Learning v Performance & Challenging Traditional Coaching

(Originally posted on another blog on 1st of June, 2014)

The text below is a section from an email exchange I had with someone a while back. I thought it might be relevant (even thought-provoking?) for others to read. As always, feedback welcome!

The thing I would like to address from the Skill Acquisition field, is distinguishing between “performance” and “learning”.

It seems many coaches have a mindset that practice activities/drills must contain a high level of success and mistakes are “bad”. So to achieve this, it is only natural that activities and drills are used that almost ensure that players have a high level of success (and lots of “blocked” repetition too). In this case the players “performance” in a drill could be very good — and coaches are interpreting that as “learning” that will translate into match conditions. This is flawed and hence leads us to the notion of “pseudocoaching” (a term brought to prominence by @innovatefc).

Digging deeper into “learning” — you cannot directly measure or observe learning per se (although advances in neuroscience may change that in the future by telling us how the brain is changing based on long-term exposure to different kinds of practice activities). Instead it can be inferred by 2 principles — “retention” and “transfer”.

Retention is concerned with how well I can perform a task some time after an exposure to practicing that task. For example, I do a drill working on receiving and controlling the ball. Towards the end of the drill I successfully control the ball 90% of the time, but what is really important is how well I can control the ball 1 week later. So 1 week later I do a “retention” test where I have 5 reps of trying to control the ball. Suddenly I can only successfully control 2, so 40% success. This is the idea of retention — how do we best design practices that result in a higher retention rate. What research tells us is that if practice tasks of controlling the ball were done in a more variable way (ie not all to feet but using different parts of the body, different length/speed/trajectories of passes), or even interspersed with other skills, is that retention rates will be higher. So whilst my performance in practice might be less because I am having to deal with more variability (my success rate might drop to 70%), in the retention test 1 week later my success rate will be higher, say 60%. This is obviously a paradox — lower performance in practice (90% v 70%) can result in higher retention rates (40% v 60%). So it is this retention rate that we should be focusing on.

The 2nd principle is “transfer” — how well does my performance in a practice task transfer to performing tasks required of me in a “real match”? Most skills in a range of human performance domains rely on practice tasks being “representative” (similar) to “real” tasks for high rates of transfer to occur. As I have tweeted/blogged about many times, this brings in the saying…

“transfer from practice to match conditions depends on the extent to which practice resembles those match conditions”

In essence humans get good at the tasks they do — evolution didn’t really have in mind doing one task to get good at doing another task! For example, dribbling around cones a lot will help us get better at dribbling around cones, not necessarily dribbling around a “live” defender. If we want to get good at the task of beating a defender, practice tasks where you are required to beat a defender! Where this principle of transfer is most violated is in the separation of ‘technique” from “perception” and decision making. Coming back to the coaches mindset of needing success in drills, incorporating decision making (via introducing opposition for example) does not align with it — “those damn defenders keep stuffing up my drill!”. As a consequence of this mindset, far too many drills are used that remove the decision making component. But now we are doing a task that is different from that required in a match, hence transfer is reduced. Reduced transfer = reduced learning (assuming our objective is learning of skills for the match environment).

So when talking about long-term player development, a major focus needs to be learning, rather than short-term performance in a drill (although there is a place for that still — depending on the individual player and their needs, and the psychological aspects like confidence and self-belief. These things need to be taken into account still).

Hope all that makes some sense?

Other evidence from the Skill Acquisition field also challenges the notion of high levels of prescriptive instruction & feedback from coaches. There is also an increasing awareness of the important role that movement variability plays when learning a skill — this is particularly important for young players. Rather than trying to drill/instruct a “textbook” technique/movement, it is potentially better to let players “explore” subtle variations in a technique/movement to find one that is best suited to them and “feels right”. Allowing young players this “exploration” can ultimately lead to development of a high level of stability AND adaptability in the technique/movement. In a dynamic game such as soccer, this is the ultimate mix that the best players possess.

So with some of the above Skill Acquisition theory in mind, I’m sure you can see a conflict — primarily that the research/evidence either refutes, or at least questions, many of the “traditional” coaching methods. This is not restricted to soccer either by the way. All team sports seem to have the same “traditions” and even Martial Arts (here is a recent blog post written on Martial Arts teaching & learning — it outlines exactly the issues I see in team sports).

Why have these traditional methods not been challenged? Well largely because coaches fall back on tradition and the “passing down” of knowledge/methods between coaches. This has been occurring for so many decades that it has become an unbelievably powerful socio-cultural force (and there are positives to come from that dynamic). Other reasons that Skill Acquisition and evidence-based methods have struggled to make an impact could be….

• practice design is “core business” for coaches — not as likely to hand over/collaborate with specialists like other Sport Science disciplines
• coaches usually lack a science/pedagogy background
• Skill Acquisition is often absent in Coach Education
• challenging to understand & apply principles within context of “real-world” practice sessions — need mentoring, hard to up-skill coaches at one-off workshops
• perceptual-cognitive skill (decision making, anticipation) was long thought of as genetic or “natural talent” that could not be enhanced by practice
• lag in being able to evaluate effectiveness of certain approaches used in practice (long-term nature of acquiring skill)
• need more robust research (longitudinal, sport-specific)

Mark Upton is co-creator of myfastestmile and the Coaching Science Manager at the English Institute of Sport.