Abstract
The purpose of this chapter is to introduce the concepts of dispersion in the aim of soccer analysis. A set of different measures have been proposed to identify the level of dispersion between teammates and between opponents. Based on that, a summary of the dispersion measures, definitions, interpretation and graphical visualization will be presented on this chapter. The measures of Stretch Index, Surface Area, Team Length and Team Width and lpwratio will be introduced throughout the chapter. The case studies presented involve two five-player teams in an SSG considering only the space of half pitch (68 m goal-to-goal and 52 m side-to-side) and another eleven-player team in a match considering the space of the entire field (106.744 m goal-to-goal and 66.611 m side-to-side) even though only playing in half pitch.
Access provided by CONRICYT-eBooks. Download chapter PDF
Similar content being viewed by others
Keywords
5.1 Stretch Index
5.1.1 Basic Concepts
The Stretch Index of a team is calculated as the average distance of a team’s players and the Geometrical Center, giving a notion of the compactness of the team.
Definition 5.1
[1] The Stretch Index, considering both axes, at a given instant t can be calculated by:
where C(t) represents the Geometrical Center of the team, N the number of players in the team, \(P_{xk}(t)\) the position along the longitudinal axis for player k at instant t, and \(P_{yk}(t)\) the position along the vertical axis for player k in instant t.
Definition 5.2
[1] The Stretch Index on a single axis, for a given instant t, is given by the following expression:
where \(C_x(t)\) represents the x coordinate for the Geometrical Center at instant t and \(P_{xk}(t)\) the position along the longitudinal axis for player k at instant t.
Remark 5.1
The same formula is applicable for the calculation along the vertical axis, only replacing \(P_{xk}(t)\) for \(P_{yk}(t)\) and \(C_x(t)\) for \(C_y(t)\).
Definition 5.3
[1] The average values of the Stretch Index both for the coordinates, along the two axes and for each axis separately, is given by the following equation:
where \(N_t\) represents the total number of time instants measured.
Remark 5.2
This same formula is applicable for \(\overline{SI_x}\) and \(\overline{SI_y}\), by replacing SI(t) for \(SI_x(t)\) or \(SI_y(t)\), respectively.
5.1.2 Real Life Examples
The results obtained by a player from Team A in the SSG are presented in Table 5.1 with intervals of 30 s and for the entire 3 min in Table 5.2.
The results obtained by a player from Team B in the SSG are presented in Table 5.3 with intervals of 30 s and for the entire 3 min in Table 5.4.
A screenshot of a representation of the Stretch Index captured from the uPATO software is displayed in Fig. 5.1.
5.1.3 General Interpretation
Stretch Index was introduced in basketball to measure the expansion and contraction of space, in both axes (longitudinal and lateral), demonstrated by a team during a match [2]. This measure represents the mean deviation of each teammate to the geometrical center [2, 3]. Stretch Index is also known as radius [4].
Dispersion of the players depends on contextual variables and mostly on the moment of the game (with or without possession of the ball) [5]. Dispersion is greater in attacking moments (with possession of the ball) and smaller in defensive pressure (without possession of the ball), in the case of soccer [6]. This follows the main idea that in attacking moments it is necessary to spread the players to attract the opponents towards the outside of the middle and in defensive moments it is necessary to keep the teammates closer to guarantee fewer spaces for opponent’s penetration [7].
This measure works in longitudinal, lateral and/or global, thus different information can be used. In the case of the Stretch Index for the longitudinal axis it can be computed to measure specific situations of counter attacks in which a greater dispersion in goal-to-goal direction can be observed. In the other hand, greater dispersions are found in side-to-side during positional attack (ball circulation). Considering the defensive moments, both axes will drastically decrease in comparison to attacking moments. However, longitudinal dispersion can be used to classify the defensive pressure against positional attacks in which large values of dispersion may suggest that forward teammates are too far away from the defensive colleagues.
5.2 Surface Area
5.2.1 Basic Concepts
The Surface Area of a team is calculated as the area of the polygon defined as the convex polygon with the least number of vertices that can encompass all of the teams’s players, and where the potential vertices are defined as the positions of the players.
Definition 5.4
[8, 9] Given a set of points, the following algorithm is applied to define the Convex Hull:
Definition 5.5
[9, 10] Given the coordinates of the n vertices that compose the Convex Hull of the team, the Surface Area is given by the following equation:
where \((x_i, y_i)\) are the coordinates of the ith vertex of the Convex Hull.
5.2.2 Real Life Examples
The results obtained by both teams in the SSG and another team in the match are presented in Table 5.5 with intervals of 30 s and for the entire 3 min in Table 5.6.
A screenshot of a representation of the Surface Area captured from the uPATO software is displayed in Fig. 5.2.
5.2.3 General Interpretation
Surface Area represents the area of a polygon constituted by all teammates (dots) [11]. The measure can be defined as the total space covered by a team considering the area within the convex hull [9]. The Surface Area measures the contraction and expansion of teams across the soccer match, as does the Stretch Index [3]. However, this measure represents all the area covered by the team, while the Stretch Index only measures the mean deviation to the Geometrical Center of the team.
Typically, the area of the teams is significantly bigger in possession of the ball than without possession [3, 5]. Moreover, the variability of Surface Area decreases across the game [4, 12], thus suggesting a stabilization of the team in attacking and defensive moments.
In a study conducted in elite Spanish soccer teams values between 800 and 2800 \(\mathrm {m^{2}}\) in attacking and [1000; 2000] \(\mathrm {m^{2}}\) in defensive moments were found [13]. Values in attacking varied between 1638 and 1831 \(\mathrm {m^{2}}\) in possession of the ball and [1277; 1369] \(\mathrm {m^{2}}\) without possession in an elite Portuguese team [14].
The visualization of Surface Area can help coaches identify the space between sectors (defensive, middle and forward) and the dispersion of the team in specific moments. The generated triangulations can provide an immediate analysis of the collective behavior, particularly in moments of positional attack and defensive ‘block’.
5.3 Team Length and Team Width
5.3.1 Basic Concepts
The Team Length and Width is defined by its most advanced and rear players for the length, and its rightmost and leftmost players for the length.
Definition 5.6
[4] Given a set of points for team player positioning along a time-series of length N, P, and where \(P_x\) and \(P_y\) represent the set of longitudinal and lateral coordinates for every player of the team on every measured time instant, the Team Length and Team Width on a given time instant i can be calculated as follows:
where \(t_l\) represents the Team Length, \(t_w\) represents the Team Width and \((P_x(i), P_y(i))\) represents the set of coordinates of the team’s players in instant i.
5.3.2 Real Life Examples
The results obtained by both teams in the SSG are presented in Table 5.7 with intervals of 30 s and for the entire 3 min in Table 5.8.
The results obtained by the team in the match are presented in Table 5.9 with intervals of 30 s and for the entire 3 min in Table 5.10.
5.3.3 General Interpretation
The Team Length represents the maximum length of a team considering the minimum and maximum position of a player in the longitudinal (goal-to-goal) direction [4]. The same application is applied in the case of Team Width (side-to-side) [4]. Therefore, the length and width measures provide information about how stretched are the two farthest players of a team in longitudinal and lateral directions.
Coaches can use this information to understand the optimal distances to readjust some tactical tasks in training sessions based on usual length and width found in official matches. An interesting approach used the width and length to suggest specific sizes to work positional attack and counter-attack in small-sided games [15]. Moreover, coaches can compare the length and width in different moments of the match and identify in which moments the extreme size can be associated with critical moments (e.g., shots, goals).
5.4 Length per Width Ratio
5.4.1 Basic Concepts
The ratio between a team’s length and width is the Length per Width Ratio (lpwratio) of that team.
Definition 5.7
[16] Given \(t_l\) and \(t_w\) as the Team Length and Team Width of a team on a given instant, the lpwratio of a team is given by the following equation:
5.4.2 Real Life Examples
The results obtained by both teams in the SSG and the other team in the match are presented in Table 5.11 with intervals of 30 s and for the entire 3 min in Table 5.12.
A screenshot of a representation of the lpwratio captured from the uPATO software is displayed in Fig. 5.3.
5.4.3 General Interpretation
The lpwratio quantifies the relationship between the length (maximum distance between the two farthest players in longitudinal direction) and width (maximum distance between the two farthest players in lateral direction) during the match [16]. The authors of this measure argue that small variations of this measure may suggest a team’s higher adherence to width and concentration on “principles of play” [16]. In the other hand, large variations of this ratio may suggest a more individual and less collectively coordinated approach to the soccer game [16].
Coaches can use this measure to classify the teams in different moments of the match. Some teams will tend to play in counter-attack, thus increasing the length and reducing the width. In the other hand, teams that opt to attack with circulation of the ball will have increased width and reduced the length. This relationship will help to understand some patterns of play. The same case can be applied for defensive moments in which teams that opt to defend in ‘block’ closer to the goal will present decreased length and teams that opt to extend the defensive ‘block’ for middle or forward zones of the pitch will have increased length.
References
Silva P, Vilar L, Davids K, Araújo D, Garganta J (2016) Sports teams as complex adaptive systems: manipulating player numbers shapes behaviours during football small-sided games. SpringerPlus 5(1):191
Bourbousson J, Sève C, McGarry T (2010) Space-time coordination dynamics in basketball: part 2. the interaction between the two teams. J Sport Sci 28(3):349–358
Bartlett R, Button C, Robins M, Dutt-Mazumder A, Kennedy G (2012) Analysing team coordination patterns from player movement trajectories in football: methodological considerations. Int J Perform Anal Sport 12(2):398–424
Duarte R, Araújo D, Folgado H, Esteves P, Marques P, Davids K (2013) Capturing complex, non-linear team behaviours during competitive football performance. J Sys Sci Complex 26(1):62–72
Clemente FM, Couceiro MS, Martins FML, Mendes RS, Figueiredo AJ (2013a) Measuring tactical behaviour using technological metrics: case study of a football game. Int J Sport Sci Coach 8(4):723–739
da Costa IT, Garganta J, Greco PJ, Mesquita I, Seabra A (2010) Influence of relative age effects and quality of tactical behaviour in the performance of youth soccer players. Int J Perform Anal Sport 10:82–97
Clemente FM, Martins F, Mendes R, Figueiredo A (2015) A systemic overview of football game: the principles behind the game. J Hum Sport Exerc 9(2):656–667
Barber CB, Dobkin DP, Huhdanpaa H (1996) The quickhull algorithm for convex hulls. ACM Trans Math Softw (TOMS) 22(4):469–483
Frencken W, Lemmink K, Delleman N, Visscher C (2011) Oscillations of centroid position and surface area of football teams in small-sided games. Eur J Sport Sci 11(4):215–223
Bourke P (1988) Calculating the area and centroid of a polygon. http://paulbourke.net/geometry/polygonmesh
Clemente FM, Couceiro MS, Martins FML, Mendes RS, Figueiredo AJ (2013b) Measuring collective behaviour in football teams: inspecting the impact of each half of the match on ball possession. Int J Perform Anal Sport 13(3):678–689
Moura FA, Martins LEB, Anido RO, Ruffino PRC, Barros RML, Cunha SA (2013) A spectral analysis of team mics and tactics in brazilian football. J Sport Sci 31(14):1568–1577 PMID: 23631771
Castellano J, Álvarez D, Figueira B, Coutinho D, Sampaio J (2013) Identifying the effects from the quality of opposition in a football team positioning strategy. Int J Perform Anal Sport 13:822–832
Clemente FM, Couceiro MS, Martins FML, Mendes RS, Figueiredo AJ (2014) Using collective metrics to inspect spatio-temporal relationships between football players. South African J Res Sport Phys Educ Recreat 36(2):47–59
Fradua L, Zubillaga A, Caro O, Iván Fernández-García A, Ruiz-Ruiz C, Tenga A (2013) Designing small-sided games for training tactical aspects in soccer: extrapolating pitch sizes from full-size professional matches. J Sports Sci 31(6):573–581
Folgado H, Lemmink KA, Frencken W, Sampaio J (2014) Length, width and centroid distance as measures of teams tactical performance in youth football. Eur J Sport Sci 14(sup1):S487–S492
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2018 The Author(s)
About this chapter
Cite this chapter
Clemente, F.M., Sequeiros, J.B., Correia, A.F.P.P., Silva, F.G.M., Martins, F.M.L. (2018). Measuring the Dispersion of the Players. In: Computational Metrics for Soccer Analysis. SpringerBriefs in Applied Sciences and Technology. Springer, Cham. https://doi.org/10.1007/978-3-319-59029-5_5
Download citation
DOI: https://doi.org/10.1007/978-3-319-59029-5_5
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-59028-8
Online ISBN: 978-3-319-59029-5
eBook Packages: EngineeringEngineering (R0)