A Three Dimensional Comparison of Elite and Recreational Ice Hockey Siap Shots

Timothy Keith Woo 119735479

A Thesis Submitted to the Faculty of Graduate Studies and Research in Partial Fulfillment of the Requirements for the Degree:

Masters of Science

Department of Kinesiology and Physical Education

Division of Graduate Studies and Research Faculty of Education McGiII University Montreal, Quebec Canada

July 2004 Library and Bibliothèque et 1+1 Archives Canada Archives Canada Published Heritage Direction du Branch Patrimoine de l'édition

395 Wellington Street 395, rue Wellington Ottawa ON K1A ON4 Ottawa ON K1A ON4 Canada Canada

Your file Votre référence ISBN: 0-494-06471-4 Our file Notre référence ISBN: 0-494-06471-4

NOTICE: AVIS: The author has granted a non­ L'auteur a accordé une licence non exclusive exclusive license allowing Library permettant à la Bibliothèque et Archives and Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par télécommunication ou par l'Internet, prêter, telecommunication or on the Internet, distribuer et vendre des thèses partout dans loan, distribute and sell th es es le monde, à des fins commerciales ou autres, worldwide, for commercial or non­ sur support microforme, papier, électronique commercial purposes, in microform, et/ou autres formats. paper, electronic and/or any other formats.

The author retains copyright L'auteur conserve la propriété du droit d'auteur ownership and moral rights in et des droits moraux qui protège cette thèse. this thesis. Neither the thesis Ni la thèse ni des extraits substantiels de nor substantial extracts from it celle-ci ne doivent être imprimés ou autrement may be printed or otherwise reproduits sans son autorisation. reproduced without the author's permission.

ln compliance with the Canadian Conformément à la loi canadienne Privacy Act some supporting sur la protection de la vie privée, forms may have been removed quelques formulaires secondaires from this thesis. ont été enlevés de cette thèse.

While these forms may be included Bien que ces formulaires in the document page count, aient inclus dans la pagination, their removal does not represent il n'y aura aucun contenu manquant. any loss of content from the thesis. ••• Canada Table of contents

Acknowledgements ...... 3

List of tables and figures ...... 4

Abstract ...... 6

Résumé ...... 7

Chapter 1: Introduction 1.1 Rational ...... 8 1 .2 Objectives and hypothesis ...... 10 1.3 Nomenclature ...... 10 1.4 Technical and theoretical considerations ...... 14 1.5 Study Limitations ...... 15 1.6 Delimitations ...... 16 1.7 Study design ...... 16

Chapter 2: Review of Literature 2.1 Ice hockey slap shots ...... 17 2.2 Kinematic analysis methods ...... 21 2.3 Three dimensional (3D) kinematics ...... 23 2.4 Joint centers ...... 24

Chapter 3: Methods 3.1 Subjects ...... 27 3.2 Testing procedures ...... 27 3.3 Joint center calibration setups ...... 28 3.4 Joint center calculations ...... 30 3.5 Data reduction ...... 34

Chapter 4: Results 4.1 Stick kinematic results ...... 36 4.2 Body kinematic results ...... 39 4.3 Swing timing and sequencing ...... 41

Chapter 5: Discussion ...... 44

References ...... 53

Appendices A - Graphs of lead shoulder kinematics ...... 56 B - Graphs of trail shoulder kinematics ...... 57 C - Graphs of elbow shoulder kinematics ...... 58 o - Graphs of pelvis and trunk rotation kinematics ...... 59

1 E - Graphs of stick displacement in frontal, sagittal and transverse planes ..... 60 F - Graphs of blade velocities ...... 61 G - Graphs of the stick rotation center distance from heel of stick ...... 62 H - Table of MANOVA values for measures ...... 63

2 Acknowledgements

1would like to acknowledge several people who have made significant contributions to the completion of this master thesis. 1would first like to thank my thesis supervisor professor David Pearsall for ail the help he has given me. His suggestion and guidance helped me tremendously throughout the whole process. Without his help 1would have not been able to finish this endeavor.

Secondly, 1would like to thank Mr. J.J. Loh for the help he provided in the

Matlab® programming. His brilliant work saved me a enormous amount of time in the processing and analyzing of the data. Aiso 1would like to thank Ms. Kristin

Leuszler and Mr. Joel Bergeron for helping me find subjects for this study. would also like to thank Professor Rene Turcotte for his suggestions and feedback. Furthermore, 1would like to thank Brad Leonard for his help in translation the French abstract. Finally 1would like to extend my deepest gratitude to ail my colleagues, friends and most of ail my family for their unconditionallove and support throughout the entire process.

3 List of tables and figures

Figure 1.01 Lead and trail side of the body

Figure 1.02 Shaded plane indicates the sagittal plane

Figure 1.03 Shaded plane indicates the frontal plane

Figure 1.04 Shaded plane indicates the transverse plane

Figure 1.05 Movement descriptions of the shoulder and elbow. A indicates vertical shoulder movements of abduction and adduction. B indicates horizontal shoulder movements of abduction and adduction. C indicates movements at the elbow of flexion and extension.

Table 2.01 Movements regarding optimal technique of the body during the ice hockey slap shot of as summarized from various coaching manuals and books. (Walker and Dann 2002; Smith 1996; Meeker 1975; Almstedt 1974)

Figure 3.21 Movements do ne by subject to calibrate joint center. A movements performed to find shoulder joint centers. B Movements performed to determine elbow joint center. C Movements performed to determine wrist joint center

Figure 3.23 Points used to calculate the center of the sphere. The intersects of the radial vectors normal to the tangential movement of the distal markers, were determined by least squared optimization.

Figure 3.41 Screen shot of fimbox®program used to collect electromagnetic sensor co-ordinates. Views shown (from top left corner, clockwise): Oblique, sagittal, frontal and transverse.

Figure 3.42 Screen shots from Matlab® program used to calculate joint and stick kinematics. A - screen shot showing oblique view. B - Screen shot showing sagittal view. C - Screen shot showing frontal plane view. o - Screen shot showing transverse plane view.

Figure 3.51 Global co-ordinate system of the setup

Table 4.01 The peak width, depth and height of the blade positions during the slapshot

Table 4.02 Blade velocities of elite and novice players at impact.

4 Table 4.03 Resultant stick angular displacement, velocity and acceleration at impact

Table 4.05 Global center of rotations x, y, z

Table 4.06 Center of rotation position on stick measured from heel of blade

Table 4.07 Stick orientations in the sagittal, frontal and transverse planes

Figure 4.20 Graph and table of elite vs recreational kinematic values of the lead arm at the start, top of swing and impact.

Figure 4.21 Graph and table of elite vs recreational kinematic values of the trail arm at the start, top of swing and impact.

Figure 4.21 Graph and table of elite vs recreational kinematic values of the trunk and pelvis rotations at the start, top of swing and impact.

Table 4.30 Percentage of the backswing part of the slap shot

Table 4.31 ANOVA table of backswing percentage

Table 4.32 Table of the times required for the backswing and downswing phase of the slap shot.

Figure 4.30 Sequence of peak joint angular velocities for elite and recreational players during the downswing phase of an ice hockey slapshot.

Figure 5.01 Stick velocities of elite and recreational players deconstructed.

Figure 5.02 Examples of elite and recreational trace downswings during a stationary ice hockey slap shot.

Figure 5.03 Figures showing movement path which (A) allows for a greater margin for error as weil as increased accu rate projection of the puck as (B) opposed to the more curvilinear motions of the stick (adapted from Kreighbaum and Barthels 1996)

Table 5.01 Elite versus novice variance (standard deviation) at the start, top and impact point of the slap shot.

Table 5.02 The position and movement of the upper body during a stationary ice hockey slap shot based on the data observed in this study

5 Abstract

The purpose of this thesis was to examine the three dimensional kinematic differences between elite and recreational ice hockey players while performing a stationary slap shot. Ten subjects, five elite-Ievel players and five recreational players, each performed five stationary ice hockey slap shots. Data were co"ected using the Ultratrak® electromagnetic system (Polhemus Inc.,

Burlington, VT, USA) at 60 Hz. Kinematics of the torso, arms and hockey stick were examined using a multiple analysis of variance (MANOVA). The results indicated that: 1) the elite subjects shot significantly (p < 0.05) faster due to the translational movement aspect of the stick; 2) the proximal to distal kinematic chain sequence of the elite subjects was better than the recreational subjects; and 3) the elite subjects showed less variability in stick movement within groups, than the recreational subjects. Further studies are needed to address kinematics of the lower limbs and of different ice hockey stick ski"s.

6 Résumé

Cette thèse avait pour but d'examiner les différences cinématiques tridimensionnelles entre les joueurs d'élite et les joueurs récréatifs lors de l'exécution en position stationnaire d'un lancer frappé de hockey sur glace.

Chacun des dix sujets, dont cinq joueurs d'élite et cinq joueurs récréatifs ont exécuté cinq lancers frappés stationnaires. Les données ont été recueillies à l'aide du système électromagnétique Ultratrak® (Polhemus Inc., Burlington, VT,

États-Unis) à 60Hz. On a examiné la cinématique du torse, des bras et du bâton de hockey, en utilisant une analyse de variance multiple (MANOVA). Selon les résultats obtenus, les lancers frappés des sujets d'élite étaient considérablement plus rapides en raison de l'aspect de mouvement de translation du bâton (p <

0.05). En outre, la séquence de mouvement, c'est-à-dire l'ordre proximal à distal de la vitesse de crête d'un axe, des sujets élites était plus coordonné que celui des sujets récréatifs. Enfin, les sujets élites ont montré moins de variabilité de mouvement de bâton à l'intérieur des groupes par rapport aux joueurs récréatifs.

Des études approfondies doivent aborder le sujet de la cinématique des membres inférieurs ainsi que diverses habiletés de bâton de hockey sur glace.

7 Chapter 1: Introduction

1.1 Rational

The stick is a distinguishing feature of the game of ice hockey. Over the years, the stick has evolved in geometric dimensions, materials, manufacturing and technical innovations (Dowbowgin, 2001). In tandem with skating, stick work skills are integral to the sport, these include: skating, positional play, shooting, checking, puck control, passing, hockey sense, desire/attitude, toughness/aggressiveness and strength (Renger, 1994).

Participation in ice hockey in Canada is very high. Over 538000

Canadians participate in organized ice hockey leagues and over 4.5 millions are involved with organized hockey in sorne fashion (Hockey Canada report, 2003).

Though not the official national game (that being lacrosse), in the minds of most

Canadians it is part of our national pastime.

ln ice hockey the primary tool for puck control and shooting is the ice hockey stick that is required to put the puck in the opposing team's net. There are several ways to shoot the puck: the two most common methods being the wrist shot and the slap shot. The wrist shot is generally accepted as a more accu rate technique for puck projection where as the slap shot's advantage is higher puck release velocities (Le. typical maximal velocities of 20 and 30 mIs for wrist and slap shots, respectively, Pearsall et al, 1999). Though numerous coaching manuals present general guides for technique development (Gendron,

1957; Randy, 1956; Haché, 1970; Dowbiggin, 2001), beyond the emphasized

8 repetitive practice drills to improve both accuracy and speed of the shot, the ability to define the underlying mechanics of the shot is relatively unknown.

ln the NHL ail star weekend, the slap shot contest is one of the most exciting events to watch. AI Maclnnis has won the event five of the last seven years as of 2003 (www.NHL.com).lsit because he is much stronger than everyone else or is it because his technique is better? This however, begs the question, of how is this skill performed? What is the difference between

Maclnnis and a novice player? There have been many coaching manuals and books (Gendron, 1957; Randy, 1956; Haché, 1970; Dowbiggin, 2001) describing shot techniques and every av id hockey fan can probably describe how to do either shot. Such anecdotal descriptions are common, but to date there has been very little quantifiable data regarding the three dimensional (3D) kinematics of the hockey slap shot describing the gross motor co-ordination patterns.

ln the never-ending search for a more competitive edge, there has been a push to use new equipment designs, mate rials and construction methods to enhance sport performance. In most sports, both kinematics and kinetics are rarely measured although they are most relevant to increase the desired performance of the equipment used. Biomechanical analysis of how the equipment is used may lead to optimal performance. Without an in depth knowledge of the stick function, how can a stick be designed or redesigned to increase the performance? Part of the answer can be revealed by a detailed 3D kinematics analysis. This can help with the design of not just hockey sticks, but of ail other sports equipment used, as an example, developing shoulder pads to

9 both protect the player and allow for the full range of motion needed to perform a slap shot.

Given that there is a desire for players, coaches and trainers to understand the optimal body joint coordination and techniques of an elite slap shot, and furthermore the desire to develop superior equipment, studying the 3D kinematics of a slap shot is both relevant and important.

1.2 Objectives and hypothesis

The objective of this study was to compare gross movement patterns between elite and recreational ice hockey players performing a stationary slap shot, more specifically; kinematics, range of motion, linear and angular velocities and segment/joint co-ordination. It is hypothesized that of the upper body, trunk and stick kinematics, differences will be found between the elite and novice groups; more specifically,

1. Significant differences will be found in gross movement patterns

between elite and novice groups;

2. Greater variability in gross movement patterns will be found in novice

group; and

3. Greater stick velocities will be found among the elite group

1.3 Nomenclature

• Left handed shot - Holds the stick with the right hand near the top of the

stick and the left hand near the middle of the stick.

10 • Right handed shot - Holds the stick with the left hand near the top of the

stick and the right hand near the middle of the stick.

• Lead side - The side of the body closest to the target at set up; left side

for a left handed shot and right side for a right handed shot. (Figure 1.01)

• Trail side - The side of the body farthest to the target at set up; left side

for a right handed shot and right side for a left handed shot. (Figure 1.01)

Trail side

Figure 1.01 Indicates lead and trail side of the body

• Sagittal plane - This global plane is orientated parallel with respect to the

net opening and 90° to the forward orientation of the subject. (Figure

1.02)

Figure 1.02 Shaded plane indicates the sagittal plane

11 • Frontal plane - This global plane is orientated at 90° to the net opening

and comprises most of the stick displacement. (figure 1.03)

Figure 1.03 Shaded plane indicates the frontal plane

• Transverse plane - This global plane is parallel to the floor surface.

(Figure 1.04)

Figure 1.04 Shaded plane indicates the transverse plane

• Shoulder vertical/horizontal ab/adduction and elbow flexion/extension­

Local joint angles were deconstructed from 3D movements with respect to

determined joint centers. (Figure 1.05)

12 A

••

c

Figure 1.05 Movement descriptions of the shoulder and elbow. A indicates vertical shoulder movements of abduction and adduction. B indicates horizontal shoulder movements of abduction and adduction. C indicates movements at the elbow of flexion and extension.

13 1.4 Technical and theoretical considerations

Two dimensional kinematic analyses of movements in sport have been widely studied. Three dimensional analyses are becoming more common with the advances in both hardware and software technologies that have made it more practical. However, both presentation and interpretation of the 3D kinematics are not intuitive. Euler angles are, for instance, commonly used to describe 3D angular orientation. However, describing euler angles are not easily understood in layman or body anatomical terms (Allard et aL, 1995; Grood &

Suntay 1983). Previously, 3D kinematic analysis techniques have mainly been restrained to optical methods. This technique involves the use of two or more cameras, surface light reflective passive markers on the body regions of interest, and a calibrated area. Strengths and weaknesses to this technique are discussed further by Niggs Cole and Wright (1998). In brief, practical problems such as video camera calibration and arrangement for volume during testing and most critically the time needed to digitize the digital videos (Le. extract x, y, z coordinate positions of the body and implement) are common to most optical setups. Technological advances in software have helped minimize these problems, such as semi and automatic marker detections. From the other perspective, using film or video images facilitates categorization of movements. ln the end, for effective communication, both qualitative and quantitative results need to be concurrently juxtaposed.

Recent advances in both software and hardware have allowed researchers to analyze kinematics using different techniques (Allard et aL, 1995).

14 One su ch recent advancement has been the use of electromagnetic tracking systems to measure kinematics. A major advantage of using this type of system in comparison to conventional optical and infrared systems is that the data received from each sensor provides both the position and orientation, thereby reducing the number of markers needed.

A widely used technique to measure kinematics has been to place surface markers over the joints and segments of interest to determine its position.

Although this makes the collection of data relatively easy, it may not necessarily depict the true movement of the joint in question. Thus, it is essential is to determine the joint centers to measure body kinematics.

1.5 Study Limitations

Experimentation was performed in the biomechanics laboratory at McGiII

University, hence some of the known limitations of the test context include:

0 0 1. The experiment took place at room temperature (22 - 24 C)

instead of ice rink temperature.

2. The subjects performed the slap shot on a polyethylene sheet (Le.

an artificial ice surface) instead of on ice.

3. The subjects did not wear their full gear (Le, no shoulder pads,

elbow pads, hockey pants, shin guards, jock strap, neck guard and

helmet) but did use their own skate and gloves.

4. In terms of physiological responses, the experiment was not

performed under a real game situation.

15 5. Ali the slap shots were performed stationary.

6. The target net was only 3.34m away.

7. Only one right and one left handed of ice hockey stick was used

(Le. wood laminate Bauer Supreme 3030) for ail subjects

8. The subjects were attached to a harness that guided the sensor

wires to the computer terminal.

1.6 Delimitations

The following lists the delimitations of the study:

1. The elite subjects were recruited fram the McGill University male

Ice Hockey team.

2. The recreational players were recruited fram the McGill University

Kinesiologyand Physical Education pragram.

3. Subjects ranged fram 18 - 30 years old.

4. Only male subjects were used in the study

5. Only the slap shot was measured.

1.7 Study design

This study used the skill level of ice hockey players as the independent variable. The dependant variables were the upper body and stick kinematics. A

Multiple analysis of variance (MANOVA) was used for statistical analysis. The specific dependant variables will be defined in the methodology section 3.5.

16 Chapter 2: Review of Literature

The following sections will review 1) prior literature regarding ice hockey shooting in addition to discussion of the various technological and 2) methodological details that need to be considered in three dimensional evaluation.

2.1 Ice hockey slap shot

There are many books and coaching manuals that describe the ice hockey

slap shot. These, like in most other sports are based on the experiences and

observations of the authors (Almstedt,1974; Meeker, 1975; Smith, 1996; Walker

and Dann, 2002). The following table summarizes fundamental qualitative

observations. (table 2.01)

Table 2.01 Movements regarding optimal technique of the body during the ice hockey slap shot as summarized from various coaching manuals and books. 'Walker and Dann 2002; Smith 1996; Meeker 1975; Almstedt 1974) Segment Initial position Backswing Top of Downswing Impact backswing Trunk Rotated to the Rotate trunk away Trunk should Trunk rota tes Trunk should target about 30· from target to about still be f1exed towards the target continue to rotate Trunk f1exed 75-80· reaching about 90· towards the target about 45· rotations in the as contact is made directions of the target Lead arm Elbow is bent Elbow straightens Arm pulls the Arm stays by side about 90· out. horizontally pommel down and of the trunk acting adductand towards the target like a pivot point for elevates to si de of the trunk to the stick shoulder level about the mid trunk level Trail arm Elbowfully Elbow f1exes to Arm pushes the Arm pushes the extended or just about 90·. stick towards the stick forward slightly bent shoulder ground and target "though" the puck horizontally "Ioading" the stick while pronating the abducts and forearm externally rota tes Stick Blade fiat on the Blade moves first Stick should The blade of the Stick should bend ice surface. away from the be almost stick makes a as it hits the ground Pommel ahead target. then up until vertical in circular motion traveling along the of the puck the stick is almost position downwards ground until contact vertical towards the puck is made with the making contact puck. with the ground about8inches before the puck

17 Many skills are used when playing ice hockey; the major categories include passing, skating, stick handling, shooting and checking (Pearsall et al.

2000). There are several types of shots in ice hockey, principally the slap and wrist (Pearsall & Turcotte, 2000). The slap shot allows players to generate the most puck velocity whereas the wrist shot praduces better accuracy (Hoerner,

1989). Shooting the puck with optimal velocity and precision is a decisive factor in the overall performance of a player (Lariviere and Lavallee, 1972).

Montgomery and colleagues (2004) performed a task analysis of the different skills used during a professional ice hockey game. Ten national hockey league teams were analyzed at the Moison center in Montreal. Video records were taken and analyzed to label, categorize the frequency and duration of skills used byeach player. The skills included types of, forward skating, backwards skating, hits, shots and passing. The slap shot was used 25% of the time for the forwards and 54% of the time for the defense for ail shots. It was hypothesized that the defensemen took more slap shots because they were generally further away fram the net than the forwards.

Over the years, only a few studies have investigated the various types of shots. Most authors primarily addressed the slap shot training and technical fundamentals with regards to performance impravement. It has been suggested that different body parts contribute different amounts of velocity to the slap shot.

For instance, the trunk, shoulders and elbow/wrist contribute 25%, 40-45% and

30-35% respectively (Hayes, 1964; Wells & Luttgens, 1976). Using EMG,

18 Emmert (1984) showed the specifie muscles involved in the performance of a slap shot. During the backswing phase, the main muscles used are the pectoralis major, deltoid and biceps brachi. In the downswing phase, the muscles used are the pectoralis major, anterior deltoid, external obliques and internai obliques. At impact and during follow through, the major muscles involved are the teres major, latissimus dorsi, obliques, triceps and anterior deltoid (Emmert, 1984).

Alexander et al. (1963) performed the first study comparing the slap and wrist shot with respect to speed and accuracy. The results showed the slap shot was faster (30.8 to 35.3 mis) than the wrist shot (26.6 to 32.6 mis). The standing slap shot was the least accu rate where as the skating wrist shot was the most accurate (Alexander et aL, 1963).

ln order to further understand the differences between the slap and wrist shot, Naud investigated the contact and release point of the snap, wrist and slap shot as weil as the puck velocities using cinematographics. Similar velocities to

Alexander et al (1963) were found for the slap and wrist shot. The study showed that the contact to release point for the wrist shot was 0.216 m and the slap and snap shot averaged 0.152 m.

ln 1999, Pearsall and colleagues studied the influence of stick stiffness on the performance of the slap shot. Ten elite and ten non-elite male subjects took part in the study. Each subject executed thee slap shots with each of three different shaft types (medium, stiff and composite). Reaction forces between the stick blade and ground surface were measured using an AMTI force plate and

19 stick kinematics were recorded using a high speed camera (Redlake, 480frlsec) as seen primarily in the sagittal plane. A radar gun was placed behind the net to record the puck velocities. Counter to conventional wisdom, the results showed that there were no significant differences between the shaft types on the puck velocities. There was a difference in puck velocities between the elite and recreational subjects (30.0 mIs and 23.3 mIs respectively).

Recently Wu et al. (2003) investigated differences in stick velocities using different shaft stiffnesses. Twenty males and twenty females partook in the study. Ten subjects from each group were considered skilled and ten unskilled.

Each subject performed the slap and wrist shot with three sticks of different constructions (wood carbon fiber and composite) and stiffness. The shot mechanics were measured using simultaneous measurements from a force plate, high speed camera (1000 fps) and a radar gun. As expected, the slap shot velocity was greater than wrist shot velocity. The slap shot was between 1.2 and

1.4 times faster than the wrist shot. The skilled and unskilled males performed the slap shots at 30.0 mIs and 23.3 mIs, respectively, and the wrist shot 19.7 mIs and 16.0 mIs respectively. The females skilled and unskilled groups performed the slap shot at 18.8 mIs and 13.3 mIs respectively and performed the wrist shot at 13.6 mIs and 9.4 mIs respectively. Furthermore they found that shot velocity was directly related to the skill level and upper body strength of the person taking the shot.

ln 2000, Meng and Zhao studied four shooting techniques in ice hockey.

The four techniques discussed were the pulling shot, reflection shot, flick shot

20 and hitting shot. These terms are not commonly used in North America and may be specific only to China. A thorough search of the internet proved futile in deciphering the types of shot described. For this study, three top hockey players from China had their shots analyzed using a high speed camera filming at 72 frames per second. Parameters measured included the body center of gravit y, puck kinematics, as weil as width of the hands relative to the length of the stick.

Puck velocities ranged from 14.68 mIs to 24.01 mIs, puck angle of departure ranged from 9.7° to 24.8°, contact time ranged from 0.04 s to 0.39 s, and the ratio of the hand grip width relative to the stick ranged from 2.09 to 2.54 (48% to 39% of the stick length). It is difficult to compare this study to other work because of the lack of knowledge of what shots were used.

There have been some previous attempts to quantify human movements during the slap shot, such as Stephen Murphy's Phd thesis from the University of

Waterloo, but it remains unavailable to public domain.

2.2 Kinematic analysis methods

The following section provides background information relevant to the measurement technique employed in this study. It includes technological information regarding the methods used to collect kinematic data as weil as the manners by which determining joint centers will be calculated.

Optical measurement methods are the conventional means to enable kinematic analysis. Traditional optical methods such as film and video recording

21 have been used to collect biomechanical information on human movements.

Chronophotography, stroboscopy and flashing light sources are examples of other such methods (Niggs and Herzog, 1998). In this approach passive markers are placed overlying anatomical surface landmarks on the segments of interest

(Inman, 1981; Krebs et aL, 1992). These passive markers are covered with a reflective material, which are easily identified on the image when a light is placed in proximity to the camera (Johanson, 1994). Analyzing the recorded series of images in the movie involve spatial identification in x, y, z (Le. digitization) of each marker. Given the latter special information and the known M between images, permit kinematics to be estimated.

Alternatively, active markers using infra-red (IR) may be used to determine the body landmark positions (e.g. Optotrak®). These are active in the sense that the markers emit IR signais from each of the markers using its own frequency therefore making the identification of each distinct marker possible (Inman,

1981). 1R sensitive cameras pairs on pods are used to obtain each markers position. There are other types of methods available to quantify human movement. Hybrids between these two approaches include the Vicon® and

Elite™ systems, using passive markers and infrared light.

Since the gross movement patterns of the ice hockey shots involve movements within three planes, one camera is insufficient to capture ail the data.

Ali these methods mentioned use combinations of multiple two dimensional (20) images to reconstruct the 30 spatial representation of the body motions. To determine the 30 co-ordinates of each marker, two or more 20 images obtained

22 from different viewpoints of the body are used in the direct linear transformation

(DL T) method (Abdel-Aziz and Karara, 1971). The DL T method uses the multiple 2D co-ordinates, synchronizes and establishes a relationship to the 3D co-ordinates. Accurate three dimensional co-ordinates depend on the accuracy of the digitized images, the quality of the DL T reconstruction, the quality of the lenses used in the cameras and the film quality (Cappozzo, 1991; Allard et aL,

1995).

2.3 Three dimensional (3D) kinematics

To facilitate comparison of studies using 3D measures, a standardized protocol has been presented. The most commonly used notation for analyzing

3D kinematics are Euler angles (Le. sequential rotations about two or three axes). In 1995, Wu and Cavanaugh published the recommendations of the

International Society 8iomechanics regarding the standardization in reporting kinematic data. In their paper they suggest that five parts are necessary to describe kinematics appropriately.

Part one requires a definition of a consistent global reference frame, using the notation (X,Y,Z). Specifically this requires:

1. a definition of the segment positions and orientation with respect to

the anatomical position;

2. the displacement of the segment in reference to the global

reference frame;

23 3. the specification of the orientation of the segments with respect to

the global reference frame; and

4. the segment to express itself relative to other segments.

Another land mark paper written by Grood and Suntay (1983) described a joint coordinate system for the clinical description of three-dimensional motions.

Their system of describing joint motion in three dimensions used the combinations of segment orientations to create a free floating vector. The combination of the free floating vector and the segment orientations allowed for the descriptions of the joint to be sequence independent. This sequence dependency is a major soùrce of variance when using Euler angles. Grood and

Suntay (1983) used the method of the free floating axis on the knee in this paper.

2.4 Joint centers

Determining the joint center of the bounding segments during kinematic analyses is crucial for accu rate measurements of the movements (Le. Surface markers may not necessarily be assumed to represent true center of rotation).

The most commonly accepted method of determining joint centers is through the spherical fitting method as described by Piazza, Okita and Cavanaugh (2001). In brief, this method of determining joint centers involves (1) moving the segments in a sequential pattern, (2) measuring the positions of a point on the segment distal to the joint desired, (3) fitting a sphere to the positions and (4) determining the center of this sphere (Cappozzo, 1984; Leardini et al, 1999; Piazza, Okita and Cavanaugh, 2001). To verify the accuracy of this technique, Piazza, Okita

24 and Cavanaugh (2001) used a mechanical linkage consisting of two rigid segments connected bya bail joint to evaluate the joint center determination.

During the trial, the linkage was bolted to the floor and a six camera Vicon 370 motion analysis system was used to collect the 3D data. Several different magnitudes were measured ranging from 15° to 30°. The mean error found was

4.4 ± 0.2 mm at 30° and 9.1 ± 1.5 mm at 15°. Reducing the range of motion from

30° to 15° was found to have statistically significant effect on error. However, reducing the number of markers used to fit the sphere did not affect the accuracy of the joint center calculation.

Similarly, Leardini et al. (1999) validated the method of determining joint centers using the spherical method. This study looked at 11 male subjects comparing joint centers found from a roentgen stereophotogrammetric analysis, which was assumed to be the true joint center and the functional spherical method. The average error found in the spherical method of determining joint centers was 13 mm. The authors concluded that the spherical method of determining joint centers was appropriate for identifying joint centers and had small and unbiased errors.

The necessity of using accu rate joint centers was discussed by Stagni et al. (2000). This study quantified the errors produced when mislocating the hip joint on gait analysis. Angles and moments were measured during normal gait for five subjects (two males and three females). The hip joint center location was determined using the spherical fitting method. The angles and moments were recalculated using an error rang of ±30mm. The results showed that the

25 mislocation of the joint substantially aftected the results of the analysis. An anterior mislocation of the hip joint produced an error by reducing the flexion/extension movement by 22%. Lateral mislocation of the hip produced an error of 15° in the abduction/adduction movement by 15%.

26 Chapter 3: Methods

3.1 Subjects

T en male subjects were recruited to perform this study. Five of the subjects (mean age: 23.0 ± 1.6 years, mean height: 184.9 ± 9.1 cm, mean mass:

88.6 ± 9.1 kg) were recruited from the McGili Redmen Ice Hockey team to form the elite (ELITE) group. Five other recreational (REC) players (mean age: 23.8 ±

3.1 years, mean height: 181.9 ± 6.9 cm, mean mass: 80.5 ± 9.7kg) who played less than twice a week during the winter season were selected to form the REC group. Ali subjects were healthy and showed no physical injury that would prevent them from performing the task.

3.2 Testing procedures

Fifteen surface electromagnetic sensors were placed on the subject and secured using 3M™ surgical tape. Specifically, sensors were placed on the dorsal part of the hands over the third mid-metacarpal, dorsal part of the wrist on the dorsal tubercle of the radius, the upper arm posterior to the olecronon fossa, acromion processes, over C7 spinous process, the greater trochanters and the lateral maleolies. The remaining two sensors were on the trail side of a wooden

Bauer Supreme 3030 ice hockey stick 24.0 cm and 145.0 cm from the heel of the blade.

An electromagnetic tracking device, the Ultratrak®, (Polhemus Inc.,

Burlington, VT, USA) was used to collect the kinematic data at 60Hz while the subjects performed the slap shot trials. Filmbox® version 1.5 software (Kaydara,

27 Montreal, CAN) was used to control the Ultratrak® data recording and MatLab®

(version 6.0.0.88 release 12.0) (MathWorks inc., Natick, MA, USA) programming scripts were used to analyze and process the data. The testing was done on a wooden platform 240 cm deep and 720 cm long that was covered with polyethylene sheets to simulate low friction ice surfaces (Pearsall et aL, 1999;

Wu et aL, 2003). Each subject performed calibrated (see section 3.3) movements to determine joint centers of the wrist, elbow and shoulder for both right and left side (Cappozzo, 1984; Leardini et al, 1999; Piazza, Okita and

Cavanaugh, 2001; States 1997). Subjects wore their own skates and gloves and took five practice shots to acclimatize themselves prior to the testing. Verbal confirmation was used to determine if the subjects were comfortable with the shooting setup.

Each trial consisted of a stationary slap shot into the designated target.

Successful completion of the trial was determined by verbal confirmation from the subject approving the slap shot, and hitting the target (1.30 m x 1.13 m) 3.34 m away with the puck. Five trials were recorded for each of the ten subjects.

3.3 Joint center calibration setups

Joint calibrations were calculated for the wrist, elbow and shoulder for both the right and left side. Each subject moved their limbs in pre specified movements which would break two planes (figure 3.21) thus enabling the determination of the joint center.

28 (?? Figure 3.21 - Movements done by subject to calibrate joint center. A movements performed to find shoulder joint centers. B Movements performed to determine elbow joint center. C Movements performed to determine wrist joint center

Moving the distal segments through the specified movements allowed kinematics to be collected so that a sphere could be calculated based on the kinematics

(figure 3.23).

Figure 3.23 - Points on the sphere can be used to calculate the center of the sphere. The intersects of the radial vectors normal to the tangential movement of the distal markers, were determined by least squared optimization.

29 3.4 Joint center calculations

It should be noted, that though the sensor may be placed superficial to a joint of interest, they do not necessarily correspond to the axis of rotation, particularly a joint with three degrees of freedom such as the gleno-humeral joint.

Concequently using surface location tracking info may lead to erroneous measures. The following algorithms were used to measure joint centers (Eberly,

2001) based on the information collected by the movement sequencing describe previously.

Given a set of points {(xj'Yj,z;)}:"m ~ 4, fit them with a sphere

(x - ay + (y - bY + (z - cy = r2 where (a,b,c) is the sphere center and ris the sphere radius, assuming that ail points are not coplanar. To do this, the following functions were used:

m

E(a,b,c,r)= I(Li -rY j=1 where Lj = ~(Xj - ay + (Yi - bY + (Zj - c). Taking the partial derivative with respect to r gives;

Setting equal zero yields

Taking the partial derivative with respect to "a" gives;

30 ôE m ôL. m ( ôL. ) -=-2L(L;-r)-' =2L (x;-a)+r-I , ôa ;=1 ôa ;=1 ôa

Taking the partial derivative with respect to "b" gives;

Taking the partial derivative with respect to "c" gives;

Then setting the tree derivatives to zero yields,

1 m 1 m ÔL. a=-Lx;+r-L-' , m ;=1 m i=l ôa

1 m 1 m ôL. b=-Ly;+r-L-' and m ;=1 m i=l ôb

1 m 1 m ôL. e=-Lz;+r-L-' m ;=1 m ;=1 ôe

)/ Replacing r by its equivalent fram ôE / ôr = 0 and using ôL; / ôa = (a - xi L; ,

ôL; / ôb = (a - y; )/ L; and ôL; / ôe = (a - Zj)/ L; , we get three nonlinear equations in a,b and c;

a = x + lIa =: F(a, b, e)

b = y + lIb =: C(a,b,e)

e = z + Ile =: H(a,b,e)

Where

x=-- 1 Lm x. m ;=1 1

31 z=-- 1 Lm z. m i=1 1

l =~ "m L. m .L..i=1 1

- 1 a-xi L =-Lm -­ a m ;=1 L. 1

- 1 b-y; L =-Lm -­ b m ;=1 L. 1

- 1 C-Z; L =-Lm -- c ;=1 L m i

Finally, fixed point iteration can be applied to solving these equations: ao = x ,

i ~ o.

The exported surface electro magnetic sensor information from Filmbox®

(figure 3.41) were transferred into Matlab® (figure 3.42) processing module to render estimates of body joint centers.

32 Figure 3.41 Screen shot of fimbox® program used to collect electromagnetic sensor co-ordinates. Views shown (from top left corner, clockwise): Oblique, sagittal, frontal and transverse.

Figure 3.42 Screen shots from Matlab® program used to calculate joint and stick kinematics. A - screen shot showing oblique view. B - Screen shot showing sagittal view. C - Screen shot showing frontal plane view. D - Screen shot showing transverse plane view.

33 3.5 Data reduction

The position of the stick was measured in the three planes: sagittal, frontal and transverse. Specifically the measures taken included:

1. Global stick shaft orientation (or lie angle) of stick and blade (x,y,z

planes);

2. Global displacement, velocity, acceleration of blade (x,y,z planes

and resultant) with respect to initial puck position;

3. Global angular displacement, velocity and acceleration of stick

(x,y,z planes and resultant);

4. Displacement of the center of rotation of stick (x,y,z) with respect to

initial puck position. (This was determined by adjacent vector

intersects of shaft's longitudinal axis);

5. Local rotation of stick (about longitudinal axis of stick shaft) with

respect to initial start orientation;

The following cosine law function was used to determine the angles measured in Matlab®. V.u cos 0" = IVI/ui

Where Vand U are vectors in the global co-ordinate system.

With respect to the kinematics, the following angles were calculated:

1. Flexion and extension of the lead and trail elbow;

2. Shoulder movements horizontal and vertical adduction/abduction;

34 3. Trunk rotation (shoulders with respect to pelvis);

4. Pelvis rotation (relative to feet position);

5. Spatial volume envelope covered by the stick

6. Blade velocity

7. Stick angular velocity

From the kinematic curve profiles (see appendices for examples), discrete curve parameters were extracted and averaged at the start, top of the backswing and at impact for statistical analysis. Furthermore, the temporal profiles were normalized from start to impact as 1 to 100 points.

ln translations, (0, 0, 0) indicates the original puck position. The "x" values indicate the depth position, the "y" position indicates the height and the "z" position indicates the width. Positive "x" values indicate movements away from the puck (behind player), positive "y" values indicate the upwards direction and positive "z" values indicate movements away from the target (figure 3.51).

y

z Figure 3.51 global co-ordinate system of the setup

The data was normalized from 1 -100 and averaged across subjects and groups. The mean data points at the start, top of swing and impact were used for statistical analysis (ANOVA).

35 Chapter 4: Results

ln general, the three dimensional tracking of the gross movement patterns of the body and stick generated a detailed record of the slap shot technique execution. Distinct differences between elite and recreational shots were observed, though few statistical significances could be identified due primarily to the large variance in parameter estimates. The following section provides details of the specifie kinematic comparisons

4.1 Stick kinematic results

The spatial volume envelope covered by the stick was smaller for the elite compared to the novice (table 4.01). For instance, the average volume of space

3 3 for the elites and recreational were 1.04 m and 1.73 m , respectively. The furthest backwards displacement of the blade was 0.46 ± 0.20 m for the elite players compared to 0.71 ± 0.24 m for the recreational players, respectively.

The greatest excursion of the blade in the frontal plane was 1.15 ± 0.11 m and

1.21 ± 0.07 m for the elite and recreational players, respectively. The highest vertical position of the blade was 1.93 ± 0.21 m and 2.00 ± 0.17 m for the elite and recreational players, respectively.

Table 4.01 The peak width, depth and height of the blade positions during the saps1 h 0 t

Maximum wldth Maximum dpeth Maximum height

elite rec elite rec elite rec mean 1 sd mean 1 sd mean 1 sd mean 1 sd mean 1 sd mean l sd 0.46 1 0.20 0.71 1 0.24 1.16 1 0.11 1.21 1 0.07 1.93 1 0.20 2.00 1 0.16

36 The velocity of the blade (table 4.02) for the elite players was significantly faster at impact than the recreational players, 29.14 mIs and 26.46 mIs (p <

0.05).

Table 4.02 Blade velocities 0 f eln e an d novice players a t impact. Impact

Elite (mis) Rec (mis)

mean 1 sd mean 1 sd Blade velocity -29.14 1.39 -26.47 0.66 1 1

However, angular displacements (table 4.03), velocities and accelerations were similar for both the elite and recreational players at impact. By deconstructing the linear and angular components of the general movement, differences between groups were observed. The velocities of the blades attributed by the angular displacement of the stick were 16.00 mIs and 17.38 mIs for the elite and novice players respectively. Thus 13.14 mIs and 9.08 mIs (p < 0.05) were attributed to the translational movements of the stick for the elite and recreational players.

Table 4.03 Resultant stick angular displacement, velocity and acceleration at impact Impact elite rec mean sd mean sd Resultant angular displacement (deg) 17.2 2.6 17.3 3.4 Resultant angular velocity (deg/sec) 1032.6 156.1 1040.4 206.4 Resultant angular 2 acceleration (deg/sec ) 61956.0 9360.0 62424.0 12384.0

The elite's stick center of rotation (table 4.05) at the top of the backswing was higher (1.18 m vs 0.89 m) and further away (0.68 m vs 0.43 m) from the target than the recreational player.

37 Table 4.05 Global cen ter 0 f ro ta f Ions x, y, Z top Impact Ellte(m) Rec(m) Ellte(m) Rec(m)

mean sd mean sd mean Sd mean sd Center of 0.26 0.11 0.268 0.23 0.55 0.08 0.51 0.06 rotation x Center of 1.18 0.36 0.886 0.79 0.79 0.08 0.74 0.10 rotation y Center of 0.68 0.10 0.428 0.14 ·1.50 0.33 ·1.91 0.64 rotation z

Alternatively expressed, the position of the center of rotation relative to the

stick is shown in table 4.06. At the top of the swing and impact the center of

rotation was significantly (p < 0.05) higher up on the stick when comparing the

elite to the recreational players

Table 4.06 C enter 0 f ro ta f Ion pOSI Ion on sf IC k measure dfrom h e el of blade

Top lm act

ellte (m) ree m) ellte m) rec m)

mean 1 std mean std mean 1 std mean std 1.74 1 0.06 1.54 0.13 1.23 1 0.02 1.16 0.03

There were no significant differences found in the stick orientation in each

of the different planes. Stick positions were similar for both the elite and

recreational players at the start, top of backswing and impact (table 4.07).

Table 4.0 7 S'tiC k orientations .ln th e sagl"ttlf a, rona tl an dtransverse planes Star! top Impact Elite deg) RecJdegj Elite degt Rec deg) Elite (deg) Rec deg) mean sd mean sd mean sd mean sd mean sd mean sd

Stick sagittal 48.4 5.8 44.8 1.5 ·104.2 9.4 ·125.5 33.0 52.5 1.5 54.6 11.5 Stick frontal 96.8 3.3 105.2 5.2 262.4 6.1 290.0 30.2 106.4 3.4 100.4 12.1 Stick transverse -7.3 3.9 -15.0 4.6 -154.5 30.2 -189.7 61.0 -20.8 3.5 -12.4 14.0

38 4.2 Body kinematic results

Descriptions of the conventions for body kinematics assessed are shown previously in figure1.05.

The following section presents the results of the joint kinematics at the start of the swing, top of backswing and at impact. Figure 4.20, 4.21 and 4.22 show the results of the lead, trail and trunk body kinematics. For further details see appendix A, Band C.

Lead elbow Flexion/Extension Lead arm Vertical Adduction/Abduction Lead arm Horizontal Adduction/Abduction elite Impact Impact elite rec ---1 Impact Start elite Start rec Topelite Top rec Impact rec elite Lead arm Horizontal Adduction/Abduction -17.0 -34.2 92.1 86.4 -21.3 -21.7 Lead arm Vertical Adduction/Abduction 59.6 53.4 81.3 64.8 61.8 47.0 Lead elbow Flexion/Extension 87.3 93.2 103.8 111.1 68.3 72.9 ------Figure 4.20 Graph and table of elite vs recreational kinematic values of the lead arm at the start, top of swing and impact.

39 Trail elbow Flexion/Extension Trail arm Vertical Adduction/Abduction Trail arm Horizontal Adduction/Abduction

Impact Impact elite rec - ---- Impact Startelite Start rec Topelite Top rec Impact rec elite Trail arm Horizontal Adduction/Abduction 65.0 73.3 15.7 -23.2 38.4 38.7

-----~-- Trail arm Vertical Adduction/Abduction 51.6 58.1 110.7 87.3 30.2 38.7 ------~ ~ Trail elbow Flexion/Extension 141.0 148.9 129.9 129.3 119.2 134.9 -----_._- . Figure 4.21 Graph and table of elite vs recreational kinematic values of the trail arm at the start, top of swing and impact.

Pelvis rotation Trunk rotation Impact Impact rec elite

Start elite Start rec Top elite Top rec Impact elite Impact re~~~! Trunk rotation 48.8 39.5 61.7 54.2 30.5 -33.6- ~ ~ 1 ------.--. '--1 Pelvis rotation 12.3 14.4 18.6 13.1 56.8~ __~~_ j Figure 4.21 Graph and table of elite vs recreational kinematic values of the trunk and pelvis rotations at the start, top of swing and impact.

40 At the shoulder the lead arm of the elite players was less elevated (Le. abducted) than the recreational players at the top and impact (64.8° vs 81.4° and

46.9r vs 61.8° respectively). The trail arm of the elite was less elevated than the recreational player at the top but more elevated at impact (87.3° vs 110.r and 38.r vs 30.2° respectively).

The elite players extended both elbows more than the recreational players did except at the top of the swing for the trail arm where both arms were flexed the same amount.

Both the elite and recreational players had comparable arm positions in the ab/adduction (horizontal) direction at impact (-21.r vs -21.3° and 38.r vs

38.4 0). The lead shoulder of the elite players were horizontally abducted more than the recreational players at the start and top of the swing (-34.2° vs -17.0° and 86.4° vs 92.1°).

The trunk of the elite players rotated less than the recreational players at the top of the swing (-54.2° vs -61. r). Furthermore, the trunk of the elite players rotated less towards the target than the recreational players. At impact, although not significant, the elite players had rotated their hips 8.2° more than the recreational player

4.3 Swing timing and sequencing

The time ratio of backswing to downswing phase for the elite players was significantly greater than those of the recreational players (72.9% vs 65.7%)

(ANOVA P < 0.00005). (table 4.30 and 4.31)

41 Table 4.30 Percentage of th e bac k sWln9pa rt 0 f th es ap s h 0 t

elite rec

rnean 1 sd rnean 1 sd Percentage of backswing 72.85 1 3.39 65.73 1 3.25

Table 4.31 ANOVA table of backswing percentage

df MS df MS

Effect Effect Error Error F p.level

2 492.1848 8 11.01389 44.68766 0.000046

The total time (table 4.32) of elite and recreational subjects for the backswing of were similar (0.74 sand 0.72 s) as was the downswing (0.32 sand

0.33 s). However, the variances of the swing times were smaller for the elite players compared to the recreational.

Table 4.32 Table of the times required for the backswing and downswing phase of the slap shot.

Backswing Downswing

mean sd rnean sd

Elite 0.74 0.026 0.32 0.039

Recreational 0.72 0.052 0.33 0.090

Both elite and recreational players exhibited different movement sequences. These sequences may be defined ln part by identifying the instant of peak joint angular velocity. Sequence in peak velocity for the elite players were; trunk rotation, pelvis rotation, lead shoulder horizontal adduction, trail shoulder vertical adduction, lead shoulder vertical adduction, trail shoulder horizontal adduction, lead elbow flexion finishing with trai! elbow extension. The average peak joint angular velocity sequence for the novice players went as follows: trail shoulder vertical adduction, trail elbow extension, lead shoulder vertical

42 adduction, trunk rotation, lead elbow flexion, lead shoulder horizontal adduction, pelvis rotation and trail shoulder horizontal adduction. (figure 4.30) -1Iunt...... - NvII ...... laIdtbouldlr ...... acIdIcIfoft 1IIItbouIdIrWllclladdudlon - ....1ItouIcI ...acIdIcIfoft - 1III1houIcIIr ...... - ...... -,...... , ......

....IhouIcIIrWdbladdudlon ...... ,. .... dlladdudlon Novfce - lad...... acIdIcIfoft - 'IIlItbouIdIr ...... "rc.1Ioft -"DaI...,-lAId ...... Figure 4.30 Sequence of peak joint angular velocities for elite and recreational players during the downswing phase of an ice hockey slapshot.

43 Chapter 5: Discussion

This study compared the movement patterns of elite and recreational ice hockey players performing a stationary slap shot. It was hypothesized that there would be significant differences in the blade velocity and of the overall gross motor pattern kinematics between the elite and recreational players. As expected and congruent with previous research, this study found a significant difference (p > 0.05) in the speed of the shot between the elite and recreational players. The stick peak velocities were 29.14 mIs to 26.46 mIs (table 4.02), respectively; similar to the velocities reported by Wu et al (2003). From the detailed kinematic evaluation, differences in joint coordination and global stick movements provide insight into the possible mechanism by which the elite players were able to generate puck velocities. These include the translational velocities of the stick, position of the stick center of rotation and the sequencing of body movements.

For instance, deconstructing the swing displacement of the stick into the rotational and translational components yielded some interesting results (figure

5.01).

44 Elite Novice

• .._t__a... __17...__ Figure 5.01 Stick velocities of elite and recreational players deconstructed. The velocity components due to pure rotation and translation were 16.00 and 13.14 mIs, respectively for the elite, and 9.08 and 17.38 mIs for the recreational

The major differences in blade velocities between groups resulted from the translation component of the stick movement (table 4.20). Comparing stick movements in the frontal plane (figure 5.02), the traces of elite players iIIustrated the typical swing motion as a downward followed by a translation towards the target, whereas the recreational players typically made more of a sweeping motion. In addition, this component of translational movement may also contribute to an increase in the accuracy of the shot (Kreighbaum & Barthels

1996). Though not measured as part of this study, a more linear movement would create a larger window for optimal projection of the puck to hit the desired target as opposed to the more curvilinear movement as seen with the recreational players (figure 5.03).

45 Elite Recreational

Figure 5.02 - Examples of elite and recreational trace downswings during a stationary ice hockey slap shot. Traces were taken at 60 flsec.

A

B

Figure 5.03 - Greater linear movements of the stick (A) allows for a greater margin for error as weil as increased accu rate projection of the puck as opposed to the more curvilinear motions of the stick (B) (adapted from Kreighbaum and Barthels 1996)

46 The greater downward motion performed by the elite players corresponds to the data reported by Pearsall et al (1999) that investigated the downward forces exhibited by hockey players during a stationary slap shot. Their findings indicated that the resulting downward force during stick to ground contact for the skilled players were 164.7 N compared to 104.7 N for the unskilled. This is also congruent with the ice hockey teaching literature that says to "Ioad " the stick into the ice and use the unbending recoil energy of the stick to increase the velocity.

This has implications in the design and construction of the stick.

Despite the contrasting shooting techniques, differences in gross motor patterns were few as assessed from the kinematic measures at the start, top of backswing and impact. However, difference in sequencing coordination of the trunk and upper limbs were found. Movement sequences performed by the elite players were more consistent with the motor controlliterature of mature movement patterns (Langerdorfer, 1987; Messick, 1991 and Haywood 1993); specifically, a proximal to distal sequence was observed: where by peak angular velocities began with the trunk, followed by shoulder then byelbows. In contrast, as seen in figure 5.04, the recreational players exhibited no proximal to distal sequence: the peak angular velocities began the movements in the trail shoulder and elbow, followed by trunk rotation, lead elbow and shoulder and finished with pelvis rotation.

Mature or "skilled" movement patterns are acquired and reinforced the more often a person executes the skill (Haywood 1993). Unlike the elite players, the recreational players did not have the same practice time. The elite players

47 practiced five times a week and played at least one game per week during the fall/winter season, whereas the recreational players were on the ice no more than twice a week. Although both groups began playing hockey at approximately the same age it can be suggested that the amount of practice elite players had played an important role in the development and maintenance of highly proficient slap shot techniques.

ln addition to the above, differences in the amount of variation within group executions were observed. The elite group tended to exhibit much less stick movement variance between players than the recreational group. Table

5.01 shows the variance of the data collected. In 19 of 21 (91%) of the stick position variables examined, the elite players showed less variation in stick position than the recreational players. In 13 of 24 (54%) of the body position variables examined, the elite players exhibited less variation. This may be indicative that there is a specific position in which the stick should be in for the most optimal slap shots. Likewise, elite players must have mastered an optimal technique of executing the ice hockey slap shot. Novice players attempt to imitate the proper technique of the ice hockey slap shot, but as they do not practice as much as the elite players they are not able to reproduce the movement as consistently as the elite players. Each subject moved in a manner that they felt was optimal to perform a slap shot.

48 Table 5.01 Elite versus novice variance (standard deviation) at the start, top and impact point of the slap shot.

Start Top Im(l8ct

Elite Rec Elite Rec Elite Rec

Stick sagittal (degrees) 5.77 > 1.52 9.37 < 33.02 1.52 < 15.11

Stick frontal (degrees) 3.32 < 5.23 6.12 < 30.19 3.45 < 12.05

Stick transverse (degrees) 3.69 < 4.61 30.24 < 61.00 3.53 < 13.96

Blade X position (meters) 0.00 < 0.01 0.21 < 0.29 0.03 > 0.02

Blade Y position (meters) 0.00 < 0.01 0.13 < 0.43 0.04 < 0.06

Blade Z position (meters) 0.00 = 0.00 0.19 < 0.22 0.01 = 0.01

lead elbow Flexion/Extension (degrees) 16.67 > 14.06 21.64 > 14.64 9.60 < 16.57 lead arm Horizontal Adduction/Abduction (degrees) 22.95 < 34.75 13.66 < 19.39 15.61 < 16.23

lead arm Vertical Adduction/Abduction (degrees) 5.64 < 11.19 13.50 < 26.05 7.65 < 10.60

Trail elbow Flexion/Extension (degrees) 10.06 > 7.74 16.19 > 17.64 14.13 > 12.55

Trail arm Horizontal Adduction/Abduction (degrees) 21.47 > 7.09 27.19 < 57.59 36.95 > 26.46

Trail arm Vertical Adduction/Abduction (degrees) 6.77 < 7.61 24.64 < 24.66 14.01 > 8.06

Trunk rotation(degrees) 0.60 < 1.17 0.95 < 1.60 2.26 < 7.17

Pelvis rotation (degrees) 1.01 > 0.60 2.5 > 2.49 3.10 > 2.20

One of the goals at the beginning of the study was to summarize the

optimal movements to perform during a slap shot. A chart was developed earlier

based on the teaching literature (table 1.01). The following chart is based on the

data collected in this study (table 5.02).

49 Table 5.02 Indicates the position and movement of the upper body during a st af lonary Ice hoc key s 1ap s h0 t b ase d on th e d aao t bd' serve ln th'IS St U d Segment Start Backswing Top Downswing "Impact Trunk Begins rotated 50° Rotate trunk Trunk should be Trunk rotates 86° Trunk should be towards target 1100 away from rotated away from rapidly towards target rotated towards target target 61° target 25° Lead arm Arm is vertically Arm vertically Arm is vertically Arm vertically adducts Arm is horizontally abducted 55° and adducts 12° then abducted 65°. rapidly 30° followed abducted -22°. horizontally abducts 20°. Arm Arm is horizontally by 10° of rapid Arm is vertically adducted -30°. horizontally adducted 85°. abduction. adducted 45°. Elbow is f1exed 92°. adducts 120°. Elbow is f1ex 108° Arm horizontally Elbow is bent 67°. Elbow extends 40° abducts 110° then begins f1exing Elbow f1exes 30° 12°. Trail arm Arm is vertically Arm vertically Arm should be Arm adducts vertically Arm should be abducted 60° from adducts 30° then vertically abducted 57° and horizontally vertically adducted trunk and abducts 57°. 87° and adducts 62°. 38° and horizontally Arm horizontally horizontally Elbow extends 50° horizontally adducted 75°. abducts 100°. abducted -25°. adducted 38°. Elbow is bent about Elbow f1exes 60° Elbow is bent 90°. Elbow is bent 140° 150° Stick Stick angled 50° in Stick rotates up Stick is angled - Stick rotates 155° Stick is angled 55° the sagittal plane. 150° in the sagittal 105° in the sagittal down in the sagittal in the sagittal Stick angled 100° plane, 160° away plane. plane, 160° forward in plane. Stick is in the frontal plane. from target in the Stick is angled the frontal plane and angled 100° in the Stick is angled 8° in frontal plane and 264° in the frontal 150° in the transverse frontal plane. Stick the transverse 150° in the plane. plane. is angled 10° in the plane. transverse plane Stick is angled transverse plane. 158° in the transverse plane.

The results shown in this chart, with some exceptions, are very similar to what is described in the hockey literature. This chart, however, provides a more quantitative set of data for analysis.

This study has given insights into the kinematics of the stationary ice

hockey slap shot, however much work in this area is still needed. Given the

large volume of information generated in three-dimensional kinematic evaluation,

ail aspects of the body movement could not be evaluated. In the future it may be

necessary to look at other specific regions of the slap shot, such as trunk

flexion/extension, forearm supination/pronation, stick rotation (longitudinal), lower

leg kinematics and center of gravity movements. Furthermore, other skills, not just the slap shot should be examined. Skills such as stick handling, passing and

other types of shots are crucial parts of the game and should not be ignored.

50 The issue of the specific brand and/or model of equipment used should not be ignored. Although there has been no evidence that the slap shot is any faster using composite rather than wood sticks (Pearsall et al. 1999, Wu et al,

2003), there may be some insights that can be gained by investigating if the kinematics change when using different types of sticks. For example, will the stick bend more? Does the swing path change? How does the glove, grip interact with the stick? These are a few questions that need to be addressed to understand how the stick affects the game.

The methodology used in this study was effective in quantifying the movement patterns of the ice hockey slap shot. The error in sensor location and position was found to be less than 1cm and 2 degrees in this study. The sampling rate (Le. 60hz) used in this study, although adequate for the gross motor pattern was not sufficient to quantify the rapid distal end of the stick movement with some aliasing measurement. Nonetheless it was encouraging that the results were comparable to the speeds found in previous research.

Hence this system sampled fast enough to measure the gross overall movement of the stick. Alternative systems such as the VICONTM system may have yielded a higher sampling rate Le. up to 500Hz. Using a system such as this would have eliminated the use of sensors attached via cables which were very cumbersome.

It would have been interesting to be able to use a portable system to collect the data on ice as weil as to incorporate other forms of measures along with the kinematics. Integrating kinematics along with other forms of measures such as

51 kinetics and EMG may provide further insight into the slap shot not obtainable by using just one technique.

As seen in previous research, upper body strength is a variable that corresponds to increased puck velocities when performing the slap shot 0Nu et al. 2003, Fergenbaum & Marino, 2004). It would have been interesting to conduct a longitudinal kinematic analysis while the subjects undertook a training program. This would have allowed us to see if there were any kinematic changes occurring as strength increased.

The polyethylene ice surface was not the same as real ice. The subjects only performed stationary slapshot and the only equipment being used were their skates and gloves. Thus studies on ice should be conducted to verity the findings of this study.

This study has given some insights into the technique of the ice hockey slap shot. Coaches teaching the slap shot should be aware that each player may have slightly different body motions, but there is a certain path the stick should follow. The translational aspect of the stick plays a large role in the shot velocity when comparing the elite and recreational players.

52 References

Abdel-Aziz & Karara, H.M. (1971) Direct linear transformation from comparator coordinates into object space coordinates in close range photogrammetry. In ASP Symposium on Close Range Photogrammetry (pp. 1-18), The University of Illinois.

Alla rd , P., Blanchi, J-P. and Aissaoui, R. (1995). Bases ofthree-dimensional reconstruction. In: Allard, P., Stokes, 1. and Blanch, J.P. eds. Three-Dimensional Analysis of Human Movement. Champaign, IL: Human Kinetics, 1995: 19-39.

Alexander, J.F., Drake, C.J., Reichenbach, P.J. & Haddow, J.B.(1964). Effect of strength development on speed of shooting of varsity ice hockey players. Research Quarter/y. 35: 101-106

Alexander, J.F. Haddow, J.8. & Sultz, G.A (1963). Comparison of the ice hockey wrist and slap shots for speed and accuracy. Research Quarterly. 34: 259-266.

Almstedt, J. (1974) Hockey development guide, 8 to 18. Canadian amateur hockey technical advisory committee.

Cappozzo, A, (1984). Gait analysis methodology. Human Movement science. 3,27-50.

Cappozzo, A. (1991). Three dimensional analysis of human walking: experimental methods and associated artifacts. Human Movement science. 10, 589-602.

Dowbiggin, B. (2001). The stick: A history, a celebration, an elegy. Macfarlane Walter & Ross. Toronto.

Eberly, D. (2001). Least squares fitting of data. www.magic-software.com

Emmert, W. (1984). The slap shot - strength and conditioning program for hockeyat Boston College. National Strength and Conditioning Journa/. 6(2): 4-6, 68,71,73.

Fergenbaum, M.A and Marino, G. W., "The effects of an upper-body plyometrics program on male university hockey players", Safety in ice hockey: Fourth Va/ume, ASTM STP 1446, D.J. Pearsall and AB. Ashare, Eds., ASTM International, West Conshohocken, PA, 2004.

Gendron, D. (1957) Coaching hockey successfully / Dennis "Red" Gendron with Vern Stenlund. Human Kinetics, Champain IL.

53 Grood, E.S. & Suntay, W.J. (1983). A joint coordinate system for the clinical description of the Three-Dimensional motions: Application to the knee. Journal of biomechanical engineering. 105: 136-144.

Haché A (1970). The physics of Hockey. The Johns Hopkins Univeristy Press, Baltimore.

Hayes, D. (1965). A mechanical analysis of the hocky slap shot. Journal of the Canadian Association for Health, Physiscal Education and Recreation. 31: 17.

Haywood, K.M. (1993). Life span motor development. Human kinetics. Champaign, II.

Hockey Canada report (2003) (http://www.hockeycanada.ca/e/aboutlindex.html)

Hoerner, E.F. (1989). The dynamic role played by the ice hockey stick. Safety in ice hockey. ASTM STP 1050, C.R. Castaldi and E.F. Hoerner Eds. American Society for Testing and Materials, Philidelphia, USA 154-163.

Inman, V.T., Riston, H.J. & Todd, F. (1981). Human walking. Baltimore: Williams and Wilkins. 1-61

Johanson, E.m. (1994). Gait laboratory: structure and data gathering. In: Rose J, Gambie J.G., eds. Human walking (2nd ed). Baltimore: Williams and Wilkins. 201-224

Krebs, D.E., Wong, D., Jevsevar, D., O'Riley, P., & Hodge, AW. (1992). Trunk kinematics during locomotor activities. Physical Therapy. 72,505-514.

Kreighbaum, E. & Barthels, K.M. (1996) Biomechanics. A qualitative approach for studying human movement. Needham Heights . Allyn & Bacon. p.350.

Langendorfer, S. (1987). Prelongitudinal screening of averarm stricking development performed under two environ mental conditions. In J.E. Clark & J.H. Humphrey (Eds.), Advances in motor development research (Vol. 1, pp. 17-47). New York: AMS Press.

Leardini, A, Cappozzo, A, Cattani, F., Toksvig-Larsen, S., Petitto, A, Sforza, V., Cassanelli, G. & Giannini, S. (1999) Validation of a functional method for the estimation of hip joint center location. Journal of Biomechanics. 32,99-103.

Meeker, H. (1975) More hockey basics from Howie Meeker. Scarborough. Prentice-Hall.

54 Mesick, J.A. (1991) Prelongitudinal screening of hypothesized developmental sequences for the overhead tennis serve in experienced tennis players 9-19 years of age. Research Quarlerly for exercise and sporl, 62, 249-256.

Montgomery, D.L., Nobes, K., Pearsall, D.J., and turcotte, RA. (2004) Task analysis (Hitting, Shootin, Passing and skating) of Professional Hockey Players. Safety in ice hockey: Fourlh Volume, ASTM STP 1446, D.J. Pearsall and A.B. Ashare, Eds., ASTM International, West Conshohocken, PA, 2004.

Nigg, B.M., Cole, G.K., & Wright, I.C. (1998). Optical methods. In: Nigg, B.M. and Herzog, W. eds. Biomechanics of the Musculo-skeletal System (2nd edition). Toronto: John Wiley & sons. 302-331

Nigg, B.M. and Herzog, W. eds. (1998). Biomechanics of the Musculo-skeletal System (2nd edition). Toronto: John Wiley & sons.

Pearsall, D.J., Montgomery, D.L., Rothsching, N. & Turcotte, RA. (1999). The influence of stick stiffness on the performance of ice hockey slap shots. Sporls Engineering. 2, 3-11.

Piazza, S.J., Okita, N. & Cavanaugh, P.R (2001). Accuracy of the functional method of hip joint center location: effects of Iimited motion and varied implementation. Journal of Biomechanics. 34, 967-973.

Randy, G. (1956) Hockey: drill solutions. FP Hendriks Publishing. Stettler AB.

Renger, R (1994). Identifying the task requirements essential to the success of a professional ice hockey player: a scouts perspective. Journal of Teaching in Physical Education. 13, 180-195.

Smith, M. (1996) Hockey drill book. Buffalo. Firefly.

States, RA. (1997). Two simple methods for improving the reliability of joint center locations. Clinical Biomechanics. 12(6).367-374.

Walker, N. & Dann, S. (2000) Hockey in action. New York. Crabtree Pub.

Wells, K.F. & Luttgens, K. (1976). Kinesiology, scientific basis of human motion. Philadelphia: W.B. Saunders Company.

Wu G. and Cavanagh P.R (1995). ISB recommendations for standardization in the reporting of kinematic data. Journal of biomechanics. 28(10), 1257-1261.

Wu, T-C., Pearsall, D., Hodges, A., Turcotte, R & Lefebvre, R (2003). The performance of the ice hockey slap and wrist shots: the effect of stick construction and player skil!. Sporls Engineering. 6(1), 31-39.

55 Appendix A

Graphs of le ad shoulder kinematics

Lead shoulder horizontal ab/adduction

120

100

80

60

UI 1\) 40 2! Cl Frecl a1\) 20 ~--e~lite

0

-20

-40

-60 Swing percentage

Lead shoulder vertical ab/adduction

100

90 80

70

60 UI 1\) 2! Cl 50 a1\) 40

30

20

10

0 <0

Swing percentage

56 Appendix B

Graphs of trail shoulder kinematics

Trail shoulder horizontal ab/adduction

100

80

60

III 40 eCI) Cl cCI) 20

0

-20

-40 Swing percentage

Trail shoulder vertical ab/adduction

120

100

80

III CI) I!! Cl 60 cCI) 40

20

0 (0

Swing percentage

57 Appendix C

Graphs of elbow shoulder kinematics

Lead elbow flexion/extension

160

140

120

100 III CIl ~ CI 80 cCIl l__~~~e J 60

40

20

0 <0 <0 <0 <0 <0 <0 <0 <0 <0 ~ <0 N N M (") ~ "II" u:; 10 èO <0 r::: r-. a; co CT> CT> Swing percentage

Trail elbow flexion/extension

180

160

140

120

.,.. 100 e '"., --elite c 80 ~

60

40

20

0 ~ ...... N ., .... 0 .... en N ., .... a ~ ~ ~ ~ N N N ;;; ~ ... ~ ~ ~ ~ :3 f8 iD ;1; 10 ~ I:! te .... ., :g co 0; ~ en '" '" ~ Swing percentage

58 Appendix D

Graphs of pelvis and trunk rotation kinematics

Pelvis rotation

80

70

60

50

40 1/1 CI) I!! Cl 30 CI) Q 20

10

0

-10

-20 Swing percentage

Trunk rotation

60

40

20

1/1 CI) 0 I!! ~ Cl CI) Q -20 L-==_~!

-40

-60

-80 Swing percentage

59 Appendix E

Graphs of stick displacement in frontal, sagittal and transverse planes

Stick angular dis placement in frontal plane

350 300 250 li! 200 I!! ~ 150 Cl 100 50 0 (0 (0 ...... (0 (0 (0 ...... (0 (0 (0 (0 (0 N N c;:; C") ~ LI) LI) <0 (0 r:: «; co 0; ()) """ "- Swing percentage

Stick angular displacement in sagittal plane

100

50

(/) Q) 0 ~ Cl Q) Cl -50

-100

-150 Swing percentage

Stick angular dis placement in transverse plane

0

-50

(/) -100 Q) ~ Cl Q) Cl -150

-200

-250 Swing percentage

60 Appendix F

Graphs of blade velocities

Blade velocity

15

10

5

0

-5 en -10 -E -15

-20

-25

-30

-35 swing percentage

61 Appendix G

Graphs of the stick rotation center distance from heel of stick

Stick center of rotation distance from heel of blade

2

1.8

1.6

1.4

1.2

E

0.8

0.6

0.4

0.2 o

Swing percentage

62 Appendix H

Table of MANOVA values for measures

Results of MANOVA df MS df MS Effect Effect Error Error F p-Ievel <0.05

1 0.1977955 8 0.6432238 0.3075065 0.5943733 Blade X acceleration im~act 1 0.0098227 8 0.0123032 0.7983872 0.3976583 Blade X acceleration start 1 0.4183372 8 0.1432235 2.9208708 0.1258157 Blade X acceleration top 1 23.212236 8 6.8469806 3.390142 0.1028506 Blade Xp_osition impact 1 0.4191895 8 0.6235038 0.6723127 0.435993 Blade X position start 1 449.81244 8 630.52771 0.7133905 0.4228585 Blade X position top 1 20.072998 8 11.344209 1.7694489 0.2201205 Blade X velocity impact 1 0.0017737 8 0.0403912 0.0439133 0.8392529 Blade X velocity start

1 0.0162689 8 1.8723251 0.0086891 0.9280245 Blade X veloci!!, t~ 1 0.0334694 8 2.1810734 0.0153454 0.9044689 Blade Y acceleration impact 1 0.0366111 8 0.0334878 1.0932646 0.3263103 Blade Y acceleration start

1 0.1175819 8 0.42933 0.2738731 0.6149277 Blade Y acceleration t~ 1 92.08094 8 45.085773 2.0423503 0.1908351 Blade Y position impact

1 0.0960218 8 0.4528398 0.2120437 0.6574286 Blade Y~osition start 1 6088.7993 8 1006.6416 6.0486269 0.0393558 YES Blade Y position top

1 17.874969 8 1.1826658 15.114134 0.0046243 YES Blade Y velocity-'m~act 1 0.1898283 8 0.1470326 1.2910632 0.2887416 Blade Y velocit}!. start 1 3.7482083 8 3.9601557 0.94648 0.3591104 Blade Y velocity top 1 0.4474029 8 1.6153219 0.2769744 0.6129653 Blade Z acceleration impact 1 0.0001403 8 0.0019382 0.0723862 0.7947022 Blade Z acceleration start 1 2.1729739 8 0.3403623 6.3842955 0.0354362 YES Blade Z acceleration top 1 0.2063429 8 0.0703412 2.9334562 0.1251165 Blade Z position start

1 4.4588103 8 0.4037642 11.043106 0.0104913 YES Blade Z~osition i~act 1 89.843033 8 432.2681 0.207841 0.6605853 Blade Z position top 1 1.6023562 8 2.0115767 0.7965673 0.3981722 Blade Z velocity impact 1 0.0031419 8 0.0065364 0.4806736 0.5077427 Blade Z velocity start 1 0.7234834 8 0.694158 1.0422459 0.3371768 Blade Z velocity top 1 35.553513 8 47.792088 0.7439205 0.4135147 Center of rotation X impact 1 300.11603 8 77.467026 3.8741133 0.0845688 Center of rotation X start 1 2.9448035 8 313.504 0.0093932 0.9251752 Center of rotation X top_ 1 4185.4482 8 2574.3987 1.6257963 0.2380668 Center of rotation Y imJ!.act 1 199.55679 8 64.59938 3.0891442 0.1168738 Center of rotation Y start 1 1534.9789 8 152.90689 10.038651 0.0132268 YES Center of rotation Y top_ 1 76.874413 8 81.362091 0.9448431 0.3595041 Center of rotation Z impact 1 470.5173 8 115.87225 4.0606556 0.0786531 Center of rotation Z start

1 2121.9031 8 3743.9116 0.566761 0.4731179 Center of rotation Z t~ 1 0.0116977 8 2.0035203 0.0058386 0.9409689 Resultant angular acceleration impact 1 0.0017212 8 0.0012215 1.4090866 0.2692582 Resultant angular acceleration start 1 0.024954 8 0.0428933 0.5817691 0.4675116 Resultant angular acceleration top

1 0.04562 8 9.2940359 0.0049085 0.9458649 Resultant af!fLular di~acement impact 1 0.272091 8 0.071545 3.8030751 0.0869724 Resultant angular dis placement start

63 1 1.0062801 8 0.8110549 1.2407051 0.297679 Resultant angular displacement top

1 0.0684093 8 2.2656434 0.0301942 0.8663668 Resultant angular veloc~impact 1 0.7690168 8 0.2012009 3.8221333 0.0863189 Resultant angular velocitv top 1 0.0060774 8 0.0089933 0.6757678 0.4348622 Resultant angular velocity start 1 90.311867 8 78.592186 1.1491201 0.3150033 Stick frontal impact 1 177.08542 8 19.193417 9.2263632 0.0161246 YES Stick frontal start 1 1894.2991 8 474.40472 3.9930019 0.0807348 Stick frontal top 1 50373.367 8 13595.107 3.7052569 0.0904289 Stick sagittal impact 1 32.331394 8 17.795521 1.8168278 0.2146133 Stick sagittal start 1 1128.0963 8 588.85614 1.9157418 0.2037119 Stick sagittal top 1 175.88509 8 103.67233 1.696548 0.2289835 Stick transverse impact 1 149.95667 8 18.175592 8.2504416 0.0207536 YES Stick transverse start

1 3105.448 8 2317.9973 1.3397117 0.2804724 Stick transverse t~ 1 169.23557 8 69.560745 2.4329176 0.1574317 Feet to pelvis impact 1 0.3408156 8 256.81754 0.0013271 0.9718329 Lead arm Horizontal Adduction/Abduction impact 1 741.15857 8 867.35822 0.8545011 0.3823185 Lead arm Horizontal Adduction/Abduction start 1 82.168617 8 281.26129 0.2921434 0.6035685 Lead arm Horizontal Adduction/Abduction top 1 548.17194 8 89.142479 6.1493912 0.0381237 YES Lead arm Vertical Adduction/Abduction impact 1 97.294556 8 78.441994 1.2403376 0.2977457 Lead arm Vertical Adduction/Abduction start 1 681.24738 8 430.34241 1.5830356 0.2438023 Lead arm Vertical Adduction/Abduction top 1 52.555431 8 183.32593 0.2866776 0.6069168 Lead elbow Flexion/Extension impact 1 85.571411 8 241.31015 0.3546117 0.5679737 Lead elbow Flexion/Extension start 1 133.80803 8 344.14346 0.3888147 0.5502837 Lead elbow Flexion/Extension top 1 11.216817 8 14.706079 0.7627334 0.4079239 Pelvis rotation start 1 75.650375 8 113.17977 0.6684089 0.4372767 Pelvis rotation top 1 0.2320546 8 1033.4292 0.0002245 0.9884112 Trail arm Horizontal Adduction/Abduction impact 1 172.7944 8 255.52107 0.6762433 0.434707 Trail arm Horizontal Adduction/Abduction start 1 3790.9844 8 2028.084 1.8692442 0.2087393 Trail arm Horizontal Adduction/Abduction top 1 181.13557 8 130.80666 1.3847581 0.2731164 Trail arm Vertical Adduction/Abduction impact 1 104.45779 8 53.47382 1.9534378 0.1997571 Trail arm Vertical Adduction/Abduction start 1 1372.7288 8 617.40979 2.2233672 0.1742741 Trail arm Vertical Adduction/Abduction too 1 620.83942 8 178.60921 3.4759653 0.0992584 Trail elbow Flexion/Extension imoact 1 159.0874 8 88.261681 1.8024515 0.2162641 Trail elbow Flexion/Extension start 1 1.0385478 8 324.5228 0.0032002 0.9562747 Trail elbow Flexion/Extension top 1 24.419172 8 68.289383 0.3575837 0.5663905 Trunk rotation impact 1 215.75572 8 160.85904 1.341272 0.2802128 Trunk rotation start

1 141.5154 8 147.98468 0.9562841 0.3567661 Trunk rotation t~

64