This article was downloaded by: [University of Massachusetts, Amherst]On: 08 October 2012, At: 01:43Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Sports SciencesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/rjsp20

Anticipatory strategies of team-handball goalkeepersMarcos Gutierrez-Davila a , F. Javier Rojas a , Manuel Ortega b , Jose Campos c & JuanParraga da Faculty of Sports Sciences, University of Granada, Granada, Spainb Faculty of Education Sciences, University of Seville, Seville, Spainc Faculty of Sciences of Physical Activity and Sports, University of Valencia, Valencia, Spaind Faculty of Humanities and Education Sciences, University of Jaen, Jaen, Spain

Version of record first published: 13 Jul 2011.

To cite this article: Marcos Gutierrez-Davila, F. Javier Rojas, Manuel Ortega, Jose Campos & Juan Parraga (2011):Anticipatory strategies of team-handball goalkeepers, Journal of Sports Sciences, 29:12, 1321-1328

To link to this article: http://dx.doi.org/10.1080/02640414.2011.591421

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form toanyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses shouldbe 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 inconnection with or arising out of the use of this material.

Anticipatory strategies of team-handball goalkeepers

MARCOS GUTIERREZ-DAVILA1, F. JAVIER ROJAS1, MANUEL ORTEGA2,

JOSE CAMPOS3, & JUAN PARRAGA4

1Faculty of Sports Sciences, University of Granada, Granada, Spain, 2Faculty of Education Sciences, University of Seville,

Seville, Spain, 3Faculty of Sciences of Physical Activity and Sports, University of Valencia, Valencia, Spain, and 4Faculty of

Humanities and Education Sciences, University of Jaen, Jaen, Spain

(Accepted 23 May 2011)

AbstractThis study seeks to discover whether handball goalkeepers employ a general anticipatory strategy when facing long distancethrows and the effect of uncertainty on these strategies. Seven goalkeepers and four throwers took part. We used a forceplatform to analyse the goalkeeper’s movements on the basis of reaction forces and two video cameras synchronised at500 Hz to film the throw using 3D video techniques. The goalkeepers initiated their movement towards the side of the throw193+ 67 ms before the release of the ball and when the uncertainty was reduced the time increased to 349+ 71 ms. Thekinematics analysis of their centre of mass indicated that there was an anticipatory strategy of movement with certainmodifications when there was greater uncertainty. All the average scores referring to velocity and lateral movement of thegoalkeeper’s centre of mass are significantly greater than those recorded for the experimental situation with biggeruncertainty. The methodology used has enabled us to tackle the study of anticipation from an analysis of the movement usedby goalkeepers to save the ball.

Keywords: Anticipation, force platform, 3D kinematics, uncertainty

Introduction

The ability to predict future events based on move-

ments of other players is one of the most relevant

perceptual abilities for performance of sporting

activities. This is particularly relevant for those

activities that require a rapid and accurate response

(Abernethy & Russell, 1987; Abernethy & Zawi,

2007). Several research studies have sought to

identify the clues related to anticipation in various

sports where an object moving at high speeds has to

be intercepted, such as tennis (Huys et al., 2009),

baseball (Ranganathan & Carlton, 2007) and bad-

minton (Abernethy & Zawi, 2007). In general, it has

been shown that players are able to anticipate the

direction and intercept the object successfully on the

basis of the information given by the opponent

moments before the release of the ball (Lidor, Argov,

& Daniel, 1998; Williams, Davids & Williams 1999).

The goalkeepers use positional strategies in team

sports such as soccer or team-handball to reduce

uncertainty about the ball’s direction. On some

occasions, they make their decisions based on the

defensive actions of the field players and, on others,

they use their bodies to close off an area of the goal to

force the throw to the free zone (Rogulj, Papic, &

Srhoj, 2005). This enables them to process informa-

tion more quickly and respond more rapidly and

efficiently (Ranganathan & Carlton, 2007). The

goalkeeper’s strategy is not limited to reducing

uncertainty. Even when he knows the direction of

the throw in advance, he must delay starting his

movement until the point when it would be difficult

for the thrower to modify the trajectory of the ball

during the throw (Savelsbergh, Van der Kamp,

Williams, & Ward, 2005; Schorer, Fath, Baker, &

Jaitner, 2007).

The importance of this has been shown from

deductions drawn from Fradet et al. (2004). They

demonstrated that the team-handball throw does not

behave like a typical structure in its temporal

segmented sequence, from the proximal to the distal

(P-D), as occurs with other throws where the aim is

to achieve high velocity. One argument to explain

this finding is that the thrower can deceive the

goalkeeper, and change the direction of the throw at

the last moment. This confirms the importance of

the strategies used by both goalkeepers and throwers.

Correspondence: F. J. Rojas, University of Granada, Faculty of Sports Sciences, Granada, Spain. E-mail: fjrojas@ugr.es

Journal of Sports Sciences, September 2011; 29(12): 1321–1328

ISSN 0264-0414 print/ISSN 1466-447X online � 2011 Taylor & Francis

DOI: 10.1080/02640414.2011.591421

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012

Moreover, goalkeepers need to use a premeditated

strategy to intercept the ball, selecting the clues

perceived in the action of each environmental

situation. Savelsbergh, Williams, Van der Kamp,

and Ward (2002) have shown that goalkeepers pick

up selected information during the throw and it is

more difficult to outguess the height of any attempt

rather than its lateral placement in penalty kicks in

soccer. Bideau et al. (2004) and Vignais et al. (2009)

used virtual reality technology to record the effects

that small changes in the thrower’s movements

produce on the goalkeepers, as well as the impor-

tance of arm movements. They suggested that,

technical execution, in addition to the existence of

a strategy, is another relevant factor in the goal-

keeper’s efficiency.

The aim of this work is to verify the possible

existence of a general strategy by handball goal-

keepers by using a methodology that measures

anticipation in a situation where the functional

coupling between perception and action is preserved.

We sought to test the effect of uncertainty in

anticipatory movements of team-handball goal-

keepers. Based on their positional strategies used in

the games, two levels of uncertainty were proposed;

throwing only to the right of the thrower, to the upper

or lower corners of the goal, two possibilities of

direction (TM2) or throwing towards one of the four

corners of the goal, four possibilities of direction

(TM4), Figure 1. We predicted that the goalkeepers

will maintain an anticipatory strategy, and they will

reduce anticipation time when uncertainty increases,

as this increases the difficulty of guessing the

direction of the throw.

Method

Participants

Eleven males participated in this study, seven

team-handball goalkeepers and four throwers. The

team-handball goalkeepers (age¼ 28+ 5 years;

body height¼ 1.86+ 0.03 m; body mass¼ 89.79+9.93 kg), due to their age, had a total experience of

Figure 1. Representation of the positions of the recording systems used.

1322 M. Gutierrez-Davila et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012

20+ 5 years playing handball, and all of them have

played in first division or higher Spanish League,

three of them having played with the national

handball team.

Four team-handball players, who were specialists

in shooting and belonged to first division teams in the

Spanish League (age¼ 24+ 1 years; height¼ 1.86+0.05 m; body mass¼ 86.36+ 6.13 kg) acted as

throwers. The study was approved by the local ethics

committee and all participants signed informed

consent.

Design

An intragroup design for repeated measures was

used with the purpose of checking the effect

produced by two situations differentiated by the

level of uncertainty (TM2 and TM4, Figure 1) on

the factors that determine the actions made by team-

handball goalkeepers, facing throws of more than

nine metres.

In situation TM2 the thrower made the perpendi-

cular throw on goal with only two possibilities for the

direction of the throw: the upper and lower corners

of the goal on the same side as the throwing arm. In

situation TM4 the thrower could throw at each of the

four corners of the goal (Figure 1).

Apparatus and data analysis

The throws were made 10 m from the goal in a zone

previously delimited by a reference system of 2.32 x

1.58 x 2 m. We used a force platform 0.8 x 0.8 m

(Dinascan/IBV Valencia, Spain), situated in line with

the centre of the goal and one metre in front of the

shooting zone. The throws were filmed using two

high-speed digital video cameras, Redlake Motion-

Scope PCI 1000S (San Diego, CA), at a frequency of

500 Hz, situated on the thrower’s dominant side at

25 m from the geometric centre of the shooting zone

and 30 m apart. This same frequency was used to

record the reaction force coming from the force

platform. To synchronise the two cameras and the

force platform, an electronic signal was used to

activate the start (Figure 1).

The accuracy of the throw was quantified using a

third digital video camera, Canon mv730 i, at a

frequency of 25 Hz, placed 15 m behind the goal and

perpendicular to it (Figure 1). Starting from the

image where the goalkeeper intercepted the ball or it

reached its target, we determined the distance

between the centre of the ball and the corner of the

goal or base of the post, calculating it with a 2D

reference system linked to the goal.

The three-dimensional coordinates of five body

points of the thrower (point of the left foot; centre of

the articulations of the hip, shoulder, elbow and

wrist) plus the point corresponding to the geometric

centre of the ball were determined for the throws

chosen for analysis. The calculations were done in

three phases: (a) the position of the six points

(landmarks) were digitalised from the image received

from the two high-speed video cameras, at a

frequency of 125 Hz, (b) the method of direct linear

transformation was used (Abdel-Aziz & Karara,

1971) to obtain the three-dimensional coordinates

and (c) Quintic spline functions were applied to the

spatial coordinates obtained (Wood & Jennings,

1979) to smooth and interpolate the spatial coordi-

nates at the same frequency as the force platform

(500 Hz) with the objective to synchronise kinetics

and kinematics data.

In order to be able to compare the positions of the

three-dimensional coordinates of the landmarks, we

made a transformation for each thrower (landmarks)

with respect to a reference system whose origin was

the coordinates of the point of the foot when the sole

was planted firmly on the ground, and the axis

coincided with the reference system represented in

figure 1.

Instructions and procedure

The four field players were instructed to throw after a

run 10 m from the goal, with the sole of the front

foot firmly on the ground, seeking to obtain

maximum velocity when releasing the ball and

adjusting the throw to the corners of the goal, but

taking into account the conditions applying to each

experimental situation (TM2 and TM4). Despite the

restrictions of the experimental situation, we tried to

reproduce the real situation. Thus, the throwers were

told that they could make their usual moves before

throwing, as well as changing direction during the

throw if they considered it opportune. The throws

were made in blocks of five, alternating the order of

the throwers for each goalkeeper and situation (TM2

and TM4). Throws were considered valid where the

thrower threw the ball at the goal, including the posts

and the ground delimiting it.

We instructed all goalkeepers to situate themselves

in their habitual position on a force platform and not

to move prior to the definitive action to save the ball.

After carrying out the usual warm-up, each goal-

keeper had to try to save 10 valid throws in each of

the two proposed situations and before each block of

five throws the goalkeepers knew the experimental

situation TM2 or TM4, so that the uncertainty was

known. The save was only considered valid when the

goalkeeper moved in the direction to save the ball,

marking it as an error when the goalkeeper moved to

the incorrect side or stood still. After recording

twenty valid actions for each goalkeeper (10 in each

situation), we selected for analysis the five most

Anticipatory strategies of team-handball goalkeepers 1323

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012

accurate throws in each experimental situation,

accuracy being determined by the proximity of the

ball to the upper or lower corners of the goal to

determine the differences between the two experi-

mental situations.

Dependent variables

The time of the throw (T(THROW)) was defined as the

period between the end of the thrower’s run up,

considered as the instant when the entire thrower’s

front foot made full contact with the ground and the

instant of the release of the ball from the thrower’s

hand.

The velocity of the ball at release from the

thrower’s hand (Vt(RELEASE)) was recorded. To

determine the instantaneous velocity at the moment

of release, the first derivative from Quintic spline

functions, with zero smoothing, was used.

The accuracy of the throw was determined by the

distance between the centre of the ball and the corner

of the goal or base of the post in the image where the

goalkeeper saved the ball or it reached its target.

The beginning of the goalkeeper’s horizontal and

vertical movement before ball release, T(START-X)

and T(START-Z) respectively, were recorded.

T(START-X) was determined by using the data

recorded from the transversal component of the

reaction force (FX). This was estimated at 0.001 s

(half of the interval) before the instant in which net

force reached a value greater than or equal to 1% of

bodyweight. To eliminate any possible systematic

errors, the baseline was determined by measuring the

mean of the first 100 samples we received from the

platform, where the goalkeeper was motionless

before the thrower initiated his prior run (Gutier-

rez-Davila, Dapena, & Campos, 2006).

Considering this behaviour as general, the instant

of the start of the final movement of the vertical

component of the centre of mass (CM) (beginning of

the acceleration impulse phase), has been considered

as the time where the vertical velocity of the CM

becomes zero (T(START-Z)). When this score does

not reach a value equal to zero, it is considered as

theturning point of the score, the time of the

previous interval where the vertical component of

the velocity of the CM obtains the value closest to

zero (Figure 2).

We recorded the velocity of lateral movement and

distance covered 100 ms before the release of the ball

(VX-100 and eX-100, respectively); velocity of lateral

movement and distance covered at the instant of ball

release (VX-REL and eX-REL, respectively); velocity of

vertical movement and distance covered 100 ms

before ball release (VZ-100 and eZ-100, respectively);

velocity of vertical displacement and distance cov-

ered at the instant of ball release (VZ-REL and eZ-REL,

respectively) and maximum velocity of the vertical

component during the anticipation period(VZ-MAX).

The instant transversal acceleration of the goal-

keeper’s CM (aX) was calculated on the basis of the

respective components of the goalkeeper’s net force

and mass. The transversal velocity (vX) and dis-

placement of the goalkeeper’s CM (eX) were

calculated from the respective functions of accelera-

tion-time, using trapezoidal integration. After nor-

malising the vertical component by eliminating the

body weight of each goalkeeper, this same procedure

was used to determine the vertical velocity (vy).

Statistics

Data were assessed for normality and homogeneity of

variance, and are expressed as mean and standard

deviation (s) for each experimental variable and

situation. We used Software Statgraphics 5.1 of

Statistical Graphics Corporation (STCS, Inc. 2115

East Jefferson Street, Rockville, Maryland, USA) to

treat the data. A one way Anova was used to quantify

the differences between the average scores in the two

experimental situations. The level for acceptance of

significance (a) was set at 0.05.

Results

The goalkeepers moved in the correct direction of

the throw in every case and managed to save the ball

in 91.1+ 9.4% of the throws, in situation TM2. In

situation TM4, the goalkeepers made errors on

17.5+ 7.6% occasions (N¼ 95) and managed to

save the ball on 66.3+ 7.5% of occasions.

Table I sets out the descriptive statistics for the

average velocity of the ball at the instant of release

from the thrower’s hand (Vt(RELEASE)). In general,

the mean velocity obtained in all the throws analysed

(N¼ 129, velocity¼ 24.57+ 1.76 m � s71), was

slightly higher than the values obtained by Fradet

et al. (2004) and Wagner, Buchecker, von Duvillard,

Figure 2. Method used to obtain the start time of the goalkeeper’s

vertical movement. In case (a), the score of the vertical component

of velocity of CM reaches zero values, in case (b) this value is not

obtained.

1324 M. Gutierrez-Davila et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012

and Muller (2010). Table I also shows the descrip-

tive statistics of accuracy and the time of the throw

(T(THROW)) for each thrower. The time taken to

make the throw varied between 183+ 16 ms for

thrower 2 and 237+ 23 ms for thrower 3, the

average of all throws being 206+ 30.3 ms.

Figure 3 shows transversal (FX) and vertical (FZ)

components of a typical example of the goalkeeper’s

reaction force during his movement to save the ball.

This figure also shows the transversal (VX) and

vertical (VZ) velocity of the goalkeeper’s CM with

respect to the time and the position of the thrower at

Table I. Descriptive statistics of the velocity of the ball at the instant of release from the thrower’s hand (Vt(RELEASE)), Accuracy, and the time

of throw (T(THROW)) of each thrower.

THROWER 1

(N¼ 35 )

THROWER 2

(N¼ 33)

THROWER 3

(N¼ 30)

THROWER 4

(N¼ 31)

Vt(RELEASE)) (m � s71)

Accuracy (m)

T(THROW) (ms)

24.39+1.39

0.24+0.12

219+22

23.88+ 2.12

0.23+ 0.13

183+ 16

25.43+ 1.44

0.23+ 0.13

237+ 23

24.71+ 1.70

0.25+ 0.13

184+ 17

Figure 3. A typical example of the transversal component (FX) and vertical component (FZ) of the goalkeeper’s reaction force, the

corresponding velocities (VX and VZ, respectively) and the positions adopted by the thrower at the most significant instants.

Anticipatory strategies of team-handball goalkeepers 1325

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012

the most significant instants. Support is treated as the

instant in which all the thrower’s front foot makes

full contact with the ground.

FX is represented in Figure 3, where initially a

certain increase in the score for the reaction force for

time is noted with a more accentuated increase from

the time closest to the release of the ball. Thus the

score for transversal velocity (VX) shows that the

goalkeeper moves initially at a relatively slow speed

until instants close to release of the ball, when he

definitively raises his velocity. The score of (FZ)

tends to be lower to bodyweight since the goalkeeper

commences his movement in the instants close to the

release of the ball. From that instant, FZ tends to

increase. This behaviour is similar to that recorded

for vertical jumps with a countermovement or in the

sprint start (Gutierrez-Davila et al., 2006).

Table II shows the mean, standard deviation and

the significance level of the goalkeeper’s variables in

the two environmental situations TM2 and TM4.

The lateral movement start (T(START-X)) between the

two conditions proposed for goalkeepers shows that

the anticipation time is significantly greater when the

goalkeeper’s uncertainty is reduced (situation TM2).

The goalkeepers started their movement an average

of 349+ 71 ms before the release of the ball for

situation TM2, and when the uncertainty increased

(situation TM4), the average anticipation time is

reduced (P5 0.001), the movement starting an

average of 193+ 67 ms before the release of the ball.

Table II shows the mean, standard deviation and

the significant differences of the transversal compo-

nent of the velocity and the space covered at the

instant of the release of the ball between the two

conditions proposed (TM2 and TM4), as well as the

corresponding score 100 ms before release (VX-REL;

eX-REL; VX-100 y eX-100, respectively). The data for

the goalkeepers reveals that their average scores are

clearly lower for TM4 (P5 0.001). Table II also

shows these same data for the vertical component

(VZ-REL; eZ-REL; VZ-100 and eZ-100, respectively). The

data for the goalkeepers shows that in both experi-

mental situations there is a clear tendency for the

goalkeeper to drop his CM during the anticipation

period (from T(START-X) to release), the average

values of VZ-100 and eZ-100, being significantly lower

in situation TM4 (P5 0.001). No statistically

significant differences were found between the two

experimental conditions in the velocity of the vertical

component at the instant of the release (VZ-REL),

while there were clear differences for eZ-REL

(P5 0.001), being lower when uncertainty was

greater (TM4). The tendency of the goalkeepers to

lower their CM during the anticipation period was

confirmed with the score obtained in the maximum

vertical velocity during that period (VZ-MAX). In both

experimental situations, the data coincide in that VZ-

MAX reaches average negative values, being signifi-

cantly less for TM4 (P5 0.001)

Discussion and conclusions

In situation TM2, the goalkeepers begin their lateral

movement (T(START-X)) 349+ 71 ms before the

thrower releases the ball. Considering that the time

of the average throw for the four throwers analysed

(T(THROW)) is 7206+ 30 ms, in all cases the start of

the goalkeeper’s movement occurred at the end of

the thrower’s run, before he initiated the throw

(Table I). This fact seems logical when the side of the

Table II. Descriptive statistics and significance level of anticipation time (TSTART-X), time where the vertical component is zero before the

final movement (TSTART-Z), velocity of lateral movement and space covered 100 ms before the release of the ball (VX-100 and eX-100,

respectively); velocity of lateral movement and space covered at the instant of the release of the ball (VX-REL and eX-REL, respectively);

velocity of vertical movement and space covered 100 ms before the release of the ball (VZ-100 and eZ-100, respectively); velocity of vertical

displacement and space covered at the instant of the release of the ball (VZ-REL and eZ-REL, respectively); maximum velocity of the vertical

component during the anticipation period(VZ-MAX), in situations TM2 and TM4.

Goalkeepers

TM2 TM4 F/P

TSTART-X (ms)

TSTART-Z (ms)

VX-100 (m � s71)

eX-100 (m)

VX-REL (m � s71)

eX-REL (m)

VZ-100 (m � s71)

eZ-100 (m)

VZ-REL (m � s71)

eZ-REL (m)

VZ-MAX (m � s71)

7349+71

68+73

0.37+0.17

0.035+0.021

0.77+0.29

0.090+0.041

70.30+0.21

70.040+0.035

70.23+0.32

70.068+0.045

70.38+0.19

7193+ 67

77+ 70

0.09+ 0.119

0.005+ 0.011

0.31+ 0.20

0.024+ 0.026

70.16+ 0.16

70.012+ 0.033

70.160+ 0.21

70.030+ 0.045

70.16+ 0.22

89.03***

0.82

82.12***

54.45***

90.43***

82.51***

13.08***

10.35**

0.95

12.62***

23.08***

Note: Results are (mean+ s); ***P50.001; **P5 0.01

1326 M. Gutierrez-Davila et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012

throw is known beforehand, and therefore, that

anticipation time cannot be thought of as proceeding

from clues given by the thrower’s movement. If the

goalkeeper begins his movement too soon, he can

enable the thrower to change the direction of the

throw easily, but that situation is unlikely, since the

goalkeeper’s CM moves relatively slowly, at least

until 100 ms before the release of the ball. The

thrower may not appreciate the goalkeeper’s lateral

movement, as evidenced by the reduced lateral

movement of the CM until ex-100. From that instant,

goalkeepers tend progressively to increase the aver-

age velocity of lateral movement of the CM until the

release of the ball (Table II). The increase in velocity

of the CM during the 100 ms before release occurs

while the throw is being executed. This makes it

more difficult for the thrower to change the trajectory

of the ball without affecting the release velocity.

As expected, the results indicate that when

uncertainty was greater (TM4), the start of the

goalkeeper’s lateral movement was delayed until

7193+ 67 ms, showing that it happened only some

instants after the start of the throw at the instant that

the thrower put his front foot on the ground. This

fact indicates that goalkeepers were able to identify

the side of the throw from clues given by the

thrower’s run, confirming the research of Canal-

Bruland and Schmidt (2009).

The research of Abernethy and Russell (1987) had

shown an anticipation period of less than 83 ms in

badminton strokes. However, in analysing the differ-

ences between their work and the present study,

in the anticipation time period, it must be remem-

bered that our methodology gives greater sensitivity

in detecting the start of the goalkeeper’s

movement. Nevertheless, our data are lower in value

(7193+ 67 ms) than those of Savelsbergh et al.

(2002) when analysing anticipation of soccer goal-

keepers employing techniques recording eye

movements synchronised temporally with the manip-

ulation of a joystick (300 ms before contact of the foot

with the ball). These differences may correspond to

the different structures of the movements analysed.

The reduced velocity and the low average displace-

ment of the goalkeeper’s CM 100 ms before release

of the ball (VX-100 and eX-100, respectively; see Table

II) shows, that until that instant at least, certain clues

about the throw can be detected, even though they are

not demonstrated as a definitive lateral movement,

coinciding with the comments of Savelsbergh et al.

(2005) that expert soccer goalkeepers wait longer

than inexperienced ones to decide on their response.

This behaviour has two possible explanations:

(a) that the clues perceived have not been

sufficiently clear up to that moment so that a

slow lateral movement would enable him to

modify the direction of his movement if he

had made a mistake in his perception of the

pre-clues or

(b) that the clues were clear when he begins his

movement but he tries to avoid giving

information to the thrower about that move-

ment before the thrower starts to execute the

temporal segmented sequence.

Our results cast doubt on this second explanation,

basically because when uncertainty is reduced

(TM2), all the average scores referring to velocity

and lateral movement of the goalkeeper’s CM are

significantly greater than those recorded for TM4

(P5 0.001; see Table II).

When the data on the vertical velocity of the

goalkeepers CM (Table II) were analysed, it could be

seen that there is a clear tendency to lower the CM,

at least until the instant when the ball is released.

The positive values or the change of tendency was

produced some instants after the release of the ball

(68+ 73 ms and 77+ 70 ms for TM2 and TM4,

respectively). The fact that the absolute value of the

velocity reached during the anticipation time is

significantly lower when uncertainty is greater

suggests that goalkeepers took certain precautions

against the possibility of having to modify their

movement or giving clear clues about their move-

ment to the thrower, at least until the release of the

ball.

According to these results, we would suggest that

goalkeepers were able to identify the clues that

indicate the side of the throw well in advance, but

they have some difficulty in doing so with the height

of the throw, starting their vertical movement

instants after the ball is released. These data confirm

the results reported by Savelsbergh et al. (2005),

stating that most errors committed by soccer goal-

keepers are associated with the incorrect perception

of the clues about height. In spite of that, we also

consider that the goalkeeper was capable of adjusting

the timing of his movement to the distance of the

throw. In this study the distance of the throw is three

metres longer than that analysed by Schorer et al.

(2007), where it was shown that at velocities similar

to those we obtained, the ball took some 297 ms to

travel seven metres. Possibly the strategies used by

goalkeepers are different depending on the distance

of the throw.

As already mentioned, in both experimental

conditions, all the average scores for the vertical

velocity of the goalkeepers’ CM, as well as the

maximum vertical velocity of the CM, indicate a

tendency to lower the CM until instants after the

release of the ball, when again it reaches positive

values or a change of direction to intercept balls

aimed at the lower part of the goal. This behaviour

Anticipatory strategies of team-handball goalkeepers 1327

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012

enables certain pre-tensing mechanisms of the

musculature to be activated before the acceleration

impulse, readying them to make the final movement

to intercept the ball; this is especially positive when

the throw is aimed at the upper corners of the goal

and a rapid upward movement of the CM is

required, as a countermovement jump. When the

throw is aimed at the lower corners and a rapid

downward movement of the CM is needed, pre-

tensing can prejudice it, as the tension created has to

be deactivated. This deactivation is not produced

instantaneously and can end by delaying the down-

ward movement of the goalkeeper’s CM (Gutierrez-

Davila et al., 2006; Neptune & Kautz, 2001), see

Figure 3.

The goalkeepers could improve their performance

trying to identify the clues of side and height of the

throws using the actions of the others players with the

objective of diminishing the uncertainty. The goal-

keepers should begin the movement slowly and wait

until they have the clues of the throw.

The methodology used has enabled us to tackle

the study of anticipation from an analysis of the

movement used by goalkeepers to save the ball. We

think that the methodology proposed will also enable

the temporal and spatial clues of the throw con-

nected with anticipation to be identified. To do this,

the contributions of Schorer et al. (2007) using

kinematics 3D analysis and hierarchical cluster

analysis for the identification of the movement

patterns of throws in handball may be an especially

relevant precedent for future research.

References

Abdel-Aziz, Y. I., & Karara, H. M. (1971). Direct linear

transformation from comparator coordinates into object space

coordinates in close-range photogrammetry. ASP Symposium on

close-range photogrammetry (pp. 1–18). Falls Church, VA:

American Society of Photogrammetry.

Abernethy, B., & Russell, D. G. (1987). The relationship between

expertise and visual search strategy in a racquet sport. Human

Movement Science, 6, 283–319.

Abernethy, B., & Zawi, K. (2007). Pickup of Essential Kinematics

Underpins Expert Perception of Movement Patterns. Journal of

Motor Behavior, 39(5), 353–367.

Bideau, B., Multon, F., Kulpa, R., Fradet, L., Arnaldi, B., &

Delamarche, P. (2004). Using virtual reality to analyze links

between handball thrower kinematics and goalkeeper’s reac-

tions. Neuroscience Letters, 372, 119–122.

Canal-Bruland, R., & Schmidt, M. (2009). Response bias

in judging deceptive movements. Acta Pychologica, 130, 235–

240.

Fradet, L., Botcazou, M., Durochel, C., Cretual, A., Multon, F.,

Prioux, J., & Delamarche, P. (2004). Do handball throws

always exhibit a proximal-to-distal segmental sequence? Journal

of Sports Sciences, 22, 439–447.

Gutierrez-Davila, M., Dapena, J., & Campos, J. (2006). The effect

of muscular pre-tensing on the sprint start. Journal of Applied

Biomechanics, 22(3), 194–201.

Huys, R., Canal-Bruland,R., Hagemann, N., Beek, P., Smeeton,

N., & Williams, M. (2009). Global information pickup under-

pins anticipation of tennis shot direction. Journal of Motor

Behavior, 41(2), 158–170.

Lidor, R., Argov, E., & Daniel, S. (1998). An exploratory study of

perceptual-motor abilities of women: Novice and skilled

players of team handball. Perceptual and Motor Skills, 86(1),

279–288.

Neptune, R. R., & Kautz, S. A. (2001). Muscle activation and

deactivation dynamics: The governing properties in fast cyclical

human movement performance. Exercise and Sport Science

Reviews, 29, 76–81.

Ranganathan, R., & Carlton, L. (2007). Perception–action

coupling and anticipatory performance in baseball batting.

Journal of Motor Behavior, 39(5), 369–380.

Rogulj, N., Papic, V., & Srhoj, V. (2005). Optimization of the

positioning of the goalkeeper in team handball. Leistungssport,

35(4.3), 50–55.

Savelsbergh, G. J. P., Van der Kamp, A. M., Williams A. M. &

Ward, P. (2005). Anticipation and visual search behaviour in

expert soccer goalkeepers. Ergonomics, 48, 1686–1697.

Savelsbergh, G. J. P., Williams, A. M., Van der Kamp, J., & Ward,

P. (2002). Visual search, anticipation and expertise in soccer

goalkeepers. Journal of Sports Sciences, 20, 279–287.

Schorer, J., Fath, F., Baker, J., & Jaitner, T. (2007). Identification

of inter-individual and intra-individual movement patterns in

handball players of varying expertise levels. Journal of Motor

Behavior, 39(5), 409–421.

Vignais, N., Bideau, B., Craig, C., Brault, S., Multon, F.,

Delamarche, P., & Kulpa, R. (2009). Does the level of

graphical detail of a virtual handball thrower influence a

goalkeeper’s motor response? Journal of Sports Science and

Medicine 8, 501–508.

Wagner, H., Buchecker, M., von Duvillard, S. P., & Muller E.

(2010). Kinematic comparison of team handball throwing with

two different arm positions. International Journal of Sports

Physiology and Performance, 5, 469–483.

Williams, A. M., Davids, K., & Williams, J. G. (1999). Visual

perception and action in sports. London: E. & F.N. Spon.

Wood, J. A. & Jennings, L. S. (1979). On the use of spline

functions for data smoothing. Journal of Biomechanics, 12,

477–479.

1328 M. Gutierrez-Davila et al.

Dow

nloa

ded

by [

Uni

vers

ity o

f M

assa

chus

etts

, Am

hers

t] a

t 01:

43 0

8 O

ctob

er 2

012