Variability in Practice

If variety is the spice of life, variability is the spice of sports performance and development. I have previously discussed how variability in snooker effects shot selection. I have discussed variability in practice design and in relation to Physical Education. Movement variability underpins an athlete's adaptability by coordinating many degrees of freedom that are available, leading to efficient movement solutions. I have written previosuly about a coaches best ability is adaptability, and I believe the same is true of a player. On the field of play, we want players to display dexterity, which is "finding a motor solution for any situation and in any condition... "The demand for dexerity is not in the movement itself, but in the surrounding conditions" (Bernstein, 1996, p21). An important aspect of developing movement dexterity is movement variability in training. If dexterity, adaptability and creativity are skills that are highly valued in performance, variability is a key ingredient in preparation.

When considering the constraints acting on a player during a chaotic game (essentially any team sport), the only constant is change. The problems that players face in the performance arena are always evolving. For example, in Gaelic football there is an attacker bearing down on goal against one defender, and the attacker has a teammate in support (2 on 1). However, if the ball carrier delays the pass, this will give the opposition a chance to get back, and soon the 2 v 1 problem might become a 2 v 2. The problems that players face never stay the same, they are always evolving, they are always alive. To make a player's competition performance robust in chaotic environments, they must train in chaotic environments.

Variability in training was one of the most challenging concepts to understand when I first embarked on my journey in skill acquisition. My thought process around coaching was repetition-based and action-focussed - I believed that in order for a player to get good at an action, they must repeat it. However, my thought process has very much changed - I now believe if a player wants to develop a skill, they must learn to adapt it. We can force players to adapt by infusing variability in training.

Variability in training - a contradictory concept?

If we take an example of a sports skill - free-taking/place-kicking in any sport (Gaelic football, hurling, soccer, AFL, rugby etc.). If a player practices from the same spot multiple times, repeatedly over multiple sessions, then they would get better at shooting from that particular spot, but it may not allow for the same improvement from other positions. The contrast is, if the player takes their free from random positions, it is very possible that they do not experience the same rate of improvement in their shooting. This creates a paradox - one strategy leads to rapid improvements, while the other strategy may not. If the goal of training is to improve, then why would a coach not choose the less variable option which leads to greater advances in performance?

At this point it is important to distinguish between performance and learning. Good performance does not necessarily mean players are learning. And when players are learning, they may not perform well. Performance and learning are two different things (although interrelated). Training with excellent performance is a great way for a player to build their confidence. In such a case, a coach could design a task for the player that is reasonably simple and allows the player to excel. However, if the coach is looking to allow the player to learn, then they must increase the difficulty of the task, which in turn will increase the number of mistakes/errors.

This is such an important point, and the challenge I had in grasping the concept around variability. Errors, while frustrating, are vital sources of information for players to develop, and vital sources of information for coaches to understand (for further explanation around the importance of errors, see here).

When learning to do anything, increasing variability may frustrate early training, but would pay off in increased transfer of what is learned.

Greater variability may initially hinder performance, but typically leads to an improved ability to learning learning to new contexts.
To give a concrete example from above, variability in shooting positions may hinder performance initially, but it will transfer to greater shooting adaptability from different positions.

A common argument against this variability just mentioned is that there is always varibility in human movement, so therefore any task has variability. There is no such thing as rote repetition, it is always repetition without repetition. For coaches, there are two key questions:

What is the goal of the task - Performance or Learning?
2) What is the challenge of the task relative to the learner. This relates to the challenge point framework. It follows the Goldilocks principle of practice design:

In the shooting example, the challenge of shooting from the same spot may be suitable for some players, and they will learn from that. However, for other players, the challenge of the task may need to be increased by incorporating variability to ensure learning takes place. The whole variability concept can be explained with this video snippet of Damien Farrow:

Integrating variability into training.

There are numerous ways that a coach can incoporate variabilty into training, aside from the aforementioned shooting example. Utilising a games-based approach, coaches will force players to continuously make decisions and adapt their skills:

By mixing players across teams in training, alternating positions, or by creating mismatches, coaches can increase the variation in training experiences for players.
Within representative tasks, getting players to change their starting position (kneeling, lying down, defender in front, defender behind, recovering defender etc.), will also vary the circumstances that players execute skills in.

Contextual Interference (CI) demands increased mental effort, which results in greater long-term retention of skills. CI generally comes in 2 forms:

Repetitive practice - also known as blocked practice
Interleaved practice - also known as random practice

High CI, even though causing immediate limited performance, leads to superior performance on retention and transfer tests. But it is important to consider the two questions above, particularly question one,

If the goal of a session is motor performance - perhaps the team have a game in 2 days and the coach wants to ensure players are confident going into it - the coach may implement more blocked practice to ensure a greater success rate as opposed to random practice. Players are unlikely to learn a huge amount in one training session prior to an important game (i.e. You can't fatten a pig the night before the mart), so it important that they are confident in the skills and abilities that they already possess.
If the goal of a session is motor learning, the coach should implement practice with greater contextual interference. This may mean that motor performance in the session is compromised, but it will ensure a greater long-term return on investment of practice time.

Based on the CLA, practitioners can easily manipulate the constraints acting on an athlete to allow for new behaviours to emerge. To encourage this adaptation, coaches can incorporate differential learning into training. Changing the task or environment (even beyond what a competitive environment consists of) will influence how the individual interacts with the task (i.e. forcing them to adapt) and a new behaviour will emerge to achieve the task outcome. Two simple examples are:

Changing the surface that athletes sprint on while running.
Change the ball shape and size during a SSG.

The general goal remains the same, but as the environment in which the athletes perform the task changes, the athletes must adapt their behaviour to complete their given task - i.e. complete the sprint, complete a pass etc.

This can be particularly useful for high-level athletes and forcing them to adapt. To challenge high-level athletes, it is necessary to continue to change the constraints acting on them during a task. There are only so many changes a coach can make to a task before it loses its representativeness i.e. the task constraints of training do not simulate the task constraints of competition. Alternatively, coaches can look at changing the equipment or environment. Even a basic task, that typically might have a low challenge point for a high level athlete, can be made more difficult by changing the environment in which the task is completed. For less skilled athletes, differential learning is possibly a step too far. As the saying goes: you should exhaust all options inside the box, before looking outside.

Key Points:

Learning does not happen in a single session. Repetitive/blocked practice can lead to rapid improvements in performance, but subsequent forgetting of these improvements can also occur. Interleaved/random practice may lead to a slower rate of acquisition but learners are more likely to maintain or develop these skills long-term.
Interleaved practice is more likely to transfers to other contexts. Learners who have experienced random practice can use the skills they have acquired more flexibly.
Affording learners time to consolidate their learning can enhance their long-term retention and performance. Learning can be consolidated through sleep, or exercise, can impact how readily available a skill is in future.
Random practice contains greater attentional demands than repetitive practice. This challenge is key to sustained learning and long-term performance (see video above)
This highlights the value of a games based approach. In a game (SSG or regular - that does not promote restrictions on the skills that can be used), players will experience random practice conditions and exercise will also be incorporated into the task. These 2 things combined highlight the potential benefits of using a games-based approach for greater learning.

Variability in Practice

Recent Posts