Content uploaded by Ricardo Duarte
Author content
All content in this area was uploaded by Ricardo Duarte on Sep 10, 2014
Content may be subject to copyright.
This material is the copyright of the original publisher.
Unauthorised copying and distribution is prohibited.
Terms and Conditions for Use of PDF
The provision of PDFs for authors’ personal use is subject to the following Terms & Conditions:
The PDF provided is protected by copyright. All rights not specifi cally granted in these Terms & Conditions are expressly
reserved. Printing and storage is for scholarly research and educational and personal use. Any copyright or other notices or
disclaimers must not be removed, obscured or modifi ed. The PDF may not be posted on an open-access website (including
personal and university sites).
The PDF may be used as follows:
• to make copies of the article for your own personal use, including for your own classroom teaching use (this includes
posting on a closed website for exclusive use by course students);
• to make copies and distribute copies (including through e-mail) of the article to research colleagues, for the personal use
by such colleagues (but not commercially or systematically, e.g. via an e-mail list or list serve);
• to present the article at a meeting or conference and to distribute copies of such paper or article to the delegates
attending the meeting;
• to include the article in full or in part in a thesis or dissertation (provided that this is not to be published commercially).
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
Sports Teams as Superorganisms
Implications of Sociobiological Models of Behaviour for Research
and Practice in Team Sports Performance Analysis
Ricardo Duarte,
1
Duarte Arau
´jo,
1,2
Vanda Correia
2,3
and Keith Davids
4
1 Laboratory of Expertise in Sports, Faculty of Human Kinetics, Technical University of Lisbon,
Lisbon, Portugal
2 CIPER –Interdisciplinary Centre for the Study of Human Performance, Lisbon, Portugal
3 School of Education and Communication, University of Algarve, Faro, Portugal
4 School of Human Movement Studies, Queensland University of Technology, Brisbane, QLD, Australia
Abstract Significant criticisms have emerged on the way that collective behaviours
in team sports have been traditionally evaluated. A major recommendation
has been for future research and practice to focus on the interpersonal re-
lationships developed between team players during performance. Most re-
search has typically investigated team game performance in subunits (attack
or defence), rather than considering the interactions of performers within the
whole team. In this paper, we offer the view that team performance analysis
could benefit from the adoption of biological models used to explain how
repeated interactions between grouping individuals scale to emergent social
collective behaviours. We highlight the advantages of conceptualizing sports
teams as functional integrated ‘super-organisms’ and discuss innovative mea-
surement tools, which might be used to capture the superorganismic proper-
ties of sports teams. These tools are suitable for revealing the idiosyncratic
collective behaviours underlying the cooperative and competitive tendencies
of different sports teams, particularly their coordination of labour and the
most frequent channels of communication and patterns of interaction be-
tween team players. The principles and tools presented here can serve as the
basis for novel approaches and applications of performance analysis devoted
to understanding sports teams as cohesive, functioning, high-order organisms
exhibiting their own peculiar behavioural patterns.
1. Introduction
Nature provides evidence that groups of co-
operating individuals can gain many functional
advantages in coordinating their actions when
working and living together. Research has dem-
onstrated the superior performance of groups
over single organisms in a wide range of human
social phenomena.
[1]
Sports teams are also com-
posed of different interacting individuals who
develop cooperative relations to achieve success-
ful performance outcomes. The collective per-
formance of sports teams has been extensively
investigated by a range of analytical performance
indicators.
[2,3]
For example, it has been demon-
strated that the percentage of ball possession in
association football teams changes as a function
of the evolving match status, game location and
CURRENT OPINION Sports Med 2012; 42 (8): 633-642
0112-1642/12/0008-0633/$49.95/0
Adis ª2012 Springer International Publishing AG. All right s reserved.
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
level of opposition,
[4]
and that the percentage of
ball possession in the opposition penalty box re-
mained high when teams used a counter-attacking
style.
[5]
However, the view that some game events
are more important than others, and that the prev-
alent methodology of notating discrete actions or
events fails to provide enough information about
the performance context, is challenging sports sci-
entists to rethink their research strategies.
[6]
With reference to tracked positional data, re-
cent studies have begun to reveal how players and
teams continuously interact during competition.
For example, teams tend to be tightly synchronized
in their lateral and longitudinal movements,
[7]
with
a counter-phase relation regarding their collec-
tive expansion and contraction movement pat-
terns,
[8]
commonly caused by changes in ball
possession.
[9]
This type of investigatory approach
shares some conceptual similarities with models
of behaviour in biological systems, which have
revealed emergent social collective behaviours
in human groups and animal societies.
[10]
Here,
we offer the view that sports teams also exhibit
emergent collective behavioural tendencies that
differ from the sum of individual aggregated
performances. Analysis of patterns of behaviour
atthecollectivesystemlevel in team sports requires
a reformation of notational analysis methods
used to study performance.
Like agents in other collective systems, sports
team performers often need to make decisions about
where to move and when, and which actions to
perform in uncertain and shifting environmental
conditions. It has been suggested both in biology
[11]
and in team sports analyses
[12]
that individuals base
their movement decisions on locally acquired
information sources such as the relative position-
ing, motion direction or changing motion direc-
tion of significant others operating in a system,
making a collective response all the more re-
markable. This finding implies an intertwined
relationship between perception, action and the
intentions of individuals functioning in a complex
biological system in order to intimately coordi-
nate their patterned behaviours. Next we discuss
the potential advantages of integrating socio-
biological models to study emergent collective
behaviours of performers in team sports.
2. A Brief Incursion into Sociobiology
Studies in biology have shown how the repeated
interactions among grouping animals (including
humans) scale to global collective system behav-
iours.
[10]
These social interactive behaviours
within a group lead to the emergence of a ‘collec-
tive’, which can be understood as a ‘new organism’
within the animal-environment system.
[13,14]
In
this sense, the actions of individual organisms
(e.g. team players) constrain and are constrained
by the actions of neighbouring organisms (e.g.
team-mates and opponents) toward the mutually
exclusive goals of the ‘collective’. The coordina-
tion and reorganization of these ongoing inter-
actions occur via externally controlled feedback
processes sustained by the continuous exchange
of information between the grouping individuals.
[15]
For example, when an ant finds a food source it
deposits a pheromone so that other members
of the colony can locate it, or when a fish swims in
a specific direction, its nearest neighbours in a
school soon follow to remain within one ‘fish’
length and preserve group security. As such re-
cruitment behaviours continue, the number of
individuals engaged in a goal-directed activity
multiplies.
[10]
Evidence is beginning to reveal that
similar processes seem to emerge during compet-
itive performance in team games in which a
player’s motions can functionally influence the
spatial-temporal characteristics of patterned move-
ments in team-mates and opponents, creating a
purposeful aggregation during specific perfor-
mance subphases.
[16,17]
What does a commitment to viewing sports
teams as complex superorganismic systems
imply? The sensitivity of biological systems to in-
formation provided in feedback loops can help us
to better understand the postulate that ‘a system
is more than the sum of its parts’. The functional
integration of individuals in highly related group-
ing organisms, such as social insects (e.g. ants,
bees, wasps), is a central aspect to consider. This
feature has been attributed to natural multi-
level selection mechanisms acting at the level of
between-group colonies and not just at the level
of genetic selection.
[18]
The evolutionary ad-
vantages of functional (re)organization through
634 Duarte et al.
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
cooperative activities have led some animal societies
to develop tightly coordinated and complementary
behaviours among group members, which im-
prove the likelihood of the entire group to be se-
lected, or to attain some functional adaptations
to their goal-directed activities.
[19]
In the perfor-
mance context of team sports, these advantages
may act to promote functional group adaptations
as a means of symbolic group ‘selection’. Here,
the term ‘selection’ can be equated with a team
succeeding when competing against another group
of individuals in sport, even though the outcomes
of performance might differ from those pursued
by a biological system like a colony, flock or
school. Despite the need for functional integra-
tion, each individual in a group is different in
terms of genetic heritage, previous experience and
specific roles in the group. It is widely accepted
that interindividual variation is a valuable process
that can lead to system variability and yield a
continual supply of new solutions to the behav-
ioural challenges that groups face.
[10]
Therefore,
complex biological systems face a complementary
interplay between functional specialization ten-
dencies (based on interindividual variation) and
functional integration tendencies.
[20]
These prop-
erties have led some scientists to advocate that
highly coordinated groups behave like ‘super-
organisms’,
[18,21-23]
since individuals possessing
high levels of interindividual variability can co-
operate together to perform as a single social
entity in order to achieve specific higher-order
task goals. For instance, team games players with
different attributes, unique skills and varied roles
(functional specialization) may work together to
collectively regain ball possession by restricting
space on the field and pressurizing the opposition
(functional integration).
2.1 Viewing Teams as ‘Superorganisms’
The ‘superorganism’ concept was first proposed
by William Morton Wheeler to describe the de-
gree to which ant colony members appear to op-
erate as a single functional unit.
[24]
The concept
has been extensively used in sociobiology al-
though some criticism has pointed to the absence
of experimental and mathematical support for this
notion.
[25]
Recent formal descriptions of group
adaptation
[19]
and intergroup competition
[26]
have
proved its utility and tempered the criticisms. An
example of successful collective system behaviours
was demonstrated by fire ants self-assembling
waterproof rafts as an adaptive evolutionary
strategy to survive floods.
[27]
The cooperative com-
plementary relations of the conspecific individuals
allowed the emergence of superorganismic behav-
iours based on the trapping of ants at the raft
edge by their neighbours. These data suggested
that the ‘superorganism’ concept can be defined
as ‘‘a group of individuals self-organized by divi-
sion of labour and united by a closed system of
communication’’,
[28]
(p. 84, in italics our emphasis).
These two main features of highly coordinated
grouping organisms –division of labour and sys-
tem of communication –might also be of interest
for performance analysis in sports teams, func-
tioning as integrated organisms.
Division of labour has been considered in stud-
ies of team sports as a key aspect expressing the
functional integration, complementarity of be-
haviour and coordination among team-mates.
[29]
The existence of a communication system is an-
other central issue also present in team sports
research. Coaches, performance analysts and re-
searchers have begun to enhance understanding
of the communication channels used by players
to support the effectiveness of teamwork.
[30]
Com-
mon actions in team games, such as passing the
ball or switching positions with team-mates, are
founded on a platform of communication or in-
formation exchange. Such actions imply the ex-
istence of informational links between players,
which allow them to detect the appropriate en-
vironmental conditions to successfully cooperate
during performance.
[31]
It has been suggested that
an important feature of superorganisms is the
existence of altruistic cooperation,
[28]
which op-
poses the existence of internal conflicts between
system agents during performance.
[19]
Despite the
absence of data on this latter issue, it is clear that
sports teams do possess altruistic forms of co-
operation between their team members. This idea
has been exemplified many times in team sports
by the unselfish play of an individual who passes
the ball to a better-placed team-mate to score,
Sports Teams as Superorganisms 635
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
while eschewing the personal glory of scoring
a deciding goal. Extending understanding from
other social neurobiological systems, we argue
that considering sports teams as functional inte-
grated ‘superorganisms’ might allow us to capture
the self-organized dynamics of complex social
interactions that shape collective behaviours in
teams. In order to progress understanding of sports
teams as superorganisms, sport science needs to
develop specific analysis methods that provide
insights into the functional collective behaviours
of such social neurobiological systems. How has
recent work developed methodological and anal-
ysis tools that can capture the superorganismic
properties of sports teams?
3. Capturing the ‘Superorganismic’
Properties of Sports Teams
The interactions of agents in sports teams,
defined as collective social systems, reveal common
underlying principles. In this respect, emergent
interactions between team players are sustained
by informational flow fields that specify each in-
dividual’s opportunities for action.
[32,33]
For ex-
ample, in the team sport of rugby union, Passos
and colleagues
[16]
showed that an attacking sub-
group with a ball, generated information for each
individual to decrease interpersonal distances with
other players and act as a single cohesive social unit,
during its approach to a subgroup of defenders.
However, each performer possessed different char-
acteristics that influenced his/her action capa-
bilities,
[34,35]
which constrained each individual
to display his/her own idiosyncratic behaviours.
Thus, conceptualizing sports teams as super-
organisms requires dedicated methodological tools
suitable for capturing the ‘division of labour’ and
the ‘communication systems’ of each collective
during the interplay between ‘interindividual var-
iation’ and ‘functional integration’ processes.
3.1 Tools to Assess ‘Division of Labour’ and
‘Communication Systems’
One approach to characterize the ‘division of
labour’ amongst individual agents in sports teams
involves measuring the area of a performer’s in-
terventions onfield (known as the major range).
[8]
The predominant area of each individual’s inter-
ventions during performance is defined by an
ellipse centred at the 2-dimensional mean loca-
tion of each performer, with semi-axes being the
standard deviations in X and Y directions, re-
spectively. Figure 1 displays the major ranges for
four backs (figure 1a) and three forward players
(figure 1b) of a goalkeeper+4+3+3 team forma-
tion during a football match, classified by de-
fending and attacking phases (as a function of
ball possession).
30
Defending phase
−50 −40 −30 −20 −10 0 1020304050
X (m)
Attacking phase
Y (m)
20
10
0
−10
−20
−30
9
7
6
7
9
6
30
ab
Defending phase
−50 −40 −30 −20 −10 0 1020304050
X (m)
Attacking phase
Y (m)
20
10
0
−10
−20
−30
10
11
3
2
2
3
11
10
Fig. 1. Major ranges for two subgroups of football players in defending and attacking phases. (a) Shows the four defensive players; (b) shows
the three forwards of the same team. Reproduced from Yue Z, et al.,
[8]
with permission.
636 Duarte et al.
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
These spatial data revealed an inverse trend
between the two subgroups of players. While de-
fenders increased their individual covered area
(i.e. range of displacement trajectories) in at-
tacking phases, the forwards enlarged their cov-
ered areas in the defending phases. This tool can
also be used to assess the coordination of labour
during performance as the game unfolds. Figure 2
presents exemplar data from a change in the
global trend of team performance, from smaller
and more proportional individual areas of inter-
vention (a) in the first 5 minutes of the match, to
highly narrow and elongated ranges (b) in the
next 5 minutes of the game.
Data from figure 2 exemplify how the co-
ordination of labour can change during perfor-
mance under the influence of natural variations
in competitive constraints, such as the opposi-
tion’s style of play
[5]
or the dynamics of ball dis-
placement onfield.
[37]
For instance, the changes
observed in figure 2 can be interpreted as indicat-
ing that the opposing team has started to impose
a direct playing style in the sport of football, with
predominant longitudinal displacements of the
ball.
Concerning the ‘communication system’ of
teams, an approach to capture the tendencies in
the relationships of team players is provided by
small-world networks.
[38]
A ball passing action
exemplifies a functional relationship between team-
mates and trends of passing relations can reveal
preferred channels of communication within a
team. Figure 3 shows an application of network
analysis to passing trends during a water polo
competition, based on simple notational data.
[31]
The strength of the relationships between pairs
of players was expressed by the width of the ar-
rows. Dominant relations and probabilities of
interaction among team-mates can be assessed
using this method. Deeper understanding of team-
work effectiveness can be obtained by constructing
ab
Fig. 2. Major ranges of ten outfield players from a football team.
(a) Shows values from the first 5 minutes of the game. (b) Displays
values from the next 5-minute segment. A grey colour was used to
distinguish the midfielders from the defenders and forwards. Re-
produced from Duarte et al.,
[36]
with permission.
ab
Player 2
Player 3
Player 6
Goalkeeper of
opposite team
Player 4 Player 5
Player 1
Player 2
Player 3 Player 6
Goalkeeper of
opposite team
Player 4 Player 5
Player 1
Fig. 3. (a) and (b) display trends for each team. Grey circles represent players involved in the units of attack. Orientation of the black arrows
indicates pass direction. Origin of the arrow represents the player who passed the ball and the arrowhead represents the player who received
the ball. Width of the black arrows denotes quantity of passes from one player to another during performance (i.e. the thicker the arrows, the
more passes occurred between specific players). Reproduced from Passos et al.,
[31]
with permission.
Sports Teams as Superorganisms 637
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
networks according to passing accuracy.
[30]
Other
actions, such as switching positions between two
players, if understood as communication pro-
cesses, can also be studied in these analyses. Thus,
networks provide a useful method to qualitatively
and quantitatively describe the communication-
based interactions that emerge among players in
team sports.
3.2 Compound Positional Variables
Based on individual positional data, other
variables have been proposed to assess specific
functional collective behaviours of sports teams.
These have been termed compound positional
variables because they integrate the individual
positions of each team player into a meaningful
description of a collective team pattern.
Examples of such compound positional vari-
ables include:
i) the surface area occupied by teams,
[39]
which
represents the covered area of the field by the
whole team in each time frame;
ii) the geometrical centre of teams,
[7,8]
which
represents a kind of ‘centre of mass’ of a team;
iii) the stretch index of teams
[9]
(also known as
radius
[8]
), which represents the mean dispersion
value of the players around the centre of each
team (i.e. the geometrical centre);
iv) the team ranges (also known as length and
width of the team
[7]
), which represent the size of
the team in the longitudinal and lateral field
directions.
These innovative compound measures of team
performance reveal meaningful collective behav-
iours from a practical perspective and can be
used to assess the idiosyncratic performance val-
ues of each team.
[36]
Table 1 in the Supplemental
Digital Content illustrates these compound posi-
tional variables from two competing teams
[40]
and
can be found online at http://links.adisonline.
com/SMZ/A10. These collective measures can
assist understanding of interactions between agents
in sports teams as having ‘superorganismic’ qual-
ities. Observing their changes on different time-
scales, due to variations in performance constraints,
such as the evolving score line or different offen-
sive playing styles, is an important aspect to
consider in future performance evaluations. An-
other important issue to consider when measuring
team performance behaviours is to discriminate
values for compound positional variables during
defending and attacking performance phases.
Measurement functions to discriminate data be-
tween these phases have been reported previously
in the literature.
[8]
Despite the merits and poten-
tial of these collective measures, they are based on
the assumption that each individual agent’s be-
haviour equally contributes to functional collec-
tive performance. However, team players may
not always have the same weight in the emergence
of the social collective system behaviours.
[41]
The
weighting of the contribution of each player may
change as a function of the evolving game context
(e.g. the place where the ball is located onfield)
and the action capabilities of each individual (e.g.
maximum movement displacement speed). The
next section proposes alternative methods that
account for the weighted contribution of each
team player during competitive performance.
4. Emerging Alternative Approaches and
Future Directions
4.1 Cluster Phase
The cluster phase method was recently pro-
posed in order to analyse synchrony within
systems with a small number of oscillating com-
ponents.
[42]
This method is based on the Kur-
amoto order parameter,
[43]
which has been used
to investigate phase synchronization in sys-
tems with large numbers of oscillating compo-
nents (e.g. emergence of collective clapping in
theatre audiences).
[44]
The Kuramoto model de-
scribes the synchronization of oscillatory move-
ment components (e.g. team players) in a single
collective parameter. Investigators have adapted
this model and showed its applicability using
a rocking chair paradigm with only six oscilla-
tory units (i.e. six individuals coordinating rock-
ing chair movements).
[42]
Specific measures of
individual and whole-group synchrony can be
obtained, which can be a useful means to quan-
tify the contribution of each team player to the
global behaviour of a team, as well as changes in
638 Duarte et al.
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
the global synchronization tendencies within
a team.
4.2 Dominant Region
The dominant region is a method for group
motion analysis proposed to analyse the spatial
interactions in team sports.
[45]
The dominant re-
gion of a team player is defined as a region of the
field space where a particular performer is likely
to arrive earlier than other players.
[46]
This method
determines a high functional and dynamic sphere
of influence around each individual by integrat-
ing data on position, speed, direction and accel-
eration. Comparing those kinematic data of all
the performers, this method specifies their move-
ment possibilities and the functional area of in-
tervention behind the control of each specific
player. The weighted contribution of each player
interacting together with his/her team-mates and
opponents originates a purposeful aggregation
(functional integration), which can collectively
express the space-time relations between teams.
Dominance diagrams integrating the sphere of
influence of all players can be visualized in a
frame-by-frame manner depending on the sam-
pling rate of the positional data.
[47]
Figure 4 presents
an illustration of the dominant region method for
a single frame. The measures provided are time
series of individual and collective dominant re-
gion areas, time occupancy rate per specific zones
and number of links with near neighbours of each
team.
4.3 Modelling
Common underlying principles of team sports
and animal collectives can form the basis for
a formal description (mathematical modelling)
of the collective behavioural dynamics of sports
teams, allowing scientists and practitioners to
make accurate predictions about team behaviours.
More than 30 years ago, biologists
[48]
developed
computer simulations, known as self-propelled
particle (SPP) models, that captured the collec-
tive behaviours of animal groups in terms of local
interactions developed.
[10]
For example, Couzin
and colleagues
[49]
proposed a model in which in-
dividual animals follow only three simple rules of
thumb: (i) move away from very nearby neigh-
bours; (ii) adopt the same direction as those that
are close by; and (iii) avoid becoming isolated.
Biological systems such as schools of fish are able
to produce different complex patterns due to
small changes in these simple localized rules. SPP
models have also been used to formalize phe-
nomena in human crowds. Treating humans as
particles that interact according to a set of ‘social
forces’, these models have been successful in
predicting specific collective behaviours such as
escape panic, walking in a busy street, the for-
mation of Mexican waves in football stadiums
and the emergence of traffic jams.
[10]
It is likely
that adaptations of these models can be success-
fully applied to capture the time-evolving dy-
namics of sports teams as functional integrated
entities or ‘superorganisms’. For example, these
models could help coaches predict how attacking
and defensive formations might change during
the course of a match as specific individuals be-
come fatigued, if weather conditions deteriorate
or if the competitive performance constraints of a
game changes from beginning to end.
[50]
This
might be an important advance for sports per-
formance analysis, given the recent criticism that
it is overly concerned with documenting discrete
performance statistics, often in specific perfor-
mance subphases.
[51]
These innovative collective
system analysis methods may support simula-
tions and accurate theoretically principled pre-
dictions about the collective behaviours of whole
teams under competitive conditions.
[52]
5. Concluding Remarks
This paper has attempted to conceptualize sports
teamsasfunctionallyintegrated ‘superorganisms’,
proposing an explanation of how highly co-
ordinated collaborating players might collectively
operate as a single social unit. From a performance
analysis perspective, coordination tendencies un-
derlying the emergence of team behaviours seem
to be governed by locally-generated informa-
tion sources from the relative positioning of other
team players, motion directions and changes in
motion. The ‘superorganism’ proposal, more
than focusing attention on compiling discrete
Sports Teams as Superorganisms 639
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
action frequencies, suggests the need to regard the
meaningful and synergistic (inter)actions within
sports teams as the appropriate focus of analysis.
The present framework suggests the reconceptu-
alization of research approaches for studying team
sport collectives, as well as performance analysis
applications. For instance, simple notation data
typically collected by performance analysts can
be interpreted in reference to small-world net-
works. Notational variables should also contain
contextual information regarding the performance
constraints surrounding the players’ behaviours,
such as their relative position on the pitch, nu-
merical relations between players in opposing
teams and relative dispersion between teams. The
development of player tracking systems, such as
electronic portable devices or multi-player video-
based systems
[52,53]
offer a novel opportunity to
improve research and sports performance ana-
lyses. The innovative tools presented here might
support these novel approaches to performance
analysis, devoted to the understanding of sports
11
9
6
a
b
4
2
73
8
7
510 2
6
10
9
5
4
811
31
1
11
9
6
4
2
73
8
7
510 2
6
10
9
5
4
811
31
1
Fig. 4. Visualization of the dominant region method in a single frame. (a) Shows players’ onfield positions with tracked courses. (b) Displays
individual (boundary lines) and each team’s dominant regions (colour contrasts), as well as intra-team links among players sharing direct/
immediate space (Taki T, exemplar unpublished data).
640 Duarte et al.
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
teams as functionally integrated high-order
organisms that exhibit their own idiosyncratic
features.
The conceptual approach presented here might
also have some important applications concern-
ing learning and training design. Changing the
local available information that guides players’
decisions and actions during practice games,
coaches can manipulate the local rules governing
interactions between neighbour team-mates, in-
ducing the emergence of new patterns of collec-
tive movement solutions. For example, changing
the numbers of players involved in different team
game practices is likely to promote functional
adaptations in the way team players coordinate
their labours, allowing them to explore different
channels of communication, which should lead to
the emergence of a distinct pattern of collective
behaviour.
[54]
In a practical and functional way,
this conceptualization can help coaches and sport
scientists gain a better understanding of processes
of team cohesion related to task and social
orientation. For example, in many team sports
there is a current tendency to use a large squad of
players to enhance and refresh the competitive
performance of the team throughout the season.
The team-as-superorganism concept could help
coaches and practitioners to understand how to
foster ‘group thinking’ so that teams can cope
with issues like injuries, illness and loss of form,
rather than to allow team players to adopt an
individualized focus, throughout such a long
performance season.
A challenging future task for researchers and
practitioners is the formal description of social
collective behaviours of sports teams. Mathemat-
ical models adapted from other biological sys-
tems, such as SPP models, may provide computer
simulations to undertake performance predic-
tions without the need to experimentally test a
whole range of team patterns onfield.
Acknowledgements
This research was partially supported by a grant
(SFRH/BD/43994/2008) awarded to Ricardo Duarte by the
Foundation for Science and Technology (Portugal). The au-
thors wish to thank Hugo Foldado, Telmo Frias and Tsuyoshi
Taki for their valuable help in some computation procedures.
The authors have no conflicts of interest to declare that are
directly relevant to the content of this review.
References
1. Kerr NL, Tindale RS. Group performance and decision
making. Annu Rev Psychol 2004; 55: 623-55
2. Hughes M, Bartlett R. The use of performance indicators in
performance analysis. J Sports Sci 2002; 20: 739-54
3. Hughes M. Notational analysis: a mathematical perspective.
Int J Perf Anal Sport 2004; 4: 97-139
4. Lago C, Martin R. Determinants of possession of the ball in
soccer. J Sports Sci 2007; 25: 969-74
5. Tenga A, Holme I, Ronglan LT, et al. Effect of playing
tactics on achieving score-box possessions in a random
series of team possessions from Norwegian professional
soccer matches. J Sports Sci 2010; 28: 245-55
6. James N. Notational analysis in soccer: past, present and
future. Int J Perf Anal Sport 2006; 6: 67-81
7. Lames M, Ertmer J, Walter F. Oscillations in football: order
and disorder in spatial interactions between the two teams.
Int J Sport Psychol 2010; 41: 85-6
8. Yue Z, Broich H, Seifriz F, et al. Mathematical analysis of a
soccer game. Part I: individual and collective behaviors.
Stud Appl Math 2008; 121: 223-43
9. Bourbousson G, Se
`ve C, McGarry T. Space-time coordina-
tion dynamics in basketball: part 2. The interaction be-
tween the two teams. J Sports Sci 2010; 28: 349-58
10. Sumpter DJT. The principles of collective animal behaviour.
Phil Trans R Soc B 2006; 361: 5-22
11. Couzin ID. Collective cognition in animal groups. Trends
Cogn Sci 2009; 13: 36-43
12. Grehaigne JF, Bouthier D, David B. Dynamic-system anal-
ysis of opponent relationships in collective actions in soc-
cer. J Sports Sci 1997; 15: 137-49
13. Marsh KL, Richardson MJ, Baron RM, et al. Contrasting
approaches to perceiving and acting with others. Ecol
Psychol 2006; 18: 1-38
14. Marsh KL, Richardson MJ, Schmidt RC. Social connection
through joint action and interpersonal coordination. Top
Cogn Sci 2009; 1: 320-39
15. Krause J, Ruxton GD, Krause S. Swarm intelligence in an-
imals and humans. Trends Ecol Evol 2010; 25: 28-34
16. Passos P, Milho J, Fonseca S, et al. Interpersonal distance
regulates functional grouping tendencies of agents in team
sports. J Motor Behav 2011; 43: 155-63
17. Correia V, Arau´ jo D, Davids K, et al. Territorial gain dy-
namics regulates success in attacking sub-phases of team
sports. Psychol Sport Exerc 2011; 12: 662-9
18. Wilson DS, Wilson EO. Rethinking the theoretical founda-
tion of sociobiology. Q Rev Biol 2007; 82: 327-48
19. Gardner A, Grafen A. Capturing the superorganism: a
formal theory of group adaptation. J Evol Biol 2009; 22:
659-71
20. Edelman GM, Gally JA. Degeneracy and complexity in
biological systems. Proc Natl Acad Sci U S A 2001; 98:
13763-8
21. Seeley TD. The honey bee colony as a superorganism. Amer
Sci 1989; 77: 546-53
Sports Teams as Superorganisms 641
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)
This material is
the copyright of the
original publisher.
Unauthorised copying
and distribution
is prohibited.
22. Seeley TD. Decision making in superorganisms: how col-
lective wisdom arises from the poorly informed masses. In:
Gigerenzer G, Selten R, editors. Bounded rationality: the
adaptive toolbox. Cambridge: MIT Press, 2001: 249-61
23. Wilson DS, Sober E. Reviving the superorganism. J Theor
Biol 1989; 136: 337-56
24. Wheeler WM. The ant-colony as an organism. J Morphol
1911; 22: 307-25
25. Crozier RH. Insect sociobiology: the genetics of social evo-
lution. Science 1989; 245: 313-4
26. Reeve HK, Ho
¨lldobler B. The emergence of a super-
organism through intergroup competition. Proc Natl Acad
Sci U S A 2007; 104: 9736-40
27. Mlot NJ, Tovey CA, Hu DL. Fire ants self-assemble into
waterproof raftsto survive floods. Proc Natl Acad SciU S A
2011; 108: 7669-73
28. Ho
¨lldobler B, Wilson EO. The superorganism: the beauty,
elegance, and strangeness of insect societies. London:
W.W. Norton, 2009
29. Eccles D. The coordination of labour in sports teams. Int
Rev Sport Exerc Psychol 2010; 3: 154-70
30. Duch J, Waitzman JS, Amaral LAN. Quantifying the per-
formance of individual players in a team activity. PLoS
ONE 2010; 5: e10937
31. Passos P, Davids K, Arau´ jo D, et al. Networks as a novel
tool for studying team ball sports as complex social sys-
tems. J Sci Med Sport 2011; 14: 170-6
32. Arau´ jo D, Davids K, Hristovski R. The ecological dynamics of
decision making in sport. Psychol Sport Exerc 2006; 7: 653-76
33. Gibson JJ. The ecological approach to visual perception.
Boston: Houghton Mifflin, 1979
34. Fajen B, Riley MR, Turvey MT. Information, affordances
and control of action in sports. Int J Sport Psychol 2009;
40: 79-107
35. Dicks M, Davids K, Button C. Individual differences in the
visual control of intercepting a penalty kick in association
football. Hum Mov Sci 2010; 29: 401-11
36. Duarte R, Frias T. Collective intelligence: An incursion
into the tactical performance of football teams. In: Jemni
M, Bianco A, Palma A, editors. Proceedings of the First
International Conference in Science and Football; 2011
Apr 15-17; Palermo: 23-28
37. Travassos B, Arau´ jo D, Vilar L, et al. Interpersonal co-
ordination and ball dynamics in futsal (indoor football).
Hum Mov Sci 2011; 30: 1245-59
38. Watts DJ, Strogatz SH. Collective dynamics of ‘small-world’
networks. Nature 1998; 393: 440-2
39. Frencken W, Lemmink K, Delleman NJ, et al. Oscillations
of centroid position and surface area of soccer teams in
small-sided games. Eur J Sport Sci 2011; 4: 215-23
40. Duarte R, Arau´ jo D, Davids K, et al. In search for dyna-
mical patterns of teams’ tactical behaviours during the
match 2011 [abstract no. 114 plus oral presentation]. VII
th
World Congress on Science &Football; 2011 May 26-30;
Nagoya
41. McGarry T. Applied and theoretical perspectives of perfor-
mance analysis in sport: scientific issues and challenges. Int
J Perf Anal Sport 2009; 9: 128-40
42. Frank TD, Richardson MJ. On a test statistic for the Kur-
amoto order parameter of synchronization: an illustration
for group synchronization during rocking chairs. Physica
D 2010; 239: 2084-92
43. Kuramoto Y, Nishikawa I. Statistical macrodynamics of
large dynamical systems: case of a phase transition in os-
cillator communities. J Stat Phys 1987; 49: 569-605
44. Neda Z, Ravasz E, Brechet Y, et al. The sound of many
hands clapping. Nature 2000; 403: 849-50
45. Taki T, Hasegawa J. Quantitative measurement of team-
work in ball games using dominant region. ISPRS J Pho-
togramm 2000; 33: 125-31
46. Taki T, Hasegawa J. Dominant region: a basic feature for
group motion analysis and its application to teamwork
evaluation in soccer games. P SPIE IS&T Elect IM 1998;
3641: 48-57
47. Taki T, Hasegawa J. Visualization of dominant region in
team games and its application to teamwork analysis. In:
Proceedings of the International Conference on Computer
Graphics; 2000 Jun 19-24; Geneva, 227-235
48. Deneubourg JL. Application de l’ordre par fluctuations a
`la
description de certaines e
´tapes de la construction du nid
chez les termites. Insect Soc 1977; 24: 117-30
49. Couzin ID, Krause J, James R, et al. Collective memory
and spatial sorting in animal groups. J Theor Biol 2002;
218: 1-11
50. Glazier P. Game, set and match? Substantive issues and fu-
ture directions in performance analysis. Sports Med 2010;
40: 625-34
51. Vilar L, Arau´ jo D, Davids K, et al. The role of ecological
dynamics in analysing performance in team sports. Sports
Med 2012; 42: 1-10
52. Barris S, Button C. A review of vision-based motion analysis
in sport. Sports Med 2008; 38: 1025-43
53. Carling C, Bloomfield J, Nelson L, et al. The role of motion
analysis in elite soccer: contemporary performance mea-
surement techniques and work rate data. Sports Med 2008;
38: 389-862
54. Correia V, Arau´ jo D, Duarte R, et al. Changes in practice
task constraints shape decision-making behaviours of team
games players. J Sci Med Sport 2012; 15 (3): 244-9
Correspondence: Ricardo Duarte, Faculdade de Motricidade
Humana, Estrada da Costa, 1495-688 Cruz Quebrada,
Portugal.
E-mail: rduarte@fmh.utl.pt
642 Duarte et al.
Adis ª2012 Springer International Publishing AG. All rights reserved. Sports Med 2012; 42 (8)