Cardiff Metropolitan University | Prifysgol Fetropolitan Caerdydd

LATE SUBMISSION WITH PERMISSION

(MARK TO BE RECORDED AS FIRST ATTEMPT)

Student Name: Student Number:

Sam Bracey 20018354

Programme: SPE Year: 3 Term: 2

Module Number & Title: SSP6050 Independent Project

Original Submission Date: 19 March 2015

Date extension requested: 9th February 2015

Programme Director:

Anna Bryant

Date agreed by PD: 20th February 2015

New Submission Date: 2nd

April 2015

Note: A completed copy of this form must be attached to your dissertation. Both dissertation and form must be submitted by the ‘new submission date’. Failure to do so will result in a maximum mark of 40% as a second attempt.

Cardiff School of Sport DISSERTATION ASSESSMENT PROFORMA: Empirical 1

Student name: Sam Bracey Student ID: St20018354

Programme: SPE

Dissertation title: The Wicketkeeper’s Role in Twenty20 (T20) Cricket: Wicketkeeper or Batsman?

Supervisor: Ray Ponting

Comments Section Title and Abstract (5%)

Title to include: A concise indication of the research question/problem. Abstract to include: A concise summary of the empirical study undertaken. Introduction and literature review (25%)

To include: outline of context (theoretical/conceptual/applied) for the question; analysis of findings of previous related research including gaps in the literature and relevant contributions; logical flow to, and clear presentation of the research problem/ question; an indication of any research expectations, (i.e., hypotheses if applicable). Methods and Research Design (15%)

To include: details of the research design and justification for the methods applied; participant details; comprehensive replicable protocol. Results and Analysis (15%) 2

To include: description and justification of data treatment/ data analysis procedures; appropriate presentation of analysed data within text and in tables or figures; description of critical findings. Discussion and Conclusions (30%) 2

To include: collation of information and ideas and evaluation of those ideas relative to the extant literature/concept/theory and research question/problem; adoption of a personal position on the study by linking and combining different elements of the data reported; discussion of the real-life impact of your research findings for coaches and/or practitioners (i.e. practical implications); discussion of the limitations and a critical reflection of the approach/process adopted; and indication of potential improvements and future developments building on the study; and a conclusion which summarises the relationship between the research question and the major findings. Presentation (10%)

1 This form should be used for both quantitative and qualitative dissertations. The descriptors associated with both quantitative and qualitative dissertations should be referred to by both students and markers. 2 There is scope within qualitative dissertations for the RESULTS and DISCUSSION sections to be presented as a combined section followed by an appropriate CONCLUSION. The mark distribution and criteria across these two sections should be aggregated in those circumstances.

To include: academic writing style; depth, scope and accuracy of referencing in the text and final reference list; clarity in organisation, formatting and visual presentation

CARDIFF METROPOLITAN UNIVERSITY Prifysgol Fetropolitan Caerdydd

CARDIFF SCHOOL OF SPORT

DEGREE OF BACHELOR OF SCIENCE (HONOURS)

SPORT AND PHYSICAL EDUCATION

2014-5

THE WICKETKEEPER’S ROLE IN TWENTY20 (T20) CRICKET: WICKETKEEPER OR BATSMAN?

(Dissertation submitted under the PERFORMANCE ANALYSIS area)

SAM BRACEY

ST20018354

THE WICKETKEEPER’S ROLE IN TWENTY20 (T20) CRICKET: WICKETKEEPER OR BATSMAN?

Cardiff Metropolitan University Prifysgol Fetropolitan Caerdydd

Certificate of student

By submitting this document, I certify that the whole of this work is the result of my individual effort, that all quotations from books and journals have been acknowledged, and that the word count given below is a true and accurate record of the words contained (omitting contents pages, acknowledgements, indices, tables, figures, plates, reference list and appendices). I further certify that the work was either deemed to not need ethical approval or was entirely within the ethical approval granted under the code entered below.

Ethical approval code: 14/5/37U

Word count: 9830

Name: Sam Bracey

Date: 9th March 2015

Certificate of Dissertation Supervisor responsible

I am satisfied that this work is the result of the student’s own effort and was either deemed to not need ethical approval (as indicated by 'exempt' above) or was entirely within the ethical approval granted under the code entered above.

I have received dissertation verification information from this student

Name:

Date:

Notes:

The University owns the right to reprint all or part of this document.

Table of Contents

List of Tables List of Figures Acknowledgements i Abstract ii

CHAPTER 1 - INTRODUCTION 1 1.1 - Rationale for Study 5 1.2 - Research Question 6 1.3 - Hypothesis 6 1.4 - Delimitations 6 1.5 - Limitations 7 1.6 - Glossary of Terms 7

CHAPTER 2 - LITERATURE REVIEW 10 2.1 - Performance Analysis Research Relating to Cricket 11 2.2 - Wicket Keeping 14

CHAPTER 3 - METHOD 18 3.1 - Research Design 19 3.2 - Equipment 21 3.3 - Participants/Sample 21 3.4 - System Development 22 3.5 - Operational Definitions 22 3.6 - Procedure 25 3.7 - Data Analysis 26 3.8 - Reliability 27

CHAPTER 4 - RESULTS 29 4.1 - Wicket-keeping perspective 31 4.2 - Batting perspective 34

CHAPTER 5 - DISCUSSION 38 5.1 - Wicket-keeping overview 39 5.2 - Batting overview 41 5.3 - Practical Implications 43 5.4 - Future Research Recommendations 45

CHAPTER 6 - CONCLUSION 47 REFERENCE LIST 50 APPENDICES Appendix A - Summary of tournament matches excluded from the current study

Appendix B - Summary of tournament matches analysed during the current study

Appendix C - The bespoke notational analysis system used during data collection

Appendix D - Extra column (win/loss) added to the data collection spreadsheet prior to data analysis, to signify winning and losing events

List of Tables

Page

Table 1. Glossary of essential terms relevant to this study 7

Table 2. Operational definitions outlined prior to data collection 23

Table 3. Kappa results from the reliability study 28

Table 4. Wicketkeeper up to the stumps vs wicketkeeper back from 32 the stumps (seam bowling only)

Table 5. Dismissals completed, missed opportunities, and byes 33 conceded by wicketkeepers in winning teams compared to losing teams (spin and seam)

Table 6. Raw batting statistics for the wicketkeeper from each team 34

Table 7. Wicketkeepers’ batting performances in winning teams 35 compared to losing teams

Table 8. Comparison of batting performance - current study vs 42 Lemmer’s (2008; 2011b) studies

List of Figures

Page

Figure 1. Flowchart illustrating the data collection process adopted to 26

fill in the notational analysis system

Figure 2. Number of deliveries where wicketkeepers stood up to the 31 stumps in winning teams compared to losing teams (seam bowling only)

Figure 3. The wicketkeepers’ batting positions in winning and losing 35 teams

Figure 4. The individual batting scores for the wicketkeepers in 36 winning and losing teams

Figure 5. The number of runs per partnership, in which 37 wicketkeepers contributed

Acknowledgements

I would like to thank Ray Ponting for his assistance and continued support throughout my final year at university.

i

Abstract

The context of the study is the current interest in T20, the newest format of the game, often considered to be the future of cricket. With cricketing priorities changing in the last decade, wicketkeepers are currently required to score runs heavily, with selectors sometimes sacrificing wicket-keeping quality. Such a philosophy is debatable, especially since the role of the wicketkeeper is largely ignored in academic research.

As a contribution to the debate, using a basic method, this study sought to provide a greater understanding of the wicketkeeper’s contribution to a winning team’s performance. Should a wicketkeeper’s batting or wicket keeping ability carry the most weight, when selecting a team for T20 cricket?

Using notational analysis, the study compared the performance of wicketkeepers from winning and losing teams, in relation to both their batting and fielding contributions, during the 2014 T20 World Cup in Bangladesh. Particular attention was paid to a specialist skill of wicket keeping, standing up to the stumps. Mann Whitney U tests identified a few significant differences between winning and losing teams.

From a wicket-keeping perspective, winning wicketkeepers stood up to the stumps (3.07  4.42 deliveries) significantly more (p = 0.023) than losing wicketkeepers (0.69  2.07 deliveries), during the powerplay overs. During the non-powerplay overs, standing back emerged as significantly better in relation to both ‘Dot Ball %’ (p = 0.015) and ‘Wicket %’ (p = 0.002).

From a batting perspective, winning wicketkeepers contributed in significantly more (p = 0.005) partnerships of 50 plus runs (8), than losing wicketkeepers (1). Results from both winning and losing teams confirmed the perception that wicketkeepers are generally expected to bat in the top order in T20 cricket.

The study did not conclusively identify whether batting or wicket keeping should be the priority for selection of the wicketkeeper in T20 cricket. As a result, future quantitative research is required, using a larger sample, and qualitative research relating to professional perceptions of the wicketkeeper’s role.

ii

ii

CHAPTER 1 INTRODUCTION

1

The precise origins of “a very old, widespread and uncomplicated pastime” (ESPNcricinfo, 2014a) which has evolved, over centuries, into the modern-day game of cricket are, historically, unclear. What is clear, however, is that the present, more complex game, developed around the basic principle of scoring more runs (see Glossary of Terms) than the opposition - categorised by Hughes and Bartlett (2002) as a striking/field game - has many components integral to successful performance.

Contested by two teams of eleven players, cricket has been played as an international sport since 1844, when Canada hosted the USA in the first official encounter (ICC, 2014a). Subsequently, the game has evolved substantially, with the game now played regularly in a professional capacity across the world. Governed by the International Cricket Council (ICC), there are currently ten, fully-affiliated, test- playing nations (England, Australia, South Africa, Sri Lanka, India, Pakistan, New Zealand, West Indies, Bangladesh and Zimbabwe), with six associate nations, who have international status in the one-day 50-over format (Ireland, Afghanistan, UAE, Scotland, Papua New Guinea (PNG) and Hong Kong) (ICC, 2014b).

International Test Cricket has long been considered the pinnacle of the sport, with each game completed over a period of five days - usually two ‘innings’ per team, testing technical, physical, and mental attributes to the maximum limit (Woolmer et al., 2008). One-day International Cricket (ODI) – presently a single ‘innings’ of 50 ‘overs’ per side - represents the medium between traditional test-match cricket, and the newest form of the game, Twenty20 (T20) – 20 ‘overs’ per side. In relation to the current study, T20 is of particular interest, with many cricket enthusiasts, including Subhani et al. (2012), considering this form as the future of the game, due to its entertainment value and ability to attract large audiences. Described as a ‘sporting statistician’s dream’ (Lewis, 2005, p804), cricket provides the perfect opportunity for significant real-time performance analysis to take place, due to its sequential stop/start nature. A common field for analysis relates to understanding the key performance variables, in the three major formats of the game: Test (e.g. Akhtar and Scarf, 2012); ODI (e.g. Petersen et al., 2008a); and, more recently, T20 (e.g. Douglas and Tam, 2010).

Many cricket-coaching experts, including Woolmer et al. (2008), have explained how there are three main components related to successful performance: batting,

2 bowling, and fielding. Bowling elements appear to have received the greatest attention, due to the advanced level of technicality (e.g. Glazier and Wheat, 2014), and the prominent injury-risk associated with first-class bowling (e.g. Hurrion and Harmer, 2009). Similarly, numerous studies have focused on batting elements from varied perspectives, including biomechanical (e.g. Stretch et al., 1998) and performance (e.g. Bukiet and Ovens, 2006). Despite being labelled by Petersen et al. (2008a) as one of three major components related to successful performance, fielding has not been investigated to the same extent as batting and bowling. This has been demonstrated in a number of studies, which have attempted to assess individual player and team performances without proper consideration of fielding aspects (e.g. Lewis, 2005; Mukherjee, 2013). Furthermore, one specific fielding role remains a largely-ignored quantity in academic research, that of the wicketkeeper, although MacDonald et al. (2013b) reviewed the wicket-keeping literature, and Lemmer (2011a) attempted to construct a performance measure to rank wicketkeepers in terms of contribution.

It is appropriate to illustrate the issues behind the wicket-keeping debate, to provide purposeful direction. Murphy (2004) explained how priorities have changed in the last decade, since wicketkeepers are currently required to score runs heavily, while sometimes sacrificing wicket-keeping quality. Whilst this is a criterion regularly adopted in the current professional game, there is widespread controversy as to whether this is in fact the correct philosophy (e.g. Lemmer, 2011a; Hotten, 2013). This philosophy has been clearly evidenced in recent times, with the recent selection of Jos Buttler as England’s test-match wicketkeeper; a decision which is unlikely to please the wicket-keeping purists (Marks, 2014).

Snell (2014) illustrated how attitudes towards the art of wicket keeping have altered in recent times, since wicketkeepers were previously selected for their outstanding ability with the gloves (e.g. Alan Knott, Jack Russell). However, a shift in approach has seen runs scored taking precedence over wicket-keeping ability, with several current international wicketkeepers beginning their international cricket careers as specialist batsman (e.g. England’s Jos Buttler, South Africa’s AB De Villiers). It is commonly agreed by cricket enthusiasts, that Australian Adam Gilchrist revolutionised the role of the wicketkeeper, with an exceptional ability to perform with both bat and gloves to a world-class standard, in all formats of the game (Woolmer

3 et al., 2008; Lemmer, 2011a). This is no mean feat, and whilst many countries try to find a player to emulate his talent, multiple solutions to the conundrum have been explored. For example, in ODI cricket many teams, including England and Australia, have tried to use their wicketkeeper to open the batting, whereas in Test cricket it is common for the wicketkeeper to bat in the middle order (data obtained from ESPNcricinfo, 2014b).

One specific area for debate has been the role of the wicketkeeper in T20 cricket. In the longer forms of the game (Test and 50-over), justification for selection can be based on the additional requirement for runs, given the length of time teams are required to bat, in order to post a match-winning total. However, in T20 this ought to be less of a consideration, due to the top order batsman often batting the majority of what is, comparatively, a short innings (Amin and Sharma, 2014); unless of course the wicketkeeper bats in the top order. Whilst multiple sides opt for batting specialism taking precedence, a few sides have demonstrated a change of approach in recent times. This was clearly evident during the 2013 Friends Life T20 competition – an England and Wales Cricket Board (ECB) county competition. Interestingly, the competition winners, Northamptonshire, regularly selected a specialist wicketkeeper, who was only required to bat in 3 of his team’s 13 matches. On the occasions in which he was required to bat, he batted no higher than number 8 in the batting order, thus suggesting selection based on his wicket-keeping capabilities. In comparison, the losing finalists, Surrey, utilised their keeper to open the batting in 11 of 13 matches (ESPNcricinfo, 2014c; 2014d). Not only does this show varying approaches, but it may also reflect varied perceptions of not only the actual role performed by specialist wicketkeepers, but also the significance of their contribution as wicketkeepers.

More recently, cricket pundits have expressed their views within the media, and suggestions have been made as to where the priorities of the T20 wicketkeeper could potentially lie in the future. Hotten (2013) questioned the requirement of having a wicketkeeper who bats to a high level in T20 cricket. Using game-based statistics to justify his viewpoint, he proposed that, when the wicketkeeper’s team is fielding, he/she is the only member of the team who has the potential to affect every delivery when using their specialist, wicket-keeping skills. A batsman, in comparison, has the opportunity to affect significantly less, with only the opening batsmen able to affect

4 an unlikely maximum of 50% of the match. It was therefore suggested that the wicket-keeping position should be used as an attacking option for the fielding side. This attacking option should include the employment of a wicketkeeper who has the ability to ‘stand up’ (close to the stumps) to high-quality seam bowling, thus putting the batsman under greater pressure, for example, to stay within the batting crease. Standing up to the stumps has been described as the “true test of glovemanship”, and the aspect of performance which separates the most gifted wicketkeepers, from those not so talented (Woolmer et al., 2008, p350). Whilst Hotten’s (2013) theory is not the result of thorough research, it challenges the theory of Woolmer et al. (2008) who suggested that the modern-day wicketkeeper must be able to bat in the top seven in all forms of the game, and have the ability to score at six an over regularly, in all forms of limited-overs cricket.

1.1 Rationale for Study

The rationale for this study was centred around both the distinct lack of research in this particular area, and the potential impact that an extended knowledge and understanding could have on team selection in the future. This could have added significance if T20 were to become the pre-eminent form of international cricket as some are forecasting (e.g. Subhani et al., 2012). A number of common themes were identified as essential factors to consider. The primary issues which required addressing were: a) The extent to which wicket keeping has been specifically considered within the literature; b) Whether a wicketkeeper should be selected in T20 cricket for his/her wicket- keeping ability or batting ability?

It was hoped that results from this study would provide coaches with a greater understanding of the extent to which wicketkeepers contribute in all facets of T20 cricket. In contrast to the specific wicket-keeping study published by Lemmer (2011a), the current study is concerned not with ranking wicketkeepers but, rather, considering the shared characteristics of wicketkeepers in winning teams, in comparison to wicketkeepers from losing teams.

5

1.2 Research Question

Taking into consideration the stated rationale and, in particular, the primary issues identified, the study aims to answer the following question:

Does a wicketkeeper’s batting have a greater impact than his/her wicket keeping in winning teams, compared to losing teams in T20 cricket?

Using notational analysis, the study compared the performance of identified wicketkeepers in relation to both their batting and fielding contributions. In particular, variables considered included: the number of runs scored, and wicketkeeper positioning in relation to the stumps. As previously alluded to, ‘standing up’ to the stumps is a feature particularly associated with specialist wicket keeping, in comparison to ‘standing back’. Wicketkeeper positioning was analysed in conjunction with relevant key performance indicators (KPIs) identified from previous research. In addition, the influence wicket keeping can have on team selection has also been considered and discussed.

1.3 Hypothesis

Null Hypothesis: There will be no significant difference between the wicket keeping contribution and/or batting contribution of wicketkeepers in winning and losing teams.

Alternative Hypothesis – There will be a significant difference between the wicket keeping contribution and/or batting contribution of wicketkeepers in winning teams compared to losing teams.

1.4 Delimitations

The sample consisted of matches from the men’s 2014 T20 Cricket World Cup hosted in Bangladesh. Although results can be used to facilitate understanding, they should not be generalised directly in relation to amateur cricket and women’s cricket. Furthermore, results should be interpreted with caution, when considering cricket in other locations (e.g. other continents).

6

1.5 Limitations

It was initially intended that all matches from the 2014 T20 Cricket World Cup would be utilised as part of the sample during the current study. However, following construction of the method, it was realised that data could not be collected reliably on 4 matches, due to the required footage being unavailable. In addition, 2 matches were excluded from analysis following data collection, due to the outcome being decided by the Duckworth Lewis (DL) method. It was believed that including these 6 matches would bring the validity of the results into question. Appendix A provides a summary of the matches not included as part of the current study.

It should also be noted at this stage that wicket-keeping performance is largely affected by the performance of his/her team’s bowling, its quality with regards to speed and accuracy of delivery. Lemmer (2011a) identified this as a limitation when considering wicket-keeping performance, but accepted that it is largely unavoidable.

1.6 Glossary of Terms

Table 1. Glossary of essential terms relevant to this study

Term Definition

Twenty20 Cricket Also referred to as T20. A game in which both teams can bat and field for a maximum of 20 overs each. (An ‘over’ is the bowling of six balls by a single bowler). This format is played both internationally and domestically in a professional capacity.

Innings The batting turn of a player or team of players.

Powerplay The opening 6 overs of a T20 innings. During this period only 2 fielders are permitted to be outside of the 30 yard inner circle before the

7

ball has been released.

Delivery The bowling of the ball, by the bowler, towards the batter. The bowler is aiming to hit the ‘stumps’/’wicket’ – a wooden target defended by the batter.

Batter The player who faces the delivery once it has been released by the bowler. It is the batter’s job to score as many runs as possible.

Run The successful running of the batter between the stumps (at each end of the pitch), after hitting the ball, without being given ’out’.

Wicketkeeper The fielder who is stationed closest directly behind the batter’s wicket. He/she is entitled to wear gloves and pads as protection.

NOTE: Given evident inconsistencies, in documents accessed, in the spellings of words relating to the wicketkeeper’s role, it was decided to use the following spellings: wicketkeeper; wicket keeping (noun); wicket- keeping (adjective).

Wicketkeeper Positioning The distance behind the batter’s wicket at which the wicketkeeper stands. ‘Standing-up’ close to the wicket is a skill commonly associated with specialist wicketkeepers in certain situations.

8

Wicket A delivery which results in a batter being dismissed – being ‘given out’ by the umpire – meaning he/she can no longer bat. Modes of dismissal include; caught; bowled; LBW; run out; stumped; hit wicket; handled ball; obstructing the ball; timed out. (‘Wicket’ can also mean the strip of short grass – in the centre of the field of play - on which bowling and batting takes place).

Dot Ball A delivery off which a run hasn’t been scored. If every ball within an over is a dot ball this is known a ‘maiden over’.

Boundary A line marked around the ground to delineate the perimeter of the field of play.

Boundaries Deliveries in which the batter hits the ball for 4 runs (to the boundary) or 6 runs (over the boundary). Once the ball touches or crosses the boundary, the ball is then ‘dead’ – no longer in play.

Economy Rate The average number of runs a bowler concedes in an over. An economy rate of 6 runs per over in limited-overs cricket is considered to be excellent.

(Adapted from Davis and Collins, 2012).

9

CHAPTER 2 LITERATURE REVIEW

10

This review outlines the depth of academic research, previously published, which appears relevant to the aim of this study. Whilst there is a paucity of literature directed specifically towards the wicketkeeper, it is important also to consider the role within the wider context of performance analysis research relating to cricket as a whole.

2.1 Performance Analysis Research Relating to Cricket

The majority of cricket research has tended to be completed under a quantitative research design, involving the analysis of statistics relating to performance, rather than qualitative research which is concerned with interviews and other subjective measures. Probably, this is because the quantitative approach may be considered to be both ‘easier’ to pursue and more objective. Performance analysis of sport requires the investigation of actual sports performance, in both match and training settings (O’Donoghue, 2010). Such investigations have encompassed a variety of forms, including time-motion analysis (e.g. Rudkin and O’Donoghue, 2008), and notational analysis (e.g. Petersen et al., 2008a; 2008b).

From a time-motion analysis viewpoint, studies have tended to focus on movement patterns, particularly during fielding performance. MacDonald et al. (2013a) defined it as the most prevalent approach to quantifying the physiological demands of cricket. Furthermore, appreciation of physiological factors should not be underestimated, given that output data can be used effectively, to construct programmes which match the physical demands of the game. Rudkin and O’Donoghue (2008) used Global Positioning Satellite (GPS) units to observe positional movement of a fielder, positioned at cover point, during a first-class cricket match. Results provided evidence of the percentage of time which fielders spend performing high-intensity activities, but the conclusions had limited value, as only one fielding position was analysed. Using a similar method, Petersen et al. (2009) quantified the time-motion characteristics of five different fielding positions (including the wicketkeeper). In comparison to Rudkin and O’Donoghue (2008), analysis was directed towards T20 cricket, due to its fast, intensive nature. Interestingly, results suggested that wicketkeepers have the lowest physiological workload of all fielders within a T20 cricket team, compared to bowlers, who have the highest workload. Petersen et al.

11

(2009) advocated the use of the results to prescribe individual conditioning programmes. Whilst these results are beneficial from a physiological viewpoint, and can contribute to a greater understanding of individual training needs, they could be considered of limited value when considering a wicketkeeper’s participatory workload. The role is very different to other players, bowlers particularly, fulfilling a role as the most critical, and demanding, of all fielders (Woolmer et al., 2008).

Franks (2004) explained how notational analysis is a statistics-based approach, beneficial to the analysis of team sports. The use of both hand and computerised notation systems can produce valuable feedback, relating to identified key performance indicators (KPIs), useful to improving effective decision-making (Drust, 2010). In relation to cricket, this is imperative, given that the game’s evolution within and between its different forms, reflecting changing philosophies, requires varying aspects of performance from its players to be addressed.

A battery of literature has investigated the key performance indicators (KPI’s) for success within cricket (e.g. Moore et al., 2012). Within performance analysis, such indicators are considered to be dependent variables which represent relevant and important aspects of performance (O’Donoghue, 2010). Given the individual disciplines within cricket (batting, bowling, and fielding), understanding the key success indicators is essential, as this can provide a clear, direct focus for player and team development. In recent times, Petersen et al. (2008a; 2008b) have headed the research in this area, with detailed investigation into the key performance components at international level. They have published analyses from both ODI and T20 major tournaments, with the aim of understanding the key to success. Additionally, Douglas and Tam (2010) have similarly investigated the key T20 performance components. There has been consistent agreement between scholars as to which performance components correlate positively with winning teams. All of those researched have utilised Cohen’s effect size method which compares the magnitudes of differences. The following variables have been attributed to success: wickets, dot balls, boundaries and economy rate (Petersen et al., 2008b; Douglass and Tam, 2010; Moore et al., 2012). Interestingly, it should also be noted that despite the familiar cliché – “catches win matches” – successful catches, dropped catches, and runs conceded as a result of a dropped catch have all been presented as statistically insignificant to success, with other related fielding components

12 discussed to a limited degree. Additionally, these studies identified the powerplay as a critical period for bowling teams, contributing significantly to the match outcome. The significance of the powerplay lies in the restricted fielding positions (see Glossary of Terms), perceived as advantageous to the batting team. This requires the fielding team to be tactically astute to reduce that advantage, with the employment of attacking field placements (Petersen et al., 2008a). The wicketkeeper standing up to the stumps would be such a placement. Whilst considering the samples utilised within the aforementioned studies, it should be noted that all these studies completed their analysis based on statistical analysis of a single, complete tournament, rather than across tournaments in different locations.

Many studies have considered, and consequently constructed, performance measures within cricket (e.g. Lemmer, 2004; 2008; 2011a; 2011b). Gerber and Sharp (2006) expressed the importance of utilising accurate performance measures, with extended discussion of how resulting data can influence team selection at international level. Additionally, they proposed that T20 cricket should be analysed in more detail, particularly in relation to the make-up of a T20 team (team selection criteria). Barr and Kantor (2004) also discussed the importance of using statistics for the purpose of team selection, and advocated the use of a measurement tool which considers a minimum of two dimensions for the analysis of batting performance, for instance, batting average and strike rate as a combined measure.

Lemmer (2008) identified the unsuitability, and unreliability, of batting averages as a measure of batting performance, where analysis relates to only a small number of matches, because of the effect that several not-out scores can have on the outcome. Consequently, he proposed a combined measure utilising average and strike rate, also adjusting values to reflect not-out scores; thus aiming to ensure a fairer measure of batting performance. Subsequently, Lemmer (2011b) supplemented his earlier measure with consideration of the batting conditions faced by each batter in each particular match. He concluded that this was a far superior measure of batting performance compared to previous traditional measures which reflected only average and strike rate components. Whilst appreciating Lemmer’s concern to produce a fairer measure, the adjustments made to reflect both not-out scores, and the perceived batting conditions might be viewed as somewhat subjective, in the absence of additional research to support the adjustments made.

13

Saikia et al. (2012) constructed a performance measure using a double-weighted tool, which assessed fielding performance in T20 international cricket. During the construction of the measure, it was identified that interpretation of fielding contribution should not be based purely on numbers, for instance number of catches, and number of balls fielded. This was referred to as the ‘preparatory measure’, used to assess actual effect. As a result, a final measure was proposed, taking into account the context of the fielding act, and weightings which reflect the importance of specific actions, e.g. a catch which involves diving and/or running is weighted higher than a simple standing catch. This final measure is identified as the ‘fairer measure’. Perhaps, this represents movement away from the pure objectivity of statistics, possibly involving subjective judgements regarding degree of difficulty. Based on the ‘preparatory measure’, the Indian wicketkeeper MS Dhoni emerged as the fielder who had the greatest effect on the outcome of the match, fielding over twice as many deliveries compared to any other fielder. Even applying the ‘fairer measure’, MS Dhoni still emerged as the third most significant player. Given these results, it seems surprising that the particular role of the wicketkeeper is not discussed in more detail, within the study, to any degree.

2.2 Wicket Keeping

Despite being labelled as the only person involved with every delivery (MacDonald et al., 2013b), the role of the wicketkeeper is an area which appears to have been sparsely researched. MacDonald et al. (2013b) conducted a literature review into the role of the wicketkeeper, where they considered a number of factors related to the role. A major strength of the article relates to the level of detail in which physical conditioning components are discussed. This is significant, given the physical expectations of modern day professional cricketers (Petersen et al., 2009). However, this article failed to discuss how the wicketkeeper’s performance with bat and gloves directly influenced the outcome of a game. This represents a serious limitation of the study, particularly given previous references to fielding performance, and how success in this area can create pressure on the opposition (Woolmer et al., 2008; Davis and Collins, 2012; Saikia et al.,2012). A qualitative psychology study published by Miles and Neil (2013) investigated the use of self-talk during elite batting.

14

Although only indirectly related to this study, high-level fielding performance and application of field changes emerged from interviews as external pressures felt by batsman (e.g. a wicketkeeper’s own positioning, and advice regarding field placing, could affect tactical decision making).

Iyer and Sharda (2009) used a neural networking model to predict player performance. Whilst wicket-keeping performance was not directly investigated, concentrating as it did on analysis of batting and bowling disciplines, the authors advised future research into the contribution of the wicketkeeper. They considered his/her ability as imperative to a team, primarily from a wicket-keeping viewpoint. With reference to future research direction, MacDonald et al. (2013b) advised most notably the analysis of the wicketkeeper from a performance perspective, with a greater appreciation of tactical factors.

A recent article published by Amin and Sharma (2014) assessed player performance in T20 cricket. In contrast to the majority of articles, the wicketkeeper was considered specifically. Unfortunately, the methodology epitomised the current perception of modern-day wicketkeepers, whereby selection of the wicketkeeper is justified purely on batting statistics, and ignored wicket-keeping performance. It did not reflect the advice of studies, referred to previously, which stated the need for due consideration to be given to the actual activity of keeping wicket. This represents a major limitation, given the way in which a wicketkeeper’s presence behind the stumps can influence proceedings. Woolmer et al. (2008) have described the wicketkeepers tactical role as one of the most critically demanding roles. A range of coaching manuals including Tennant (2009) have described the role of the wicketkeeper, making reference to what specialist ‘glovemen’ offer in comparison to non-specialists. Woolmer et al. (2008) recognised the significance of the specialist wicketkeeper’s role, tactically. They drew attention to how such wicketkeepers apply pressure through exceptional skills, particularly standing up to the stumps. This tactic can impact on the batsman, resulting in a batting mistake (e.g. loss of wicket).

In comparison to many studies which have focussed on measuring performance, Lemmer (2011a) did identify the need for individual analysis of the wicketkeeper, recognising that a sharp performance behind the wickets can impact significantly on the outcome of a match. Using a simple method, the study attempted to combine

15 both batting and wicket-keeping criteria into a measure of describing wicket-keeping performance; a step forward on previous, related studies which based analysis on batting statistics alone. Lemmer (2011a) argued that it would be inappropriate to base a wicketkeeper’s performance only on dismissal rates and byes, but not batting, given that these statistics are often influenced significantly by the quality of the bowlers. Consequently, he constructed a bespoke measure, for the analysis of ODI and test-match wicketkeepers. He reasoned that “the two most important criteria, the batting performance measure (BP) and the dismissal rate (D), can be combined into a measure of wicket keeping” (Lemmer, 2011a, p90). He included byes conceded, in the test-match context. However, limitations of the study are reflected within the construction of the measure. Firstly, it is questionable whether ‘dismissal rate’ (even with byes) is a sufficient factor in determining the wicket- keeping contribution; other contributions require consideration. Secondly, byes were not included during the analysis of ODI wicket keeping, despite being a common (and reasonable) factor in wicketkeeper assessment (Woolmer et al., 2008). Additionally, from a bald statement of this measure, it could be assumed that batting and wicket keeping should be considered as of equal significance; however, the measure does require wicket-keeping performance to be calculated at three times the priority over batting performance. This recognises the significance of the wicketkeeper’s wicket-keeping role over his/her batting contribution, but it is unclear why it carries a tariff three times that of batting. Unfortunately, for this study, Lemmer’s proposals did not reflect the Twenty20 format of the game.

A broader issue is that the study by Lemmer (2011a) was primarily concerned with ranking wicketkeepers in terms of competency, with Adam Gilchrist emerging as the number-one-ranked wicketkeeper in test matches, followed by Brad Haddin and Kamran Akmal; whereas Gilchrist, Haddin, and MS Dhoni formed the top three ranked wicketkeepers in ODI cricket. Whilst this is interesting to know, it does not distinguish specific attributes successful wicketkeepers possess, in comparison to unsuccessful wicketkeepers. Additionally, it does not consider how wicketkeepers contribute specifically in winning sides, compared to losing sides.

Whilst it is easy to assess batting performance from an objective viewpoint (e.g. averages, strike-rates), objectively assessing a wicketkeeper’s performance could be considered more difficult. This presents the greatest challenge when it comes to

16 team selection (Gerber and Sharp, 2006). The use of a research design which incorporates notational analysis, and an appreciation for statistics, has been advocated (MacDonald et al., 2013a).

17

CHAPTER 3 METHOD

18

3.1 Research Design

The study’s aim was to contribute to increased understanding of whether a wicketkeeper’s wicket-keeping or batting capabilities should take precedence in team selection. The study focussed on the contribution of the wicketkeepers in a recent international T20 cricket tournament. Lemmer (2011a) had utilised a performance measure relating to wicketkeepers, to individually rank them, but this was related to Test match and ODI cricket only. Lemmer’s measure was considered inappropriate for this study which had a different focus, to understand the significance of the wicketkeeper’s contribution in winning sides, compared to losing sides, in T20 cricket. Consequently, the researcher created a bespoke system of notational analysis to collect quantitative data on selected variables, to produce relevant results. Both identified aspects of wicketkeeper performance, wicket keeping and batting, were analysed.

Wicket Keeping

Given that standing up to the stumps to seam bowling has been highlighted as ‘the true test of glovemanship’ (Woolmer et al., 2008, p350), and the aspect of the role which separates the exceptional from the rest, the impact of wicketkeepers standing up to seam bowling has been presented as an individual concept. This study considered the influence of wicketkeeper positioning at various stages of the game, specifically during powerplay (overs 1-6) and non-powerplay (overs 7-20).

Although standing up to seam bowlers has been discussed as the criteria for excellence, Woolmer et al. (2008) also alluded to the challenge faced when completing dismissal opportunities delivered by spin bowlers. As a result, dismissals completed, missed opportunities, and byes conceded by wicketkeepers in winning teams compared to losing teams (context of both spin and seam bowling), were considered.

The variables used for measurement were those which have been identified as the KPIs in T20 cricket (Petersen et al., 2008b; Douglass and Tam, 2010; Moore et al., 2012), which are dot-balls, wickets, boundaries, economy rate.

Despite being proposed as insignificant by Petersen et al. (2008b), catches were also included within the performance assessment criteria, along with stumpings,

19 given that Lemmer (2011a) identified rates of such dismissals as an important aspect of wicketkeeper analysis. Byes were analysed as a final factor, given their inclusion has been considered a common factor in wicketkeeper assessment (Lemmer, 2011a).

The method centred on analysing the effect of wicketkeeper positioning – specifically the specialist skill of standing up to the stumps, in comparison to the less-advanced skill of standing back from the stumps. Wicketkeeper performances in winning and losing teams were compared, distinguishing between powerplay and non-powerplay overs.

Analysis, on the basis of wicketkeeper positioning, was related to the following:

 Number of deliveries up to the stumps.  Impact on opposition’s batting performance in relation to identified KPIs.  Wicketkeeper dismissals and byes conceded.

Batting

Taking into consideration Barr and Kantor’s (2004) recommendations, regarding measurement tools, that batting performance should include a minimum of two dimensions (aspects of batting performance), the study used six dimensions for measurement variables. Five performance indicators (PIs) were used: deliveries faced, dot-balls, boundaries, wickets, individual averages and strike rates. Also, the positioning of the wicketkeeper in the batting order was noted.

Previous studies have discussed the value of partnerships in excess of 25 runs, but particularly the importance of partnerships in excess of 50 runs, which has been consistently presented as a KPI correlated with winning teams (Douglass and Tam, 2010; Petersen et al., 2008a; Petersen et al., 2008b). So, additional to individual batting performance, the wicketkeeper’s contribution as part of significant batting partnerships was analysed.

Analysis of batting was related to the following:

 Raw batting statistics in relation to identified KPIs.  Number of runs per batting partnership.

20

 Wicketkeepers’ positions in teams’ batting orders.

3.2 Equipment

Obtaining Footage:

 Sanyo DVD recorder DVR – V100E (used to record live matches shown on Sky Sports on to blank discs)  Panasonic DMP-BD79 blue ray disc player (used to check that all footage had been successfully transferred on to disc, and was therefore available for use during the study)

Data Collection:

 LOGIK L22FED13 22" TV - DVD Player (used to retrieve the required data from the recorded footage)  Apple iPad MD786B/A - 16GB (used to retrieve the required data from ESPNcricinfo)  HP Pavilion 15-p273na 15.6" Laptop  Microsoft Excel 2010

Data Analysis:

 IBM SPSS (Software) - Version 20.0 for Windows

3.3 Participants/Sample

Previous studies directed towards T20 cricket have tended to use a complete major tournament as the sample for analysis (e.g. Petersen et al., 2008b; Douglass and Tam, 2010). Following this trend, this study retrospectively analysed performances from 29 matches of the 2014 T20 World Cup held in Bangladesh. Not only did this provide a large sample size, but the sample also represented diversity, with international teams from various continents and at differing levels within the ICC World Rankings. The wicketkeeper from each team was identified prior to data collection commencing. Appendix B provides a breakdown of the matches used

21 during the current study, with identification of the wicketkeepers analysed in each match.

3.4 System Development

A pilot study was completed using footage from a pre-tournament warm-up match between England and India. Its aim was to test the effectiveness of the bespoke hand-notation template, whilst also checking for data retrieval issues, which could potentially impact upon reliability. Initially, it was intended that scorecards and ball- by-ball commentary, published by ESPNcricinfo, would be used during data collection to retrieve the necessary ball-by-ball data. However, following early research, it emerged that this source did not consistently provide the data required to fulfil the aim of the study (particularly relating to wicketkeeper positioning). This presented a major reliability issue, so an alternative method for data retrieval was explored. All World Cup matches covered by Sky Sports, were recorded on to disc via a Sanyo DVD recorder DVR – V100E. Once recorded, a Panasonic DMP-BD79 blue ray disc player was used to check that all footage was available for use, and matches were copied on to individual blank discs for external use. Ball-by-ball data was collected using a combination of visual lapsed-time analysis (to establish wicketkeeper positioning), and ESPNcricinfo-published commentary (to record delivery outcome).

Results from the pilot study suggested that the proposed bespoke template (See Appendix C) and amended retrieval procedure would fulfil the purpose of the study. No further modifications were required from a design viewpoint. A significant strength of the system was the extensive data that could be collected, relating to each single delivery, without compromising the reliability and validity of data entry.

3.5 Operational Definitions

Before data collection commenced, operational definitions were set to avoid ambiguity and consequent inaccuracy during data entry. O’Donoghue (2010) expressed the importance of outlining precise operational definitions prior to data

22

collection, so that there is a shared understanding of the variables used. These definitions (Table 2) relate to the nine key points for consideration in data collection. Furthermore, as a consequence of the pilot test, the appropriate/reliable data retrieval source to code each variable has been identified (in italics) as part of the operational definitions.

Table 2. Operational definitions outlined prior to data collection

Consideration Point Operational Definition 1. Wicketkeeper Positioning Recorded footage was watched in order to obtain this data accurately.

Up – If the wicketkeeper is standing within immediate reach of the stumps at the point of delivery. Back – If the wicketkeeper is not standing within immediate reach of the stumps at the point of delivery.

2. Over Number Objective information obtained from ESPNcricinfo published commentary.

3. Bowler Type Objective information was provided by Sky Sports prior to each over commencing. Recorded footage was also watched in order to ensure accuracy of data, for verification. Bowler types relevant to this study: ‘Right-arm seam’, ‘Left- arm seam’, ‘Off spin’, ‘Leg spin’, ‘Left-arm leg spin' and ‘Slow left arm’.

4. Delivery Outcome Objective information obtained from ESPNcricinfo published commentary.

23

5. Mode of Dismissal Objective information obtained from ESPNcricinfo published commentary.

6. Missed Chance by Recorded footage was watched in order to obtain this data Wicketkeeper accurately.

Dropped Catch – If the batsman hits the ball, without bouncing, to the wicketkeeper and he drops it. The chance is confirmed as missed when the ball hits the ground. Missed Stumping – When the batsman is out of the crease, misses the ball, and the wicketkeeper fails to remove the bails (with ball in hand). The chance is confirmed as missed when the batter regains his ground without the bails being removed in time.

7. Wicketkeeper on Strike Objective information obtained from ESPNcricinfo published commentary.

8. Number of Runs Objective information obtained from ESPNcricinfo published commentary.

9. Batting Position Objective information obtained from ESPNcricinfo published commentary.

24

3.6 Procedure

Sequential ball-by-ball data for each innings, within each match, was collected chronologically (Figure 1) and inputted into the finalised notation system, reflecting operational definitions:

Relating to wicket keeping

1. Fielding team’s wicketkeeper positioning was identified, i.e. ‘Up’ or ‘Back’ 2. Identification of over number, i.e. ‘1-20’ 3. Bowler Type, i.e. ‘Right arm seam’, ‘Left arm seam’, ‘Off spin’, ‘Leg spin’, ‘Slow left arm’, ‘Left arm leg spin’ 4. Delivery outcome was coded appropriately, i.e. ‘No-ball 1-5’, ‘Dot’, ‘Out’, ‘Wide 1-5’, ‘Leg-bye 1-5’, ‘Bye 1-5’, ‘1-6’(Runs), ‘Penalty Runs (5)’ 5. Where applicable, mode of dismissal was then recorded, i.e. ‘LBW’, ‘Bowled’, ‘Double Hit’, ‘Hit Wicket’, ‘Keeper Catch’, ‘Keeper Stump’, ‘Catch Other’, ‘Run- Out’ 6. Missed opportunities by the wicketkeeper were identified, i.e. ‘Dropped Catch’, ‘Missed Stumping’

Relating to batting:

7. Batting team’s wicketkeeper on strike, i.e. ‘Yes’ or ‘No’ 8. Number of runs per delivery recorded, i.e. ‘Dot’, ’1-6’, ‘Out’ 9. Batman’s position in batting order, i.e. ‘1-11’

25

‘Up’, ‘Back’

‘1’, ‘2’, ‘3’, ‘4’, ‘5’, ‘6’, ‘7’, ‘8’, ‘9’, ‘10’, ‘11’ ‘1-20’ 1. Wicketkeeper Positioning

’Right arm seam’, ‘Left 9. Batting Position 2. Over arm seam’, ‘Off spin’, ‘Dot’, ‘1-6’, ‘Leg spin’, ‘Slow left ‘Out’ arm’, ‘Left arm leg spin’

8. Number of 3. Bowler Type Runs

7. Wicketkeeper 4. Delivery on Strike Outcome

‘No-ball 1-5’, ‘Dot’, ‘Out’, ‘Wide 1- ‘Yes’, ‘No’ 6. Missed Chance 5. Mode of 5’, ‘Leg-bye 1-5’, ‘Bye 1-5’, ‘1- by Wicketkeeper Dismissal 6’(Runs), ‘Penalty Runs (5)’

‘Dropped Catch’, ‘‘LBW’, ‘Bowled’, ‘Double Hit’, ‘Hit ‘Missed Wicket’, ‘Keeper Catch’, ‘Keeper Stumping’ Stump’, ‘Catch Other’, ‘Run-Out’

Figure 1. Flow chart illustrating the data collection process adopted to fill in the notational analysis system

3.7 Data Analysis

Following collection, all raw data was collated into a single Excel spreadsheet, before being uploaded into SPSS in preparation for statistical analysis. To ease analysis, an additional column was added to the combined Excel spreadsheet, and each row of data was labelled as either winning or losing, from the bowling team’s perspective (Appendix D). Variables were cross-tabulated to generate descriptive statistics, and to allow for the calculation of dot-ball percentages, boundary

26 percentages and wicket percentages of relevance to the KPIs. Using O’Donoghue’s (2012, p129) test decision tree, an appropriate statistical test was selected, for consideration of the study’s hypothesis from a statistical viewpoint. Given that the data collected during the study were of a non-parametric nature, and that the hypothesis was concerned with identifying differences between wicketkeepers, in winning and losing teams (two independent samples), the Mann Whitney U test was chosen. This was conducted in relation to all variables, so that significant differences between winning and losing teams (beyond chance) could be reported.

3.8 Reliability

In all forms of performance analysis, assessment of reliability is considered an essential measurement for determining accuracy of results (Gratton and Jones, 2010). Particularly appropriate to this study is O’Donoghue and Longville’s (2004) observation that reliability is critically important in relation to coaching contexts, because of the effect that data could have on both player and coach decision- making.

Given that data collection was based primarily upon events with objective operational definitions, and that lapsed-time analysis was conducted, so allowing for uncertainties to be reviewed, near perfect levels of reliability could be expected. However, to ensure this was the case, both intra-observer reliability and inter- observer reliability were tested, applying a critical procedure recommended by O’Donoghue (2007), to check for data entry errors. Such errors were classified, by James et al. (2002), as a common source of inaccuracy related to reliability.

This study’s author collected data on the first innings of a selected match (Hong Kong v Bangladesh), before re-collecting data on the same passage of play a week later, to avoid any memory and recall of events (O’Donoghue, 2012) – a test of intra- observer reliability. Using the operational definitions provided, a trained observer also collected data on the selected passage of play - a test of inter-observer reliability. Prior to this, informed discussions (aided by example video sequences) took place between the two observers to ensure shared understanding of operational

27

definitions (particularly wicketkeeper positioning), as proposed by O’Donoghue (2007).

A Kappa test was selected to measure levels of reliability, as recommended by O’Donoghue (2012) when collecting data on nominal variables. Results generated from the Kappa test are presented in Table 3. Altman (1991, p404) stated that Kappa values of between 0.81 and 1.00 represent a ‘very good’ level of reliability. Given that all of the results expressed in Table 3 indicate a perfect level of agreement (1.00), it can be assumed that the system selected for the current study is reliable, and that all results produced as a result can be deemed reliable.

Table 3. Kappa results from the reliability study

Data Collection Component Intra-Observer Reliability Inter-Observer Reliability 1. Wicket Keeper Positioning 1.00 1.00 2. Over 1.00 1.00 3. Bowler Type 1.00 1.00 4. Delivery Outcome 1.00 1.00 5. Mode of Dismissal 1.00 1.00 6. Missed Chance by Wicketkeeper 1.00 1.00 7. Wicket Keeper Runs 1.00 1.00 8. Number of Runs 1.00 1.00 9. Batting Position 1.00 1.00

28

CHAPTER 4 RESULTS

29

The data on the role of wicketkeeper in winning and losing teams are presented below. The results are presented in two parts: ‘Wicket keeping perspective’, which considers the effects that positioning (up/back) had on the opposing team’s batting success, and ‘Batting perspective’, which covers the contributions made by the wicketkeeper in terms of batting performance. In addition to the basic performance measures (reflecting the KPIs previously discussed), p values from Mann Whitney U tests are presented to indicate statistical significance. Given that the study’s hypotheses are concerned with assessing the difference between winning and losing teams, the p values presented reflect results from a two-tailed test.

30

4.1 Wicket-keeping perspective

Data from both winning and losing teams (seam bowling only) were split into four separate groups; powerplay up, powerplay back, non-powerplay up, and non- powerplay back. Figure 2 compares wicketkeeper positioning in relation to seam bowling.

12

10

8 *

6

4

2 Deliveries up to the stumps the to up Deliveries 0 Winning Losing Winning (Non- Losing (Non- (Powerplay) (Powerplay) Powerplay) Powerplay) Team Type and Period

* Significantly different, statistically, from ‘Losing (Powerplay)’ (p < 0.05)

Figure 2. Number of deliveries where wicketkeepers stood up to the stumps in winning teams compared to losing teams (seam bowling only)

Figure 2 shows that, during the powerplay, there was a significant difference (p = 0.023) between the wicketkeepers in winning teams standing up (3.07  4.42) compared to wicketkeepers in losing teams (0.69  2.07). There was no significant difference for the non-powerplay overs (p > 0.05).

Table 4 shows the effect that wicketkeeper positioning had on selected variables.

31

Table 4. Wicketkeeper up to the stumps vs wicketkeeper back from the stumps (seam bowling only)

Variable Powerplay (Overs 1-6) Non-Powerplay (Overs 7-20) Up Back Significance Up Back Significance Dot Ball (%) 57.07  19.20 50.69  13.04 0.259 22.24  18.06 34.42  11.67 0.015* Boundary (%) 15.48  17.17 17.89  9.41 0.325 15.69  17.08 17.29  8.15 0.238 Wicket (%) 6.78  8.78 2.98  2.95 0.814 2.21  3.41 6.05  4.18 0.002* Economy Rate 5.67  4.20 6.87  2.40 0.179 8.52  4.57 7.98  2.44 0.942 (Average runs per over) * Statistically Significant (p < 0.05) Bold - identifies the better of the two results within each variable

For the powerplay overs all ‘Up’ values were more positive than ‘Back’ ones, but there were no statistically significant differences. For the non-powerplay overs, results were inconsistent. However, ‘Dot Ball (%)’ was statistically significant with ‘Back’ (34.42  11.67) being better than ‘Up’ (22.24  18.06). Similarly, ‘Wicket (%)’ showed a statistically significant difference, with ‘Back’ (6.05 

4.18) being better than ‘Up’ (2.21  3.41).

32

Table 5 gives the immediate contribution made in terms of dismissals, missed opportunities and byes conceded by the wicketkeepers.

Table 5. Dismissals completed, missed opportunities, and byes conceded by wicketkeepers in winning teams compared to losing teams (spin and seam)

Variable Winning Losing Significance Catches 0.79  0.86 0.52  0.91 0.107 Dropped Catches 0.03  0.19 0.07  0.26 0.557 Stumpings 0.34  0.48 0.24  0.51 0.289 Missed Stumpings 0.07  0.26 0.10  0.31 0.643 Byes 0.97  1.78 1.10  2.26 0.647 Bold - identifies the better of the two results within each variable

There was no significant difference between winning and losing teams in relation to these parameters (p > 0.05). However, all variables for the winning teams’ wicketkeepers were more positive.

33

4.2 Batting Perspective

Table 6. Raw batting statistics for the wicketkeeper from each team

Team (Wicketkeeper) Innings Runs Deliveries Wickets Average Strike Rate

Afghanistan (Mohammad Shahzad) 3 74 73 3 24.67 101.37 Australia (Brad Haddin) 3 24 21 2 12.00 114.29 Bangladesh (Mushfiqur Rahim) 5 118 103 5 23.60 114.56 England (Jos Buttler) 3 43 42 3 14.33 102.38 Hong Kong (Jamie Atkinson) 3 44 44 3 14.67 100.00 India (MS Dhoni) 3 50 44 1 50.00 113.64 Ireland (Gary Wilson) 1 0 2 1 0.00 0.00 Nepal (Subash Khakurel) 3 86 88 3 28.67 97.73 Netherlands (Wesley Barresi) 5 108 92 4 27.00 117.39 New Zealand (Luke Ronchi) 2 7 7 2 3.50 100.00 Pakistan (Kamran Akmal) 4 70 77 3 23.33 90.90 South Africa (Quinton De Kock) 5 64 67 5 12.80 95.52 Sri Lanka (Kumar Sangakkara) 4 69 64 3 23.00 107.81 West Indies (Denesh Ramdin) 2 5 12 2 2.50 41.67 Zimbabwe (Brendan Taylor) 1 59 46 1 59.00 128.26 Overall 47 821 782 41 20.02 104.99

34

Figure 3 shows the wicketkeepers’ batting order positions in each innings for winning and losing teams.

Winning Losing

8 8

7 7 6 6 5 5 4 4 3 3

2 2 Batting Position Batting Batting Position Batting 1 1 0 0 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Innings Number Innings Number

Figure 3. The wicketkeepers’ batting positions in winning and losing teams

The mean wicketkeeper batting position for winning teams was (3.30  1.87), in comparison to losing teams (4.04  2.17). No significant difference was found (p > 0.05).

Table 7 compares the batting performance of wicketkeepers from winning and losing teams against selected variables.

Table 7. Wicketkeepers’ batting performances in winning teams compared to losing teams

Variable Winning Losing Significance Deliveries Faced 20.90  17.15 13.56  11.03 0.168 Runs Scored 21.85  21.19 14.22  15.00 0.400 Dot Balls 9.40  6.95 6.37  4.73 0.148 Boundaries 2.60  2.82 1.89  2.24 0.458 Wickets 0.85  0.37 0.89  0.32 0.696 Bold - identifies the better of the two results within each variable

35

Results presented in Table 7 show differences, but these were not significant (p > 0.05).

Figure 4 elaborates upon the ‘Runs Scored’ variable shown in Table 7 by showing the wicketkeepers’ individual batting scores from each innings (the inserted line at 10 is to highlight differences in individual batting scores below 10 runs, whereas the lines at 30 and 50 runs are to highlight differences in higher scores).

Winning Losing

70 70

60 60 50 50 40 40 30 30

Number of Runs Number Number of Runs Number 20 20 10 10 0 0 0Figure5 5. 10Shows15 the wicketkeepers’20 25 30 individual batting0 scores5 10for each15 innings20 25 30 Innings Number Innings Number

Figure 4. The individual batting scores for the wicketkeepers in winning and losing teams

There was no significant difference between winning and losing teams, in relation to individual scores of under 10 runs, individual scores of 30 or more, and individual scores of 50 or more (p > 0.05).

Wicketkeepers’ batting performances are shown, in terms of partnerships, in Figure 5 (the inserted lines at 25 runs and 50 runs, are to highlight differences in longer partnerships).

36

Winning Losing

125 125

100 100

75 75 50 50

25 25

Number of Runs of Number Number of Runs Number 0 0 0 10 20 30 40 0 10 20 30 40 Number of Partnerships Number of Partnerships

Figure 5. The number of runs per partnership, in which wicketkeepers contributed

Though Figure 5 appears to show an obvious difference, with mean wicketkeeper partnership length for winning teams being higher (28.69  27.22) than losing teams (17.02  19.63), this was not significant. However, the number of partnerships of 50 runs or more in which wicketkeepers contributed was significantly higher (p = 0.005) for winning teams (8), compared to losing teams (1).

37

CHAPTER 5 DISCUSSION

38

Various media reports (Hotten, 2013; Marks, 2014; Snell, 2014) and cricket writers (Woolmer et al., 2008), have questioned the role of the wicketkeeper and discussed whether selection criteria should be based on wicket keeping ability or batting ability. The current study aimed to investigate the impact of a wicketkeeper’s contribution in winning teams, compared to losing teams, from both wicket-keeping and batting perspectives.

5.1 Wicket-keeping overview

In previous works, coaches, including Woolmer et al. (2008), have alluded to the positive impact that wicket keeping can have on matches, particularly when standing up to seam bowling. Given that it appears that no studies have previously investigated the impact of wicketkeeper positioning, this study analysed the effect of wicketkeeper positioning on identified KPIs.

Figure 2 shows that during the powerplay overs, winning wicketkeepers stood up to seam bowling (3.07  4.42) for significantly more deliveries (p = 0.023), than losing wicketkeepers (0.69  2.07). Table 4 appears to show advantages of the wicketkeeper standing up to seam bowling during the powerplay overs of a T20 match. Although inconclusive (no significant differences were found), all variables considered (dot ball percentage, boundary percentage, wicket percentage, and economy rate) were positively affected by the wicketkeeper standing up to the stumps during the powerplay, compared to when the wicketkeeper was standing back. A larger sample would show whether or not, during the powerplay, both the significance of winning wicketkeepers stood up to seam bowling increases, and the difference between standing up and standing back, in terms of performance variables, becomes significant. Whilst the powerplay overs only make up 30% of the innings, the importance of this period has consistently been alluded to by scholars (Petersen et al., 2008a; Petersen et al., 2008b; Douglass and Tam, 2010), which enhances the importance of these results.

During the non-powerplay overs, there was no significant difference between winning and losing wicketkeepers, in terms of number of deliveries for which they stood up to seam bowling. Interestingly, however, a significant difference (p = 0.015) was found

39 for dot ball percentage during this period, with standing back (34.42  11.67) being better that standing up (22.24  18.06). Furthermore, a significant difference (p = 0.002) was also found during these overs for wicket percentage, with standing back (6.05  4.18) being better than standing up (2.21  3.41). Although there appears to be no peer-reviewed literature either to support or question these findings, they seem surprising, given that it is standing up which has been identified by coaches as the specialist skill, which can have a negative influence on batting performance (Woolmer et al., 2008; Russell, 2014), not standing back. However, this finding may reflect suggestions made by Petersen et al. (2008b), who advised the employment of defensive (run restrictive) field placements during the non-powerplay overs, and attacking, wicket-taking, fielding placements during the powerplay overs. Hotten (2013) identified the wicketkeeper standing up as an attacking fielding position. A future study could investigate this particular finding in more depth. Perhaps, by splitting the innings into three periods (first 6 overs, middle 8 overs, and final 6 overs), in comparison to the powerplay/non-powerplay divide chosen during the current study, greater clarity might be obtained as to these findings’ significance.

Consistent with findings from previous studies (Petersen et al., 2008a; Petersen et al., 2008b; Douglass and Tam, 2010), results suggested that completed dismissals (catches and stumpings), and missed opportunities (dropped catches and missed stumpings), had no significant impact on the outcome of matches, supporting the null hypothesis. Despite this finding, it would be inappropriate to neglect further discussion into the impact of wicket keeping dismissals, given that wickets have been consistently identified as a KPI for T20 cricket. Lemmer (2011a) considered dismissal rate as one of the two most important criteria when considering wicket keeping performance. When Saikia et al. (2012) undertook a study of fielding performance, a double weighted measure was used during calculation, which recognised the context and difficulty of the fielding act; an assessment not undertaken during this study. However, such an approach is to be commended, and any future study might well benefit from assessing the quality of wicket-keeping acts (particularly dismissals) in a similar way, by utilising a measure which accounts for the difficulties faced when completing dismissals standing up to seam bowlers, compared to when standing back. This might make a meaningful contribution to discussions regarding the criteria employed in wicketkeeper selection in the future.

40

5.2 Batting overview

The study identified that, throughout the tournament, the mean batting position for wicketkeepers in winning teams was (3.30  1.87), in comparison to the losing teams mean of (4.04  2.17). This is not a significant difference, so supporting the null hypothesis. However, the finding does confirm the current perception of wicketkeeper batsman in international T20 cricket, whereby winning and losing teams both adopted strategies by which their wicketkeeper was regularly employed to bat in the top order (batting positions 1-5). This indicates a consistent, notable appreciation of the wicketkeepers’ batting ability. However, it is worth noting that the wicketkeepers in winning teams only batted in 20 of the 29 matches.

Whilst considering the wicketkeepers’ batting performances (see Table 7), there was no significant difference between winning and losing teams in terms of deliveries faced, runs scored, dot balls faced, boundaries scored, and times dismissed, thus supporting the null hypothesis. It could be argued that these results undermine the significance of batting in a wicketkeeper’s contribution to a winning team. However, when considering batting performance from a team perspective (see Figure 5), wicketkeepers in winning teams contributed significantly (p = 0.005) to more partnerships of 50 runs or more, than wicketkeepers in losing teams; supporting the alternative hypothesis. It should also be noted that, proportionally, the winning wicketkeepers participated in 8 out of 28 partnerships of 50 runs or more, compared to 1 out of 16 for losing wicketkeepers. This is of particular importance given that previous research into T20 cricket has consistently recognised and discussed the importance of partnerships of 50 plus runs (Petersen et al., 2008a; Petersen et al., 2008b; Douglass and Tam, 2010).

There have only been a limited number of peer-reviewed studies published which have measured individual T20 batting performance (e.g. Lemmer, 2008; 2011b), making it difficult to determine with certainty whether or not wicketkeepers from both winning and losing teams performed well as batsman in the current study. However, using results presented by Lemmer (2008; 2011b), in studies which similarly measured the T20 batting performance of several international players using a combined measure which included the consideration of averages and strike rates, it is possible to calculate and compare results against those produced as a result of

41

the current study (see Table 8). Despite concerns raised by Lemmer (2008) regarding the suitability of using batting averages alone to measure batting performance, it can be considered acceptable, in this instance, as a large combined sample is being considered. Lemmer’s studies considered the 46 (Lemmer, 2008) most effective batsmen from the 2007 T20 world cup, and the 47 (Lemmer, 2011b) most effective batsmen from the 2009 Indian Premier League (IPL), respectively. The criterion for selection in both studies was for each batsman to have batted a minimum of 3 innings. Therefore, the comparison shown in Table 8 (current study) selected only those wicketkeepers who each completed a minimum of 3 innings. Mann Whitney U tests were used to identify significant differences between average and strike rate means from the current study and Lemmer’s studies.

Table 8. Comparison of batting performance - current study vs Lemmer’s (2008; 2011b) studies

Current Study* Lemmer (2008)** Lemmer (2011b)** Tournament 2014 World Cup 2007 World Cup 2004 IPL

Batting Average 23.10  10.72 30.84  16.19 31.20  12.07†

Strike Rate 105.05  8.95 136.46  25.40†† 124.83  16.90†† * Combined data from both winning and losing wicketkeepers ** Studies considered all batsmen (not only wicketkeepers) † Statistically significant (p < 0.05) †† Statistically highly significant (p < 0.001)

Although, the current study was concerned with assessing only the wicketkeepers’ batting performance, it is appropriate to compare their performance against that of specialist batsmen, given the frequent references which have suggested that current wicketkeepers need to be good batsmen, capable of making valuable runs, potentially in the top order (Woolmer et al., 2008; Lemmer, 2011a; Amin and Sharma, 2014). Results in Table 8 show several significant differences, with Lemmer’s values being more positive than those presented in the current study; three instances out of four statistically, significantly different. Based on this knowledge, it can be argued that the wicketkeepers underperformed as batsman during the 2014 T20 World Cup. It is reasonable to assume that a number of

42 batsmen-wicketkeepers within this present study should be within a ‘top 47’ as used by Lemmer. Perhaps this raises further questions regarding the significance of batting in a wicketkeeper’s required contribution. Of course, this situation may reflect dissimilar performance contexts, for instance, different location for World cup, different situations as IPL is not the same as a World Cup. A suitable study would require consideration of the batting performances of wicketkeepers in the other competitions, or the batting performances of the ‘top 47’ in the 2014 T20 World Cup.

Wicketkeepers from winning teams outperformed wicketkeepers from losing teams for the majority of the variables considered (balls faced, runs scored, boundaries scored and wickets conceded). However, the fact that, as a combined group, they so clearly underperformed, in comparison to the specialist batsman considered in Lemmer’s (2008; 2011b) studies, suggests that teams could benefit from specialist batsmen batting in the top order, instead of the wicketkeeper. Such a strategy might be considered consistent with Petersen et al. (2008b), who advised the selection of specialist batsman to bat in the first 6 overs. It could be that the added pressures of completing dual roles in a high pressure environment impacts negatively upon batting performance, which can only have a detrimental effect on team performance. Given the small margins for error in T20 cricket (Petersen et al., 2008b; Russell, 2014), it could be argued that teams should consider the selection of specialists in this format; a theory which has been supported by Smith (2014).

5.3 Practical Implications

Having reviewed and discussed the results from both wicket-keeping and batting perspectives, it is appropriate to consider how the current findings could impact upon future practice, such as selection criteria. Evidence from both wicket-keeping and batting perspectives supports the importance of the wicketkeeper’s role in winning teams. However, the question still remains as to whether batting or wicket-keeping ability should be the priority in the selection of a wicketkeeper.

From a wicket-keeping perspective, the evidence presented justifies the requirement to have a wicketkeeper capable of standing up to all bowling types during the powerplay overs. Given both the positive impact this aspect of performance appears to have on the identified KPIs, and the view that standing up to seam bowling is the

43 most challenging aspect of wicket keeping (Woolmer et al., 2008), it could be argued that wicketkeepers should be selected for T20 matches based on their ability to stand up to the stumps, regardless of their batting ability. In the absence of previous empirical literature to support or reject this suggestion, it is worth referring to alternative sources, of an anecdotal nature, who have discussed the matter. Hotten (2013) advised the selection of a specialist wicketkeeper as an attacking option for a fielding side, to create pressure which can negatively impact upon batting performance, reducing a batsman’s scoring options. During research, prior to the commencement of this study, ex-professional wicketkeepers Russell (2014) and Scott (2014) discussed the advantages that can be gained by playing a specialist wicketkeeper. The latter made a particular reference to Michael Bates’ outstanding performance, including a last-ball stumping, when standing up to the stumps. This contributed significantly to Hampshire winning the 2012 Clydesdale Bank 40 final (a major domestic 40-over tournament in England and Wales). While such suggestions and observations are of interest, they lack academic rigour. Consequently, to pursue Hotten’s suggestion, it would be appropriate for future researchers to consider the impact of wicketkeeper positioning on the variables selected for this study, over a larger sample, possibly over a complete season of domestic T20 cricket, or across all T20 internationals played during a calendar year.

It is also worth considering how the results impact on the selection of a wicketkeeper for his batting ability, given the suggestions that coaches favour this approach in the current international game. Since wicketkeepers from winning teams faced more dot balls per innings than wicketkeepers from losing teams, it could be argued that dot ball percentages are of questionable relevance when selecting a wicketkeeper for batting ability. Although not directly related to the role of the wicketkeeper, previous research into T20 cricket has identified boundary striking (Moore et al., 2012) and strike rate (Lemmer 2008; Douglass and Tam, 2010; Lemmer, 2011b) as the primary criteria for the selection of a batsman. Of particular note, from this study, is the statistical significance of the wicketkeeper’s contribution to partnerships greater than 50 runs. Despite the fast scoring nature of T20 cricket, when selecting a wicketkeeper for his/her batting ability, coaches may need to consider carefully what specific criterion carries the greatest weight.

44

5.4 Future Research Recommendations

Given the paucity of research specifically directed towards wicket keeping, it is appropriate to consider where future research should be directed, if understanding of the wicketkeeper’s role is to be significantly enhanced. This study has identified that the wicketkeeper plays an important role in winning teams from both a wicket- keeping and batting perspective, but only a few variables were found to have statistically significant difference between wicketkeepers in winning and losing teams. It may be that a similar study using larger samples might identify additional, significantly important variables.

The significance of the wicketkeeper’s batting contribution might also be better understood if, following Lemmer (2008; 2011b), performances in a given competition were compared with those of the top batsmen, also playing in that competition.

Applying principles outlined by Saikia et al. (2012), relating to the measurement of fielding performance, the quality of wicket keeping acts (particularly dismissals) could be assessed, by utilising a measure which accounts for the difficulties faced when completing dismissals, both standing up and standing back to seam bowlers.

In addition to the variables considered in this study, there are other ways in which a wicketkeeper’s importance can be justified, from a wicket-keeping perspective. Snell (2014) discussed the added value that specialist wicketkeepers can bring to a fielding unit in terms of his tactical ability; specifically the ability to move fielders to the correct angles, and to use the Decision Review System (DRS) successfully. The latter, particularly, represents a critical area for consideration given that DRS is now used regularly at international level, and ‘wickets taken’ has been labelled as a KPI in all formats of the game (Petersen et al., 2008a; Petersen et al., 2008b; Douglass and Tam, 2010). Woolmer et al. (2008) have also alluded to the importance of the wicketkeeper as a motivator, and the negative impact that body movements and occasional sledging (verbal communication with opposing batsman) could have on the success of the opposition. Snell (2014), an ex-professional wicketkeeper himself, described these as unwritten aspects of the job, which are difficult to teach prospective wicketkeepers. Although these aspects of performance do not appear to influence the scoreboard directly, making it difficult to analyse from a quantitative viewpoint, they could be investigated from a qualitative perspective, using a

45 framework similar to that adopted by Miles and Neil (2013); a study which retrospectively considered performance using an interview procedure (aided by video sequences to prompt thought and discussion). Results from such studies would allow the characteristics of successful wicketkeepers to be further understood, providing coaches with information on the added tactical advantages that specialist wicketkeepers can provide. This could supplement quantitative data describing a wicketkeeper’s more obvious contribution (e.g. data produced as a result of the current study).

Pursuing the analysis of wicket keeping from a qualitative viewpoint, one area of interest not previously documented, but which would be worth pursuing, is player/captain/coach perceptions of the role of the wicketkeeper. This seems surprising, given the contrast in opinions as to whether a wicketkeeper should be selected for his wicket-keeping ability (Hotten, 2013), or his batting ability (Woolmer et al., 2008).

46

CHAPTER 6

CONCLUSION

47

The results do not allow for a precise answer to the research question originally posed for this study; they are inconclusive. However, this study did show that wicketkeepers in winning teams stood up to the wicket (3.07  4.42) for significantly (p = 0.023) more deliveries, during the powerplay overs (a skill often related to the specialist wicketkeeper), than the wicketkeepers in losing teams (0.69  2.07). Whilst the winning wicketkeeper’s performance during this period appears to have been more positive than the losing wicketkeeper’s, across all identified variables, it was not significantly so.

Of particular note is that, there was no significant difference in wicketkeepers’ positions in the batting order, between winning teams (3.30  1.87) and losing teams (4.04  2.17). This finding confirms the previously stated perception that wicketkeepers are generally expected to bat in the top order. However, the lack of any difference between winning and losing teams’ wicketkeepers, challenges the perception that the wicketkeeper’s batting position is a particular contribution to winning teams. The fact that in the winning teams, the wicketkeepers batted in only 20 of the 29 matches, might beg the question as to the weight given to the wicketkeeper’s batting ability when selecting for T20 cricket. Importantly, the primary batting contribution of the winning wicketkeeper, and significantly so (p = 0.005), was participation in partnerships of 50 or more runs; important because this has not been identified previously, in the wicket-keeping literature.

These issues are worth pursuing further, as is attempting to understand the significance of standing back during the non-powerplay overs as the ‘Dot Ball’ (p = 0.015) and ‘Wicket’ (p = 0.002) variables were significantly better for standing back, than for standing up during this period.

Lemmer (2011) concluded that future research should direct a focus towards developing performance measures for T20 cricket, so that players can be accurately ranked, based on their batting and wicket-keeping performance combined. Given that previous authors have discussed the value of performance measures in aiding team selection (Lewis, 2005; Iyer and Sharda, 2009), it is important that any measures constructed should be user-friendly for coaches, both to monitor player performance, and, where used for selection, to select players accurately according to their contribution.

48

For such measures to properly influence selection and resulting performance, the identification, and application, of appropriate KPIs in relation to successful wicket keeping, and the nature of the wicketkeeper’s significant batting contributions, need to be understood more clearly, and justified by hard evidence rather than assumption. This study has raised a number of relevant areas for future consideration. These include: further consideration of the benefits of the wicketkeeper standing up during the powerplay overs, attempted understanding of the value of standing back during the non-powerplay overs, further evaluation of the wicketkeeper’s role as a batsman in terms of both batting position and particular skills required.

49

REFERENCE LIST

50

Akhtar, S. and Scarf, P. (2012) ‘Forecasting test cricket match outcomes in play’, International Journal of Forecasting, 28(3), pp. 632-643.

Altman, D.G. (1991) Practical Statistics for Medical Research. London: Chapman and Hall.

Amin, G. R. and Sharma, S. K. (2014) ‘Cricket team selection using data envelopment analysis’, European Journal of Sport Science, 14(1), pp. 369-376.

Barr, G. D. and Kantor, B. S. (2004) ‘A Criterion for Comparing and Selecting Batsmen in Limited Overs Cricket’, The Journal of the Operational Research Society, 55(12), pp. 1266-1274.

Bukiet, B. and Ovens, M. (2006) ‘A Mathematical Modelling Approach to One-Day Cricket Batting Orders’, Journal of Sports Science and Medicine, 5(4), pp. 495-502.

Davis, M. and Collins, S. (2012) Bowling: How to play, coach and win. London: Bloomsbury.

Douglas, J. M. and Tam, N. (2010) ‘Analysis of team performances at the ICC World Twenty20 Cup 2009’, International Journal of Performance Analysis in Sport, 10(1), pp. 47-53.

Drust, B. (2010) ‘Performance analysis research: Meeting the challenge’, Journal of Sports Sciences, 28(9), pp. 921-922.

ESPNcricinfo (2014a) A brief history of cricket. Available at: http://www.espncricinfo.com/ci/content/story/239757.html (Accessed: 24th February 2015).

ESPNcricinfo (2014b) Statistics. Available at: http://www.espncricinfo.com/ci/content/stats/index.html (Accessed: 24th February 2015).

ESPNcricinfo (2014c) Records/2013 - Northamptonshire/Twenty20 matches/match results. Available at: http://stats.espncricinfo.com/ci/engine/records/team/match_results.html?class=6;id= 2013;team=1272;type=year (Accessed: 6th March 2015).

51

ESPNcricinfo (2014d) Records/2013 - Surrey/Twenty20matches/match results. Available at: http://stats.espncricinfo.com/ci/engine/records/team/match_results.html?class=6;id= 2013;team=1409;type=year (Accessed: 6th March 2015).

Franks, I. M. (2004) ‘The need for feedback’, in Hughes, M. and Franks, I. M. (eds.) Notational Analysis of Sport: Systems for Better Coaching and Performance in Sport. London: Routledge, pp. 9-17.

Gerber, H. and Sharp, G. (2006) ‘Selecting a limited overs cricket squad using an integer programming model’, South African Journal for Research in Sport, Physical Education and Recreation, 28, pp. 81-90.

Glazier, P. S. and Wheat, J. S. (2014) ‘An integrated approach to the biomechanics and motor control of cricket fast bowling techniques’, Sports medicine, 44(1), pp. 25- 36.

Gratton, C. and Jones, I. (2004) Research Methods for Sports Studies. London: Routledge.

Hotten, J. (2013) Sarah Taylor, Michael Bates, and the future of wicketkeeping. Available at: http://www.theguardian.com/sport/blog/2013/feb/06/cricket- wicketkeeping-future-sarah-taylor (Accessed: 4th November 2014).

Hughes, M. D. and Bartlett, R. M. (2002) ‘The use of performance indicators in performance analysis’, Journal of Sports Sciences, 20(10), pp. 739-754.

Hurrion, P. and Harmer, J. (2009) The Fast-Medium Bowler: Sports Biomechanics and Technical Analysis Model. Available at: http://www.quintic.com/education/case_studies/Cricket%203.htm (Accessed: 7th November 2014).

ICC (2014a) History of the International Cricket Council. Available at: http://www.icc- cricket.com/about/62/icc-organisation/history (Accessed: 11th November 2014).

ICC (2014b) Associate/Affiliate ODI Ranking Table. Available at: http://www.icc- cricket.com/associate-affiliate-rankings (Accessed: 11th November 2014).

52

Iyer, S. R. and Sharda, R. (2009) ‘Prediction of athletes’ performance using neural networks: An application in cricket team selection’, Expert Systems with Applications, 36(3), pp. 5510-5522.

James, N., Jones, N. M. P. and Hollely, C. (2002) ‘Reliability of selected performance analysis systems in football and rugby’, Proceedings of the 4th International Conference on Methods and Techniques in Behavioural Research, pp. 116-118.

Lemmer, H. (2004) ‘A measure for the batting performance of cricket players’, South African Journal for Research in Sport, Physical Education and Recreation, 26(1), pp. 55-64.

Lemmer, H. (2008) ‘An analysis of players’ performances in the first cricket Twenty20 World Cup Series’, South African Journal for Research in Sport, Physical Education and Recreation, 30(2), pp. 71-77.

Lemmer, H. (2011a) ‘Performance measures for wicket keepers in cricket’, South African Journal for Research in Sport, Physical Education and Recreation, 33(3), pp. 89-102.

Lemmer, H. (2011b) ‘The single match approach to strike rate adjustments in batting performance measures in cricket’, Journal of Sports Science and Medicine, 10(4), pp. 630-634.

Lewis, A. J. (2005) ‘Towards fairer measures of player performance in one-day cricket’, The Journal of the Operational Research Society, 56, pp. 804-815.

MacDonald, D., Cronin, J., Mills, J. and Stretch, R. (2013b) ‘Wicket-Keeping in Cricket: A Literature Review’, International Journal of Sport Science and Coaching, 8(3), pp. 531-541.

MacDonald, D., Cronin, J., Mills, J., McGuigan, M. and Stretch, R. (2013a) ‘A review of cricket fielding requirements’, South African Journal of Sports Medicine, 25(3), pp. 87-92.

53

Marks, V. (2014) Jos Buttler is ahead of the game but must be shown patience. Available at: http://www.theguardian.com/sport/blog/2014/jul/22/jos-buttler-england- wicketkeeper-test-cricket-matt-prior (Accessed: 3rd March 2014).

Miles, A. and Neil, R. (2013) ‘The use of self-talk during elite batting performance’, Psychology of Sport and Exercise, 14(6), pp. 874-881.

Moore, A., Turner, D. J. and Johnstone, J. A. (2012) ‘A preliminary analysis of team performance in English first-class Twenty-Twenty (T20) cricket’, International Journal of Performance Analysis in Sport, 12(1), pp. 188-207.

Mukherjee, S. (2013) ‘Quantifying individual performance in Cricket — A network analysis of batsmen and bowlers’, Statistical Mechanics and its Applications, 393(1), pp. 624-636.

Murphy, P. (2004) Stumpers or Stoppers? Available at: http://www.espncricinfo.com/wisdenalmanack/content/story/154381.html (Accessed: 7th August 2014).

O’Donoghue, P. G. (2007) ‘Reliability Issues in Performance Analysis’, International Journal of Performance Analysis in Sport, 7(1), pp. 35-48.

O’Donoghue, P. G. (2010) Research methods for sports performance analysis. London: Routledge.

O’Donoghue, P.G. (2012) Statistics for Sport and Exercise Studies: An Introduction. London: Routledge.

O’Donoghue, P.G. and Longville, J. (2004) ‘Reliability Testing and the Use of Statistics in Performance Analysis Support: A Case Study from an International Netball Tournament’, in O’Donoghue, P.G. and Hughes, M.D. (eds.) Performance Analysis of Sport 6. Cardiff: CPA Press, pp. 1-7.

Petersen, C., Portus, M. R. and Dawson, B. (2009) ‘Quantifying positional movement patterns in Twenty20 cricket’, International Journal of Performance Analysis in Sport, 9(2), pp. 165-170.

54

Petersen, C., Pyne, D. B., Portus, M. R. and Dawson, B. (2008b) ‘Analysis of Twenty/20 Cricket performance during the 2008 Indian Premier League’, International Journal of Performance Analysis in Sport, 8(3), pp. 63-69.

Petersen, C., Pyne, D. B., Portus, M. R., Cordy, J. and Dawson, B. (2008a) ‘Analysis of performance at the 2007 Cricket World Cup’, International Journal of Performance Analysis in Sport, 8(1), pp. 1-8.

Rudkin, S. T. and O’Donoghue, P. (2008) ‘Time-motion analysis of first-class cricket fielding’, Journal of Sports Science and Medicine in Sport, 11(6), pp. 604-607.

Russell, J. (2014) Email from Jack Russell, 2 September.

Saikia, H., Bhattacharjee, D. and Lemmer, H. (2012) ‘A Double Weighted Tool to Measure the Fielding Performance in Cricket’, International Journal of Sports Science and Coaching, 7(4), pp. 699-713.

Scott, C. (2014) Email from Chris Scott, 15 July.

Smith, E. (2014) ‘Number Crunch’, The Sunday Times, 10 August, p. 4.

Snell, S. (2014) ‘Cricket Wicketkeeping such a key role in the modern game’, Evening Post (Bristol edn), 18 July, p. 68.

Stretch, R., Buys, F., Du Toit, E. and Viljoen, G. (1998) ‘Kinematics and Kinetics of the drive off the front foot in cricket batting’, Journal of Sports Sciences, 16(8), pp. 711-720.

Subhani, M. I., Hasan, S. A. and Osman, A. (2012) ‘Will T20 clean sweep other formats of Cricket in future?’, European Journal of Scientific Research, 86(1), pp. 98- 102.

Tennant, A. (2009) Cricket Manual: The official guide to playing the game. Yeovil: Haynes Publishing.

Woolmer, B., Noakes, T. and Moffett, H. (2008) Bob Woolmer’s art and science of cricket. London: New Holland.

55

APPENDICES

APPENDIX A - SUMMARY OF TOURNAMENT MATCHES EXCLUDED FROM THE CURRENT STUDY

Match (Venue) Stage of Date Justification for exclusion Tournament UAE v Netherlands Preliminary March 17th 2014 Required footage was unavailable. Group Stage

Zimbabwe v Netherlands Preliminary March 19th 2014 Required footage was unavailable. Group Stage

Ireland v UAE Preliminary March 19th 2014 Required footage was unavailable. Group Stage

Zimbabwe v UAE Preliminary March 21st 2014 Required footage was unavailable. Group Stage

England v New Zealand Group Stage March 22nd 2014 The game was reduced from a 20 over game, and eventually decided via the Duckworth Lewis (DL) method. It was believed that including this would affect the validity of the results.

Sri Lanka v West Indies Semi Final April 3rd 2014 The game was reduced from a 20 over game, and eventually decided via the Duckworth Lewis (DL) method. It was believed that including this would affect the validity of the results.

A-1

APPENDIX B - SUMMARY OF TOURNAMENT MATCHES ANALYSED DURING THE CURRENT STUDY

Match Stage of Winning Team (Identified Losing Team (Identified Date Venue Tournament Wicketkeeper) Wicketkeeper) 1. Preliminary Bangladesh Afghanistan March 16th Mirpur Group Stage (Mushfiqur Rahim) (Mohammad Shahzad) 2014

2. Preliminary Nepal Hong Kong March 16th Chittagong Group Stage (Subash Khakurel) (Jamie Atkinson) 2014

3. Preliminary Ireland Zimbabwe March 17th Sylhet Group Stage (Gary Wilson) (Brendan Taylor) 2014

4. Preliminary Afghanistan Hong Kong March 18th Chittagong Group Stage (Mohammad Shahzad) (Jamie Atkinson) 2014

5. Preliminary Bangladesh Nepal March 18th Chittagong Group Stage (Mushfiqur Rahim) (Subash Khakurel) 2014

6. Preliminary Nepal Afghanistan March 20th Chittagong Group Stage (Subash Khakurel) (Mohammad Shahzad) 2014

B-1

7. Preliminary Hong Kong Bangladesh March 20th Chittagong Group Stage (Jamie Atkinson) (Mushfiqur Rahim) 2014

8. Preliminary Netherlands Ireland March 21st Sylhet Group Stage (Wesley Barresi) (Gary Wilson) 2014

9. Group Stage India Pakistan March 21st Mirpur (MS Dhoni) (Kamran Akmal) 2014

10. Group Stage Sri Lanka South Africa March 22nd Chittagong (Kumar Sangakkara) (Quinton De Kock) 2014

11. Group Stage Pakistan Australia March 23rd Mirpur (Kamran Akmal) (Brad Haddin) 2014

12. Group Stage India West Indies March 23rd Mirpur (MS Dhoni) (Denesh Ramdin) 2014

13. Group Stage South Africa New Zealand March 24th Chittagong (Quinton De Kock) (Luke Ronchi) 2014

B-2

14. Group Stage Sri Lanka Netherlands March 24th Chittagong (Kumar Sangakkara) (Wesley Barresi) 2014

15. Group Stage West Indies Bangladesh March 25th Mirpur (Denesh Ramdin) (Mushfiqur Rahim) 2014

16. Group Stage England Sri Lanka March 27th Chittagong (Jos Buttler) (Kumar Sangakkara) 2014

17. Group Stage South Africa Netherlands March 27th Chittagong (Quinton De Kock) (Wesley Barresi) 2014

18. Group Stage West Indies Australia 28th March Mirpur (Denesh Ramdin) (Brad Haddin) 2014

19. Group Stage India Bangladesh 28th March Mirpur (MS Dhoni) (Mushfiqur Rahim) 2014

20. Group Stage South Africa England 29th March Chittagong (Quinton De Kock) (Jos Buttler) 2014

B-3

21. Group Stage New Zealand Netherlands 29th March Chittagong (Luke Ronchi) (Wesley Barresi) 2014

22. Group Stage India Australia 30th March Mirpur (MS Dhoni) (Brad Haddin) 2014

23. Group Stage Pakistan Bangladesh 30th March Mirpur (Kamran Akmal) (Mushfiqur Rahim) 2014

24. Group Stage Sri Lanka New Zealand 31st March Chittagong (Kumar Sangakkara) (Luke Ronchi) 2014

25. Group Stage Netherlands England 31st March Chittagong (Wesley Barresi) (Jos Buttler) 2014

26. Group Stage West Indies Pakistan 1st April 2014 Mirpur (Denesh Ramdin) (Kamran Akmal)

27. Group Stage Australia Bangladesh 1st April 2014 Mirpur (Brad Haddin) (Mushfiqur Rahim)

B-4

28. Semi-Final India South Africa 4th April 2014 Mirpur (MS Dhoni) (Quinton De Kock)

29. Final Sri Lanka India 6th April 2014 Mirpur (Kumar Sangakkara) (MS Dhoni)

B-5

APPENDIX C - THE BESPOKE NOTATIONAL ANALYSIS SYSTEM USED DURING DATA COLLECTION

C-1

APPENDIX D - EXTRA COLUMN (WIN/LOSS) ADDED TO THE DATA COLLECTION SPREADSHEET PRIOR TO DATA ANALYSIS, TO SIGNIFY WINNING AND LOSING EVENTS

D-1