Improvement of Early Recognition of Attacks in Karate Kumite Due to Training in Virtual Reality

In a Virtual Reality training, young karate athletes divided in two groups (intervention vs. control group) responded to attacks of a virtual opponent. For the analysis, the first reaction of the responding karate athletes was detected. From that point three reaction times were subtracted to analyze the attack of the virtual opponent at the time of the recognition of the real athlete. The attacks were divided into four movement stages. Analysis of Variance (ANOVAs) with repeated measures and estimation of effect sizes as well as Bonferroni post-hoc tests were applied to calculate interactions between time (PRE to POST), group (intervention vs. control) and reaction time (150 ms vs. 255 ms vs. 370 ms). We found significant effects for time and time x group interactions for the attacks Gyaku-Zuki and Kizami-Zuki as well as an effect for time x reaction time in Gyaku-Zuki (all p 0.05). Paired t-tests showed significant improvements in attack recognition from PRE to POST for the intervention group, but not for the control group. At the pretest all athletes responded to late movement stages (extension of the pushing arm) while the intervention group responded to early movement stages (preparing steps and reduction of distance before the attack) at the posttest due to the Virtual Reality training. Early steps for the preparation of the attack and the reduction of distance seem to be important signals for attack recognition.