Date Published: July 14, 2017
Publisher: Public Library of Science
Author(s): Angel Ric, Carlota Torrents, Bruno Gonçalves, Lorena Torres-Ronda, Jaime Sampaio, Robert Hristovski, Tiago M. Barbosa.
The analysis of positional data in association football allows the spatial distribution of players during matches to be described in order to improve the understanding of tactical-related constraints on the behavioural dynamics of players. The aim of this study was to identify how players’ spatial restrictions affected the exploratory tactical behaviour and constrained the perceptual-motor workspace of players in possession of the ball, as well as inter-player passing interactions. Nineteen professional outfield male players were divided into two teams of 10 and 9 players, respectively. The game was played under three spatial constraints: a) players were not allowed to move out of their allocated zones, except for the player in possession of the ball; b) players were allowed to move to an adjacent zone, and; c) non-specific spatial constraints. Positional data was captured using a 5 Hz interpolated GPS tracking system and used to define the configuration states of players for each second in time. The configuration state comprised 37 categories derived from tactical actions, distance from the nearest opponent, distance from the target and movement speed. Notational analysis of players in possession of the ball allowed the mean time of ball possession and the probabilities of passing the ball between players to be calculated. The results revealed that the players’ long-term exploratory behaviour decreased and their short-term exploration increased when restricting their space of interaction. Relaxing players’ positional constraints seemed to increase the speed of ball flow dynamics. Allowing players to move to an adjacent sub-area increased the probabilities of interaction with the full-back during play build-up. The instability of the coordinative state defined by being free from opponents when players had the ball possession was an invariant feature under all three task constraints. By allowing players to move to adjacent sub-areas, the coordinative state became highly unstable when the distance from the target decreased. Ball location relative to the scoring zone and interpersonal distance constitute key environmental information that constrains the players’ coordinative behaviour. Based on our results, dynamic overlap is presented as a good option to capture tactical performance. Moreover, the selected collective (i.e. relational) variables would allow coaches to identify the effects of training drills on teams and players’ behaviour. More research is needed considering these type variables to understand how the manipulation of constraints induce a more stable or flexible dynamical structure of tactical behaviour.
The analysis of performance in sport through the collection and subsequent processing of data has been widely used to provide useful information for coaches . This performance analysis has sought to obtain indicators of execution, such as offensive technical actions or successful defensive events, and has ranked them using statistical procedures to characterize football performance [2–5]. Although these previous studies have led to advances in football performance, the notation of discrete actions and/or events has not provided information about the certain performance contexts . For instance, ball location relative to the scoring targets constrains the emergence of spatiotemporal coordinated team behaviours . Various studies have employed other measures such as players’ trajectories, interpersonal distances, relative angles between players or velocities as state variables [8–11]. They have served to define the coordination system states (patterns) at different levels of analysis (i.e., player, team or game). Hence, the challenge in performance analysis is to capture key contextual information that helps to describe and model the varied game scenarios explored by the team and players during both training and competition [12,13].
This study aimed to present the influence of spatial restrictions on players’ movements on the exploratory dynamics of tactical behaviour, ball flow dynamics and performance contexts when players were in possession of the ball. The main finding suggests that restricting players’ movements out of their home sub-area enhanced their exploration in a short timescale, but not their long-term exploratory behaviour. Furthermore, results did not reveal that spatial restrictions help players to possess the ball in advantageous conditions, i.e. far from the opponent in advanced zones. However, allowing players to move out of their home sub-area fostered the ball flow dynamics and the inter-player passing relations.
Performance analysis can be improved by providing information about the dynamics of football games. The combination of ball events and positional data can help coaches to understand the effect of a task constraint on individual and collective behaviour. Moreover, the combination of relational variables can give a clear picture about the tactical performance of players and its dynamics. Network analysis is widely used during football matches, but its use in training settings is still limited. Here, it has been reported new knowledge on how ball flow dynamics could be constrained through pitch positioning restrictions. Spatial restrictions did not stimulate the long-term players’ exploratory breadth, but increased the rate of exploration to perform different tactical solutions on a shorter timescale. Dynamic overlap might be considered as a potential order parameter for performance analysis in sport. The depiction of relational variables also provide relevant information for tactical performance in a macroscopic level. Overall, coaches can foster players’ exploration and/or to stabilize concrete coordinative states by constraining the players’ space of interaction behaviour. In this sense, depending on the training goals of the coaches i.e. enhancing the exploration of task solutions or focusing on smaller set of task solutions, they can manipulate the constraints to attain each of them.