其他摘要：In the last few years team sports are being studied within a complex adaptive systems approach. From this perspective, the importance of game complexity, team self-organization, non-linear and multidimensional performance models has been growing from micro to macro issues. One interesting topic to investigate is team self-organization in a way to detect behavioural transitions that may define periods of stability and instability in game dynamics. This study has the aim of identifying the variability in confronting teams' efficacy in several different game conditions (outcome, location and final score differences). Play-by-play data was gathered for all 34 regular season games played by one Spanish first level basketball team (2004/2005). Data was introduced in BSK software in order to calculate partial in-game efficacy (points scored per 100 ball possessions) according to all studied contextual factors for confronting teams. Relative phase was calculated by Hilbert transform, as we have considered teams¿ efficacies as two time-series of two coupled oscillators and quantified their interaction by measuring the phase difference between the signals. Data reliability was high (above 0.90). The obtained results allowed for establishing different stability and instability periods across conditions. Although there was a clear tendency for in-phase behaviour, describing synchronicity between teams' efficacies, there was a large variation between conditions. Most variability in relative phase was identified in the first and in the fourth game periods. There was a contrast between home and away games and winning and losing games. Also, there was a clear interaction between game outcome and game final differences. Although preliminary, the results may suggest that basketball teams¿ self-organization influences game performance across the different game conditions.
