A critical look at a constraints-led approach to training

Anyone reading HMMR media this month will be well informed on the work of Frans Bosch. Bosch is perhaps the world’s leading proponent of constraints-led learning and applying complex systems theory in the realm of athletic development, as he detailed in his book. As a strength and conditioning coach from a more classic strength training background, I have always found his work interesting and have enjoyed reading his ideas and listening to him talk. But before I jump on the bandwagon, I still see the need to take a critical look at a few some of the concepts, and, more importantly, how I see them being implemented in training.

The theory of constraints-led learning

Before I challenge some of the applications of constraints-led learning and Bosch’s training philosophy in the strength and conditioning world, I think it’s worth noting that I see the huge merit of this approach to motor learning. Summarizing this approach in a sentence doesn’t do it justice, but it revolves around finding the attractors of a skill, identifying the fluctuations around a skill and then developing the attractors by challenging the ability to maintain them against variations in the environment, the task, or the organism. By adding variation the attractors become more stable and the ability to execute the technique is improved. This is nothing new: self-organization theory has been well documented for more than half a century, and complex systems theory has been documented for decades.

» Learn more: This month we are taking an in-depth look at the training methods of Frans Bosch and John Pryor, including an interview with both on the GAINcast, our video webinar on hip lock with Bosch and a robust running webinar with Pryor.

In skill development, having the body to learn a skill through trial and error allows for the body to develop the key pillars to technique. This, in essence, is self-organization: the body will learn how to perform a movement more effectively through its own errors and motor control develops in response to the feedback from these errors. The motor cortex controls motor control, and it essentially learns from this feedback and develops better pathways. Over time developing pillars/attractors of technique through the systematic application of variability has been shown to be a very effective way to learn motor skills.

Where is the context?

The theory has merit. What concerns me is that, in my opinion, many strength coaches misinterpret, misunderstand or incorrectly apply this approach without scrutiny. The problems I most often see relate to missing the context in which these exercises are used.

The value of variability has been taken to the extreme by some coaches and expanded to the point that the whole of the time spent on physical development is spent on adding variability to techniques at the expense of more traditional strength and conditioning. I see little point in adding variety for variety’s sake. Coaches must first understand why they are using variety and if it is required. Even if it is required, they then need to understand the movement, the attractors, and the role variety will play in helping the athlete. Too often, these steps are skipped. Variety without context just means randomness.

Related, when coaches watch others use Bosch-inspired exercises, they need to understand the context other athletes are using the exercises in. This criticism could be applied to any number of training methods, but too many coaches see a cool exercise on Instagram and try to apply it to their athletes without understanding why it was being used in the first place. For instance, most commonly seen is the development of “hip lock”. If you don’t understand hip lock, you will probably execute the hip lock movements you see with the wrong intent and miss any potential benefit from them. It may look exciting on Instagram, but it won’t help your players on the field.

» Related content: Frans Bosch explains hip lock and how to train it with the wrong intent in our latest video lesson.

Where is the evidence?

The  more I read about constraints-led learning theories, the more I find myself constantly wanting to see more evidence of their effectiveness. If it is so effective, why isn’t there any evidence to show the effects? Take sprinting as an example: much has been written about differential learning and de-stabilizing in order to stabilize attractors and improving sprint speed as a result. I see the logic in the theory, particularly for games players who operate in a variable environment, but where is the evidence that it works?

As a coach who runs the strength and conditioning program for a pro rugby team, I appreciate that it often isn’t feasible to run a randomly controlled trial on your own athletes. It can also be difficult to measure some aspects of the game, or even find enough elite athletes to run a study. But it’s certainly possible to measure some change, isn’t it? In the past decade I haven’t seen a single set of objective data of the results of this approach, not even a simple case study.

While evidence is not always there to support all the traditional methods–just look at how periodization has struggled to show its effectiveness–there is some evidence to back it up. Classic strength training theorists will point to classic physics:

  • Force = mass * acceleration
  • Power = force * velocity

Focussing on developing these variables in the neuromuscular system has consistently been shown to improve performance. The criticism from motor learning theorists will be that you’re improving the ability to perform a test and there isn’t a transfer into competition. Again, the theory has some merit, particularly for games players, but it still remains that, a theory.

Consider organizing the training of a 100-meter runner: classic physics would say it’s a combination of the magnitude of force you produce relative to body mass and the rate at which you can produce it that limits performance. Increasing force relative to mass and the rate at which you produce it will allow you to accelerate better, improving reactive strength will allow you to achieve a greater force output with a minimal contact time to increase max velocity.

There’s plenty of research that shows the importance of applying this force in the right vectors, therefore optimizing your technical model is important. Resisted sprint training is one method which has proven to be very effective at doing this. You could argue that this is a constraint in itself, although I’d argue that it’s the neuromuscular adaptation that’s behind the effect, more than any motor learning.

When training this theoretical 100-meter runner, do I think developing maximal force and power through resistance training or overloaded sprinting will work? Clearly it will. Do I think focussing on variability in order to stabilize the attractors of the movement will work? Maybe, but maybe not. I find it hard to justify dismissing the classic strength training methods in favor of a method which is yet to fully prove its effectiveness. However could there be a role for both?

Why not use both approaches?

Developing force, power, velocity has been shown to improve the performance of a whole range of athletics tasks. Traditional methods should not be so quickly overlooked as they have helped produce some great results. Even on last week’s GAINcast, Frans Bosch and John Pryor stated that traditional strength training still has a role to play. For me, therefore, it is hard to move away from this approach which has stood the test of time and has such a huge evidence base behind it.

However, if you have athletes who produce high levels of force, can express force correctly and are robust and move well, then do we reach a point of diminishing returns? Is this a time when we look at add degrees of freedom to a movement? To give the athlete more options and increase their ability to execute technique in the highly variable, unplanned environment that is competition? Can we increase the transfer of general training effects to the specificity of the sporting field? Bosch’s idea around attractors and the addition of variation to movement would make a lot of sense here. However I would question how many athletes who focus solely on constraints-led learning in their strength and conditioning reach this point?

Too often in debating topics like this it comes down to an either-or approach. Everyone loves to debate hamstring training, with the most recent debate being over whether we should do eccentrics or isometrics. Can’t we do both? The same goes here. A constraints-led approach is one tool for a coach, but no matter the situation you are working in, you need more than one tool. Putting together training is not an either-or proposition; coaches might need to use elements of traditional and constraints-led training methods.

Through my 11 years as a professional coach, I have witnessed huge variation in my athletes, how they run, how they move, how they jump and it makes sense that we all produce and express force differently. The relationship between the weight room to sprinting isn’t a direct correlation. Furthermore, add in some unplanned movement and this correlation weakens further. Clearly every athlete differs in their ability to transfer their neuromuscular abilities developed in the gym to the field. Adding some variability to challenge a technical model is one way that could increase this transfer, however it isn’t the only option.

Putting it in context

To finish, I am not questioning the role of a constraints-led approach to the development of motor control or skill development in sport. My question is on its application to the world of athletic development and S&C. For me, constraints-led learning is misunderstood and applied incorrectly in a lot of coaches I’ve seen. If it is effective in athletic development, then it is part of a bigger process. Develop force qualities, work on how force is applied in the context of your sport and then constraints-led learning can maybe help increase the transfer of force from a weight room onto the sports field/track/court. For me, it is a tool that can be used as part of a far bigger training process, it isn’t the process itself. However, show me some harder evidence on its effectiveness and I might change this opinion!