ABSTRACT

Behavioural studies in sports have mostly used analytical methods. Racket sports are no exception. Notably, a detailed scrutiny of squash players revealed a relationship between the types of shot and its outcome (win or loss) (McGarry et al., 1994, 1996). Results showed that some shot sequences were adopted most frequently and tended to lead to a winning point. These sequences defined a behavioral invariant repertoire for an individual player, which coaches should take into account to optimize the tactical preparation. Such invariance, however, was only relevant in reference to one single pair of players, a clear shortcoming as players are known to change their playing style as a function of their opponent. Moreover, the huge amount of data that has to be processed in studies using these kinds of procedures renders any medium or long term prediction quite unlikely. By and large, analytic methods appear then to be inept for addressing the complexity of dual sports.