On his very good blog, Harri Mannonen wrote about the systemization of skills. His major point was that a coach should “base everything you teach on current games” rather than on the coach’s execution of the skill when he or she played. Mark Upton, on his very good blog, also wrote about practice resembling the game.
Despite their major themes being very similar, the way that I understood their instructions or examples differed. Mannonen described the way that skills often are learned through a slow and inefficient process, and provided the antidote:
Base everything you teach on current games. Systematize those skills – if you teach a move, also teach counter moves because the opponent will adjust to what you’re doing. Be ready to change your systematization when something new comes up.
The blog did not describe the entire process of the systemization. By inferring from what was written, and what I see elsewhere on the Internet, I imagine the coach watching the games, breaking down specific skills, and creating a specific progression to teach a specific skill (something like the video below).
This is what I tried to do when I trained a lot of players. I never tried to train a player to play like me or to make a move like me. I use to teach moves called the AI, the Steve Nash Hook, the Will Bynum, etc. These were moves that I saw in games made by NBA players. I watched the moves, broke down the elements, and attempted to teach the moves through this systemization of the elements. Then, I added the counter to the move.
For instance, the Steve Nash Hook now is known more popularly as the Rondo (above). The counter would be to use a second fake and step-through for a lay-up (below).
Upton’s article focused on representative task design. He used an article from Dan Coyle’s blog that focused on repetitions, in this case prior to the Navy SEAL raid on Osama. Upton highlighted the descriptions of the practice: full-scale mock-ups, full-scale replica, every possible scenario, and brains for the job. It was not just repetitions, but specific repetitions based as close as possible to the eventual task.
Upton cites Magill’s textbook Motor Learning and Control (1999) which says that “transfer of practice to match conditions depends on the extent to which the practice matches the game.” When we practice, our goal is transfer – coaches want players to practice a drill and use the skill in the same way in a game. If I teach a player the Rondo, when the player is in the game, I want him to use the move to create an open shot for him or herself. Otherwise, we are practicing for the sake of practice (there are reasons for practicing beyond transfer; I believe there are some drills, some occasions, and some reasons to practice for purposes of developing confidence, coordination, and other qualities).
Upton used this example to introduce representative task design. The idea behind representative task design is that practice represents the whole task. In the Navy SEAL example, they did not practice on buildings; they practiced on an exact replica of Osama’s compound. They did not practice one scenario; they imagined every potential scenario and practiced these scenarios.
In terms of developing a skill like the LeBron James set-back, Mannonen’s article depicted a systemization or a step-by-step process for teaching a specific skill in a specific way. In this philosophy (Not Mannonen’s per se but the generalized approach), there would be one correct way to execute one specific skill.
In the Upton example, the coach would not teach one specific skill in one specific way. If the goal was to teach a move related to the LeBron James step-back, the representative task design would be to design an environment where players would use similar skills. For instance, a coach could set up a 1v1 game where players had to shoot outside the key and could use only two dribbles. By placing these constraints on the performance, the coach creates an environment for the player to learn a specific skill. However, that specific skill is not one specific move – the skill is how to create a shot in 2 dribbles.
Now, one could argue that this is not representative of the game environment, as there is no rule to prohibit shots in the key or limit the number of dribbles. This is correct. However, is this not a skill or scenario that occurs frequently in games? The other four defenders pinch in, effectively taking away the key, while also limiting the number of dribbles available, leaving a small area for a player to create his or her shot against a defender? I may not spend a lot of time on this skill or task with youth players, but for varsity and college players especially, the ability to create a shot quickly within a small area of space is an important offensive skill.
In this environment, the player can explore different solutions to the task rather than trying to repeat a specific set of steps, as with the systemization approach. Through this trial-and-error and exploration approach, the player can learn which moves work for him or her. The player may not master the LeBron James step-back jump shot, but may develop his or her own move that works better for him or her. Also, because the 1v1 game couples perception (reading the defense, deciding which move to make, etc) with action (making the move, shooting), the transfer of the skill to the game should be improved in comparison to the systemization approach. In the systemization approach, the player would have to master the steps, and then play in a 1v1 environment where he or she could read the defense and make the appropriate move or counter.
In a decision-training or play practice philosophy of coaching, the coach would start with the 1v1 (or similar game-like activity). If the player struggled, then the coach would return to a more systemized approach of teaching a specific move for a specific situation. For instance, if I start a practice with a transition drill, and suddenly we cannot make a lay-up, I may take a step backward and practice uncontested or semi-contested lay-ups before returning to another scrimmage activity. However, the remedial practice is based specifically on earlier practice in a game form rather than a belief that we have to do more training activities or more systemized practice.
These are two ways of skill development. Trainers tend to use the systemization approach, and the best trainers separate themselves with the relevance of their moves and the sophistication of their system of their teaching. The representative task design is used more by coaches. The best approach likely combines the two systems together. The question for coaches and trainers is to figure out the magic ratio of systemized skill practice versus representative task design or constraints-based practice.