2020 8th International Conference on Information Technology

and Multimedia (ICIMU)

A Software Engineering Approach in Netball

Performance Analysis: Training and Activities
Features for Automatic Players Position Selection
Abdul Hadi Mohamad Ramona Ramli Amira Farisa Ramli
Department of Information Technology College of Computing and Informatics College of Computing and Informatics
Saudi Electronic University Universiti Tenaga Nasional Universiti Tenaga Nasional
Riyadh, Saudi Arabia Kajang, Selangor Kajang, Selangor
[email protected] [email protected] [email protected]

Abstract—Player performance analysis and player selection seasonal phases, and assess game readiness. While
are two critical activities that should be carried out by coaches laboratory and field testing are important in sports,
to form sports teams. In netball, testing, training, and activities laboratory test protocols often provide a controlled
will be performed by coaches to collect player performance for environment for comprehensive physiological and
analysis and player selection. Albeit its importance, coaches performance data in players and tests can often omit sport-
are still recording this data on paper-based sheets. The time- specific movement sequences due to restrictions in
consuming and tedious manual evaluation process will be then laboratory size. Furthermore, the time and financial costs
conducted based on predefined criteria which likely leading to associated with various laboratory testing can also create a
improper team formation, lack of transparency, and
barrier for routine player assessment, particularly in large
favouritism. Recent research in literature has led to player
classification and physical performance evaluation using match
team squads. In turn, sport-specific, field tests overcome
data. While this is highly depending on coach experience, these issues, permitting closer replication of game demands
problems may occur which coaches expressed the demands of in team sports and efficient use of resources for repeated
having a more systematic and guided approach to assist in testing.
player selection. In giving some guidance for performance Sport-specific tests assessing game-relevant fitness and
evaluation and player selection, this study highlighted performance have been developed in various team sports,
computer-based system features represented in flowcharts, including basketball and soccer. In netball, traditionally,
considering inputs from practitioners and researchers using a
coaches record players' performances for each activity and
mixed approach of data gathering. While algorithms are
proposed to evaluate time and distance achieved by each player
training session on paper-based sheets. Then, coaches will
for team selection, two key factors in measuring player evaluate each player based on their game position and make
performance and quality are determined to assist computer decisions based on designated criteria to form a team
system in analysing and making decisions for automatic manually. Generally, this manual process is a tedious job and
selection. The results show that on average, around 35 per cent time consuming for coaches to analyse players separately [2],
of errors are found upon pilot path testing. A systematic fixing which may lead to improper team formation, lack of
is performed to improve the processes, and optimum results transparency [3], and favouritism.
are achieved consequently. Initial findings obtained in this
Studies drawn in the literature have shown that growth of
study may lead to further investigation of the algorithms in
actual sports settings and environment.
a netball-specific test replicating key game movements is
limited [4]. With little attention given to the development of
Keywords—sports computing, netball performance analysis, computer-based decision support systems for automated and
automated player position selection, software engineering guided netball selection based on training and activities data,
approach other research was mainly focusing on using match data and
coach experience in evaluation players performance and
I. INTRODUCTION selection [5], [6], which these factors are imperative for
Netball is a team-based game that consists of seven main netball team formation. Furthermore, although various
positions and several reserve players. Selecting players for decision support systems are available for sports, they were
each position to form a team is essential before start playing. designed for other sports, not for netball in particular. Netball
Usually, coaches will form teams by choosing players is a kind of sports with a variance of dimensionalities such as
according to their performances and familiar position [1]. movement, arrays of training and activities, and performance
Moreover, a set of training, testing, and activities will be indicators. Inherently, such issues, including the lack of
conducted to evaluate players performances and capabilities. netball analysis available in the literature are caused by little
All the results of the activities are usually recorded on paper understanding of the dynamics of netball game-play [7].
sheets.
II. RELATED WORK
Specifically, selected tests are used by coaching staff and Match performance data is utilised with evidence that
practitioners to select players with the greatest potential for football player role can be determined using performance
further physical and technical development, quantify player data relating to player actions [5]. The study introduces a
responsiveness to training interventions, monitor player decision tree model to classify players roles based on initial
fitness and performance longitudinally across different match data analysed using statistical analysis approach. It is
found that based on the model, classifications of key
This work was supported by the UNIIG 2018 (J510050876) under
Universiti Tenaga Nasional.

978-1-7281-7310-8/20/$31.00
Authorized licensed use limited to: University©2020 IEEE Downloaded371
of Gothenburg. on December 18,2020 at 13:49:12 UTC from IEEE Xplore. Restrictions apply.
 2020 8th International Conference on Information Technology
and Multimedia (ICIMU)

forwards, midfielders, and general defenders the easiest to Predict

Talent
Player availability,
classify objectively. At the same time, the dynamic midfield- Other features matches
enhancement
communication, and
forward role was more difficult to define. In another results promotion
research, cognitive and physical execution components of
football player decision making are analysed using live- TRS recommends cricket players for specific roles in a
match situation [6]. It is suggested that this kind of notational team, considering players past performances by clustering
system is a practical approach, but also depending on coach similar players. This system uses the SVM model with linear
experience in providing high-quality training for players. and non-linear poly and RBF. Furthermore, this system
Cricket players physical factors and analysis such as muscle adopted a framework which predicts the outcome of the
strength and shoulder rotators were highlighted to measure matches and performs team analysis and recommends the
players performance. Additionally, other factors, such as player role by extracting the statistics about the cricket game
training and experience, may also contribute to selection and and players from various websites [1]. C-TIES is a web-
players development [8]. based system, namely Cricket Talent Identification,
Enhancement and Selection. This system has three main
A complex systems model using work domain analysis characteristics which are talent identification, talent
was developed to measure key components of netball match enhancement and talent selection, specifically for cricket. C-
performance and relationship between them. Since netball TIES applies OWA Aggregation Operator, Relative Fuzzy
has been characterised as a complex game, and include Linguistic Quantifier (RFLQ) and Normalized Adequacy
measures such as passing and possession, cognitive Coefficient (NAC) in its engine to identify the weaknesses of
performance, and measures related to physical activity each player and select most talented n-players from a larger
deemed significant for analysing players performance [9]. group of players. The system also helps in selecting players
The quantitative approach has been adopted in the study of without bias and help them in identifying their weakness
identifying potential junior netball players. Respondents (i.e., [12]. Contrarily, KlubPro is an online service that allows
the coaches) expressed the demands of having a more official clubs, via social media messaging, graphics team-
systematic and guided approach to assist in player selection. selection and increased club involvement with players,
On the other hand, the researchers found that coaches members, sponsors, and communities, to carry out very
experience may be a necessary component for selecting efficient player communication and accessibility
netball players [10]. Arguably, coaches should have some management. This system supports local clubs of Australia
assistance guided by proper sports mechanism, regardless of such as Rugby Union, football, Australian Rules Football
their experience level. There has been an attempt to study and Hockey. It has features of player availability and
position-specific and general indicators that can be used by communication, team selection and sponsor promotion [13].
coaches to identify individual positional weaknesses [11] and
to assess the performance of individuals or teams [7], III. RESEARCH METHODOLOGY
accordingly. This study reveals that several indicators, such
as passing, and shooting, among others, are compelling for
player improvement and selection. While there is a limited
study drawn in the literature related to players selection for
team formation using training and activities data, Table 1
lists several systems available commercially, evaluated using
commonly computer-based systems attributes observed in
this study.

TABLE I. COMPARISON OF VARIOUS SPORTS SYSTEMS

Criteria TRS [1] C-TIES [12] KlubPro [13]

Mobile & Web- Fig. 1. Research Methodology
Platform Web-based Web-based
based
Football, Cricket, The study focuses on a multitude of approach in its
Rugby, Hockey and methodology. Figure 1 shows the overall research
Intended for Cricket Cricket
Netball mainly in
Australia
methodology, embedding qualitative and design science
activities that serve in parallel as Software Engineering
Origin India India Australia
methodology in the computing field. The qualitative
OWA
SVM
Aggregation
approach that is useful in this study domain contemplates
model with input from other researchers and practitioners (university
Operator and
Technique/ linear and Information not team coach) in this field [14]. Similar to requirements
Relative Fuzzy
Algorithm non- linear available
poly and
Linguistic gathering activities, the activity is conducted using the
RBF
Quantifier unstructured interview to understand existing problems
(RFLQ) related to the experts' experiences while conducting training
Training/ performance analysis of netball players conventionally.
No No Yes
Fitness Record
Player The activity of analysis and design takes the requirements
Yes Yes Yes
Analysis gathered through qualitative means and conduct a thorough
Player system analysis. In this activity, requirements are modelled
Yes Yes Yes
Selection according to various specifications such as behavioural and
Team
Yes No Yes
architectural, including the system workflow. User interfaces
Selection and features are then built based on requirements models and

Authorized licensed use limited to: University of Gothenburg. Downloaded372

on December 18,2020 at 13:49:12 UTC from IEEE Xplore. Restrictions apply.
 2020 8th International Conference on Information Technology
and Multimedia (ICIMU)

combined them into a working prototype. Validation of these decision making. Several activities listed in Table 4 [15] are
designs is performed by considering workflows and suggested to evaluate players' development and physical
functions to measure correctness rapidly. performance. These activities deem helpful for researchers
and practitioners to evaluate further physical qualities
TABLE II. USER FUNCTIONS AND ACCESSIBILITY required for successful netball performance.
Users Description Accessibility
TABLE V. POSITION-BASED MEASUREMENT INDICATORS FOR
Manager Manage the system and player data Full access NETBALL PLAYERS
Coach / Manage the coaching of netball activities Partial access
assistant

Wing Defence (WD)

Goal Shooter (GK)

Wing Attack (WA)

Goal Defence (GD)

Goal Keeper (GK)

Goal Attack (GA)
Player Manage personal account Limited access

Main
TABLE III. A NON-EXHAUSTIVE LIST OF SYSTEM FEATURES

Indicator

Centre (C)
Description Manager Coach Player
General system access ✓ ✓ ✓
View profile ✓ ✓ ✓ Shooting accuracy
Manage profile ✓ ✓ ✓ and proportion of ✓ ✓
Manage team ✓ team total
Manage team activity ✓ ✓ Shot length ✓ ✓

Shooting
Shot position on court ✓ ✓
This study implements a framework which focuses on
Shooter movement to
three components: (a) recording player profiles, (b) fitness ✓ ✓
be available for ball
performance and (c) training activities in its features for Opposition shooting
numerous users as listed in Table 2 and Table 3. The data accuracy ✓ ✓
incorporated in the database shall calculate and present On attack from own
✓ ✓
Rebounding

player performance based on activities that deem suitable, missed shots (%)
suggesting ranking and also suitable positions for players, as On attack from other
✓ ✓
shown in Figure 2 to form netball teams. shooter's shots (%)
On defence (%) ✓ ✓
Passing accuracy and
✓ ✓ ✓ ✓ ✓ ✓ ✓
proportion of team
Passing

Total pass position on

✓ ✓ ✓ ✓ ✓ ✓
court
Opposition passes as
proportion of team ✓ ✓ ✓ ✓ ✓
total
Feeding accuracy and
✓ ✓ ✓ ✓ ✓ ✓
Feeding

proportion of team
Feeding position on
court ✓ ✓ ✓ ✓
Centre
Passes

Proportion of centre
✓ ✓ ✓ ✓
passes received
Intercepts and Tips

Fig. 2. Sample of players data of ranking and position in the database Number of intercepts ✓ ✓ ✓ ✓ ✓ ✓ ✓
Intercept position on
A. Players Performance Assessment and Recommended court ✓ ✓ ✓ ✓
Activities for Assessments Number of tips ✓ ✓ ✓ ✓ ✓ ✓ ✓

TABLE IV. ACTIVITIES FOR PHYSICAL PERFORMANCE EVALUATION

Tip position on court ✓ ✓ ✓ ✓

Tests Descriptions
Number of penalties ✓ ✓ ✓ ✓ ✓ ✓ ✓
Committed
Penalties

Countermovement jump Measure of lower-body explosive Type of penalty

(Offensive/Defensive) ✓ ✓ ✓ ✓ ✓ ✓ ✓
performance
5- and 10-m sprint Evaluation of acceleration and short-sprint Penalty position on
✓ ✓
performance. court
Isometric mid-thigh pull The measure of isometric lower-body
Errors

Number of Errors
strength, which is strongly correlated with ✓ ✓ ✓ ✓ ✓ ✓ ✓
Made
jumping, sprinting, and changing direction.
1RM back squat Measure of maximum muscular strength,
which as described is significantly related TABLE VI. USER FUNCTIONS AND ACCESSIBILITY
to jumping, sprinting, and changing
Evaluation Description Evaluation Mechanism
direction. This should only be included
Score General score 1 – 5 marks * 2
once an players technique is of sufficient
standard. Level Progressive level 1 – 8 level + 2
Time Time taken to complete Higher marks either for
an activity shorter or longer time
Netball players performance were assessed iteratively Distance Distance completed for Higher marks either for
according to several activities, including movement, ball an activity shorter or longer distance
work, shooting, defending attacking, speed, fitness, and

Authorized licensed use limited to: University of Gothenburg. Downloaded373

on December 18,2020 at 13:49:12 UTC from IEEE Xplore. Restrictions apply.
 2020 8th International Conference on Information Technology
and Multimedia (ICIMU)

Diversely, Table 5 lists measurement indicators possible 20: Add 4 to H

to be utilised by coaches based on players positions. Each 21: else if G > E + 2 then
22: Add 2 to H
main measurement has various specific indicators which may 23: else if G < E and G <= E – 2 then
be beneficial to be assessed to netball players based on their 24: Add 8 to H
court positions. The activities can be calculated according to 25: else
several evaluation types such as score, level, time, and 26: Add 10 to H
distance, as listed in Table 6, and each type carries ten marks 27: endif
accordingly. Fig. 4. Distance evaluation algorithm pseudocode

Time and distance are crucial criteria for evaluating B. Players Position Selection
players performance and quality. Time indicates the time
taken by players to finish monitored activities, and they will The study inclines to assist coaches in evaluating players
be awarded marks according to their performance, as performance and unbiased selection method to form a netball
envisage in Figure 3. The coach can set target time and type team based on numerous selection criteria. The input for
of each activity. The former indicates different types of selection is analysed based on the criteria according to each
activities such as high marks for either short or longer time netball player position collected from expert researchers and
activity accomplishment. professional coach.
Algorithm 1: Time Evaluation Algorithm Typically, there are seven positions in netball such as
1: A ← target time set by coach goal shooter (GS), goalkeeper (GK), goal defence (GD), goal
2: B ← higher marks for either (1) short or (2) long time attack (GA), wing attack (WA), wing defence (WD), and
3: C ← completed distance by the player centre (C) [16]. In the netball position selection algorithm
4: D ← initial time mark presented in Figure 5, players are selected for seven main
5: if B(1) then
6: if C is equal to A then positions, and three players will become reserves of the team.
7: Add 6 to D In the algorithm, there are four categories of players
8: else if C > A and C <= A + 2 then proposed such as Defender, Attacker, Centre and Wing
9: Add 8 to D Attack, and Reserve. The proposed algorithm works on
10: else if C > A + 2 then multistage processing and players who have been assigned to
11: Add 10 to D
12: else if C < A and C <= A – 2 then
positions will be ignored, and only the remaining players will
13: Add 4 to D be analysed for position selection. Players in defending
14: else category are assessed according to their ability in making
15: Add 2 to D decision and fitness performance, labelled as "Decision" and
16: else if B(2) then "Fitness", accordingly. On the other hand, attacking players
17: if C is equal to A then
18: Add 6 to D
are observed based on their "Attacking" ability in the game.
19: else if C > A and C <= A + 2 then Lastly, "Ball Work" is the element that measures centre and
20: Add 4 to D wing players in the team.
21: else if C > A + 2 then
22: Add 2 to D Algorithm 3: Position Selection Algorithm
23: else if C < A and C <= A – 2 then 1: [All] ← a set of all players
24: Add 8 to D 2: [Def] ← a set of players of Defender category
25: else 3: [Att] ← a set of players of Attacking category
26: Add 10 to D 4: [Cwa] ← a set of players of Centre & Wing Attack category
27: endif 5: [Res] ← a set of players of Reserve category
6: Set 7 main positions, three reserves ∈[All]
Fig. 3. Time evaluation algorithm pseudocode 7: //set Defender players of GK, GD & WD
8: From remaining [All], get top 3 players for Defender ∈[Def]
The latter measures distance achieved when players are 9: for all elements of [Def] do
performing activities. As shown in Figure 4, target distance 10: if any element has the highest Decision point then
is normally set by coaches, and the players will be awarded 11: set as Goalkeeper (GK)
12: else if remaining elements have the highest Fitness point then
marks based on types of activity achievement, either higher 13: set as Goal Defence (GD)
marks for short or long-distance activity. 14: else
15: set as Wing Defence (WD)
Algorithm 2: Distance Evaluation Algorithm
16: end if
1: E ← target distance set by the coach 17: end for
2: F ← higher marks for either (1) short or (2) long distance 18: //set Attacking players for GA & GS
3: G ← completed distance by the player 19: From remaining [All], get top 2 players for Attacking ∈[Att]
4: H ← initial distance mark 20: for all elements of [Att] do
5: if F(1) then 21: if any element has the highest Attacking point then
6: if G is equal to E then
22: set as Goal Attack (GA)
7: Add 6 to H 23: else
8: else if G > E and G <= E + 2 then 24: set as Goal Shooter (GS)
9: Add 8 to H
25: end if
10: else if G > E + 2 then 26: end for
11: Add 10 to H 27: //set Centre & Wing Attack players
12: else if G < E and G <= E – 2 then
28: From remaining [All], get top 2 players for Centre & Wing
13: Add 4 to H
Attack ∈[Cwa]
14: else
29: for all elements of [Cwa] do
15: Add 2 to H
30: if any element has the highest Ball Work point then
16: else if F(2) then
31: set as Centre (C)
17: if G is equal to E then
32: else
18: Add 6 to H
33: set as Wing Attack (WA)
19: else if G > E and G <= E + 2 then

Authorized licensed use limited to: University of Gothenburg. Downloaded374

on December 18,2020 at 13:49:12 UTC from IEEE Xplore. Restrictions apply.
 2020 8th International Conference on Information Technology
and Multimedia (ICIMU)

34: end if
35: end for
36: //set reserve players of Shooter, Defender, Attacker
37: From remaining [All], get three players for Research ∈ [Res]
38: for all elements of [Res] do
39: if any element has the highest Shooting point then
40: set as reserve shooter
41: else if remaining elements has highest Defend point then
42: set as reserve defender
43: else
44: set as reserve attacker
45: end if
46: end for
Fig. 5. Position Selection algorithm

IV. RESULTS AND DISCUSSIONS

A set of pilot testing activities are conducted as part of
testing strategies in this study. Basic path testing is adopted
in this study due to its ability to automatically generate test
data which whole search space to satisfy path coverage. This Fig. 8. Manage Team flow
is another type of white-box testing methods that finds
linearly independent paths of execution represented in a
computer system [17]. This method is useful to automatically
generate test data which whole search space to satisfy path
coverage [18], which beneficial to support the structural
testing with adequate knowledge of software internal
structure [19]. In this process, it is a key challenge to search
for a set of test data in the whole search space to satisfy path
coverage. Thus, finding an efficient method for generating
test data automatically is a key issue in software testing [18].

Fig. 9. Manage Team Activity flow

TABLE VII. USER FUNCTIONS AND SEQUENCE ACCESS

Features ID Sequence Path Reference
General 1 1,2,3,4,5,6 Figure 6
access 2 1,7,2,3,8,9,10,6
3 1,2,3,8,11,12,6
Manage 4 1,2,3,4,5,6 Figure 7
Fig. 6. General flow Profile 5 1,2,3,4,7,6
Manage 6 1,2,3,4,5,6,7,8,9 Figure 8
Team 7 1,2,3,4,7,6.8.9
8 1,2,8,1
9 1,10,11,12,13
10 1,10,14,15,16,17,18,19
11 1,10,14,15,16,17,19,1
Manage 12 1,2,3,4,5,6,7,8,1 Figure 9
Team 13 1,2,3,8,1
Activity 14 1,9,10,11,12,13,14,15,16,17,
18,15,16,17,18,19,15
15 1,9,10,11,20,21,22,1

As enumerated in Table 7, several path sequences are

generated according to available and accessible paths served
as test cases for testing purposes.
Fig. 7. Manage Profile flow

Authorized licensed use limited to: University of Gothenburg. Downloaded375

on December 18,2020 at 13:49:12 UTC from IEEE Xplore. Restrictions apply.
 2020 8th International Conference on Information Technology
and Multimedia (ICIMU)

TABLE VIII. PATH TESTING RESULTS such as time and distance achieved. Since this is a
preliminary study of its kind, as set of pilot testing activities
Path IDs
Pre Post are conducted in testing several system features that are
E1 E2 E3 E1 E2 E3 proposed in this study. The testing analyses both pre and post
Path 1 ✓ ✓ x ✓ ✓ ✓ feedback given by researchers and experts upon examining
Path 2 ✓ ✓ ✓ ✓ ✓ ✓ flows that are generated using the given requirements. The
Path 3 ✓ x ✓ ✓ ✓ ✓ results show that around 35 per cent of errors are initially
Path 4 x ✓ ✓ ✓ ✓ ✓ found after the establishment of the flows. The errors were
Path 5 x ✓ ✓ ✓ ✓ ✓ then harmonised and deliberated to improve the existing
Path 6 ✓ x ✓ ✓ ✓ ✓ flow. Finally, the improved flows that are drawn in this study
Path 7 ✓ ✓ ✓ ✓ ✓ ✓ are ready for practical system implementation. Further
Path 8 ✓ ✓ ✓ ✓ ✓ ✓ examination and empirical testing are proposed to assess the
Path 9 ✓ ✓ ✓ ✓ ✓ ✓ effectiveness of the introduced time, distance, and position
Path 10 ✓ ✓ ✓ ✓ ✓ ✓ selection algorithms in an actual sports environment to gauge
Path 11 ✓ ✓ ✓ ✓ ✓ ✓ the impact of such mechanism in practical sports settings.
Path 12 ✓ ✓ x ✓ ✓ ✓
Path 13 ✓ ✓ x ✓ ✓ ✓ ACKNOWLEDGMENT
Path 14 x ✓ ✓ ✓ ✓ ✓ This research is conducted under Universiti Tenaga
Path 15 ✓ ✓ ✓ ✓ ✓ ✓ Nasional Internal Grant (UNIIG2018) project code:
Path with errors (per J510050876.
cent)a 47 33 27 0 0 0
Correct Path (per cent)b 53 67 73 100 100 100 REFERENCES
Errors (per cent)c 44 31 25 0 0 0
Correctness (per cent)d 56 69 75 100 100 100 [1] S. B. Jayanth, A. Anthony, G. Abhilasha, N. Shaik, and
a.
Σ path with errors by individual researcher / Σ overall paths examined by an individual researcher
b.
G. Srinivasa, "A team recommendation system and
Σ correct path by individual researcher / Σ overall paths examined by an individual researcher
c.
Σ path with error s / Σ overall errors
outcome prediction for the game of cricket," J. Sport.
d.
Σ correct paths / Σ overall correct paths Anal., vol. 4, no. 4, pp. 263–273, 2018.
Figure 6-9 show flows of several features that were [2] M. B. Burhanuddin, M. S. Suhaimi, and M. T. Isa,
examined by measuring its correctness. The flows are "Football Player Selection Using Fuzzy Logic," in
generated based on the requirements given by the ASEAN Conference on Information System (ACIS2015),
practitioner, converting those requirements into a sequence 2015.
of process drawn in flowchart format. Consequently, these [3] C. M. E. Colomer, D. B. Pyne, M. Mooney, A.
flows were studied for testing activities, examined using McKune, and B. G. Serpell, "Performance Analysis in
expert sampling methodology by select several key persons Rugby Union: a Critical Systematic Review," Sport.
with experience and expertise in the area [20]. The testing Med. - Open, vol. 6, no. 1, 2020.
was done in two stages: pre and post. The pre sessions [4] Q. Tan et al., "Comparing patient and healthcare worker
intended to get some feedback after the establishment of the experiences during a dengue outbreak in Singapore:
flows. The feedback was in the form of errors, general
Understanding the patient journey and the introduction
feedback, and suggestions to change according to the logical
sense of the expert. The results of the testing that are of a point-of-care test (POCT) toward better care
summarised in Table 8 show during the pre-sessions, each delivery," BMC Infect. Dis., vol. 17, no. 1, pp. 1–16,
experts found around 35per cent of errors exist in the 2017.
generated paths, which the correct paths are around 65 per [5] S. McIntosh, S. Kovalchik, and S. Robertson,
cent. Due to the errors, the correctness is found around the "Examination of player role in the Australian Football
same figure, and a systematic flows fixing were planned League using match performance data," Int. J. Perform.
considering experts feedback. Then, the feedback was Anal. Sport, vol. 18, no. 3, pp. 451–462, 2018.
harmonised and deliberated for change. The latter sessions [6] M. Lorains, K. Ball, and C. MacMahon, "Performance
discussed the improvement made in the flows after analysis for decision making in team sports," Int. J.
accommodating all feedback given. Finally, optimum results Perform. Anal. Sport, vol. 13, no. 1, pp. 110–119, 2013.
were obtained, which all paths were correct and ready for [7] L. Bruce, E. R. Brooks, and C. T. Woods, "Team and
implementation. seasonal performance indicator evolution in the ANZ
Championship netball league," J. Sports Sci., vol. 36,
V. CONCLUSION
no. 24, pp. 2771–2777, 2018.
This study highlights some important features of [8] R. November and L. Leach, "Physical and performance
managing players activities and automatic selection of measures of university cricket players," African J. Phys.
players of netball games in the field of sports computing. A Act. Heal. Sci., vol. 22, pp. 1140–1156, 2016.
multitude of approaches is adopted in gathering [9] S. Mclean, A. Hulme, M. Mooney, G. J. M. Read, A.
requirements, including inputs from researchers and Bedford, and P. M. Salmon, "A systems approach to
practitioners who are the coach of a national university team performance analysis in women's netball: Using work
coaches. Several useful activities gathered from literature are domain analysis to model elite netball performance,"
suggested in measuring player performance and quality to
Front. Psychol., vol. 10, no. FEB, pp. 1–13, 2019.
assist computer system in analysing and making decisions to
perform an automatic selection. Consequently, the selection [10] S. F. Mungovan, P. J. Peralta, G. C. Gass, and A. T.
will assist coaches and managers in forming solid teams Scanlan, "The test-retest reliability and criterion validity
based on the quality attributes in activities set by the coaches, of a high-intensity, netball-specific circuit test: The Net-
Test," J. Sci. Med. Sport, vol. 21, no. 12, pp. 1268–

Authorized licensed use limited to: University of Gothenburg. Downloaded376

on December 18,2020 at 13:49:12 UTC from IEEE Xplore. Restrictions apply.
 2020 8th International Conference on Information Technology
and Multimedia (ICIMU)

1273, 2018. Strength Cond. J., vol. 39, no. 4, pp. 10–21, 2017.
[11] C. Rose, "The comparison of Position-Specific [16] C. M. Young, P. B. Gastin, N. Sanders, L. Mackey, and
Performance Indicators between top half and bottom D. B. Dwyer, "Player load in elite netball: Match,
half teams in British National Super League Netball," J. training, and positional comparisons," Int. J. Sports
Sports Sci., vol. 1, pp. 1–23, 2013. Physiol. Perform., vol. 11, no. 8, pp. 1074–1079, 2016.
[12] G. Ahamad, S. K. Naqvi, M. M. S. Beg, and T. Ahmed, [17] J. Chen, Y. Wang, Y. Guo, and M. Jiang, A
"A web based system for cricket talent identification, metamorphic testing approach for event sequences, vol.
enhancement and selection (C-TIES)," Procedia 14, no. 2. 2019.
Comput. Sci., vol. 62, no. Scse, pp. 134–142, 2015. [18] S. M. Mohi-Aldeen, R. Mohamad, and S. Deris,
[13] “KLUBPro System.” [Online]. Available: "Application of Negative Selection Algorithm (NSA)
https://www.klubpro.com/. [Accessed: 01-Apr-2020]. for test data generation of path testing," Appl. Soft
[14] N. W. Abu Bakar, A. H. Mohamad, P. K. Comput. J., vol. 49, pp. 1118–1128, 2016.
Jatavakabhulla, and S. Musa, "Effect of Lean [19] N. Jain and R. Porwal, Automated test data generation
Systematic Method on Specimen Handling Activities applying heuristic approaches—A survey, vol. 731.
for Patients in Medical Wards," Int. J. Eng. Technol., Springer Singapore, 2019.
vol. 5, pp. 134–139, 2018. [20] I. Etikan, "Comparison of Convenience Sampling and
[15] C. Thomas, P. Comfort, P. A. Jones, and T. Dos' Purposive Sampling," Am. J. Theor. Appl. Stat., vol. 5,
Santos, "Strength and Conditioning for Netball: A no. 1, p. 1, 2016.
Needs Analysis and Training Recommendations,"

Authorized licensed use limited to: University of Gothenburg. Downloaded377

on December 18,2020 at 13:49:12 UTC from IEEE Xplore. Restrictions apply.