International Journal of Sports Physiology and Performance, 2011, 6, 51-57

© 2011 Human Kinetics, Inc.

Which Jump Variables Should Be Used to

Assess Explosive Leg Muscle Function?
Warren Young, Stuart Cormack, and Michael Crichton

Purpose: The main purpose of this study was to determine the relationships
between countermovement jump (CMJ) variables and acceleration and maximum
speed performance. Methods: Twenty-three elite Australian football players were
tested on a CMJ, which yielded several kinematic and kinetic variables describing
leg muscle function. A 40 m sprint was also conducted to assess acceleration (10
m time) and an estimate of maximum speed (flying 20 m time). Players from one
Australian Football League (AFL) club were tested and Pearson correlations for
CMJ variables and sprint performance were calculated. Results: Jump height, peak
velocity, peak force, and peak power had less than 50% common variance, and
therefore represented independent expressions of CMJ performance. Generally,
the correlations between CMJ variables and sprinting performance were stronger
for maximum speed (small to large effect sizes) than for acceleration (trivial to
moderate sizes). The variable that produced the strongest correlation with accelera-
tion was jump height (r = –0.430, P = .041) and with maximum speed was peak
power/weight (r = –0.649, P = .001). Conclusions: The results indicate that if an
integrated system comprising a position transducer and a force platform is available
for CMJ assessment, jump height and peak power/weight are useful variables to
describe leg muscle explosive function for athletes who perform sprints.

Keywords: Australian football, power, testing, speed, muscle function

Countermovement jumps (CMJ) are widely used by sport scientists and

coaches to assess the explosive qualities of the leg muscles in athletes.1–4 A CMJ is
essentially a vertical jump without an arm swing. From a standing position, a dip
or countermovement is performed before extension of the legs. The hands may be
held on the hips or a stick or bar may be held on the shoulders. A range of loads can
be used to obtain an incremental load power profile.3 With the advent of portable
force platforms and position transducers, there is a multitude of variables that can
be obtained from a CMJ. Some examples include jump height, bar velocity, force
production, and power output, and can be recorded as peak values, mean values and
may be expressed in absolute terms or relative to body mass. Further, if CMJ are
performed with a variety of loads on a bar, the best value across these loads may be

Warren Young is with School of Human Movement and Sport Sciences, University of Ballarat, Ballarat,
Victoria, Australia. Stuart Cormack is with Essendon Football Club, Essendon, Victoria, Australia.
Michael Crichton is with the School of Human Movement and Sport Sciences, University of Ballarat,
Ballarat, Victoria, Australia.

51
52   Young, Cormack, and Crichton

described as the “optimum,” this is, optimum power.5 There has been a considerable
amount of research relating to the reliability and validity of CMJ assessment3,5,6 as
well as determination of optimum power.5,7,8 It appears that the jump height, veloc-
ity, force, and power variables obtained from a CMJ yield acceptable reliability.3,6,9
When testing an athlete with the CMJ, the analyst must understand how to
interpret the multiple variables. For example, are these different variables expres-
sions of the same underlying physical quality or do they represent somewhat dif-
ferent characteristics of explosive muscle function? Based on an unloaded CMJ,
the correlation between peak power and jump height in the same jump was r =
–0.12.9 This represents a shared or common variance of only 1.4%, and suggests
that these variables measure different characteristics of CMJ performance. The
problem for the analyst is to determine which of these variables is more important
to focus on when attempting to monitor leg muscle function or when using the
CMJ to determine individual strengths and weaknesses. Since “power” is described
as a fundamental quality required for many sports,10 it is tempting to focus on the
mechanical power output (force × velocity) measured in watts. However, an impor-
tant question is whether this expression of explosive force obtained from a CMJ is
the most relevant expression to the sport of interest. One solution to this problem
could be to identify which of the variables obtained from a CMJ is most correlated
to (and therefore has most in common with) the sports performance of interest.
The relationships between power output and jump height obtained in a CMJ
with 20 m sprint performance was investigated by Hori et al.9 They reported
correlations of r = –0.58 and r = –0.69 for peak power/weight and jump height,
respectively, suggesting that jump height had somewhat more in common with sprint
performance. Another way to relate CMJ variables to performance is to determine
which variables discriminate higher and lower levels of performance. Sheppard et
al3 tested two groups of volleyball players with a CMJ and recorded jump height,
peak velocity, force, and power. They reported that the higher performance group
who had international experience, performed better in all variables but the biggest
differences were in peak power and power/weight. Further, these two variables were
more sensitive to a 12 wk power training program than the other variables. The above
findings collectively indicate that the CMJ variable that correlates best with perfor-
mance is not clear, and may depend in part on the sports performance in question.
There were two aims of the present study. The first was to determine the rela-
tionships among CMJ variables to identify whether they are essentially measuring
the same characteristics or if they represent independent qualities. A second aim
was to determine the correlations between selected CMJ variables and sprinting
performance (both acceleration and estimated maximum speed). The outcomes
of this study may be used to assist with the interpretation of CMJ assessment of
athletes who perform sprinting.

Methods
Subjects
The participants were 23 male professional Australian football players from one
club who participate in the Australian Football League (AFL). The mean ± SD
number of games, age, height, and body mass are shown in Table 1. The players,
 Jump variables and muscle function 53

Table 1 Mean ± SD descriptive results

Item Sample (n = 23)
Number of AFL games 29.0 ± 37.6
Age (y) 22.3 ± 2.1
Height (cm) 185.6 ± 7.1
Body mass (kg) 84.3 ± 9.0
10 m time (s) 1.73 ± 0.06
20–40 m time (s) 2.22 ± 0.06
Jump height (cm) 37.9 ± 0.05
Peak bar velocity (m⋅s–1) 2.56 ± 0.26
Peak force (N) 1936 ± 370
Peak force/weight (bw) 2.3 ± 0.4
Peak power (W) 4967 ± 632
Peak power/weight (W⋅kg–1) 59.1 ± 5.9

who were free of injury or illness, volunteered for CMJ and sprint testing. The study
was approved by the University Human Research Ethics Committee.

CMJ Assessment
The CMJ was performed with a self-selected countermovement while holding a light
fiberglass pole on the shoulders. The participants were instructed to hold the pole
firmly on the shoulders and attempt to be as explosive as possible while jumping
for maximum height. Since the majority of the players were not experienced with
loaded CMJ, only unloaded CMJ were assessed. This meant that calculation of
the “optimum power” was not possible. Using unloaded CMJ was not considered
a problem because unpublished observations of CMJ assessment of these athletes
typically showed no increase in power output with added loads.
The jumps were analyzed according to the recommendations of Dugan et al.8
This involved using an integrated system comprising a position transducer (PT5A,
Fitness Technology, Australia) and a force platform (400 Series Performance Force
Plate, Fitness Technology Australia) interfaced to computer software (Ballistic
Measurement System, Fitness Technology, Australia). The best score from three
trials was retained for analysis. Jump height and peak bar velocity were determined
from the position transducer, peak force was determined from the force platform
recording, and power was calculated from a combination of both devices.8 The test
and retest reliability of jump variables using these procedures has previously been
reported to be good, with a coefficient of variation of less than 10%.3,4

Sprint Testing
A 40 m sprint was conducted on an indoor wooden floor. The participants were
required to use a stationary start with the toe placed on the start line. An electronic
54   Young, Cormack, and Crichton

timing system (Fusion Sport, Queensland, Australia) was used to obtain split times
at 10 and 20 m. The 10 m time was used as a measure of ability to accelerate
and the split between 20 and 40 m (flying 20 m time) was used as an estimate of
maximum speed.11,12 The best times from three trials were retained for analysis.
The test and retest reliability for a 40 m sprint test has been shown to be good in
elite junior rugby league players, based on an intraclass correlation of 0.97 and a
coefficient of variation of 1.2%.13
Since acceleration and maximum speed are independent speed qualities,11,12
it was expected that the correlations with the CMJ variables would be different for
the two speed qualities. The sprint testing was conducted 2 d following the CMJ
testing. The participants were tested during the AFL preseason during January.

Statistical Analyses
For the purpose of determining the relationships among the CMJ variables and
between the CMJ variables and sprint performance, Pearson correlations were cal-
culated. The percent shared or common variance between variables was described
with the coefficient of determination (r2 × 100). Statistical significance was set at P
< .05, and the correlation coefficients were classified as 0.0–0.1 = trivial, 0.10–0.3
= small, 0.3–0.5 = moderate, 0.5–0.7 = large, and 0.7–0.9 = very large.14

Results
The descriptive results are shown in Table 1. The correlation for 10 m time and
the 20–40 m time was r = .673 (P = .001), representing a common variance of
45%. When the common variance is less than 50%, the variables can be said to be
independent,15 and therefore the sprint measures represented independent speed
qualities. The interrelationships for the variables generated from the CMJ are shown
in Table 2. These results indicate common variances of less than 50% (3–44%) for
jump height, peak bar velocity, peak force, and peak power, indicating that these
variables represent independent characteristics of CMJ performance.
The interrelationships between the speed qualities and the CMJ variables are
shown in Table 3. Generally, the negative correlations are stronger with the 20–40
m time than the 10 m time, indicating that explosive muscle function is somewhat
more important for maximum speed. The greatest negative correlations with

Table 2 Correlation matrix for CMJ variables (n = 23)

Variable Velocity Force Force/wt Power Power/wt
Height 0.666 (44) –0.163 (3) –0.036 (0.1) 0.411 (17) 0.741 (55)
Velocity –0.506 (26) –0.501 (25) 0.530 (28) 0.845 (71)
Force 0.845 (71) 0.185 (3) –0.410 (17)
Force/wt –0.166 (3) –0.305 (9)
Power 0.575 (33)
Note. The number in parentheses is the percent common variance (r2 × 100).
 Jump variables and muscle function 55

Table 3 Relationships Between Speed Qualities and CMJ Variables

(n = 23)
Variable 10 m 20–40 m
Height –0.430 (moderate) –0.357 (moderate)
Peak velocity –0.394 (moderate) –0.478 (moderate)
Peak force 0.166 (small) 0.378 (moderate)
Peak force/weight –0.079 (trivial) 0.168 (small)
Peak power –0.040 (trivial) –0.179 (small)
Peak power/weight –0.425 (moderate) –0.649 (large)
Note. Significant correlations (P < 0.05) are shown in boldface type. The qualitative descriptor for the
strength of the correlation is shown in the brackets.

acceleration were for jump height (r = –0.430, P = .041) and peak power/weight
(r = –0.425, P = .043) and with maximum speed were for peak power/weight (r =
–0.649, P = .001) and peak velocity (r = –0.478, P = .021).

Discussion
Interrelationships Between CMJ Variables
Table 2 shows that there were common variances of less than 50% for jump height,
peak bar velocity, peak force, and peak power, indicating that these variables rep-
resent independent characteristics of leg muscle function in a CMJ. This finding
supports the notion that CMJ variables cannot be used interchangeably. There is a
need to determine what specific characteristics are being assessed by each variable.
Conceivably, each variable could represent different portions of the force-velocity
relationship with some being more force dominated and some being more speed
dominated. Whether a variable is expressed in absolute terms or relative to body
mass may also influence the characteristics of the measure. For example, the
common variance between peak power and peak power/weight was 33%, indicat-
ing that these expressions of power are independent.
The different characteristics of the CMJ variables as indicated by their statisti-
cal independence, highlights a problem for the analyst conducting CMJ testing. If
an athlete improves explosive muscle function of the leg muscles from training,
one CMJ variable might be highly reflective of this while another could be quite
insensitive to any training-induced changes. Therefore, it is an advantage to focus
on one variable that best represents the muscle function that is desired, such as for
sprint performance.

Interrelationships Between CMJ Variables and Speed

Qualities
Table 3 indicates that generally the correlations between the CMJ variables are
slightly stronger for maximum speed (small to large) compared with acceleration
(trivial to moderate). The greatest statistically significant (P < .05) correlation
with acceleration was jump height. A possible reason for this is that jump height
56   Young, Cormack, and Crichton

is generated by the accumulated vertical velocity over the entire jump. Therefore,
it may reflect a slightly slower explosive force quality than the other variables,
and this may be a more similar characteristic to the muscle function required for
acceleration sprinting, for example, relatively longer contact times.
The only large and statistically significant correlation with maximum speed was
for peak power/weight (Table 3). Since this variable occurs earlier in the CMJ (that
is before the instant of takeoff), this may reflect faster explosive force production
(compared with jump height), which would be expected to be more similar to the
muscle function needed for maximum speed sprinting with relatively short contact
times. Although this notion is speculative, it is consistent with previous research1 that
has related the duration of explosive force measures to contact times in sprinting.
Although the negative correlation between peak power/weight and maximum
speed was large (r = –0.649), the corresponding relationship with peak power was
only small (r = –0.179). This suggests that peak power relative to body mass is
more important than absolute peak power for maximum speed, and similar findings
have also been shown by others.2,7,9

Practical Applications and Conclusions

A practical consideration for those interested in implementing CMJ tests is the
equipment that is needed to conduct a meaningful assessment. When jump height
or peak velocity are good predictors of sprinting performance, a position trans-
ducer could be used alone. In the present study, jump height produced a moderate
correlation with 10 m time, but a lower nonsignificant correlation with maximum
speed. Other studies have found relatively high statistically significant correlations
between unloaded CMJ height and measures of speed.1,2,9
Peak power/weight appears to be a useful variable to represent CMJ perfor-
mance, especially as it relates to maximum speed. Although power output can be
measured with the use of a position transducer alone, this practice is not recom-
mended due to inaccuracies arising from data manipulation.5,8 Therefore, if peak
power/weight is to be recorded, an integrated system involving a force platform to
directly measure force and a position transducer to record velocity would be recom-
mended. The user would need to weigh up the benefit obtained by using this vari-
able of CMJ performance against the cost and added complexity of this procedure.
The results of this study tend to suggest using peak power/weight from a CMJ as
it relates to both acceleration and maximum speed. Likewise, peak velocity and peak
power/weight have common characteristics to maximum speed sprinting. However,
using all these variables in CMJ assessment may create more confusion than clar-
ity. Further research is needed to determine if the relationships described here are
stable and consistent when other samples and sporting populations are evaluated.
For example, would similar relationships exist in the same sample after years of
sprint and power training? If particular CMJ variables can be shown consistently to
correlate to performance, the interpretation of CMJ assessment will be advanced.

References
1. Young W, McLean B, Ardagna J. Relationship between strength qualities and sprinting
performance. J Sports Med Phys Fitness. 1995;35:13–19.
 Jump variables and muscle function 57

2. Cronin JB, Hansen KT. Strength and power predictors of sports speed. J Strength Cond
Res. 2005;19(2):349–357.
3. Sheppard JM, Cormack S, Taylor K, McGuigan MR, Newton RU. Assessing the force-
velocity characteristics of the leg extensors in well-trained athletes: the incremental
load power profile. J Strength Cond Res. 2008;22:1320–1326.
4. McGuigan MR, Cormack S, Newton RU. Long-term power performance of elite Aus-
tralian Rules football players. J Strength Cond Res. 2009;23(1):26–32.
5. Cormie P, McBride JM, McCaulley GO. Validation of power measurement techniques
in dynamic lower body resistance exercises. J Appl Biomech. 2007;23:103–118.
6. Alemany JA, Pandorf CE, Montain SJ, Castellani JW, Tuckow AP, Nindl BC. Reli-
ability assessment of ballistic jump squats and bench throws. J Strength Cond Res.
2005;19(1):33–38.
7. Baker D, Nance S. The relation between running speed and measures of strength and
power in professional rugby league players. J Strength Cond Res. 1999;13(3):230–235.
8. Dugan EL, Doyle TLA, Humphries B, Hasson CJ, Newton RU. Determining the opti-
mal load for jump squats: A review of methods and calculations. J Strength Cond Res.
2004;18(3):668–674.
9. Hori N, Newton RU, Andrews WA, Kawamori N, McGuigan MR, Nosaka K. Does
performance of hang power clean differentiate performance of jumping, sprinting, and
changing of direction? J Strength Cond Res. 2008;22(2):412–418.
10. Bompa TO. Power Training for Sport. Plyometrics for Maximum Power Development.
Gloucester, Ontario: Coaching Association of Canada; 1994.
11. Little T, Williams AG. Specificity of acceleration, maximum speed, and agility in
professional soccer players. J Strength Cond Res. 2005;19(1):76–78.
12. Young W, Russell A, Burge P, Clarke A, Cormack S, Stewart G. The use of sprint tests
for assessment of speed qualities of elite Australian Rules footballers. Int J Sports
Physiol Perform. 2008;3:199–206.
13. Gabbett T, Kelly J, Ralph S, Driscoll D. Physiological and anthropometric characteristics
of junior elite and sub-elite rugby league players, with special reference to starters and
non-starters. J Sci Med Sport. 2009;12:215–222.
14. Hopkins W. A Scale of Magnitudes for Effect Statistics. A New View of Statistics. Avail-
able at: http://www.sportsci.org/resource/stats/effectmag.html. Accessed 14 December
2009.
15. Thomas JR, Nelson JK. Research Methods in Physical Activity. Champaign, Ill: Human
Kinetics; 1990.