Marshall, RN. Magnitude and decay of stretch-induced enhancement of power
output. Eur J Appl Physiol
84: 575-581, 2001.
14. Cronin, JB, McNair, PJ, and Marshall, RN. Force-velocity analysis of strength-training techniques and load: Implications for training strategy and research. J Strength Cond Res
17: 148-155, 2003.
15. Falvo, MJ, Schilling, BK, and Weiss, LW. Techniques and considerations for determining isoinertial upper-body power
. Sports Biomech
5: 293-311, 2006.
16. Lander, JE, Bates, BT, Sawhill, JA, and Hamill, J. A comparison between free-weight and isokinetic bench pressing. Med Sci Sports Exerc
17: 344-353, 1985.
17. Lawton, T, Cronin, J, Drinkwater, E, Lindsell, R, and Pyne, D. The effect of continuous repetition training and intra-set rest training on bench press strength and power
. J Sports Med Phys Fitness
44: 361-367, 2004.
18. Lyttle, AD, Wilson, GJ, and Ostrowski, KJ. Enhancing performance: Maximal power
versus combined weights and plyometrics training. J Strength Cond Res
10: 173-179, 1996.
19. Mayhew, JL, Ware, JS, Bemben, MG, Wilt, B, Ward, TE, Farris, B, Juraszek, J, and Slovak, JP. The nfl-225 test as a measure of bench press strength in college football players. J Strength Cond Res
13: 130-134, 1999.
20. Meng, XL, Rosenthal, R, and Rubin, DB. Comparing correlated correlation coefficients. Psychol Bull
111: 172-175, 1992.
21. Newton, RU, Kraemer, WJ, Hakkinen, K, Humphries, BJ, and Murphy, AJ. Kinematics, kinetics, and muscle activation during explosive upper body movements. J Appl Biomech
12: 31-43, 1996.
22. Vincent, WJ. Statistics in Kinesiology
. Champaign, IL.: Human Kinetics, 1995.
23. Wilson, GJ, Newton, RU, Murphy, AJ, and Humphries, BJ. The optimal training load for the development of dynamic athletic performance. Med Sci Sports Exerc
25: 1279-1286, 1993.