An infographic designed by @NickGearing1 (Twitter). See the paper by Ribeiro et al (2021) attached below.
This post has many layers to it, focusing on the skill end of the continuum above. While the heading of the post refers specifically to training design, the "repetition-representative continuum" can easily be reframed as the "information-processing (IP) - ecological dynamics (ED) spectrum", or possibly an element of the IP-ED debate. Rote repetition is born from the IP theory of skill acquistion, while representativeness / representative learning design (RLD) comes from the ED theory of skill acquisition (alternatively referred to as skill adaptation). When viewing spectums and continuums, the natural thing for everyone to do is relate it to other spectrums relevant to their life and identify comparisons. There have been three that immediately spring to mind:
Introvert-extrovert spectrum
Political spectrum: Far left to far right.
Force-velocity (F-V) curve
With the first two mentioned, the majority of people exist somewhere in the middle. With regards to personality types, some people are more introverted, some are more extroverted, but the context of each situation plays a major role (how a person behaves among close family and friends versus at a networking event or at a party). The majority of people would be considered ambiverts, which is a combination of introvert and extrovert, somewhere in the middle of the spectrum. The book Quiet by Susan Cain, discusses introversion and extroversion in huge depth, and is one of my favourite reads of all time. The political spectrum obviously has far bigger consequences, but it has similarities to the introvert-extrovert spectrum. Most people would consider themselves to be somewhere in the middle, especially as politics covers a huge range of topics. Few people would hold a far left view on every topic and few people would hold a far right view on every topic. Similar to personality types, the context of a discussion can greatly influence people's political views.
(*I have not and do not claim to be a personality or a political expert.)
The final example is the F-V curve. Given my background in athletic development, this would be a more familiar metaphor for me. The F-V refers to the inverse relationship between force and velocity and the impact this has on exercise selection. For example, a 1 repetition maximum for a squat requires maximum force, and therefore the velocity is slow. For a squat jump, the velocity is high and the force is lower. The force-velocity curves distinguishes 5 distinct training zones, ranging from maximum strength (high force, low velocity) to maximum velocity (low force, high velocity). Again, depending on the context, practitioners can target different zones. A common term regarding the force-velocity curve is "surf the curve", and that means to train all aspects of the curve at the appropriate times.
There has been much discussion online regarding how the two approaches, IP and ED, merge. Some have taken an all or nothing approach by suggesting that it doesn't make sense to mix and match, while others have been much more open to the contextual factors and how both approaches are a part of the overall solution. It must be added that they present their arguments with much more clarity than I do here. But my current understanding is that they exist on a spectrum, and I am going (to try) to explain my views through a training task design lens. With this post, I am going to dive into the repetition-representativeness continuum, and the periodisation of task difficulty.
Repetition-Representative continuum
The more representative a task is, the less repetitions of a target skill it will allow for (as the example above shows). As representativeness is reduced, a greater number of repetitions can be performed. Sometimes it is both useful and necessary to reduce complexity of tasks to allow for key components of the game to be developed.
Repetition
It is a common belief that skilled performance emerges from repetition after repetition. More than this, repetition of a "correct" technique leads to skilled performance. As Rob Gray states in his book "How we learn to move", repetition after repetition has often been spoken about as the path to elite status as a player. The easiest way to ensure repetition of a partiular skill in a game (passing, shooting, tackling etc.), is to remove representativeness. In a fully representative version of a game (i.e. a full game), players will have numerous options regarding the action they choose. When in a scoring position, they may opt to pass. When a pass is a suitable option, a player may opt to carry or dribble the ball. Representativeness and, in particular, contextual interference is not conducive to repetition of a particular skill.
Removing representativeness (in particular the perception-action coupling) would link to task decomposition as opposed to task simplification (which I discuss here), which is more indicative of an IP theory of skill acqusition, rather than an ED one. The aim of this information-processing, linear pedagogy is to teach technical proficiency, and a contemporary example of this contrast is the debate between fundamental movement skills and functional movement solutions (which I have discussed here). An alternative to this repetition after repetition approach is what Nikolai Bernstein termed "repetition without repetition." As Rob Gray states, we repeat a action outcome, but not by repeating the movement that produced it. A extract from Rafael Nadal's book:
No shot is identical. Repetition is key to performance not because of the repeatability, but because of the variability. Repetition without repetition does not encourage performers to repeat a movement solution, but it encourages performers to find a movement solution. A key challenge for coaches when designing practice tasks is to "reduce without impoverishing" (@CalJonesJudo). Can coaches create tasks to allow for higher repetitions, without reducing the task's link to the competitive environment.
Representativeness/Representative Learning Design (RLD)
What is RLD? This refers to the level at which the training environment "represents" the competition environment. Are players seeing, hearing and feeling the same things in training that they would do in a game? Ensuring a task is representative ensures that the task contains relevant information to allow for players to become more attuned to relevant information sources so they can regulate functional behaviour within that environment. A key focus with RLD is to maintain a strong perception-action coupling. Key information sources (i.e. specifying information) such as the positioning of opposition players, the positioning of teammates, position on a field of play or the speed at which the ball is travelling all will have an impact on a player's decision making. I have discussed RLD previously in other posts.
Examples along the spectrum
U7s Gaelic Football Session
This Gaelic football session was taken from twitter (@TCroninNap). As can be seen from the session plan, it is a Gaelic football wall ball session, with emphasis on ball contacts and enjoyment. This is a session that covers almost the full repetition-represntative spectrum. To start with, skill station 1 is a simple exercise where players are kicking the ball against the wall and catching. Low on representative information, but high amount of repetitions (without repetition - different angle, power, speed - minute differences in every kick). For the rest of the session, players are involved in games-based activity where they are working with or against other players, while developing movement and sport specific skills/solutions - the remaining stations contain representative information such as opponents, teammates and spacial constraints. This is a 50 minute session (not including transition time between stations), and for 80% of the session, players are invloved in games based activities. I think this is an outstanding session, and balanced right between repetition (without repetition), RLD, and activity levels for under 7s.
Periodisation of Representativeness
2 key questions emerge as I write this:
Should players only complete repetitive/representative tasks?
How much reduction of representativeness should occur?
The challenge for practitioners is to identify what an appropriate reduction on representativeness is for each given context. In the same way that strength coaches want players to surf their F-V curve, I think its important for players to train across the repetition-representative continuum, and coaches need to be agile and sensitive to what players need. Like the political spectrum, or the personality spectrum, the context will drive the content. Specifically, where does each individual training session (or station within a session) exist on the spectrum - what is the aim of this session/station/activity? Some key considerations include:
Timing of the session - How many days after the previous game has it been? How many days until the next game?
Tactical implications - What did the team struggle with previously? What specific preparation does the team need for the upcoming fixture?
Player availability - Is it a full squad session? Are there many injuries? Are there reduced player numbers due to selection on other teams (international window or intercounty season (for GAA)).
Loading implications - Intensity (depending on player numbers), physical demand (more neuromuscularly or metabolically demanding), cognitive demand (high or low task complexity). See image below for an example of how isolated sprint training (representativeness removed) can ensure players reach peak velocity in training. If the aim of isolated sprint training is to develop tissue tolerance to game demands, then it works. If the aim is to enhance the ability of players to extract relevant information from the performance environment, then it doesn't. Context drives content.
An image first highlighted by Des Ryan on twitter - a mixed methods approach seems best to prepare players for the physical demands of games.
This question of periodisation was something that I looked at for my MSc thesis. It was also a question that crossed my mind when Stephen Casey was presenting on the MSA webinar series a number of weeks back (full recording of the webinar can be found here). But from an athletic development point of view, it is important to have a range of tools to prepare athletes for their competition environment. As I have mentioned previously here, these tools include:
Identify and develop relevant athletic qualities (develop capacities).
Develop these athletic qualities in a contextual manner (develop skills).
Humility - to understand that further athletic development for a player may not be the best use of time.
Being an agile coach - to be able to design tasks that are suitably complex and at the appropriate challenge point for athletes.
Conclusion
While it can be argued that in order for training to transfer to competition, the perception-action link must be maintained. However, depending on the session aim, I think enhancing or reducing representativeness can be useful. A fully representative session will ensure that the physical and cognitive demands of the session are high, but this will result in limited repetitions of a certain skill and the output will be variable and hard to control (thinking specifically of a players returning from injury). If a player is aiming to develop a specific skill (or physical capacity), representativeness can be reduced to allow for a greater number of repetitions.
Further Reading
Presentation by Prof. Keith Davids, Sheffield Hallam University – How academic research has informed sports performance – an ecological dynamics perspective – case examples in individual sports