This article was downloaded by: [Ecole Hautes Etudes Commer-Montreal] On: 10 August 2015, At: 13:44 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place, London, SW1P 1WG

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

Perception of spin and the interception of curved football trajectories a

Remy Casanova , Olivier Borg

ab

& Reinoud J. Bootsma

a

a

Institut des Sciences du Mouvement UMR 7287, Aix Marseille Université, CNRS, Marseille, France b

Oxylane Research, Villeneuve d’Ascq, France Published online: 16 Feb 2015.

Click for updates To cite this article: Remy Casanova, Olivier Borg & Reinoud J. Bootsma (2015) Perception of spin and the interception of curved football trajectories, Journal of Sports Sciences, 33:17, 1822-1830, DOI: 10.1080/02640414.2015.1013052 To link to this article: http://dx.doi.org/10.1080/02640414.2015.1013052

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Journal of Sports Sciences, 2015 Vol. 33, No. 17, 1822–1830, http://dx.doi.org/10.1080/02640414.2015.1013052

Perception of spin and the interception of curved football trajectories

REMY CASANOVA1, OLIVIER BORG1,2 & REINOUD J. BOOTSMA1 1 2

Institut des Sciences du Mouvement UMR 7287, Aix Marseille Université, CNRS, Marseille, France and Oxylane Research, Villeneuve d’Ascq, France

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

(Accepted 26 January 2015)

Abstract Using plain white and chequered footballs, we evaluated observers’ sensitivity to rotation direction and the effects of ball texture on interceptive behaviour. Experiment 1 demonstrated that the maximal distance at which observers (n = 8) could perceive the direction of ball rotation decreased when rotation frequency increased from 5 to 11 Hz. Detection threshold distances were nevertheless always larger for the chequered (decreasing from 47 to 28 m) than for the white (decreasing from 15 to 11 m) ball. In Experiment 2, participants (n = 7) moved laterally along a goal line to intercept the two balls launched with or without ±4.3 Hz sidespin from a 30-m distance. The chequered ball gave rise to shorter movement initiation times when trajectories curved outward (±6 m arrival positions) or did not curve (±2 m arrival positions). Inward curving trajectories, arriving at the same ±2 m distances from the participants as the non-curving trajectories, evoked initial movements in the wrong direction for both ball types, but the amplitude and duration of these reversal movements were attenuated for the chequered ball. We conclude that the early detection of rotation permitted by the chequered ball allowed modulation of interception behaviour without changing its qualitative characteristics. Keywords: spin, perception, interception, reversal movements, football

Introduction Spinning balls are encountered in the course of many different sports. Imparting spin onto a ball affects not only its rebound characteristics but also its flight trajectory. The latter effect is due to the force created by the pressure differences surrounding a revolving ball moving through the air (Magnus, 1853). Operating in a direction perpendicular to the axis of rotation, the Magnus force deflects the ball from its original trajectory. In the present contribution, we concentrate on the effects of sidespin, giving rise to lateral deflections in the path of ball flight. We will refer to these as curved ball trajectories. While the mechanical effects of spin on the resultant ball flight trajectories have been studied extensively (Bray & Kerwin, 2003; Carré, Asai, Akatsuka, & Haake, 2002; Kao, Sellens, & Stevenson, 1994; Tsukada & Sakurai, 2008), the behavioural effects on the characteristics of the receiving player’s interceptive actions have received much less attention. Yet, given that imparting spin on a ball absorbs part of the energy transferred during contact, a spinning ball travels at a somewhat reduced speed compared to a non-spinning ball (Asai, Carré, Akatsuka,

& Haake, 2002; Carré et al., 2002; Whiteside, Alderson, & Elliott, 2010). Because this provides the receiving player with more time to respond, one is faced with the question of why players would prefer to use curved trajectories. A preliminary answer to this question was provided by Craig, Berton, Rao, Fernandez, and Bootsma (2006; also see Craig, Goulon, Berton, Rao, Fernandez, & Bootsma, 2009). Using a Head Mounted Display (HMD) based virtual-reality setting, they demonstrated that even top-level football players made systematic judgement errors when asked to indicate, from the goalkeeper’s perspective, whether a curved ball trajectory would end up in the goal or not. This finding, attributed to the visual system’s relative insensitivity to (lateral) acceleration, indicates that for curved ball trajectories goalkeepers cannot accurately predict the future position of the approaching ball. However, contrary to what has traditionally been assumed, interceptive actions need not rely on accurate predictions of when the ball will be where (i.e., on predictive control). For interceptive movements of sufficiently long duration, actions are in fact characterised by the pursuit of particular states of the agent-environment

Correspondence: Reinoud J. Bootsma, Institut des Sciences du Mouvement UMR CNRS 7287, Aix Marseille Université, 163 avenue de Luminy – CP 910, 13288 Marseille cedex 9, France. E-mail: [email protected] © 2015 Taylor & Francis

Interception of curved football trajectories

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

interaction that guarantee (i.e., are lawfully related to) the future achievement of the goal. Thus, in a prospective control scheme, the unfolding movement is based on time-evolving information with respect to what the agent must do so as to ensure interception, without requiring precise knowledge of when and where this will occur (Chardenon, Montagne, Laurent, & Bootsma, 2005; Craig, Bastin, & Montagne, 2011; Fajen & Warren, 2007; Ledouit, Casanova, Zaal, & Bootsma, 2013; McBeath, Shaffer, & Kaiser, 1995; McLeod, Reed, & Dienes, 2003; Michaels & Oudejans, 1992). Intercepting curved ball trajectories The few existing studies of the interception of curved ball trajectories reveal that, notwithstanding the aforementioned perceptual judgment errors, balls travelling along curved trajectories can be intercepted. Following up on the perceptual judgement tasks, in two studies, Craig and colleagues examined goalkeepers’ whole-body and/or hand movements during the interception of curved (and non-curved) ball trajectories in a similar HMD-based virtual-reality setting. In the first study (Craig et al., 2011), participants were situated 5 m behind the goal line and had to intercept the ball by moving a (virtual) rectangular bat along the goal line. The participant controlled the position of the bat by moving a handheld stylus over a graphics tablet. In the second study (Dessing & Craig, 2010) participants were situated on the goal line (along which they could freely move) and had to intercept the ball with their hands. Because wearing the HMD prevented direct vision of the hands, the latter were tracked and visually overlaid on the scene in the HMD displays. Both studies demonstrated that, for particular combinations of ball departure position, ball arrival position and spin direction (clockwise or counter clockwise), curved trajectories gave rise to reversal movements. That is to say, under these specific conditions, participants would start moving in a direction that led them away from the ball’s future arrival position before reversing movement direction and eventually (more or less successfully; see Craig et al., 2011) intercepting the ball when it crossed the goal line. These patterns of result led Craig et al. (2011) and Dessing and Craig (2010) to suggest that interception was prospectively controlled on the basis of current ball heading. In a real-world volleyball setting, Lenoir, Vansteenkiste, Vermeulen, and De Clercq (2005) studied how expert and novice players moved to receive curved (and non-curved) machine-launched serve trajectories. Focussing on balls arriving close to the receiver’s initial position, they too found that in the presence of spin, receivers first moved away from

1823

their initial position in the direction of the ball’s early heading – to their left for trajectories with clockwise spin and to their right for trajectories with counter clockwise spin – before reversing movement direction and subsequently intercepting the ball. All three studies thus demonstrated reversal movements, evoked by curved ball trajectories. Perception of ball rotation While trajectory characteristics apparently do not allow players to infer the future arrival position of a spinning ball (Craig et al., 2006, 2009, 2011; Dessing & Craig, 2010; Lenoir et al., 2005), detection of ball rotation itself might prove to be useful. Recent studies have indicated a role of prior knowledge in the organisation of interceptive actions (Diaz, Phillips, & Fajen, 2009; López-Moliner, Field, & Wann, 2007; Morice, François, Jacobs, & Montagne, 2010). According to the existing literature, the human visual system would at first glance seem to be quite good at perceiving ball rotation, with a threshold for the discrimination of rotational velocity on the order of 5% (Barraza & Grzywacz, 2002; Kaiser & Calderone, 1991). Perception of rotational velocity has been reported to rely primarily on temporal frequency characteristics, such as edge rate (Kaiser & Calderone, 1991), although texture density also appears to play a role (Barraza & Grzywacz, 2002; Kaiser, 1990; Kaiser & Calderone, 1991). However, perhaps because of a reliance on computer-generated displays, these psychophysical studies have in fact only explored the low range of object rotation velocities; indeed reference rotation velocity was around 75 deg · s–1, corresponding to a rotation frequency of 0.2 Hz. Observations in football indicate much larger ball rotation rates, up to and even slightly over 10 Hz (Craig et al., 2006; Griffiths, Evans, & Griffiths, 2005; Tsukada & Sakurai, 2008). Furthermore, while in the psychophysical studies, the horizontally rotating objects subtended a visual size of 4.5 degrees, football players are confronted with much smaller visual sizes: a standardsized 70-cm circumference (i.e., 22-cm diameter) ball gives rise to visual sizes of 1.26, 0.63, and 0.42 degrees at distances of 10, 20, and 30 m, respectively. Thus, it is presently not clear whether players can actually detect the rapid rotation of a real football located at such realistic distances (Alcock, 2010; Dupeux, Cohen, Le Goff, Quéré, & Clanet, 2011; Grant, Williams, & Reilly, 1999). We know of only one previous contribution that speaks to this issue. In their volleyball study, Lenoir et al. (2005) in fact evaluated the effects of four differently patterned balls on the receivers’ motion patterns. Interestingly, they found that the amplitude of the

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

1824

R. Casanova et al.

reversal movements evoked by curved trajectories (7.7 ± 0.5 Hz ball rotation frequency in their study) was significantly larger for the (no longer used) “white” ball compared to the other three balls officially approved by the International Volleyball Federation (FIVB). The latter differed not only in colour pattern (green-redwhite, blue-yellow-white, red-blue-white) but also in texture pattern. Thus, whether the observed effects were indeed due to colour patterns, as suggested by Lenoir et al. (2005), or (also) to texture patterns remains unclear for the moment. Nevertheless, the demonstration by Lenoir et al. (2005) that visual characteristics of the ball influenced the expert and novice receivers’ interceptive actions, with larger reversal movements for the curved trajectories of the poorly textured white ball, suggests that receivers were able (a) to detect that the coloured/textured balls were spinning and (b) to use this information to some extent in the organisation of their actions. In the present contribution, we provide experimental evidence in support of both these hypotheses. In the first experiment, we evaluated observers’ sensitivity to ball rotation, for two differently textured balls, using real footballs rotating at frequencies between 5 and 11 Hz. In the second experiment, we examined movement behaviour during the interception of real balls, following curved and non-curved trajectories using the same balls as examined in the first experiment. Our choice of using real rather than simulated balls in the present experiments was inspired by the need to ensure ecological validity of the conclusions for real-life sports settings. Indeed, the evaluation of observer sensitivity to ball rotation at realistic, sportsrelevant frequencies (Experiment 1) necessitated the use of real balls: Simulations on screens with limited refresh rates simply do not allow veridical presentation of higher rotation frequencies. Of the few existing studies on the characteristics of interception movements in response to curved ball trajectories the majority has used virtual-reality settings (Bastin, Craig, & Montagne, 2006; Craig et al., 2011; Dessing & Craig, 2010). Building on the result of Experiment 1 and following up on the only existing study using real balls (Lenoir et al., 2005), in Experiment 2, we sought to provide evidence within a real-world setting for some of the phenomena observed in virtual environment settings. For both experiments reported in the present contribution, participants provided written consent before participation. The study was approved by the local institutional review board (IRB) of the Institute of Movement Sciences (Comité Ethique de l’Institut des Sciences du Mouvement d’AixMarseille Université) and conducted according to University regulations.

Experiment 1: spin perception Method Participants. Eight male sport science students, 20.6 ± 4.2 years old (M ± s), voluntarily participated. They were recreational football players in the sense that they played occasionally without participating in any official competition. All had a visual acuity of 10/10 for each eye, as determined by a Monoyer vision test at a distance of 5 m. Apparatus. For the purposes of the present experiment, we designed a special-purpose machine for the assessment of observers’ sensitivity to ball rotation. It allowed the controlled in-place rotation (in clockwise and counter clockwise directions) around the vertical axis of footballs up to 12 revolutions per second (i.e., 12 Hz rotation frequency). A variable speed 24 V brushless DC motor with integrated drive electronics (McLennan BLDC58-50 MK2, Ash Vale UK) provided up to 50 W continuous output power. Motor rotation speed was controlled via a potentiometer and set to the required value using a numerical display of rotation frequency (in Hz, two decimal stroboscopicly-calibrated precision) created with LabviewTM software running on a Dell XPS L702X laptop computer. Balls were presented within a 70 cm wide by 65 cm high black frame, placed on a table top directly above the electrical motor hidden by black cloth. A 28 mm diameter aluminium cylinder fixed to the rotation axis of the vertically oriented motor extended into the visible part of the frame. The ball was placed on a 35 mm diameter suction cup fixed to the top of the cylinder and held in place by a second, freely rotating cup that could be lowered onto the ball from above. The two different balls tested are presented as the participants saw them in Figure 1. Both were 32panel balls and had been assembled according to the same procedure. For the “white” ball, all panels were white. For the “chequered” ball, the 12 pentagonal panels were black and the 20 hexagonal panels were white. Logo markings on the white panels were covered with mat white tape. The 22-cm diameter balls corresponded to the official regulations of the International Football Association Board. The rotation device was positioned at the end of a 2.5-m wide, 55-m long darkened hallway. Balls were illuminated by a LED lamp providing continuous (non-pulsed) white light against a dark background formed by a 1-m2 black sheet positioned directly behind the ball. These conditions gave rise to a mean contrast ratio of 400:1 with a minimal luminance of 1 cd · m–2.

1825

Figure 1. Participant’s view of the white (left panel) and chequered (right panel) balls within the rotation device of Experiment 1. The upper part of the motor’s elongated axis of rotation is visible just below the ball. The cups used to keep the ball in place are visible at the top.

Procedure. During preparation of each trial, the experimenter, ball and rotation device were hidden behind a black curtain that was opened by the experimenter at the onset of each trial. Participant sensitivity to ball rotation was measured using the method of limits (Kantowitz, Roediger, & Elmes, 2009). In the descending condition, the participant was initially positioned at the maximal distance of 50 m. After onset of the trial, the participant was to report the direction in which the ball was spinning (clockwise or counter clockwise). If he could not perceive the direction of spin at the initial 50-m distance, the participant gradually advanced towards the ball until he was certain, stopped and verbally reported the spin direction. In the rare case that the participant’s response was incorrect, the trial was rerun. The distance remaining was determined by reading the front of the participant’s foot position from a 50-m long measuring tape with cm markings. In the ascending condition, the participant began at a distance of 5 m from the ball, allowing perception of spin direction under all experimental conditions. The participant then gradually moved backward until he could no longer perceive the ball’s rotation direction. At this point, he stopped and verbally indicated having lost accurate rotation direction perception; distance was then measured in the same way as for the descending condition. No temporal constraints were imposed in any of the conditions. Balls could rotate in the clockwise and counter clockwise directions at rotation frequencies of 5, 7, 9 and 11 Hz. Each ball was thus presented under 16 experimental conditions, combining the 2 response methods (ascending/descending) × 2 rotation directions × 4 rotation frequencies. The order of experimental conditions was randomised within a block of 16 trials.

threshold distances were determined for each rotation frequency as the mean of the 4 trials (descending/clockwise rotation, descending/counter clockwise rotation, ascending/clockwise rotation, ascending/ counter clockwise rotation) for each ball at each rotation frequency. These threshold distances were submitted to a repeated-measures analysis of variance (ANOVA) with factors Ball Type (white and chequered) and rotation frequency (5, 7, 9, and 11 Hz). Where appropriate, post-hoc tests were performed using the Neman–Keuls procedure. The level of significance was set at 0.05.

Dependent variables and statistical analyses. In application of the method of limits, individual participant

Figure 2. Threshold distance for detection of rotation direction as a function of rotation frequency for the white and chequered balls.

Results and discussion As can be seen from Figure 2, threshold distance varied as a function of Ball Type and Rotation Frequency. This was confirmed by the ANOVA, revealing significant main effects of Ball Type (F(1, 7) = 208.54, P < 0.001, η2p = 0.97) and White ball Chequered ball

50 Threshold distance (m)

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

Interception of curved football trajectories

40

30

20

10

0

5

9 7 Rotation frequency (Hz)

11

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

1826

R. Casanova et al.

rotation frequency (F(3, 21) = 80.40, P < 0.001, η2p = 0.92), as well as an interaction between the two (F(3, 21) = 25.40, P < 0.001, η2p = 0.78). Posthoc analyses of the interaction revealed that for both balls, threshold distance decreased with increasing rotation frequency. This effect of rotation frequency was stronger for the chequered than for the white ball, with threshold distance varying, respectively, from 46.84 to 27.58 m and from 14.96 to 10.76 m. Nevertheless, threshold distance remained larger for the chequered ball than for the white ball even at the highest rotation frequencies (all ps < 0.001). Earlier psychophysical studies on the perception of rotation have been limited to computer-generated objects subtending around 4.5° visual angle and rotating at frequencies around 0.2 Hz (Barraza & Grzywacz, 2002; Kaiser, 1990; Kaiser & Calderone, 1991). To explore perceptual sensitivity to ball rotation under conditions pertinent to sports settings, we determined the maximal distances at which participants could detect the direction of rotation of real footballs in the 5 to 11 Hz rotation frequency range. Threshold distances were found to decrease when rotation frequency increased. Even if all 32 panels used to assemble a real ball were homogeneously white, the assembly seams allowed the direction of rotation to be perceived at distances varying from around 15 m for a 5-Hz rotation to around 11 m for a 11-Hz rotation. When the ball was textured, with 12 black pentagonal panels and 20 white hexagonal panels, the maximal distance at which rotation direction could be perceived was considerably larger, varying from around 47 m for the 5-Hz rotation to around 28 m for the 11-Hz rotation. Experiment 2: intercepting curved trajectories Although the threshold distances reported in Experiment 1 were determined under optimal viewing conditions (i.e., at a stationary position, without time pressure, and with balls presented under high luminance and contrast in a dark environment), the results indicated that the direction of rotation of the chequered ball may be perceived at a much larger distance than that of the white ball. In Experiment 2, we evaluated whether the perception of ball rotation direction could be used in the organisation of interception behaviour, as suggested by Lenoir et al. (2005). To this end, we compared participants’ interception behaviour when they were confronted with the white and chequered balls following curved and noncurved trajectories. Balls were launched from a distance of 30 m using a moderate ball rotation frequency (slightly below 5 Hz), allowing the potential use of information on ball rotation direction in the organisation of interception behaviour to be tested

under reasonable conditions. More specifically, we expected that under these conditions, participants could rapidly perceive the direction of rotation for the chequered ball but not for the white ball. If such information were useful, this should result in early differences in the patterns of interception behaviour. Method Participants. Seven male recreational football players (age: 27.1 ± 3.2 years, height: 1.78 ± 0.05 m) voluntarily participated in the experiment. All had normal or corrected-to-normal vision and reported to be in good health. None of them had participated in Experiment 1. Task and procedure. The experiment took place in a gymnasium with daylight illumination. Moving laterally along a goal line on the floor participants were to intercept balls delivered by means of a JUGS Soccer Machine (JUGS Sports, Tualatin OR, USA), from a height of 40 cm and a perpendicular distance of 30 m from to the participant’s lateral plane of motion. The launching device consisted of two independently controlled wheels turning in opposite directions, allowing to vary ball speed and spin. Preliminary testing allowed us to determine settings that propelled balls to passing distances along the participant’s displacement axis at 0 m (no spin = NS), +4 m (clockwise spin = CS) and −4 m (counter clockwise spin = CCS). Average flight times for the spinning and non-spinning ball trajectories were, respectively, 1.92 ± 0.11 s and 1.80 ± 0.06 s. Fitting an aerodynamical model including gravity, drag and Magnus-Robins lift forces and literature-based coefficients (see Craig et al., 2006, for details) indicated a spin rate of 4.3 ± 0.5 Hz for the CS and CCS trajectories. Participants’ displacements along the goal line were filmed with a Sony HDR-XR200 video camera (1920 × 1080 pixel resolution) operating at 25 Hz. The camera was fixed on a tripod at a height of 1.8 m, positioned halfway between the launching machine and the participant’s displacement axis. Participants wore a tight-fitting white swimming cap on the head. They were dressed in black and wore dark gloves to facilitate off-line tracking of their head movement during interception against the gymnasium’s green-grey background. For each trial, the plopping sound produced by the ball when it was propelled out of the launching machine was used to determine the onset of ball flight (with t0 compensated for the 15-m distance between the launching machine and the camera). Using ImageJ software, head position was determined in each video frame from t0 until the ball was intercepted or passed the goal line.

Interception of curved football trajectories

A IP –4

B IP

–2

0

+2

+4

NS CCS

CCS

CCS

CS CS

CS

NS

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

1827

–6

NS

–2

+2

+6

IP JUGS machine Figure 3. Schematic representation of the experimental set-up of Experiment 2. A: The JUGS machine launched the balls over a distance of 30 m to three different arrival positions along the participant’s displacement axis. NS: no spin – ball arrival at 0 m; CS: clockwise spin – ball arrival at +4 m; CCS: counter clockwise spin – ball arrival at −4 m. IP: initial participant position at −2 or +2 m. B: Experimental conditions from the participant’s perspective (IP) with straight (solid black), inward curving (solid grey) and outward curving (dash-dot grey) trajectories.

Before each trial, the participant was instructed to position himself at one of two possible initial positions, located 2 m to the left and to the right of the launching machine (see Figure 3, left panel). The two initial participant positions combined with the three different ball trajectories (NS, CS, CCS) gave rise to six different conditions. The order of conditions was randomised within a block of six trials. Participants performed five blocks for each ball type, for a total of 6 × 5 × 2 = 60 trials. Notwithstanding all precautions taken (precision of wheel-speed settings, control of ball orientation when launching, etc), ball trajectories were quite variable. As trajectory variability increased over flight duration and we were interested in the early effects of the type of ball (white vs. chequered), participant movement behaviour was only analysed over the first second of ball flight. Position data extracted from the films were first calibrated and then filtered using a Butterworth filter with a cut-off frequency of 5 Hz. Movement velocity was obtained by differentiating the position time series using the first central difference method. From the participant’s point of view (Figure 3, right panel), balls could arrive at distances of 6 m and 2 m to the left and to the right of their initial position. Balls arriving at −6 m (left) and +6 m (right) followed outward curving trajectories. Balls arriving at −2 m and +2 m could follow straight or inward curving trajectories. Reversal movements were defined as participants first moving away from the ball’s future arrival position before reversing movement direction to move

towards the interception position. Operationally, identification of a reversal movement required that the movement in the “wrong” direction reached a velocity above 20 cm · s–1. Dependent variables and statistical analyses. The moment of movement initiation was determined using a velocity threshold of 20 cm · s–1. This movement initiation time, calculated for the full set of conditions, was analysed using a repeated-measures ANOVA with the factors Ball Type (white or chequered), Launching Side (left or right) and Trajectory (inward, straight or outward). Newman– Keuls post-hoc tests were performed where appropriate to examine significant effects in more detail. For interception movements demonstrating reversals in movement direction, we extracted peak excursion (i.e., amplitude of movement in the “wrong” direction) and its moment of occurrence. As reversal movements were only observed for inward-curving trajectories, we averaged these kinematic characteristics over the left and right launching sides (demonstrating no significant effects) to increase statistical power and analysed the effect of Ball Type with paired t-tests. As in Experiment 1, significance level was set at 0.05. Results and discussion Participants caught all balls following straight or inward curving trajectories. Interception was more difficult for the outward curving trajectories

1828

R. Casanova et al.

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

0.30 s) or outward curving (average 0.31 s) trajectory than when it followed an inward curving trajectory (average 0.36 s). No such trajectory effects were observed for the white ball that was characterised by an average initiation time of 0.35 s. Hence, compared to the while ball, the chequered ball gave rise to earlier movement initiation for the straight and outward curving trajectories but not for the inward curving trajectories. Reversal movements. Contrary to the outward curving and straight trajectories, the inward curving trajectories gave rise to reversal movements, where participants first moved away from the ball’s future arrival position before reversing movement direction. Because such movement reversals were observed in 79.2% of all inward curving trajectory trials, they can also be discerned in the ensemble average position and velocity profiles of Figure 4. The peak excursion in the “wrong” direction was significantly (t(12) = 1.90, P < 0.05) larger for the white ball (28.6 ± 13.9 cm) than for the chequered ball (16.3 ± 9.95 cm). This peak also occurred significantly (t(12) = 3.60, P < 0.01) later for the white ball (0.74 ± 0.14 s) than for the chequered ball (0.61 ± 0.10 s). Largely replicating and extending the findings of Lenoir et al. (2005), the results of Experiment 2 confirmed the expected effects of ball trajectory and ball type on interception behaviour. First, contrary to straight and outward curving trajectories, inward curving trajectories gave rise to reversal movements. Second, these reversal movements were attenuated in both amplitude and duration when participants were confronted with the chequered ball as compared to the white ball. Third, participants initiated their interception movement earlier when confronted with the chequered ball following straight and outward curving trajectories. This latter result is somewhat different from the pattern reported by Lenoir et al. (2005) who found a main effect of ball trajectory (and no significant interaction with ball type). This difference is most likely because in the present study, we used the head position time-series to derive all relevant kinematic variables. While Lenoir et al. also used head movement to determine the amplitude of

Figure 4. Ensemble averages of participants’ position (A) and velocity (B) over the first second of ball flight for the six experimental conditions. Fat light grey: outward curving trajectories; Black: straight trajectories; Dark grey: inward curving trajectories. Continuous lines: white ball; Intermittent lines: chequered ball.

requiring the participant to move 6 m: under these conditions, they succeeded in catching or touching the balls on, respectively, 37.8% and 27.9% of the trials. Figure 4 presents the ensemble averages of the position and velocity profiles for each experimental condition over the first second of ball flight. Movement initiation. Movement initiation times for the six trajectories and the two ball types are summarised in Table I. The ANOVA revealed significant main effects of Ball Type (F(1, 6) = 6.29, P < 0.05, η2p = 0.51) and Trajectory (F(2, 12) = 5.24, P < 0.05, η2p = 0.47), as well as a significant interaction between the two (F(2, 12) = 4.08, P < 0.05, η2p = 0.41). Launching Side did not affect movement initiation (P > 0.1 and η2p < 0.1 for the main effect and interactions). Post-hoc analysis of the Ball Type × Trajectory interaction revealed that for the chequered ball, participants initiated their movement earlier when the ball followed a straight (average

Table I. Movement initiation times (in s, standard deviations between parentheses) for the six trajectories and the two ball types. Trajectories Inward

White Chequered

Straight

Outward

−6 m

+6 m

−2 m

+2 m

−2 m

+2 m

0.36 (0.04) 0.36 (0.08)

0.37 (0.05) 0.36 (0.11)

0.34 (0.04) 0.30 (0.06)

0.38 (0.05) 0.31 (0.04)

0.34 (0.05) 0.30 (0.07)

0.36 (0.05) 0.32 (0.05)

Interception of curved football trajectories reversal movements, they relied on the (force platform registered) ground reaction forces exerted by the feet to determine movement initiation (also see Benerink, Bootsma, & Zaal, in press).

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

General discussion Interception of curved ball trajectories has mostly been studied to identify the information used in the control of action. Indeed, both the systematic errors in perceptual judgments of the future arrival position of a ball following a curved trajectory (Craig et al., 2006; 2009), as well as the systematic differences in movement behaviour when intercepting balls arriving at the same position via different curved and non-curved trajectories (Bastin et al., 2006; Craig et al., 2011; Dessing & Craig, 2010; Lenoir et al., 2005) have provided evidence in favour of a prospective control strategy based on information with respect to current ball heading. However, with the exception of Lenoir et al. (2005), all these studies have relied on virtual reality settings. As noted by Zaal and Bootsma (2011), virtual reality certainly provides a powerful tool to study the organisation of behaviour underlying our interactions with the environment, but also requires regular reality checks. One of the goals of the present contribution was thus to evaluate the interception of curved ball trajectories in a real-world setting. Importantly, interception behaviour in such a realword setting revealed the same phenomena observed in virtual-reality settings: Balls following straight and inward curving trajectories converging onto the same interception location not only gave rise to different, trajectory-dependent movement patterns, but the inward curving trajectories also gave rise to reversal movements (see Craig et al., 2011; Dessing & Craig, 2010; Lenoir et al., 2005; for comparable observations). This pattern of results thus validates, at least as far as the above-mentioned phenomena are concerned, the use of virtual-reality settings to study the effects of ball trajectory characteristics on interception behaviour. Yet, however powerful a tool virtualreality may be prove to be, due to its technological limitations, it simply cannot address all issues of interest. One such issue is the perception of rotation for balls spinning at the high rates encountered in sports settings. Indeed, realistic spin rates on the order of 3 to10 Hz cannot be properly represented on a monitor operating at a limited refresh rate. A chequered ball with 12 black (pentagonal) panels and 20 (hexagonal) white panels, like the one used in the present contribution, has a repetitive black-and-white texture pattern. Thus, during a full turn, the same pattern appears several times. Assuming that rotation might be detected from 10 consecutive different

1829

images (Kaiser & Calderone, 1991), simulating rotation of a ball with 5 pattern repetitions on a 100-Hz refreshment rate monitor does not provide correct perception of ball rotation at rates above 2 Hz. Most virtual-reality settings currently operate at 30 Hz, providing an upper limit of 0.67 Hz. When object rotation is displayed above this critical rate, veridical perception is no longer possible; rotation in the other direction may even be perceived as illustrated by the well-known wagon-wheel effect, (Finlay, Dodwell, & Caelli, 1984; Purves, Paydarfar, & Andrews, 1996). In this framework, it is also useful to realise that watching a football match on television therefore does not provide a veridical picture of ball rotation. To evaluate whether perception of spin direction may be used to modulate interception behaviour, we proceeded in two steps. In Experiment 1, we evaluated observer sensitivity to spin direction by determining the maximal distance at which they could determine the direction of rotation of real balls rotating at frequencies between 5 and 11 Hz. Two differently textured balls were tested, one with 32 white panels and one with 12 black (pentagonal) panels and 20 (hexagonal) white panels. While for both balls, threshold distance decreased with increasing rotation frequency, observers could perceive the direction of rotation of the chequered ball at much larger distances than that of the white ball. In Experiment 2, we evaluated whether the early detection of spin direction permitted by the chequered ball affected interception behaviour. In line with the earlier work of Lenoir et al. (2005), our results indicated that this was indeed the case: the reversal movements evoked by the inward curving trajectories were attenuated both in amplitude and in duration when participants intercepted the chequered ball as compared to when they intercepted the white ball. We also found that participants responded earlier to the chequered ball than to the white ball for straight and outward curving trajectories. These latter results may be understood as resulting from the (useful) perception of no spin for the straight trajectories and of outward-driving spin for the outward curving trajectories. The absence of an effect of ball type on the moment of movement initiation for the inward curving trajectories is most likely related to the difficulty of assessing where the ball is initially heading with respect to the participant’s initial position. As noted above, for the inward curving trajectories, an effect of ball type did nevertheless appear somewhat later, in the course of the reversal movement. Thus, even though the detection of the direction of ball rotation does not allow players to accurately predict where the ball will end up, it does seem to provide a perhaps more global, qualitative indication with respect to the ball’s flight trajectory that can be used to modulate interception behaviour.

1830

R. Casanova et al.

Overall, the present contribution thus demonstrated that the distance at which a player can perceive the direction in which a ball is rotating depends both on the rate of rotation and on the way the ball is textured. Moreover, early detection of rotation direction was found to modulate a receiving player’s interception behaviour without changing its qualitative characteristics. Given the omnipresence of spin in ball sports, we conclude that the visual design of a ball may have non-negligible effects on the way in which players perceive and act upon play situations.

Downloaded by [Ecole Hautes Etudes Commer-Montreal] at 13:44 10 August 2015

References Alcock, A. (2010). Analysis of direct free kicks in the women’s football World Cup 2007. European Journal of Sport Science, 10, 279–284. Asai, T., Carré, M. J., Akatsuka, T., & Haake, S. J. (2002). The curve kick of a football I: Impact with the foot. Sports Engineering, 5, 183–192. Barraza, J. F., & Grzywacz, N. M. (2002). Measurement of angular velocity in the perception of rotation. Vision Research, 42, 2457–2462. Bastin, J., Craig, C., & Montagne, G. (2006). Prospective strategies underlie the control of interceptive actions. Human Movement Science, 25, 718–732. Benerink, N. H., Bootsma, R. J., & Zaal, F. T. J. M. (in press). Different temporal bases for body and arm movements in volleyball serve reception. Scandinavian Journal of Medicine and Science in Sports. doi:10.1111/sms.12384 Bray, K., & Kerwin, D. G. (2003). Modelling the flight of a soccer ball in a direct free kick. Journal of Sports Sciences, 21, 75–85. Carré, M. J., Asai, T., Akatsuka, T., & Haake, S. J. (2002). The curve kick of a football II: Flight through the air. Sports Engineering, 5, 193–200. Chardenon, A., Montagne, G., Laurent, M., & Bootsma, R. J. (2005). A robust solution for dealing with environmental changes in intercepting moving balls. Journal of Motor Behavior, 37, 52–64. Craig, C. M., Bastin, J., & Montagne, G. (2011). How information guides movement: Intercepting curved free kicks in soccer. Human Movement Science, 30, 931–941. Craig, C. M., Berton, E., Rao, G., Fernandez, L., & Bootsma, R. J. (2006). Judging where a ball will go: The case of curved free kicks in football. Naturwissenschaften, 93, 97–101. Craig, C. M., Goulon, C., Berton, E., Rao, G., Fernandez, L., & Bootsma, R. J. (2009). Optic variables used to judge future ball arrival position in expert and novice soccer players. Attention Perception & Psychophysics, 71, 515–522. Dessing, J. C., & Craig, C. M. (2010). Bending it like Beckham: How to visually fool the goalkeeper. PLoS ONE, 5(10), e13161. Diaz, G. J., Phillips, F., & Fajen, B. R. (2009). Intercepting moving targets: A little foresight helps a lot. Experimental Brain Research, 195, 345–360. Dupeux, G., Cohen, C., Le Goff, A., Quéré, D., & Clanet, C. (2011). Football curves. Journal of Fluids and Structures, 27, 659–667. Fajen, B. R., & Warren, W. H. (2007). Behavioral dynamics of intercepting a moving target. Experimental Brain Research, 180, 303–319.

Finlay, D., Dodwell, P. C., & Caelli, T. (1984). The waggonwheel effect. Perception, 13, 237–248. Grant, A. G., Williams, A. M., & Reilly, T. (1999). Analysis of goals scored in the 1998 World Cup. Journal of Sports Sciences, 17, 826–827. Griffiths, I., Evans, C., & Griffiths, N. (2005). Tracking the flight of a spinning football in three dimensions. Measurement Science and Technology, 16, 2056–2065. Kaiser, M. K. (1990). Angular velocity discrimination. Perception & Psychophysics, 47, 149–156. Kaiser, M. K., & Calderone, J. B. (1991). Factors influencing perceived angular velocity. Perception & Psychophysics, 50, 428–434. Kantowitz, B. H., Roediger, H. L., & Elmes, D. G. (2009). Experimental psychology (10th ed.). Stamford, CT: Cengage Learning. Kao, S. S., Sellens, R. W., & Stevenson, J. M. (1994). A mathematical model for the trajectory of a spiked volleyball and its coaching application. Journal of Applied Biomechanics, 10, 95–109. Ledouit, S., Casanova, R., Zaal, F. T. J. M., & Bootsma, R. J. (2013). Prospective control in catching: The persistent angleof-approach effect in lateral interception. PLoS ONE, 8(11), e80827. Lenoir, M., Vansteenkiste, J., Vermeulen, J., & De Clercq, D. (2005). Effects of contrasting colour patterns of the ball in the volleyball reception. Journal of Sports Sciences, 23, 871–879. López-Moliner, J., Field, D. T., & Wann, J. P. (2007). Interceptive timing: Prior knowledge matters. Journal of Vision, 7(13), 11.1–8. Magnus, G. (1853). Ueber die Abweichung der Geschosse, und: Ueber eine auffallende Erscheinung bei rotirenden Körpern [On the deviation of projectiles, and: On a striking phenomenon among rotating bodies]. Annalen Der Physik Und Chemie, 164, 1–29. McBeath, M. K., Shaffer, D. M., & Kaiser, M. K. (1995). How baseball outfielders determine where to run to catch fly balls. Science, 268, 569–573. McLeod, P., Reed, N., & Dienes, Z. (2003). Psychophysics: How fielders arrive in time to catch the ball. Nature, 426, 244–245. Michaels, C. F., & Oudejans, R. R. D. (1992). The optics and actions of catching fly balls: Zeroing out optical acceleration. Ecological Psychology, 4, 199–222. Morice, A. H. P., François, M., Jacobs, D. M., & Montagne, G. (2010). Environmental constraints modify the way an interceptive action is controlled. Experimental Brain Research, 202, 397–411. Purves, D., Paydarfar, J. A., & Andrews, T. J. (1996). The wagon wheel illusion in movies and reality. Proceedings of the National Academy of Sciences, 93, 3693–3697. Tsukada, T., & Sakurai, S. (2008). Changes of ball spin during flight of kicked soccer ball. ISBS Conference Proceedings Archive. Retrieved from https://ojs.ub.uni-konstanz.de/cpa/article/view/ 1988 Whiteside, D., Alderson, J., & Elliott, B. (2010). “Bend it like Beckham”: Ball rotation in the curved football kick. In R. Jensen, W. Ebben, E. Petushek, C. Richter, & K. Roemer (Eds.), Proceedings of the 28th conference of the international society of biomechanics in sports. Marquette: Northern Michigan University. Zaal, F. T. J. M., & Bootsma, R. J. (2011). Virtual reality as a tool for the study of perception-action: The case of running to catch fly balls. Presence-Teleoperators and Virtual Environments, 20, 93–103.