練習設計における制約操作、代表的学習、機能的変動を情報化するためのスポーツ分析に関する展望

As the prevalence of technology and data use increases in sports, new opportunities exist to support practitioners by informing practice design. In turn, analytical techniques which leverage this data can be used to further bring life to frameworks of skill acquisition such as the constraints-led approach. Machine learning in particular presents as a viable method to reveal detailed insights, as it can consider multivariate and non-linear relationships. In the field of skill acquisition, a variety of different algorithms are well suited to help operationalise principles of constraint manipulations, representative learning design and functional variability. Specifically, decision trees or random forests may benefit coaches by predicting how constraints can be manipulated to facilitate player behaviour. Rule association can identify interacting constraints within the competition environment which can then be replicated in practice. Clustering techniques may be beneficial to assign similar player movements, or activity repetitions, into groups, allowing coaches to manipulate training variability by prescribing movement types from various groups. These techniques are proposed as methods to support coaches and applied sport scientists' use of technology and data, as well as enhance their decision-making regarding practice design.