1

The Role Of Reaction Time and Anticipation in

Motor Control

Presented by: Zinat Ashnagar

2

Introduction

Reaction TimeReaction time (RT) is a measure of the time from the

arrival of a suddenly presented and unanticipated signal to the beginning to response to it.

(motor control &learning)

Electromyographic reaction time (ERT) is the time measured between the presentation of a stimulus and execution of an appropriate response to that stimulus. It has two fractions;

3

Premotor time which is the time interval between the onset of the stimulus and the onset of the electromyographic (EMG) activity of the relevant muscle and is thought to represent central processes involved in making the response (e.g. perception, decisions).

Motor time which is the time interval between the onset of

EMG activity and completion of the actual muscular task and represents processes associated with

musculature itself.(Botwinick and Thompson, 1966& Motor control and learning)

4

5

Motor time (MT) reflects the duration of the electromechanical transduction within muscular fibers, whereas premotor time (PMT) reflects the duration of all earlier stages of information processing.

(Audiffren &Tomporowski,2008)

6

Anticipation

Spatial, or Event AnticipationOne way in which performers can anticipate future activities is by knowing what kind of stimuli are going to be presented and

what kind of responses will be required.

Temporal AnticipationThe anticipation of when a given stimulus will arrive or when a

movement is to be made

ForeperiodThe interval between a warning signal and the stimulus to

respond

7

Simple Reaction Time (SRT)Reaction time from a task in which a single known response is

produced when a single stimulus is presented.

Choice Reaction Time (CRT)Reaction time for a task in which each response to be made is

associated with a different stimulus.

Choice RT task is often employed, preventing the subject from knowing what stimulus is going to occur; or foreperiods are

randomized so that the subject cannot predict when the stimulus will arrive.

8

Reaction time measures are common in many sport settings; an example is the interval between the starter’s gun and the first movement in a swimming race. Reaction time measures are also studied extensively in the laboratory as measures of information-processing speed.

Reaction time (RT) is a highly sensitive and objective parameter reflecting cognitive and motor function and was used in several studies (Milner, 1986; Turhanoglu and Beyazova,2003).

9

Most of the studies use computer based methods to measure RT instead of electromyographyLord et al., 1993; Lord and Castell, 1994; Arcelin et al., 1998; Brisswalter et al., 1997; Arcelin and Brisswalter, 1999; Collardeau et al., 2001a,b; Whitehurst, 1991.

Such techniques yield total reaction time rather than premotor time. (Crabtree and Antrim, 1988)

As premotor time reflects cognitive function, it is preferable to measure it selectively, apart from motor time. For this

purpose EMG can be used. (Taskiran &gunendi,2008)

10

Reaction time measures are very common in research on skills, for two basic reasons.

First, RT measures are components of real life tasks (e.g.,Sprint starts).

A more important reason is that RT presumably measures the time taken for mental events, such as stimulus

processing ,decision making, and movement programming. These two motivations for using RT measures differ

considerably.In the first case, RT is a measure studied for it’s own sake.In the second case, RT allows the researcher to understand

the kinds of mental processes that lead to movement.(e.g. ,posner,1978)

(Motor control & learning)

11

Reaction times and anticipatory skills of karate athletes

Shuji Mori a,*, Yoshio Ohtani b, Kuniyasu Imanaka a

a Department of Kinesiology, Graduate School of Science, Tokyo Metropolitan University,

1-1 Minami- Ohsawa, Hachioji, Tokyo 192-0397, Japanb Faculty of Engineering and Design, Kyoto Institute of Technology,

Matsugasaki,Sakyo- ku, Kyoto 606-8585, Japan

Human Movement Science 21 (2002) 213–230

www.elsevier.com/locate/humov

12

Successful performance in sport requires not only efficient execution of motor behavior but also a high

level of perceptual ability

In sport science, two types of perceptual abilities have been considered relevant to the player’s successful performance. One is primitive, basic sensory functions which are not specific to particular types of sport expertise

The other type of perceptual processing is sport-specific perceptual skills

13

Research has demonstrated that expert sport players are superior to novices in perceptual skills, such as detecting the presence of a ball in briefly presented sport scenes(Allard & Starkes, 1980; Starkes, 1987); making efficient search for relevant, informative parts of the opponent’s body and fields (Abernethy & Russell, 1987; Goulet, Bard, & Fleury, 1989; Ripoll, Kerlirzin, Stein, & Reine, 1995; Williams, Davids, Burwitz,& Williams, 1994); anticipating the ball direction and the opponent’s action from advance information (Abernethy, 1990; Jones & Miles, 1978; Paull & Glencross,1997; Williams & Davids, 1998);

The expert advantage in perceptual skills has been typicallyinvestigated by sport-specific realistic stimuli and tasks (e.g., judging future actions in filmed sequences of real play).

14

Karate is a good example of a competitive sport with high levels of temporal and spatial constraints which require fast reactions.

The need to offend and defend against the opponents should lead karate athletes to develop their perceptual abilities, i.e. karate-specific perceptual skills and/or non-specific basic sensory functions, to make fast reactions

A few studies have used simplistic stimuli and tasks to assess the possible advantage of karate athletes in basic sensory functions

A couple of studies by Williams and his colleagues have used more realistic stimuli and tasks to examine expertise anticipation of karate athletes, which is one important aspect of perceptual skill (Scott, Williams, & Davids, 1993; Williams &Elliott, 1999).

In both studies, the stimuli were dynamic film displays of karate athletesperforming offensive attacks against the viewer.

15

Scott et al. (1993) found that karate athletes showed faster RTs than novices for both verbal and action response modes, while the athletes showed higher accuracy only with the action mode.

In contrast, Williams and Elliott (1999) showed, in a similar experiment using the action response mode, that the expert karate athletes were no faster than the novices, while the athletes were more accurate than the novices.

there appears to be no coherent picture concerning the advantage of expert karate athletes in RT tasks.

The present study was designed to obtain a better understanding of the expertise advantage of karate athletes

For this purpose, we followed a so-called multi-task approach (Williams et al., 1999), in which the performances of expert and novice players are examined across multiple types of tasks and stimuli.

16

The multitask approach has been proved quite useful to reveal expert superiority in various motor and perceptual abilities, the latter including basic sensory functions and perceptual skills (e.g., Helsen & Starkes, 1999; Reilly, Williams, Nevill, & Franks,2000; Starkes, 1987)

In the first experiment of the present study, we measured RTs from the same group of expert athletes and novices, using two types of stimuli, realistic and simplistic, and two types of response tasks, choice RT and simple RT.

The realistic stimuli were dynamic displays of karate athletes performing offensive actions against the viewer

The target positions of the offensive actions were either the upper (‘joh-dan’) or middle (‘chu-dan’) level of the opponent (viewer)

The simplistic stimuli were presentations of a single dot at either an upper or lower location in the stimulus display

17

In the choice RT task, the participants were requiredto indicate as soon and as accurately as possible the target location of the attack,or the presentation location of the dot. In the simple RT task, the participants responded as soon as possible when the video stimulus started to move from a static display, or when the dot appeared on the monitor

realistic (i.e., video-stimulus choice-RT task)

simplistic (i.e., dot- and video-stimulus simple RT tasks and dot-stimulus choice RT task)

In the second experiment, we attempted to examine anticipation of the karate athletes in a paradigm other than an RT task.

18

Chamberlain and Coelho (1993) suggested that in sport-specific situations, novices might be less confident than experts and require a longer time for making decisions

This suggests that the RT task in the first experiment may include a possible advantage of the karate athletes in decision making, in addition to their superior anticipatory skill.

To evaluate this possibility, we used a temporal occlusion method, where the participants were given unlimited time to judge as accurately as possible the target positions of the attacks which were occluded before their completion

If the results of this experiment corroborate the RT data of the first experiment, showing a higher accuracy of judgments by the athletes, this would constitute strong evidence that the karate athletes are superior to the novices in anticipatory skills regarding the attack.

19

20

21

Experiment 1 Method

Participants :Six collegiate karate athletes (mean age of 21 in the range of 19–22) and seven novices (mean age of 28 in the range of 21–43), all males and right handed, participated in this experiment. Four of the athletes had 4–6 yrs of experience of karate training, all with black belts, and the other two had about 1 yr of experience. The novices, including the first and the second authors (ages 39 and 43, respectively), had no experience of karate training but possessed basic knowledge of karate

Apparatus and stimuliAll stimuli were presented on a high-resolution color monitor (SONY

MultiscanGDM-17SE2T), with a screen size of 31 23 cm, controlled by a personal computer (COMPAQ DeskPro 5100) with a color graphic system which also controlled the experimental timing, RT measurement and recording from a custom-made response box.

22

For the video stimuli, offensive actions of three karate athletes with black belts were recorded by a digital video camera (SONY DCR-TRV10)

the video camera was positioned at the height of 1.7 m and approximately 2 m away from the athlete’s initial position.This camera position simulated the eye level and viewing distance of karate athletes in real sparring situations, so that the video-recorded pictures would replicate actual scenes of the offensive actions against the viewer (Scott et al., 1993; Williams& Elliott, 1999).

In recording, the three athletes performed separately. Facing the camera, the athletes first took a ready stance, then executed the offensive action against a virtual opponent standing at the camera position.

23

They demonstrated thrusting punches and kicks several times, from both left and right sides, each action being aimed either at the upper or the middle level. From the recording of their performance, 50 actions (half of them aimed at the upper level, the other half at the middle level) were chosen as the stimuli to be presented in the experiment.

Each stimulus was edited in a form of successive pictures, or frames, and presented at 33 ms per frame on the monitor.

the numbers ranged from 13 to 20 frames which corresponded to stimulus durations of 430–660 ms. They showed no systematic difference between the actions aimed at the upper level and at the middle level.

For the dot stimuli, a black filled circle (1.5 cm in diameter) was presented on a white background.

24

In the choice RT task, the circle was presented at either, approximately, 7 cm above or below the center of the screen

In the simple RT task, the circle was always presented at the center of the screen.

ProcedureThe experiment was conducted in a dark booth, where the participants were seated with their head fixed on a chin rest while viewing the CRT screen binocularly. There were four conditions, two stimulus types (video and dot)2 RT tasks (choice and simple).

In the video-stimulus choice-RT condition, each trial started with 1-s presentation of a fixation point at the center of the screen, followed by 1-s presentation of the first frame of a stimulus, showing a static display of the athlete’s ready stance.

25

the action started with successive presentations of the frames (beginningfrom the first frame) at the rate of 33 ms/frame. The participants were asked to decide, as soon and accurately as possible, whether the offensive action in the stimulus would come to the upper or the middle level of their body. The response was made on a response box, pressing one key by a finger of the right hand for the upper level and the other key by a finger of the left hand for the middle level (cf., Scott et al., 1993; Williams & Elliott, 1999).RTs were measured from the onset of the second frame to the key press. The stimulus presentation was terminated by the key press and replaced with a blank screenThere were five experimental sessions of 50 trials each, with the 50 different stimuli presented once in each session in a random order. For each participant, sessions with less than 90% correct responses were discarded and replaced with new sessions. As a result, four of the athletes group and three of the novice group had to perform one or two extra sessions

26

In the video-stimulus simple-RT condition, the stimulus presentation was identical to that of the choice-RT condition, except that the duration of the static ready stance display varied randomly from 1 to 3 s. This served as a random foreperiod as typically used in simple RT tasks (Luce, 1986).

The key press was made by either the left or right hand, in an approximately equal number of trials within a session. There were two sessions of 50 trials each, with random presentations of the 50 stimuli in each session.

In the dot-stimulus choice-RT condition, each trial started with an 1-s blank interval of white background, after which the dot was presented for 33 ms at either an upper or a lower location on the screen, followed by a blank white background.

27

The participants were asked to report the location as soon and accurately as possible, by pressing one key with the right hand for the upper location or the other key with the left hand for the lower location.

RTs were measured from the onset of the dot presentation to the key press. There were two sessions of 125 trials each.

In the dot-stimulus simple-RT condition, the duration of the blank interval varied from 1 to 3 s randomly across trials, after which the dot was presented for 33 ms at the center of the screen, followed by a blank screen.

The participant’s task was to press a key as soon as the dot was presented. This condition was employed in one session of 100 trials.

28

29

Results The most notable is the very slow RTs of the video-stimulus choice-

RT condition; they are more than twice as slow as RTs of any other condition.

The second slowest is the dot-stimulus choice-RT condition. On visual inspection there appears to be no difference between the two simple-RT conditions. For all the conditions, the mean RTs were faster for the athlete group than for the novice group.

For both groups, the choice RTs were significantly slower than the simple RTs in the two stimulus conditions

For both groups, the choice RTs were slower in the video than in the dot condition but there was no difference in simple RT

30

The choice RTs were faster for the athlete group than for the novice group in the video and the dot conditions, but the simple RTs did not differ between the two groups for either stimulus condition

In the choice RT conditions, proportions of correct responses (PCRs) for eithertype of stimulus were higher than 90% for each participant, and mean PCRs were identical for the two participant groups: both groups yielded 94.7% for the video stimuli and nearly 100% for the dot stimuli

The lack of difference in PCR between the two groups and the higher PCRs of both groups for the dot stimuli (with the faster RTs) than for the video stimuli clearly indicate the absence of contamination of speed-accuracy trade-off in the choice RT data.

31

Additional analysis First, possible learning effects on the choice RTs

for the video stimuli Second, differences due to the hands used for key

pressing

An inspection of individual data showed that the choice RTs tended to be faster for later sessions.

The lack of interaction indicates that the learning effects were not different between the two groups

32

the participants were all right-handed and used their right hands to respond to the upper location throughout the choice RT tasks, the RT data may have been contaminated by possible advantages of right-hand key presses and differences in the advantage between the athletes and the novices.

The choice RT data of both groups may have been contaminated by right-hand advantages of about 10 ms, indicated by their simple RTs, although the conclusions due to their differences in choice RT are not subject to the right-hand advantage, which did not differ between the two groups.

33

DiscussionDespite the relatively small number of participants in the athlete and the novice groups, the results are clear. The superiority of the athletes was evident in the video stimulus choice-RT condition. The athletes were faster but no more accurate than the novices, consistent with the results of Scott et al. (1993) with the verbal action mode. The approximately 100-ms advantage of the athletes is ascribed to their superior anticipation of the opponent’s attacking position

It thus seems that the novices, as well as the karate athletes, anticipated the target area of the opponent’s attack, although the 100-ms delay in choice RT reflects inferiority of the novices’ anticipatory skills.

34

The athletes were also slightly but significantly faster in the dot-stimulus choice-RT condition. This suggests that, in addition to the superior anticipatory skill, they have a better ability to make quick discriminations of two vertical locations.

Why the athletes are better at the discrimination, and whether their better discrimination is specific to vertical locations or ubiquitous in space and other dimensions (e.g., Nougier et al., 1992; Whiting & Hutt, 1972) are not clear.

One possibility is that the karate athletes have learned offensive and defensive techniques in terms of vertical levels (upper, middle, lower), so they may have become sensitive to vertical differences in space.

In contrast, the lack of significant difference in simple RT between the two groups, for either the dot or the video stimuli, implies that the athletes are no better than the novices in visual detection of a single dot or of any change in the static display of the opponent’s ready stance.

35

Experiment 2Method

The apparatus was the same as in Experiment 1. The stimuli were identical to the 50 video stimuli of Experiment 1, except that their presentation was cut off at the seventh frame.

This cut-off point was chosen in consideration of results of a preliminary experiment, which showed that the presentation of the first seven frames were sufficient for the karate athlete to make correct decisions in more than 90% of trials.

The time sequence of one trial was similar to that of Experiment 1 with the video stimuli, except that after the seventh frame, the stimulus was replaced with a blank screen.

After the stimulus presentation on each trial, the participants were given unlimited time to decide as accurately as possible whether the offensive action would be aimed at the upper or the middle level

36

Results and discussion

The mean PCR of the athlete group was 96% which was significantly higher than that of the novice group, 80%.This corroborates the results of Experiment 1 and supports that the faster RTs of the athletes in the video-stimulus choice-RT condition was due to their anticipatory skills, not to their decision making.

anticipatory skills consist of (at least) two processes; extracting information from the early motion of the opponent’s attack and estimating from that information the target area (upper or middle) of the attack.

37

In this experiment, the information extracted by the first process was limited to the first seven frames of the video stimuli.

That the mean PCR of the athlete group is close to their value in the video-stimulus choice-RT condition of Experiment 1 (94.7%; for the four black-belts, 95.2%) suggests that their extraction process in the choice RT task was also confined to around the first seven frames, even though the subsequent frames were presented in the experiment

On the other hand, the mean PCR of the novices was lower than their value in the choice RT task (94.7%, identical to the athletes’).

38

In conjunction with the 100-ms delay of the choice RT from the athletes, this suggests that the extraction process of the novices worked on the first 10 frames, which corresponded to three extra frames, or a 100-ms delay (33 ms per frame), from the seventh frame suggested for the karate athletes.

Accordingly, presenting only the first seven frames lowered their PCR in this experiment.

Therefore, the main difference between the athletes and the novices seems to be in the information-extraction process, and not in the estimation process, of their anticipatory skills.

39

additional pointsFirst, cutting off every stimulus at the seventh frame may not be adequate because, as noted in Experiment 1, the stimuli may differ in minimum viewing time (or frames) necessary for the correct decisions.

The termination at the seventh frame might be too short, or too long, for some of the stimuli.An alternative is to cut off each stimulus at the frame corresponding to its minimum viewing time. This is difficult to do in practice, however, because finding the minimum time for each of the 50 stimuli requires a series of experiments with careful manipulation of the viewing time and, even if we do this, the minimum viewing times would differ among participants.

40

Second, because the same participants had taken part in Experiment 1, their prior exposure to the same set of stimuli may have contributed to the result of this experiment.

One possibility is that the karate athletes may have better memory of karate-related events than novices, and the athlete’s clearer memory of the video stimuli, acquired from Experiment 1, may have helped their decisions in this experiment.

Better recalls of sport-specific items by expert players have been reported in karate (Hodge & Deakin, 1998) as well as in other sports(Allard et al., 1980; Williams & Davids, 1995), and the results of Experiment 1showed learning effects for the athletes as well as the novices.

Further research will be needed to explore the possible effects of specialized memory possessed by karate athletes.

41

General discussionExperiment 1 showed that:

the athletes responded faster than the novices in the video-stimulus choice-RT condition, which simulated the real stimuli and task of karate, indicating superior anticipation of the athletes regarding the opponent’s attack (Scott et al., 1993; Williams & Elliott,1999).

The athletes were also slightly but significantly faster in the dot-stimuluschoice-RT condition, which was probably due to their better ability of vertical discrimination

The athletes were not different from the novices in simple RT to either dot or video stimuli, suggesting that the athletes were no better in visual detection not specific to karate.

42

These results are consistent with a general consensus of sport research that the advantage of expert players is most evident in tasks with realistic settings that assess sport-specific perceptual skills, while their advantage is relatively small or non-existent in simplistic tasks used to examine basic sensory functions (Borgeaud & Abernethy, 1987; Helsen & Starkes, 1999;Starkes,1987).

Experiment 2 provided supporting evidence for the athletes’ superior anticipation.In the temporal occlusion paradigm, the athletes were more accurate than the novices in decision making regarding the opponent’s attack. This result not only supports that the athletes’ faster RTs in Experiment 1 was due to their anticipation based on the early part of the opponent’s attack, rather than to their decision making (Chamberlain & Coelho, 1993), but it also suggests that the main advantage of the athletes may be in extracting necessary information from the opponent’s motion.

43

These findings of the athletes’ anticipation fit well with practical aspects of karate, which place a great emphasis on the role of anticipation. Anticipation is particularly important in defence, for avoiding the opponent’s attack and for taking a proper position prior to the contact (Wilk et al., 1983), which reduces significantly the damage from the attack. Some training is focused on anticipating the opponent’s attack (Egami, 1976). It has been suggested that experience in training and games would enhance the athletes’ karate-specific knowledge base, resulting in their superior anticipation (Williams & Elliott, 1999)

some issues remain unanswered:

One concerns cues in the opponent’s action that are used by karate athletes to anticipate the target location of the attack.

Such cues may include changing positions of body parts (head, hands,feet etc.), concerting movements of those parts, or even gaze directions

44

Williams and Elliott (1999) recorded the participant’s eye movement during the RT task and found that both the athletes and the novices tended to fixate on the head and the chest of the attacking athletes in the video rather than on the arms and the fists.

They suggested that the athletes, as well as the novices, used the central regions (i.e., head and chest) of the opponent’s body as ‘‘visual pivots’’ to distribute their attention over peripheral regions (also see Ripoll et al., 1995). Although this is an interesting suggestion regarding search strategy for anticipatory cues, it is not yet entirely clear what spatial and temporal aspects of the opponent’s motion serve as reliable cues for quickly and accurately anticipating the location of the forthcoming attack.

45

In Experiment 2 of the present study, we proposed that the athletesextract critical information from the first seven frames of the video stimulus, while the novices required 10 frames. This proposal fits the present data but needs to be verified with further experiments.

Particularly, the exact number of frames necessary for them to reach a certain level of accuracy should be specified.

One way to answer these questions is to manipulate the spatial and temporal occlusions of the videotaped opponent’s actions in the occlusion paradigm, which has been used in studiesof other sports to identify advance cues utilized by expert players (e.g., Abernethy, 1990; Abernethy & Russell, 1987; for a review, see Williams et al., 1999).

46

Another issue is the differences between RTs in laboratory settings and actions in real situations. In Experiment 1, the athletes’ choice RTs to the video stimuli (mean 552 ms) are comparable to the completion time of videotaped offensive actions (430– 660 ms), which means that their reactions would be too slow in real defensive situations.

This points to important differences between the laboratory settings of the present study and the conditions of real situations.

One is the response mode to the opponent’s attack.In the present study the athletes responded by a key press, while in real situations they would make defensive actions (blocking), which would be more natural for them to act against the opponent’s attack and probably faster than the key press.

The other is the presentation sequence of the offensive actions. In the present study, as well as in the study by Williams and Elliott (1999), the participants were randomly presented with separate actions of the three opponents.

47

This is quite different from real situations, where each athlete is faced with one opponent for some period of time, so that the previous actions of the opponent may provide useful information for forthcoming attacks and facilitate defensive actions in terms of speed and accuracy

Another possibility is that in real situations, it may be too late once an opponent starts an offensive action.

It is emphasized in the practice of karate to keep a proper distance (‘maai’) from an opponent, in order to prevent the opponent from making attacks.

Anticipation alone may be insufficient and need to be coupled with other defensive strategies, such as keeping a proper distance.

These points will be investigated in field studies or in laboratories with more realistic settings.

48

Visual search, anticipation and expertise in soccer

goalkeepersGEERT J.P. SAVELSBERGH,1,2* A. MARK WILLIAMS,3 JOHN VAN

DER KAMP1 and PAUL WARD3

1Research Institute for Fundamental and Clinical Human Movement Sciences, Vrije Universiteit, Amsterdam,The Netherlands, 2Centre for Biophysical and Clinical Research into Human Movement, Department of Exerciseand Sports Science, The Manchester Metropolitan University, Alsager, UK and 3Research Institute for Sport andExercise Sciences, Liverpool John Moores University, Liverpool, UK

Journal of Sports Sciences, 2002, 20, Journal of Sports Sciences, 2002, 20, 279± 287

49

Expert soccer goalkeepers demonstrate highly skilled and well coordinated behaviour when diving or jumping to catch the ball. Such skilled behaviour requires many years of practice (Ericsson et al., 1993), allied to a considerable amount of ability (Singer and Janelle, 1999).

It is now accepted that successful performance in such sports requires skill in perception as well as the efficient and accurate execution of movement patterns (see Williams et al., 1999).

A player’ s ability to use advance postural cues is particularly important in fast ball sports where the speed of play and ball velocity dictate that decisions must often be made in advance of the action.

High-speed film analysis indicates that players who react to the ball, as opposed to anticipating its intended destination, are unlikely to be successful (Glencross and Cibich, 1977).

50

In most previous studies, static slide displays have been used (e.g. Tyldesley et al., 1982) and participants were required to perform discrete rather than continuous actions in response to these stimuli (e.g. Williams and Burwitz, 1993; Franks and Hanvey, 1997).

The information contained in static displays has been shown to be less sensitive and discriminating than that provided by dynamic film or live models (Bourgeaud and Abernethy, 1987).

In particular, the relative motions presented in dynamic action sequences may be crucial to anticipation in sport (Abernethy et al., 2001; Ward et al., 2002). Similarly, the typical response requirements have varied from pressing a button (Tyldesley et al., 1982; Neumaier et al., 1987; Franks and Hanvey, 1997) to written (McMorris et al., 1993; Williams and Burwitz, 1993; McMorris and Colenso, 1996) or verbal responses (Salmela and Fiorito, 1979).

51

In such circumstances, corrections or modifications to the response are not possible as the display unfolds, unlike the real performance setting where players have the opportunity to modify their actions continuously.

The ability to make ongoing corrections has gained increasing importance since the recent rule change that allows goalkeepers to move across the goal line before the ball is struck by the penalty taker.

The present study embraces new technology in an attempt to determine key differences in visual search behaviour and anticipatory performance between expert and novice soccer goalkeepers.

The study was undertaken from a perception-action perspective, where information is presumed to evolve over time, and where action is continuously coupled to the perceptual information presented (see Savelsbergh and Van der Kamp, 2000).

52

In this innovative paradigm, visual information is picked up in a continuous rather than discrete fashion and the response is not measured by a button press, but by means of a joystick linked to a potentiometer to ensure continuous data sampling.

This procedure allows corrections to be made to the response in an ongoing manner as the flow of information changes across early and late periods in the penalty kick.

53

MethodsParticipantsFourteen players provided informed consent before participating in the study. The expert group consisted of seven soccer goalkeepers aged 29.9 ± 7.1 years who had played semi-professional football (second division of the National League) in the Netherlands for a minimum of 10 years. The novices included seven goalkeepers aged 21.3 ± 1.4 years who had played soccer less frequently for recreation.

Test film

Ten professional youth players aged 18.9 ± 1.5 years were filmed from the goalkeeper’s perspective taking penalty kicks. The film clips were recorded using a digital video camera positioned in the middle of the goal at a height of 1.77 m

54

A sailcloth (2.42 ´ 1.50 m) was hung from a regulation crossbar to indicate the area to which the players had to shoot. Six different target areas (0.81 ´ 1.50 m) were painted on the sailcloth, as highlighted in Fig. 1.

Fig. 1. The goal divided into six areas for placement of penalties and joystick movements.

55

The players were told to try and disguise the intended destination of the penalty kick, as they would in normal competition.

Each film clip included the penalty taker’ s approach to the ball, their actions before and during ball contact, and the first portion of ball flight.

Two penalties were recorded in each target location for every player, providing a total of 120 trials.

A microphone was attached to the sailcloth to indicate the moment at which the ball crossed the goal line; a second microphone was positioned near the penalty spot to record the moment of ball± foot contact.

These two temporal measures were used to calculate the flight time andvelocity for each penalty kick. The average ball flight time was 648 ms, whereas the mean ball velocity was 16.8m´s-1.

56

ApparatusThe film clips were back-projected (EIK CC-7000),using a reflective surface to increase image size, onto a large screen (2.29 ´ 2.27 m) positioned 3.45 m from the participant. The experimental layout is represented graphically in Fig. 2

Fig. 2. A side view of the experimental set-up.

57

The image of the penalty taker subtended a visual angle of approximately 8° atfoot± ball impact, thereby closely simulating the real image size and distance between the goalkeeper and the penalty spot.

The response movements performed by the participants were recorded using a hand-held joystick.

The joystick (Dual Axis Farnell M11Q61P) was positioned at waist height just in front of the participant. The joystick signal was amplified and stored on computer by means of LABVIEW (version 5.1).

Visual search behaviours were recorded using an eye± head integration system that included an Applied Science Laboratories (ASL) 4000SU eye-tracker and an Ascension Technologies magnetic head-tracker (model: 6DFOB).

58

They had to anticipate the direction of each penalty kick quickly and accurately by moving the joystick as if to intercept the ball.

Procedure

The participants were allowed to use the joystick to make corrections to their initial decision as the penalty kick evolved.

Before the penalties were presented, a non-task-specific test was undertaken to determine whether there were `baseline’ differences in simple reaction time between the two groups.

59

After familiarization and habituation, 30 film clips were presented; five penalties in each location.

Half of the film clips involved a left-footed penalty taker; the remaining trials involved right-footed players.

60

Dependent variables and analysis

Penalties saved

Correct side

Correct height

Proportion of corrections.

Time of initiation of joystick movement.

Reaction time.

61

ResultsExperts Novices

Penalties stopped 35.7 ± 11.8 25.9 ± 10.8

Correct height 42.6 ± 8.9 32.6 ± 8.2

Correct side 83.8 ± 11.8 71.4 ± 8.2Proportion of corrections

26.3 ± 4.9 38.5 ± 15.3

Time of initiation of joystick movement (ms)

296 ± 46.6 480 ± 29.2

Reaction time (ms) 258 ± 33.1 237 ± 46.4Table 1. The dependent measures recorded on the

anticipation test across groups (mean ± s)

62

Discussion

As predicted, the expert goalkeepers generally demonstrated better performance on the anticipation test than their novice counterparts.

The experts were also more accurate than the novices in predicting penalty kick side (83.8 vs 71.4%) and height (38.5 vs 26.3%).

Both groups of players were more accurate in predicting the correct side rather than height of the penalty kick (77.6 and 37.9%, respectively).

63

The expert goalkeepers initiated their joystick movements later or nearer to foot± ball contact than the novices and made corrections on fewer trials

Rodrigues et al. (1999) showed that skilled table tennis players are able to `buy extra time’ relative to less skilled performers by taking the ball later during its approach, compensating for this delay by adopting shorter movement times

64

The novices spent longer fixating on the trunk, arm and hip regions as the penalty kick evolved, whereas the experts preferred to fixate their gaze on the kicking leg, non-kicking leg and ball areas.

The expert goalkeepers also spent longer fixating on the head region, particularly early on, although this may reflect a tendency to try and recognize facial characteristics early in the action sequence.

Franks and Hanvey (1997) argued that the non-kicking foot is positioned so that it points towards the ball’s likely destination, while Williams and Burwitz (1993) proposed that the angle of the foot relative to the ball during the downswing phase provides a strong indication of intended ball placement.

65

both groups of players spent long periods fixating on `unclassified’ areas of the display.

This suggests that participants chose to anchor the fovea close to these key locations so that they could use the parafovea and visual periphery to pick up relevant information.

The effective use of such `visual pivots’ by experts has been demonstrated in a variety of sports (see Ripoll et al., 1995; Williams and Davids, 1998; Williams and Elliott,1999).

66

Another important area for future consideration is whether goalkeepers’ anticipatory performance at penalty kicks can be improved through perceptual training.

Video simulation may prove particularly effective as a method of developing perceptual skill, particularly when coupled with appropriate instructional techniques

(McMorris and Hauxwell, 1997; Williams and Burwitz, 1993)

67

Expert goalkeepers have superior perceptual skills to their novice counterparts.

Experts are quicker and more accurate in anticipating the likely destination of a penalty kick and systematic differences in visual search behaviours were apparent across groups.

although the paradigm used in this study is more realistic than in previous research, further innovations are required to accurately simulate the performance setting.

These innovations could include whole-body response measures and the manipulation of emotional states such as anxiety and motivation.

68

Testing and Training Anticipation Skills in

Softball Fielders

Tim Gabbett1, Martin Rubinoff1, Lachlan Thorburn1

and Damian Farrow2

1Queensland Academy of Sport,2Australian Institute of Sport, Canberra Australia

International Journal of Sports Science & Coaching Volume 2 · Number 1 · 2007

69

70

71

72

73

74

Studies have shown that the anticipatory skill of team sport athletes can be learned through the use of video-based perceptual training.

Introduction

However, while several studies have assessed anticipation skills in the laboratory, few studies have determined whether these improvements in anticipatory skill transfer to performance improvements in the real-world environment of a game-specific situation.

A measure of transfer is essential to determine whether the improvements observed in the laboratory setting actually transfer back to the game situation.

75

Studies often neglect to perform a retention test to determine whether improvements in anticipatory skill can be retained after the training stimulus is removed.

Softball is a sport that requires superior anticipatory skill, both from an offensive (i.e,pitch recognition) and defensive (i.e., fielding) perspective.

while several studies have investigated perceptual skill of these athletes from an offensive perspective (i.e., pitch recognition), no study has trained anticipation skills in softball fielders in an attempt to improve the ability to identify specific cues of a batter that might indicate the direction that the ball may be hit.

76

The Purpose of this study:

develop an ecologically valid anticipation test for softball in-fielders

assess the validity and reliability of the anticipation test

Determine whether a video-based perceptual training stimulus could improve anticipatory skill in softball fielders,

whether these improvements in anticipatory skill transferred tothe field environment.

77

MethodTesting was performed in three parts:

Firstly, the validity of the test was determined bytesting national, state, and club level softball players, and by comparing results across different playing positions.

Secondly, the reliability of the test was determined.

Finally, players participated in a four-week training program to determine if fielding anticipation skills could be learned through the use of a video-based training stimulus.

78

SOFTBALL ANTICIPATION TEST

The athlete’s decision accuracy and decision time were measured in the laboratory using a ‘life-like’ video-based test of a batter shown from the perspective of the in-fielder.

For the softball fielding anticipation test, a data projector (NEC LT245) was positioned on the floor 5 m back from a wall. A video image of a batter was projected on to the wall, so that the image of the player was approximately 1.80 m tall and reflective of a typical elite softball player’s physical stature.

Six occlusion periods were used: t1 (-120 ms), t2 (- 80ms), t3 (- 40ms), t4 (0 ms), t5 (+ 40 ms), and t6 (+ 80 ms).

79

A high-speed video camera interfaced with a video recorder, was positioned 5 m behind the subject in order to record the subject’s change of movement direction relative to the moment of ball contact (display occlusion).

Video was sampled at 200 Hz so that the number of frames between the moment of display occlusion and the player’s movement initiation enabled the player’s decision-making time to be recorded to within ± 5 ms for each trial.

80

RELIABILITY AND VALIDITY

Ten state players performed the anticipation test on two occasions, two days apart, to determine the reliability of the test.

The validity of the anticipation test to discriminate among players of different playing ability was evaluated by testing national (n = 21), state (n = 10) and novice (n = 9) softball players.

In addition, the validity of the anticipation test was determined by comparing different playing positions (in-fielders, pitchers, and out-fielders).

81

TRAININGTwenty-five elite female softball players were randomly allocated to: (i) a video-based perceptual training group (n = 9); (ii) a placebo training group (n = 8); (iii) a control group (n = 8).

The perceptual training group watched temporally occluded video footage focused primarily on identifying specific cues of a batter that might indicate the direction that the ball may be hit

82

Athletes performed 30 trials each session. At the commencement of each session, athletes were instructed to focus their attention on a particular aspect of the batter (e.g. hands, hips, or bat head, etc).

During training sessions, athletes watched the occluded video once, and then the same un-occluded video immediately afterwards, to provide feedback on where the ball was actually hit. Twelve training sessions, each of 10 minutes duration, were conducted over a four-week period between the pre- and post-testing sessions.

The placebo training group watched a series of arrows that randomly directed them to move left or right.

Athletes were instructed to make a small movement in the direction the arrow was pointing as soon as they saw the arrow.

The control group performed no video-based training in the intervention period.

83

TRAINING TRANSFER TEST

The athlete’s decision accuracy and decision time were measured in the laboratory pre- and post-intervention using the ‘life-like’ video-based test.

In addition, the athlete’s decision accuracy and decision time were measured in the field pre- and post-intervention to determine if anticipation skills could be transferred to the field setting.

RETENTION TESTThe athlete’s decision accuracy and decision time were measured in the laboratory and the field following a four-week period where no video-based training was performed. (i.e. retention test).

84

RESULTSRELIABILITY

The Test and Re-Test measurements of decision accuracy and decision time proved to be reliable.

VALIDITY

The decision accuracy and decision time results for the national, state, and club level softball players are presented in Figure 1.

85

86

87

CHANGES WITH VIDEO-BASED PERCEPTUAL TRAINING

88

Figure 3. Fielding Anticipation Skills for the Video-Based Perceptual Training Group, Placebo Group, and Control Group, Before Training, After 4 Weeks of Training, and Following a 4-week Non-Training (Retention) Period.

The laboratory test results are presented in panel A, while the field test results are presented in panel B. Data are reported as mean ± SE.

* Significantly different (P<0.05) from pre-training.† Significantly different (P<0.05) from placebo and control group. ‡ Significantly different (P<0.05) from control group.

89

DISCUSSIONthese results demonstrate that video-based perceptual training improves decision accuracy and decision times in a game-specific task, with these improvements in anticipatory skill transferring to the field environment.

These findings suggest that video-based perceptual training should be used in conjunction with field training to enhance the anticipatory skill of softball fielders.

The finding of superior anticipatory skill in the national level players of the present study suggests that these players had a greater ability to extract relevant information earlier in the visual display, by identifying relevant postural cues presented by the batter, and disregarding irrelevant sources of information.

90

the placebo group showed no meaningful changes in decision accuracy or decision times.

These findings demonstrate that the use of a non-sport specific stimulus (i.e. directional arrows) that randomly direct players to move left or right does not improve the anticipatory skill of softball fielders, and is consistent with previous perceptual training research that has highlighted the importance of providing a sport specific stimulus when attempting to improve an athlete’s perceptual skill.

Video-based perceptual training improved fielding anticipation skills by improving both decision accuracy and decision time in the laboratory task.

These findings are in agreement with others who have reported improvements in perceptual skill following short-term training in tennis, squash, baseball, American football, soccer, volleyball, and basketball players.

91

While decision accuracy of the video-based perceptual training group improved from pre-training (74.1%) to post-training (93.8%), no meaningful changes were detected for decision times.

These findings suggest that more than 4 weeks of training may be required to elicit improvements in decision time in softball fielders.

Furthermore, while we found that these improvements in anticipatory skill were retained following a four-week, non-training period, without continual training, no further improvements were made.

These findings suggest that regular perceptual training is required to elicit further improvements in anticipatory skill in softball fielders.

92

Thanks For Your Attention