Scott, J. P. (1962). Critical periods in behavioral development. Science, 138, 949-958.

Certain periods in the life of young children are marked by times of particular sensitivity. For example, in McGraw's (1935) attempts to modify the behaviors of identical twins by teaching them a number of physical activities, some credence to the "appropriate times for learning" postulation was presented.

1. The onset of walking was not affected by preemptory practice or help. It is a phylogenetic behavior that is largely "programmed" into the natural development timing of the youngster. It cannot be "speeded-up."

2. Roller skating, an unnatural activity but closely allied to walking, developed almost in concert with walking itself.

3. A number of other activities were actually made worse by early practice because of bad skill habits developed or the negative occurrences associated with the learning experience.

Implication
Starting a sporting experience at a very young age is not necessarily advantageous. It would seem that if one was to design development in a sport, the following are appropriate:

Provide a wide variety of activities so that generalized basic gross skills are developed.

Pay little attention to skill intricacies, instead being satisfied with gross motor movement patterns.

Provide much activity that leads to successful outcomes.

Avoid at all costs, the implementation of adult rules and sport dynamics, instead providing activities appropriate for the social, intellectual, and development stages of the participants.

There are critical periods for learning that vary from sport to sport. For each kind of coordinated muscular activity there is an optimum for rapid and skillful learning.

[McGraw, M. B. (1935). Growth: A study of Johnny and Jimmy. New York, NY: Appleton-Century-Crofts.]