Designing observational systems to support top-level teams in game sports
Abstract:Any observational system can be considered as the result of a model building process. In order to validate this process one has to ask whether the model contains the essential attributes of the original (here: the game sport observed). Or does the model contain abundant features which do not correspond to properties of the original? Mathematically spoken, one has to examine the validity of the bi-directional relation between model and original.
Only recently, game sports research has become aware, that another aspect of the model building process had perhaps not been paid enough attention to: the purpose of the model. In short this means that there is not the model of tennis or soccer, but for different purposes different models serve best. There is a wide variety of purposes for observational systems in game sports: Measuring individual performance, physical loads, tactical solutions for special situations, studies of the interactions between and within the teams, discover talents, decision making of coaches or theoretical interest in the structure of games.
In order to support top-level teams two purposes are dominant: preparation against an opponent and gaining information to support onès own training process. An analysis of these tasks reveals that in order to become practically relevant a multi-step process has to be passed, which relies to a great extent on the quality of inferences drawn. This basically interpretative processing of observational data in order to achieve the goals in top-sports has led to the development of "Qualitative Game Analysis", which applies the principles of qualitative research methodology to game observation.
Deriving conclusions from behaviour in games documented on videos is seen much in the same way as deriving the central statements of an author from a large number of pages in several books. Today's technology allows for the first time the processing of pictures almost in the same way as text sources.
A hard- and software solution for "Qualitative Game Analysis" was developed. The preparation of top-level German beach-volleyball teams was accompanied for three years. A large amount of experience was gained in the way how to gain information for onès own training process and in preparation for matches to come. The bronze medal surprisingly won by a German male team at Sydney has seen an intensive support with "Qualitative Game Analysis" before and especially during the Olympics which of course does not allow any inferences on the scientific quality of the method!