Breaking the rules to build better learning

How do you build on the players’ understanding of what they already know without confusing them and losing what they’ve previously learned?

By Sarah Cottingham, Educational Neuroscience MA with Dan Cottrell.

England’s Ben Youngs during training, Andrew Matthews/Pool via REUTERS

Let’s say you start the season with a plan to improve passing from the base of a ruck. You might have players who’ve been doing that for a long time, but they have a weak hand, or you want to develop other players who can do this skill.

Now, you want to develop that skill. How does learning science explain how to make those developments work effectively, and not confuse the players in the meantime?

Confuse the players? You don’t want to blur their memories of the original skill, so, when that’s the best outcome, they’ve forgotten how to use it.

Our question is, therefore: What are the best conditions for helping players build on what they know?

We will look at updating “memory” through a process of reconsolidation and what this might mean for coaching your players.


If our memories are going to serve us well (that is, make accurate predictions), they need to stay relevant. This means they need to update when the environment suggests updating is necessary.

What are we talking about when we say updating memory? Updating is just another way of saying we are changing what we know, i.e., learning.

For our purposes here: memory updating = learning.

For example, with the pass from the base of the ruck, you have told the players to “sweep” the ball from the ground and towards the intended target.

That’s a simple enough start. But we’ve been passing to a static target, probably around about the same distance away every time.

Now the player needs to pass to a moving target and perhaps the ball is blocked by a foot or leg too.

The initial “memory” is the first set-up. The memory now needs updating because the environment has changed. How do we do this effectively?


Let’s start with the initial memory – the sweeping motion. New information begins life as a fragile memory trace. Over time and with repeated access and usage, it becomes less fragile and more stable.

The process of stabilising memory is called consolidation. Consolidation is storing memory in a more permanent form.

Let’s say you introduce it in week one of training. In the next three training sessions, the players practise the sweeping motion of the pass from the ground. You consolidate the memory.

After week three, you (and the players you hope) know they need to have more than one way to pass from the base of a ruck. How do we go about updating what we know?

When you reactivate a memory, you may make it fragile again. This means (under some circumstances) you can alter it: strengthen, weaken, build upon, distort.

In other words, you can change your knowledge and understanding. And with time, the memory restabilises again. This is why the process of updating memory is called re-consolidation.

I like a plasticine analogy here:

Imagine a piece of plasticine is a schema (network) of consolidated memories. When you roll the plasticine in your palms it becomes soft and malleable, capable of merging with another piece of plasticine. This is like memory when it’s reactivated. Reactivation may destabilise memory again, making it malleable and prone to changes such as adding related knowledge. When reactivation ceases, the memory restabilises – just like when plasticine is left alone, it hardens again.

In our passing scenario, the original set-up was the network of consolidated memories where the players knew the feel of the skill: the sweeping motion of the pass. By challenging the players to think about the different ways the ball might be presented or that the receiver might be moving, we are reactivating and therefore destabilising the memory. They might adjust their set-up and find a new solution.

If they now remember their original solution and their new solution, then we have a good outcome. It’s even stronger if they know how they found the solution.


However, reactivating memories may not be the only necessary condition for reconsolidation. Reconsolidation is pretty labour intensive for the brain. If all it took to update memory was reactivation, our brains would update all the time. Logically, our brain only updates memories if the information it is experiencing in the environment is different from what is expected.

Another way of saying this is that our current schema (networks of consolidated knowledge) lead us to make predictions about what will happen. If incoming information mismatches a prediction, we experience a prediction error. A prediction error tells our brain it needs to update.

In training scenarios, we need to create variability in our exercises. For example, with the pass away from a ruck, where will the first receiver take the pass? Will they be moving, standing still, how far away, what angle?

Initially, we have built a sweeping-motion pass from the ground, with the passer understanding the best foot positions, ways to hold the ball and the motion of the hands, arms and body. Once that memory has been consolidated, then we can mismatch a prediction. The receiver isn’t in the same place and the passer must adjust.

As a general rule then, reactivation + mismatch may lead to reconsolidation and updating.

You may wonder if there is a point at which memories cease to update. In other words, does reconsolidation ever end? The answer seems to be no. So long as the memories continue to be activated and the environment is not exactly the same as predicted, updating could happen.

This is good news for coaches: the opportunity is always there to build upon what players know.


(1) Reactivate and create mismatch

The concept of reconsolidation appears to highlight two necessary criteria: reactivation and mismatch.

First, reactivation. In simple terms, this can be done through building new skills onto previous skills, and understanding why each fits together.

For example, “We have our basic passing set-up. Let’s adjust it for a moving receiver. Why do we need to do that?”

Secondly, mismatch.

The most obvious usage of this on the pitch is corrective feedback. Players need to see the gap between what they thought was right and what the answer actually is. This mismatch supports updating.

This mismatch is especially pronounced for incorrect answers a player had a lot of confidence in. Why? Because high confidence = greater prediction error when they find out the right answer. Greater prediction error = better updating.

In a sense, we might be setting up a training scenario in which the player might fail when they thought they might achieve success. We need to manage that carefully, but if you have an environment where failure is seen as part of a positive learning process, then this should work well.

We may also be able to alert players to incoming mismatches when we introduce new material too:

Show players that what they are about to learn is related to, but different to (i.e. extends upon) what they already know. “We normally set up like this for the pass, but we are going to change our set-up to match the conditions.”

Van Kesteren and Meeter call this operating at an ‘optimal “distance” from the schema’. They relate this to the concept of desirable difficulties:

Desirable difficulties describe learning conditions that make it harder for the learner to perform well initially, yet increase learning over the longer term (Bjork, 1994). Retrieval practice is a good example. Initially, it is harder for pupils to retrieve information than to re-read it. Retrieval often leads to decreased performance in the short term. But over the long term, retention is better for retrieved than restudied material (Roediger & Karpicke, 2006).

What’s the desirable difficulty in creating mismatch?

In pointing out the difference between what players know and the new material is introduced, players need to think to connect their prior knowledge to the new material.

How might this be achieved?


“We’ve worked on passing to a static receiver. How about if the receiver is moving? How do we change the pass? Let’s look at what we can do instead.”

Supporting players to understand how what they are being taught relates to what they already know and yet differs, could help create a sense of mismatch.

Furthermore, we tend to be more curious to learn new information if we think we already know something about it. And curiosity may provide a boost to learning.

Therefore, communicating to players –

a) you know something about the new information I’m about to coach you, but

b) you don’t know everything about it…

could be a good recipe for piquing curiosity.


You can only reliably build upon study foundations. This means building on consolidated knowledge because this knowledge is stable.

What does this mean for coaches?

We can’t introduce too much in one lesson.

For example, if we introduce new information AND try to get the players to understand a conflicting idea in the same session, this may result in confusion. The new knowledge may need to consolidate first before we introduce conflicting ideas.

For example, teaching players a new skill AND how to do it differently in another situation might be too confusing. They need time to consolidate the skill first before learning how to use it flexibly.

The sweeping pass works if there are no obstacles in the way of the pass. The power and pace of the pass needs to be adjusted if the receiver is moving. You can’t do both of those in the same session. You need to focus on one, consolidate it, and then the other.


This reminds us of the stages and characteristics of expertise: Novice learners need rules.

Novices don’t have the cognitive capacity or knowledge to flex these rules. Think of the new coach who sticks rigidly to their session plan when they ought to change course in response to players’ answers.

Only once you know the rules can
you learn how to break them.

Sequence learning to build on consolidated knowledge. Consolidation takes time and usually rest/sleep. Too much in one lesson is likely to lead to confusion.


Memory: make accurate predictions. “I pass with a sweeping motion”.

Memory updating: changing what we know, otherwise known as learning. “I need to pass differently if the path from the ground is blocked.”

Consolidation: storing a memory in a more permanent form. “Through practice, I feel comfortable passing with a sweeping motion”.

Reconsolidation: the process of updating the memory. Destabilising a previous memory, often to alter or strengthen it, before it undergoes consolidation again. “I can pass from the ground, with both a sweeping motion, and, additionally, by picking the ball up and moving towards the target with my body”.

Mismatch: using variability in training to mess with the predictions. This can only be done once the original memory (prediction) is consolidated.


Bjork, R.A. (1994). Memory and metamemory considerations in the training of humanbeings. In J. Metcalfe and A. Shimamura (Eds.), Metacognition: Knowing about knowing (pp.185-205). Cambridge, MA: MIT Press. Cited in Bjork, R. A., & Bjork, E. L. (2020). Desirable difficulties in theory and practice. Journal of Applied research in Memory and Cognition, 9 (4), 475-479.

van Kesteren, M. T. R., & Meeter, M. (2020). How to optimize knowledge construction in the brain. npj Science of Learning, 5(1), 1-7.

Thanks to Greg Mannion, Academy Coach with Leicester Tigers for reviewing this article.

Find out more about Sarah’s learning adventures and for further references to this work in particular, visit

Share this
Follow us