The Science of Learning


Professor Barry Hymer, already known to our readers for his excellent and popular Science-related blog posts, is now the Chief Science Officer for Chessable.

To commemorate this news, Professor Hymer has produced a White Paper for Chessable on The Science of Learning, which can be viewed, downloaded and shared from here and also from our dedicated Chessable Science page.

We present some of the highlights from the document in order to whet your appetite for absorbing the full piece, which contains greatly extended versions of the material presented below. 

Over to Professor Hymer…

Professor Barry Hymer - Chessable's Chief Science OfficerProfessor Barry Hymer - Chessable's Chief Science Officer
Professor Barry Hymer

The Science of Learning

How do you describe chess to your friends and family? It’s an age-old question, of course. Chess is variously described as a game, a sport, an art, and a science, and of course each of these descriptors can claim at least a measure of validity. Even at the micro-level of playing styles, we can make a case for symbolic exemplars in each category: as a game, we can invoke the mindgames of Lasker, the competitive spirit of Korchnoi, or the coffeehouse genius of Carlsen at play.

Sport, Art and Science

As a sport, we can cite the athletic dynamism of Fischer’s, Judit Polgar’s, or Gelfand’s play. As an art, we could enthuse about the left-field, right-hemisphere creativity of Tal, Speelman, or Ivanchuk. And as a science, we can point to the stern dogmatism of Steinitz or Tarrasch, the steely logic of engineer Botvinnik, or the ultra-pragmatism of Karpov.

However we choose to characterise chess, what is clear is that science is at the heart of its mastery (or at least the pursuit thereof), just as science underpins excellence in game-playing, sport and even art: when it comes to games, the mathematical models of game-theory decision-making are a high-end example (a field in which the former chess professional IM Alexander Matros has built his academic reputation). In sport, science is applied widely – knowingly or not – eg via the mechanics of motion or Bernoulli’s Principle in baseball, cricket, swimming, and yachting. And science is in play in art at the level of technique and composition of course, but also conceptually – indeed both art and science are attempts at portraying and explaining the human condition, albeit in very different ways.

Science at the Chessable Core

Little wonder then that Chessable, with its aim to make chess improvement as effective and enjoyable a process as possible, has put the science of learning at its core from the outset. We offer a skeletal summary of the key scientific processes here, but in this document I would like to offer a slightly fuller account for users interested in unpacking the broader science behind our platform. Because science privileges curiosity and scepticism, and progresses via a continual process of refutation and/or refinement of existing theory, Chessable will continue to evolve over time as existing features are phased out in response to new insights and understandings, and new features are introduced in their place.

So what do we know (currently) about the science of learning, and how do we see this made manifest in the Chessable platform? Within the constraints of ‘present purposes’ I propose to outline a succinct response to this question under the following headings:

Learning is (or should be) Visible

Within the field of educational research and practice over the past two decades, the work of Prof John Hattie has commanded an extraordinarily strong influence. Hattie is a psychometrician whose life’s work has been focused on examining very many thousands of published studies in pursuit of an answer to the question, ‘What are the most (and least) powerful influences on educational achievement?’ He has done this (and continues to do this) by conducting meta-reviews of studies – combining multiple studies in particular areas and extruding ‘effect sizes’ for these. Many of the most powerful influences (the highest effect sizes) turn out to be incorporated – mostly by dint of design but sometimes by good fortune – into the Chessable platform. Among these we can include the notion of high quality and timely feedback (about the task, the process of doing the task, and the self-regulation aspects of the task) and the provision of formative evaluation (real-time judgments and pointers towards improved performance). The spaced repetition system on which Chessable courses are built derives much of its power from its capacity for providing immediate and targeted feedback to users, and a coherent structure to assist progression.

Science of Chess Learning

Science of Chess Learning

Learning Involves Both System and Psychology

As much as Chessable’s offer is reliant on efficient and finely-grained systems, with the small army of developers and engineers that support this, it is worth emphasising that there is a strong – indeed crucial – psychological dimension to the learning process. Whilst we know so much more about learning now than we did decades ago, there remains the task of actually embracing learning technologies consistently. This requires the conversion of will – eg the desire to improve at chess – into the volitional elements of willpower, the focus of much of Roy Baumeister’s academic energies.

Learning is Learned

In a recent Chess24.com interview, former world champion Vladimir Kramnik asserted that “I think the definition of talent, and of talent in chess, is the ability to learn.” Kramnik’s view represents a far bolder assertion than might seem apparent at first, but it chimes with currently dominant views that learnability trumps ability (or the metaphor of moving from ability to transformability in the work of a network of researchers based in the Education Department at Cambridge University – Hart et al). It is also reinforced by research in the field of developing expertise, inspired by the work of K. Anders Ericsson, outlined briefly earlier. But it conflicts with a previous orthodoxy, which was the legacy of the Victorian polymath Francis Galton, which attributed talent to genetic privilege. The Galtonian legacy was played out in the 20th century in such domains as psychometric testing for intelligence and ‘aptitude’, segregated schooling and intra-school segregation (eg ability-grouping), alongside patently sinister theories and social practices like eugenics and racial stereotyping and profiling.

Chess Learning is Chemical

Chess Learning is Chemical

Learning is Chemical

With our emphasis above on the systemic and process-related elements underpinning the science of learning, let us conclude with a brief affirmation of Plato’s famous assertion that all learning has an emotional base. This observation foreshadowed our understanding many years later of the neuro-chemical architecture of our emotions, and resultant therapeutic interventions in such fields as psychiatry and mental health. For present purposes, let us point out simply that in embedding itself in an enjoyable, stress-free, game-based format, Chessable supports the release of dopamine and endorphins to enhance learning (see for example Fiorillo, 2011 and other references for further reading in this area). In fact it is this game-based element rather than the accumulation of extrinsic rewards per se that is the major reason for including prizes and other extrinsic motivators in the platform.

Please refer to the full version of the White Paper for the complete version, which greatly expands on the above topics. It can be downloaded and shared from here and also from our dedicated Chessable Science page.

Stay tuned for the next Science post from Professor Barry Hymer, which will be posted exactly one week from now!

You May Also Like