The specificity of motor learning tasks for skating development in older school-age children has not been sufficiently explored. The main objective was to compare the effects of training programs using change-of-direction (COD) speed exercises and partial skating task (SeqT) training on speed and agility performance in U12 ice hockey players. Thirteen young ice hockey males (13 ± 0.35 years, 41.92 ± 9.76 kg, 152.23 ± 9.41 cm) underwent three straight speed (4 and 30 m with and without a puck) and agility testing sessions before and after six weeks of COD training and then after a six-week intervention involving partial skating task (SeqT) training. The statistics were performed using magnitude-based decision (MBD) analysis to calculate the probability of the performance change achieved by the interventions. The MBD analysis showed that COD training had a large effect (11.7 ± 2.4% time decrease) on skating start improvement (straight sprint 4 m) and a small effect (-2.2 ± 2.4%) on improvement in agility with a puck. Partial skating task (SeqT) training had a large effect (5.4 ± 2.5%) on the improvement of the 30-m sprint with a puck and moderate effect on agility without a puck (1.9 ± 0.9%) and likely improved the 30-m sprint without a puck (2.6 ± 1.3%). COD training on the ice improves short starts and agility with a puck, while partial skating tasks (SeqT) target longer 30-m sprints and agility without a puck. Therefore, both types of training should be applied in accordance with motor learning tasks specific to current training needs.
