Scand J Med Sci Sports 2011: 21: 79–84 & 2009 John Wiley & Sons A/S

doi: 10.1111/j.1600-0838.2009.01024.x

The influence of speed of approach and accuracy constraint on the

maximal speed of the ball in soccer kicking
T. B. Andersen, H. C. Dörge
Department of Sport Science, University of Aarhus, Aarhus, Denmark
Corresponding author: Thomas Bull Andersen, MSc, PhD, Department of Sport Science, University of Aarhus, Dalgas Avenue
4, Dk-8000 Aarhus C, Denmark. Tel: 145 8942 9150, E-mail: tbull@[Link]
Accepted for publication 24 June 2009

The kicking skill of the individual player is of great the subjects, (III) with a straight-line approach of different
importance in soccer, and two parameters can be identified velocities ranging from 0 m/s to maximal running speed.
as being important in soccer kicking – achieving high speed The maximal speed of the ball was between 28.60 and
of the ball and accuracy of performance. The purpose of the 34.48 m/s. An accuracy constraint caused the speed of the
current study was to examine the influence of different ball to decline to 85%. With a straight-line approach, the
speeds of approach on the maximal speed of the ball when maximal speed of the ball was between 25.64 and 32.26 m/s.
kicking a stationary ball, and to examine the influence of When the subjects approached the ball at speeds other than
accuracy of performance on the maximal speed of the ball the self-selected speed of approach, the speed of the ball
relevant to a penalty kick. Seven male soccer players kicked declined, indicating a subject-specific optimal speed of
for maximal speed of the ball (I) with a self-selected angle approach.
and speed of approach, (II) with accuracy demand placed on

The most widely studied kicking motion is the soccer tion acting on that segment or link but also by the
instep kick to a stationary ball when kicking for motion of adjacent, linked segments. Therefore, the
maximal velocity of the ball (see reviews by Kellis & control and regulation of fast kicking movements
Katis, 2007 and Lees & Nolan, 1998). The velocity of must take this segmental interaction into account
the ball depends on the mass of the ball, the initial (Dörge et al., 1999).
velocity of the ball, the velocity of the kicking foot, Because the segments of the kicking leg are influ-
the effective striking mass of the leg and the coeffi- enced by the transfer of the speed of approach
cient of restitution, which is a measure of how well through the pelvis when the support leg is placed
the velocity of the kicking foot is transferred to the beside the ball, the speed of approach might also
ball (Bull Andersen et al., 1999). The speed of the influence the control and regulation of movement
foot before impact is most important for achieving and thereby the maximal speed of the ball. The speed
the highest ball speed (Bull Andersen et al., 1999). of approach is normally implicit in the inverse
In order to maximize the speed of the kicking foot dynamics equations (Winter, 1990) or the motion-
at the time of impact with the ball, the soccer place dependent equations (Putnam, 1991) used to analyze
kick is executed using a proximal-to-distal sequen- the kicking action, but the relationship between the
cing of the segments in the lower extremity (Putnam, speed of approach and the release speed of the ball
1983; Dunn & Putnam, 1988; Anderson & Sidaway, has not yet been studied and presented explicitly.
1994). Previously, the intersegmental kinematics and Knowledge of the influence of speed of approach
kinetics during the execution of the maximal velocity on the maximal speed of the ball is needed (I) in
instep kick in soccer have been studied (Zernicke & order to evaluate results from laboratory studies
Roberts, 1978; Luhtanen, 1988; Dörge et al., 1999) as involving standardization of the speed and (II) in
well as the impact between the kicking foot and the real game situations in order to verify whether an
ball (Plagenhoef, 1971; Asami et al., 1976; Bull instant change of normal selected speed of approach
Andersen et al., 1999). For a mechanical analysis, can result in a higher speed of the ball using the
the kicking leg is often modelled as a linked-segment instep place kick to a stationary ball with a running
system. As such, the motion of a given segment in the approach.
lower extremity is influenced not only by the forces It has been shown that a running approach results
generated by the muscles and gravitational accelera- in a higher speed of the ball than a stationary kick

79
Andersen and Dörge
(Opavsky, 1988), but this study did not report the Materials and methods
speed of approach of the kicker nor did it reveal the Pilot investigation
influence of different speeds of approach on the The use of double-beam photocells to record running speed is
maximal speed of the ball. a common method in real-life situations (athletics). Double-
A simulation study (Sørensen et al., 1999) showed beam photocells have also been used in laboratory experi-
that the speed of approach and take-off angle are ments, and the accuracy of such systems has been evaluated
using center of mass speed estimates from three-dimensional
important in long jump, but an increase of the (3D) video analysis as criteria and has been reported to have
preferred speed of approach will not instantly lead an error of speed of 0.1 m/s with a distance of 1.6 m between
to a longer jump, indicating an optimal subject- two sets of photocells (Yeadon et al., 1999).
specific speed of approach related to the current When performing an instep kick to the ball in soccer, the
anthropometry and performance profile of the sub- arms of the kicker are known to perform vigorous movements
during the approach, and placing the double beam in height of
ject. We hypothesize that a similar optimal subject- the estimated center of mass during the approach is therefore
specific speed of approach is present in soccer kick- not an optimal solution.
ing. Furthermore, it has been suggested that an A study of five subjects performing a maximal velocity
angled approach is superior to a straight-line ap- instep soccer place kick with a straight-line approach was
proach (Isokawa & Lees, 1988). This study, however, performed. A high-speed camera (400 f.p.s.) placed 10 m
perpendicular to the line of action delivered film for analysis.
only showed a trend; no significant differences were This analysis revealed that the head did not move significantly
found. An angled approach allows the kicker to in the horizontal direction in relation to the estimated center
extend the leg further without touching the ground of mass on the subjects during the approach.
with the kicking foot. This could enhance the possi- On the basis of this information, we concluded that it was
bility of transferring linear velocity to the foot. optimal to place the double-beam photocells at the head level
for running speed measurement during the study.
In real-life kicking actions on the soccer field, the
player is met with a speed as well as an accuracy
demand when kicking the ball. The speed–accuracy Participants
trade-off in fast movements has been intensively
Seven skilled male soccer players ranging in age from 20 to 33
investigated in relation to simple upper extremity years served as subjects in this study. The subjects represented
movements (e.g. Plamondon & Alimi, 1997; Etnyre, the Danish fourth to sixth league and they all trained three
1998) but in relation to complex lower extremity times per week and on average played one match per week.
movements the literature is limited. Dörge et al. Their mean height was 1.82 m (range 1.74–1.88 m) and their
(1999) have shown that when kicking for maximal mass was 78.3 kg (64–96 kg). During the experiments, the
subjects wore indoor soccer shoes without cleats. All subjects
speed of the ball, experienced players have no pro- gave their informed consent to participate and the local ethics
blem hitting the ball exactly when the linear velocity committee approved the study.
of the foot is at its maximal, indicating a remarkable
intersegmental coordination accuracy during fast
kicking movements, but often the players are met Design, apparatus and data analysis
with another source of constraint with the demand of The experimental setup consisted of a size 5 FIFA-approved,
hitting an external target with the ball, while main- standard inflated ball (Select Brillant Super, Select Sport A/S,
taining high speed of the ball, e.g. a long pass to a Copenhagen, Denmark) placed on the wooden floor of an
player or kicking for a goal. indoor gym. The mass of the ball was 0.445 kg and it was
inflated to 1 atm.
Teixeira (1999) examined the kinematics of kicking One meter behind the ball, a set of two ALGE double-beam
as a function of different sources of constraint on photocells was placed 1.6 m apart connected to an ALGE S4
accuracy. The subjects were required to powerfully receiver measuring the subject speed of the approach at the
kick balls of two sizes under conditions with defined head level (ALGE-timing GmBH, Lustenau, Austria). 0.5 m in
and undefined targets. Analysis indicated that the front of the ball, another set of ALGE double-beam photocells
(ALGE-timing BallSpeed System, 0–40 cm high photocells,
defined target condition led to lower movement 1 m apart) was placed and connected to a Comet receiver
speeds. A similar finding has been presented by Asami measuring the speed of the ball. The photocells measured the
et al. (1976), who found that the speed of the ball speed of the ball by dividing the distance between the photo-
declined to 80% of the maximal speed when external cells (1 m) by the time elapsed between penetration of the light
accuracy demands where introduced to the subjects. beams. The clocks were sampled at 1000 Hz, which only
allowed the apparatus to measure discrete values, e.g.
Accordingly, the purpose of this study was (I) to 32.26 m/s, followed by 33.33 m/s (as can be deduced from
examine the influence of different speeds of approach the results presented, Table 1). Accordingly, at the maximal
on the maximal speed of the ball when kicking a kicking velocities, the maximal possible error was approxi-
stationary ball and (II) to examine the influence of an mately 3.2%. At lower velocities, the error was lower.
accuracy constraint, relevant to the penalty kick, on Eleven meters in front of the ball, a large mat (5  4 m) was
placed to absorb the momentum of the ball after the kick. Only
the maximal speed of the ball. Furthermore, (III) the balls hitting the mat were analyzed, securing the path of the ball
influence of a limitation on the angle of approach as an extension of the line of action of the kicker and thereby
was investigated. ensuring the validity of the ball speed measurement. A kick

80
 Speed of approach
Table 1. Measured and calculated data for all the subjects

Subject 1 2 3 4 5 6 7

Maximal running speed (after 19 m) (m/s) 7.65 7.56 7.72 8.39 8.23 7.59 8.48
Optimal speed of approach* (when kicking for 5.94 4.19 4.41 4.39 4.68 4.63 4.50
maximal speed of the ball) (m/s)
Optimal speed of approach in % of maximal running speed (%) 77.65 55.50 57.12 52.35 56.85 60.99 53.03
Maximal speed of the ball (standing position) (m/s) 20.41 20.41 23.26 23.26 20.41 22.22 21.74
Maximal speed of the ball (straight approach) (m/s) 27.78 25.64 32.26 31.25 26.31 32.26 27.78
Maximal speed of the ball (angled approach) (m/s) 28.60 29.41 34.48 31.25 29.41 33.33 29.41
Maximal speed of the ball (standing) in % of 73.46 79.59 72.1 74.42 77.55 68.89 78.26
maximal speed of the ball (straight) (%)
Maximal speed of the ball (straight) in % of 97.23 87.17 93.55 100.00 89.47 96.77 94.45
maximal speed of the ball (angled) (m/s)
Maximal speed of the ball (accuracy) in % of 82.72 79.75 88.91 81.36 92.74 86.74 91.93
maximal speed of the ball (angled) (%)

*This also proved to be the preferred speed of approach in all subjects.

hitting close to the edge of the mat would cause the ball speed 11 times (range 8–15) to cover the selected approach velocities
to be measured 2.5% too low, because the speed measurement within  5%. One kick from each subject was represented in
was made only in the sagital direction. Hence, no correction of each speed category.
the direction was made. Furthermore, it was visually observed The speeds of the ball in each group were compared using a
during the experiments that no valid kicks hit the mat close to Friedman analysis of variance on ranks test, with significant
its edges but no objective measurements were made. differences indicated by P-values of o0.05. Furthermore, to
A 1.5  1.5 m target was drawn on the middle of the mat investigate to what extent the ball speed is a sum of the
starting 20 cm above the floor level. It was intended that the approach velocity and the ball speed in the standing trial
target should model a successful penalty kick, a kick in one (0%), the ball speed in the standing trial was added to the
side of the goal with sufficient ball speed to prevent the different approach velocities and graphed in the same graph as
goalkeeper from interfering with the path of the ball before the other data (Fig. 1).
the ball was inside the goal.
After warmup and several kicking trials, the experiment
consisted of three different procedures.
Results
(I) The subjects were told to kick for maximal speed of the
ball, with a self-selected angle and speed of approach. Maximal speeds of the ball between 28.60 and
The subjects kicked three times and the maximal speed 34.48 m/s were obtained, when the subjects kicked
of the ball was used in the analysis. with a self-selected angle and speed of approach with
(II) The subjects were told to kick for accuracy (hitting the
no constraints on accuracy. The intrasubject stan-
target) but attaining the highest possible speed of the ball.
The subjects approached the ball with a self-selected dard deviation (SDintra) was between 0.46 and
angle and speed. The individual subject continued to 3.41 m/s and no failed kicks were observed.
kick until three subsequent kicks hit the target. The maxi- Accuracy constraint corresponding to a penalty
mal value of the speed of the ball in the three subsequent kick caused the speed of the ball to decline to
kicks was registered for each subject. A paired t-test with
between 79.75% and 92.74% of the maximum ball
a significance level at P 5 0.05 was used to compare these
results with the maximal kicking velocity. speed (Po0.001, SDintra 5 0.00–2.09%).
(IIIa) The subjects were asked to kick for maximal speed of When the approach was limited to a straight-line
the ball using a straight-line approach with a self- approach, the maximal speed of the ball was between
selected length and speed of approach. 25.64 and 32.26 m/s (SDintra 5 0.78–2.40 m s 1, no
(IIIb) The subjects were then asked to run as fast as possible
failed kicks). The maximal speed of the ball during
in a straight line from 19 m behind the ball, and the
maximum possible speed of approach just before the the straight-line approach was obtained during the
position of the ball was measured. self-selected speed of approach in all subjects. The
(IIIc) The subjects were then told to perform a series of kicks SD values for the running velocities (measured in
for maximal speed of the ball with a straight-line percent of the optimal running speed) varied between
approach of not o8 m. The subjects were told to
5.25% and 12.23%. See Table 1 for a survey of the
approach the ball with velocities corresponding to 0%
(standing, without moving the support foot), 25%, subject data.
50%, 75%, 125% and 150% of their self-selected When the subjects were forced to approach the ball
approach velocity (100%). The order of the approach in a straight line at speeds other than the self-selected
velocities was randomized. Before each kick, the subject speed of approach, the speed of the ball declined.
had one or more attempts to aim at the target speed;
With no approach (standing), the ball speed was
these attempts were made without kicking the ball.
21.67 m/s (SD 5 1.30 m/s), a speed of approach of
The speed of approach and the release speed of the ball 25% of the optimal subject-specific speed resulted in
were registered for all kicks. On average, each subject kicked a speed of the ball of 21.04 m/s (SD 5 2.5 m/s), a

81
Andersen and Dörge
34 analysis or to measure speed with the use of photo-
32 cells.
30 Our results reveal that the maximal speed of the
28 ball declines to approximately 95% with a straight-
26 line approach compared with an angled approach
m s–1

24 with a self-selected speed of approach. This is prob-

22 ably due to the reduced rotation of the pelvic girdle
20
before the initial stage of the forward motion of the
18
thigh, and the inability of the lower extremity to be
Ball speed
further straightened just before impact in the
16 Sum of approach speed and ball
speed (standing) straight-line approach, resulting in a decrease in the
14
0 20 40 60 80 100 120 140 160 speed of the foot before impact compared with the
Approach speed (% of optimal approach speed) angled approach.
The approach in an instep soccer place kick is used
Fig. 1. The relation between the speed of approach and the
to accelerate the body to a selected velocity, to place
speed of the ball normalized to each subject’s optimal speed
of approach, at which the maximal speed of the ball was the support foot at a desired angle and distance from
obtained. The dotted line indicates the sum of the approach the ball and to transfer the velocity of the approach
speed and the ball speed when kicking from a standing to the kicking leg. Because the maximal speed of the
position. The vertical bars show the standard deviation. ball mainly is dependent on the speed of the foot, it
could be speculated that the faster the straight-line
approach, the higher the maximal speed of the ball.
speed of approach of 50% resulted in a speed of the But our results prove the existence of an optimal
ball of 24.05 m/s (SD 5 3.12 m/s) and at the speed of submaximal subject-specific speed of approach,
approach of 75%, the speed of the ball reached where the maximal speed of the ball can be obtained.
26.48 m/s (SD 5 2.94 m/s). When the subjects in- This optimal speed of approach is equal to the
creased their speed of approach to 125% and 150% normally self-selected speed of approach. An instant
of the optimal speed of approach, the speed of the increase or a decrease in the speed of approach,
ball declined to 25.44 m/s (SD 5 2.52 m/s) and compared with the optimal speed of approach, will
19.74 m/s (SD 5 3.85 m/s), respectively. All ball lead to a decrease in the speed of the ball. If this
speeds were significantly different from (and lower optimal speed of approach is a result of training or
than) the maximal speed of the ball (Po0.001). For whether training could influence it is beyond the
one subject, two failed kicks were observed at speeds scope of this study.
of approach of 125% and 150%. The speed of approach is dependent on the step
When using a speed of approach of 50% of the length and the step frequency of the kicker (Hay,
optimal speed of approach, the speed of the ball was 1993). A faster speed of approach due to increased
equal to the sum of the speed of approach and the step length can result in an increased joint angle at the
speed of the ball, kicking from a standing position. hip joint of the kicking leg during placement of the
When using a speed of approach of 75% and 100%, support leg beside the ball. It can be speculated that
the speed of the ball was higher than the sum. When an increased joint angle will cause additional length-
using a speed of approach of 125% and 150%, the ening of the muscles about the hip joint and thereby
speed of the ball was lower than the sum (Fig. 1). enable the hip muscles of the kicking leg to perform
more work. Because the kicking action can be
Discussion described as an open-chain movement consisting of
a series of stretch-shortening-cycle muscle actions
Many different types of kicks are utilized in various about the hip and knee joint (Dörge et al., 1999),
games and sports. In some types of kicks, the kicking there is a potential for utilizing energy from the
action is limited to a one-plane action were the line of parallel- and serial-elastic components of the mus-
approach is directly in line with the intended flight of cle–tendon complexes about these joints during the
the ball like in the punt kick and in the soccer toe kicking motion as well (Fenn, 1924; Asmussen &
kick. In others, the kicking action is clearly 3D, with Bonde-Petersen, 1974; Cavagna, 1977). A faster
a considerable rotation of the pelvic girdle and an speed of approach due to an increase in step fre-
angled approach to the ball-like in the soccer place quency can result in a faster coupling time between
kick. the eccentric–concentric muscle action about the hip,
In the laboratory, it is often necessary to limit the and thereby facilitate the utilization of elastic energy
approach and the intersegmental kicking action of a (Cavagna, 1977).
normally 3D movement skill to a 2D performance, Another possible explanation as to why the ap-
e.g. in order to perform affordable high-speed video proach velocity influences the ball velocity is that the

82
 Speed of approach
placement of the support foot will cause the pelvis to external target accuracy relevant to a penalty kick in
accelerate both posteriorly and vertically. This will soccer. This result is in agreement with the findings of
affect the movement of the whole leg through the hip Asami and Nolte (1983), who reported a decrease to
joint. At low approach velocities, this effect could be 80%.
very limited and at high approach velocities (higher The decrease in the maximal speed of the ball
than the optimal velocity), it is possible that limita- observed when hitting for accuracy could be related
tions in the maximal muscle force of the support leg to the control and regulation of the intersegmental
as well as limited time to perform the muscle actions movement of the lower extremity and optimal place-
of the support leg will influence the movement of the ment of the kicking foot during impact to ensure the
kicking leg negatively. control of the path of the ball after impact. The
The results from this study show that the speed of influence of accuracy constraints on the maximal
the ball obtained around 50–60% of the optimal speed of the ball in the future needs to be related to
speed of approach is approximately as much as the other relevant situations in soccer, e.g. passing and
sum of the approach speed and the speed of the ball heading.
obtained during kicking from a standing position We have shown the existence of a subject-specific
(approach speed 5 0 m/s). optimal speed of approach in relation to the genera-
When the speed of approach is 75% and 100% of tion of maximal speed of the ball during the soccer
the optimal subject-specific speed of approach, the instep place kick to a stationary ball with a straight-
speed of the ball is faster than the sum of the line approach.
approach speed and the speed of the ball obtained Furthermore, we observed that when an accuracy
during kicking from a standing position. This indi- demand relevant to the soccer penalty kick is placed
cates that additional work has been performed on the on the player, there is a reduction in the maximal
foot of the kicking leg; however, only end-point velocity of the ball to about 85%, compared with the
variables were measured, and so a specific investiga- maximal velocity at a self-selected approach angle
tion on how this phenomenon occurs is not possible and length.
with this study. Additionally, we confirmed that an angled ap-
Using a speed of approach of 125% or 150% will proach should be used if the aim is to kick a ball
result in a speed of the ball below the sum of the with maximal speed.
approach speed and the speed of the ball obtained
during kicking from a standing position, and a speed
of the ball less than could be obtained at the 100%
optimal speed of approach. We speculate that this Perspectives
could be due to (I) the difficulty of optimal placing of
the support foot to secure the optimal path of the The results from the present study can be used in
kicking foot, (II) an inability of the support leg to soccer matches by a defending player. In a situation
withstand the high momentum in the support phase, where the defender is chasing an attacking player
(III) the influence of the approach velocity on the who is in possession of the ball, the defender has to
intersegmental control and regulation of the kicking decide whether or not to make a tackle, knowing that
leg and (IV) an inability to perform the optimal there is a possibility of not hitting the ball. In some
kicking motion during the time available before the cases – as the results reveal – it can be more optimal
whole body would have moved too far horizontally. just to force the attacker to run at a very high speed.
In real-life kicking actions, the player is often met In this case, the ball velocity of any shot will be much
with a speed as well as an accuracy demand when lower.
kicking the ball. Our results reveal a decrease to 85%
of the maximal speed of the ball when kicking for Key words: soccer, kicking, biomechanics.

References
Anderson D, Sidaway B. Coordination Asami T, Togari H, Kikichi T. Energy Cavagna G. Storage and utilization of
changes associated with practice of a efficiency of ball kicking. In: Komi PV, elastic energy in skeletal muscle. Exerc
soccer kick. Res Q Exerc Sport 1994: ed. Biomechanics V-B. Baltimore: Sport Sci Rev 1977: 5: 89–129.
65: 93–99. University Park Press, 1976: 135–140. Dörge H, Bull Andersen T, Sørensen H,
Asami T, Nolte V. Analysis od power- Asmussen E, Bonde-Petersen F. Storage Simonsen E, Aagaard H, Dyhre-
ful ball kicking. In: Matsui H, of elastic energy in man. Acta Poulsen P, Klausen K. EMG activity
Kobayashi K, eds. Biomechanics Physiolog Scand 1974: 91: 385–392. of the iliopsoas muscle and leg
VIII-B. Champaign, IL: Bull Andersen T, Dörge H, Thomsen F. kinetics during the soccer place kick.
Human Kinetics, 1983: Collisions in soccer kicking. Sports Eng Scand J Med Sci Sports 1999: 9:
695–700. 1999: 2: 121–125. 195–200.

83
Andersen and Dörge
Dunn E, Putnam C. The influence of Lees A, Nolan L. The biomechanics of Putnam CA. A segment interaction
lower leg motion on thigh deceleration soccer: a review. J Sports Sci 1998: 16: analysis of proximal-to-distal
in kicking. In: de Groot G, Hollander 211–234. sequential segment motion patterns.
AP, Huijing PA, van Ingen Schenau Luhtanen P. Kinematics and kinetics of Med Sci Sports Exerc 1991: 23:
GJ, eds. Biomechanics XI-B. maximal instep kicking in junior soccer 130–144.
Amsterdam: Free University Press, players. In: Reilly T, Lees A, Davids K, Sørensen H, Simonsen E, van den Bogert
1988: 787–790. Murphy W, eds. Science and football. A. Influence of approach velocity
Etnyre B. Accuracy characteristics of London: E & FN Spoon, 1988: on long jump performance. In:
throwing as a result of maximum force 441–448. Herzog W, Jinha A, eds. International
effort. Perceptual and Motor Skills Opavsky P. An investigation of linear society of biomechanics XVIIth
1998: 86: 1211–1217. and angular kinematics of the leg congress – book of abstracts. Canada:
Fenn W. The relation between the in two types of soccer kick. In: University of Calgary, 1999: 166.
work performed and the energy Reilly T, Lees A, Davids K, Teixeira L. Kinematics of kicking as a
liberated in muscular contraction. Murphy W, eds. Science and football. function of different sources of
J Physiol (London) 1924: 58: 373–395. London: E & FN Spoon, 1988: constraint on accuracy. Percept Mot
Hay J. The biomechanics of sports 460–467. Skills 1999: 88: 785–789.
techniques, 4th edn. Englewood Cliffs, Plagenhoef S. The patterns of human Winter DA. Biomechanics and motor
NJ: Prentice Hall, 1993. motion. Englewood Cliffs, NJ: Prentice control of human movement. New
Isokawa M, Lees A. A biomechanical Hall, 1971. York; Wiley, 1990.
analysis of the instep kick motion in Plamondon R, Alimi A. Speed/Accuracy Yeadon M, Kato T, Kerwin D.
soccer. In: Reilly T, Lees A, Davids K, trade-offs in target-directed Measuring running speed using
Murphy W, eds. Science and football. movements. Behav Brain Sci 1997: 20: photocells. J Sports Sci 1999: 17:
London: E & FN Spoon, 1988: 279–303. 249–257.
449–455. Putnam C. Interaction between segments Zernicke R, Roberts E. Lower extremity
Kellis E, Katis A. Biomechanical during a kicking motion. In: Matsui H, forces and torques during systematic
characteristics and determinants of Kobayashi K, eds. Biomechanics VIII- variation of non weight bearing
instep soccer kick. J Sports Sci Med B. Champaign, IL: Human Kinetics, motion. Med Sci Sports Exerc 1978:
2007: 6: 154–165. 1983: 688–694. 10: 21–26.

84