Training Your Metacognition

A recurring theme in these articles on betting psychology has been that, as evolutionary flawed thinkers, we tend to see outcomes as binarily predetermined; they will either happen, or they won’t.  We struggle to recognise that the outcome of future events – until they happen – are undecided, and that the future, rather than being fixed, is, in fact, an ever-shifting complex and chaotic web of possible outcomes, each of which has a probability attached to it, a probability that moves in reaction to other outcomes.

Even writing that last sentence about the multiplicity of the future caused my brain to overheat, so it’s little wonder we find it hard to engage with non-binary uncertainty in our everyday lives.  Much easier – and perhaps more secure, too – is to believe that the future already exists, and that our role is simply to move towards it with as little fuss as possible.

Deep Rooted Beliefs

Deborah Kelemen and Evelyn Rosset, psychologists at Boston University, among others, have found that this tendency to believe in a predetermined future is deep rooted in all of us, and is strongest in subjects with lower education levels, or those who are placed in stressful conditions.  It’s what is called our teleological bias, a hard-to-shift belief that things have a purpose, an end, and that things happen for a reason.  It’s why our evolutionary ancestors danced to sun gods and made sacrifices to rain gods, believing that the outcome of harvests was controlled by a mighty being.  It’s what Blaise Pascal called our “god-shaped hole”.

And it’s part of the same type of flawed cognition that makes us view history in such over-simplistic terms: our so-called hindsight bias.  This prejudice, although long written about, was first scientifically demonstrated in 1975 by Baruch Fischhoff and Ruth Beyth, in their landmark paper, I Knew It Would Happen: Remembered Probabilities of Once-Future Things.  In experiments with a few hundred students at two Israeli universities, participants were asked to make predictions as to the outcome of President Nixon’s upcoming diplomatic visits to China and Russia.  After Nixon’s return, the same participants were asked to recall their predictions.  Unsurprisingly, students typically believed that their predictions were in line with how the political missions had played out, even when they had originally suggested different outcomes.

Punters will recognise this I-knew-it-all-along tendency.  How many times have we watched a sporting event and been struck by how obvious its eventual outcome was?  Of course that horse was going to break slowly at the start.  Of course that rookie player was going to wilt under the pressure of the big game.  Of course the conditions were going to favour Team X.  The problem with punting, though, is that we often have an indelible record – a losing bet – that reminds us that, in fact, the future wasn’t so obvious to us after all.

Our emotional response to this is crucial if we are to sustain a profitable mindset.  Losing bets can make us feel we’ve been duped.  Or that we’re stupid.  But this kind of reaction is dangerous.  Not only is the thinking upon which it is based flawed – the future was never that obvious to us, because the future never is – but it means our next actions will be driven by emotions of anger and revenge, neither of which tend to lead to sound betting decision making.

These twin challenges – that we inherently believe the future is already written and yet feel cheated when it wasn’t written as we had expected – create a tricky psychological starting point for any punter that needs to be managed.

Tackling Metacognition Head On

The first step in doing this is to rationally unpick our deterministic leanings.  We need to remind ourselves – with a note pinned to our walls saying “the future is not written yet” if necessary – that probabilistic thinking underpins profitable punting.  The role of the successful punter is to identify what those future probabilities are and bet accordingly.  If you can back an outcome at odds that imply it has less of a chance of occurring than you know it has, then you have found betting’s holy grail – value – and can be secure in knowing that, if all your bets share that quality, you may lose in the short-term, but will achieve long-term profitability.

Most punters can grasp this concept intellectually.  Toss a coin and ask them what the fair odds are for heads or tails and most can tell you it’s $2.00.  Offer odds of $2.20 on heads and most would recognise that as a value bet.  However, remove many of those same punters from an abstract world of hypothetical coin tosses and, faced with added complexity, their ability to think probabilistically in this way evaporates: their hardwired determinism reasserts itself.

New research into the psychological field of metacognition may give us a tool to stave off this cognitive bias, though.

Practical Example

Try this.  When reaching the point where you are about to place a bet, ask yourself a simple question:

On a scale of 1-4, how confident am I that this bet will win?

Now record the outcome on your judgement on a column in a spreadsheet that tracks the bets you place.  Do all of this before committing to the actual bet.

By doing this, we immediately shift our thinking, because the question of how confident we are presupposes that the outcome is not necessarily certain.  It also forces us to think probabilistically.  This shift is largely driven by how unsatisfactory the scale is to use: the scale translates to a series of percentage chances of an event occurring – 0%-33%-66%-100% – and, when asking ourselves the question, our most likely response is going to be frustration that our confidence in the bet we are about to place lies somewhere in between the percentages on the scale.  As we start engaging with that kind of thought process, then, we simultaneously start associating an event occurring to betting odds, which are just another mathematical expression of percentage chance.  And as soon as we start doing that, we can start seeking odds that are more favourable than that percentage chance implies: we’re back in our abstract world of hypothetical coin tosses.

[To convert odds to a percentage chance, divide 100 by the odds (e.g. 100 ÷ $1.61 = 62.11%).  To convert a percentage chance to odds, divide 100 by the percentage chance (e.g. 100 ÷ 62.11% = $1.61).]

The benefits of beginning to use this rating system goes beyond the way it forces us to engage with probabilistic thinking, though.  A team at University College London, led by Dr Stephen Fleming, who runs their Metacognition Lab, found in 2018 that provided we receive feedback on our decisions along the way, the accuracy of our confidence-ratings improves.

Fleming and his team asked participants to assess which, out of a series of images, was brightest against a dark background.  Crucially, though, they also asked how confident – on a scale of 1-4 – participants were in their answer.  Participants were then split into two: a test group who received feedback on their confidence judgement; and a control group who received none.  Over the course of eight further sessions, over a two-week period, the control group’s performance in similar tasks did not improve, and neither did their ability to judge their own performance.  The test group fared better, though, improving individuals’ performance at judging the brightness of images, and their ability to assess their own performance.  Further, the test group also showed a similar improvement in their ability to judge their own performance when performing different tasks, suggesting this metacognitive skill (being able to assess our own thinking) is transferrable.

In short, Fleming’s work suggests two things.  First, that by actively engaging our minds in assessments of how well we’ve performed a task, we improve our ability to perform that task in the future.  And second, that by doing so, our ability to objectively assess our own performance also increases.  Provided we receive feedback along the way, of course.


Fleming’s work – like most research done by psychologists in a lab – needs to be treated cautiously.  There’s a big difference between the neat certainty of a brightness test performed in controlled conditions over short bursts of 20-minute sessions and the more complex decisions, with messier outcomes, that face us in everyday life.  Nonetheless, punters would do well to experiment with Fleming’s protocol.

After all, when looking for a “real-world” situation that mirrors his lab conditions, punting comes pretty close: you are presented with a single puzzle, a sporting event, that you have to find an answer to; and you receive feedback on the event soon after.

Inserting a step in between the decision and the bet placement, where you assess the confidence you have in your selection, is straightforward.  If nothing else, it slows down the process of placing a bet – which is almost always a good thing – but it has several further benefits, too.  It forces you to think probabilistically, for one, meaning that you’re much more likely to engage with the concept of whether the bet represents value or not.  But it also allows you to start training your metacognition – your ability to figure out how well you are thinking – which means your judgement as a punter will improve over the long-run.  I’m certain of it.

Related Articles

Chimp Paradox and Confirmation Bias

Meeting a professional punter for the first time is a disappointing experience for most people. Expectations of cigar-smoking braggarts, ...

Cognitive Biases

Imagine the scenario. You’re a doctor in an emergency department. You’ve been working for 14 hours that day and ...

Avoiding Tribal Thinking

In September 2000 I travelled to Dublin, in Ireland, to visit Leopardstown Racecourse and watch the Grade 1 Irish ...