Judo coaches aim to develop left-handed techniques in right-handed judoka due to a potential frequency-dependent strategic advantage during competition. Thus, easily completed tasks are needed to track the progression of the lateral preference index (percentage as right-handed actions). Thirty naïve volunteers performed two tasks, an unopposed repetition-based dynamic task (yakusoku geiko) and an opposed-bouts task (randori), before and after eight-week training protocols differing in the executing side. Training protocols consisted of a control group (without any instruction regarding the practice side), a bilateral group (practicing with both dominant and nondominant sides), and a nondominant group (practicing exclusively on the nondominant side). ANOVA-type tests were used for analyzing the suitability of the unopposed repetition-based dynamic task for detecting changes in the lateral preference index and for analyzing the concordance between tasks. Additionally, concordance was tested with Spearman’s rank correlations and Bland-Altman plots. Lower lateral preference indices (i.e., reductions of executions as right-hander) were observed after training in the nondominant group and in comparison with the control group and bilateral group on the posttest (p < 0.05). ANOVA-test for concordance analysis revealed no differences between tasks (p > 0.05) with significant correlations (p < .05) and low bias during the pretest (ρ = 0.563; bias: 4.29, 95% LoA: -36.32-27.74%) and the posttest (ρ = 0.718, bias: 0.69, 95% LoA: -39.37-40.77%). In conclusion, an unopposed repetition-based dynamic task detected reductions in lateral preference index after an eight-week nondominant training period. These results agreed with those observed with the opposed-bouts task. This indicates the eligibility of yakusoku geiko for coaching delivery due to general suitability of the unopposed repetition-based dynamic task to track the maintenance or progression of left-handedness in judo trainees.