Modeling of interactions and processes in karate matches using a Bayesian network

The extracted processes concerning the player A1 indicate that the player succeeds in a punching offense at the timing of stopping opponent`s forward step by the player`s rapid footwork or the former punch. We can say that the player A1 is good at taking advantage of such the offensive chance. According to the other extracted process, it can be said that the player A2 excels in punching offence at the same time of blocking opponent`s kick. In Japanese martial arts, there is an old teaching that "The timing of a successful attack is immediately after the beginning or end of the opponent's movement". The results of the experiment lend support to this teaching. Next, the result on the case that the player B succeeds in any offensive actions shows that the elite players seldom repeat the same failure because the probability of the same failure is low. The proposed method has worked correctly as a tactical actions extractor. We can safely say that the function of the method is a data mining from the data base of action sequences.
© Copyright 2008 World Congress of Performance Analysis of Sport VIII. Julkaistu Tekijä Otto-von-Guericke-Universität, Department of Sports Science. Kaikki oikeudet pidätetään.

Aiheet: kamppailu-urheilu karate mallintaminen taktiikka liike toiminta tekniikka analyysi ohjelmisto teoria
Aihealueet: tekniset ja luonnontieteet valmennusoppi kamppailu-urheilu
Julkaisussa: World Congress of Performance Analysis of Sport VIII
Toimittajat: P. O'Donoghue, A. Hökelmann
Julkaistu: Magdeburg Otto-von-Guericke-Universität, Department of Sports Science 2008
Sivuja: 600-608
Julkaisutyypit: kongressin muistiinpanot
Kieli: englanti (kieli)
Taso: kehittynyt