The dynamical systems approach is increasingly adopted in performance analysis in sports. Mainly, the oscillatory behaviour of players in racket sports has been under investigation (Lames, 2006; McGarry et al., 1999; Palut & Zanone, 2005). However, team sports like rugby, basketball and soccer have not been subject of these analyses for various reasons. It is hypothesized that in soccer, attacking and defending is strongly coordinated between teams. Therefore, the aim of this study was to explore small-sided soccer games as a dynamical system by means of 2 collective variables, i.e. centroid position and surface area (Frencken & Lemmink, 2008). In addition, we tried to identify patterns in the development of goal-scoring opportunities through visual inspection. Methods Ten young male elite athletes (17.3 years ± 0.7) played three small-sided games (4-a-side plus goalkeepers) of 8 minutes on a 28x36 m pitch. Player positions were recorded at 45 Hz per player by means of innovative technology, i.e. local position measurement (LPM) system. In total, three measures were calculated per team for the centroid position (forward/backward displacement, lateral displacement and radial displacement) and surface area (length, width and surface area). Correlation coefficients (r) were calculated for these measures to analyze coherence of time series data of both teams. Results Correlation coefficients of centroid measures were higher for the centroid position (.57 < r < .95) compared to surface area (-.06 < r < .39). Highest correlation coefficients were found for forward/backward displacement of the centroid (r > .93). For 10 out of 19 goals, a crossing of the centroid positions in forward/backward direction was present prior to the goal being scored. Discussion In this study, we analyzed small-sided soccer games from a dynamical systems perspective. Results indicate inphase oscillations for the centroid position. The strongest inphase pattern is found for the forward-backward displacement of the centroid position. Teams coordinate attacking and defending primarily in this direction. No specific oscillatory pattern was found for all surface area related measures. Although rules of these small-sided games were similar to 11 vs. 11, players behaviour is constrained differently when compared to 11. vs. 11. This results in the absence of a clear pattern. Finally, we showed that deviations of the inphase pattern of centroid position indicate goal-scoring opportunities in small-sided games. We conclude that the small-sided soccer games can be considered a dynamical system and this approach opens up new types of analyses in performance analysis in team sports.
© Copyright 2009 14th annual Congress of the European College of Sport Science, Oslo/Norway, June 24-27, 2009, Book of Abstracts. Published by The Norwegian School of Sport Sciences. All rights reserved.