A little while back we had a look at confirmation bias, and how it can affect your decision making when betting.
The human brain being what it is, there’s many different ways in which we process information inefficiently; confirmation bias is just one of what is known as ‘cognitive biases’, and there’s a very long list of them if you’re interested!
Very basically (in line with my knowledge of neuroscience!), they’re short-cuts that the brain uses when processing information, which means that each and every judgement or decision we make is not always completely objective in isolation.
Given successful betting relies on us making objective decisions and consistently on each individual bet, the dangers that cognitive biases pose to the punt are fairly obvious.
Today we’ll look at one of these: hindsight bias.
In simple terms, hindsight bias refers to a tendency for us to look back at an event in the past and see it as having been easily predictable, when in fact there wasn’t a strong objective reason for predicting it.
Applied to racing, the effect may look like this: ask a successful punter to re-do their own form study and market after a race, and in line with hindsight bias, often they will assign a higher probability (ie, lower odds) to the winner than they did before the race.
Their pre-race market may have been (and hopefully was) completely objective: they assessed the race using their proven methods and produced ratings that over time have been shown to be reliable.
Yet with the ‘benefit’ of hindsight, they may well concentrate far more heavily on different factors in their quest to determine why the winner saluted, and why their own more-fancied runners struggled.
It’s naturally very difficult for a person to look back on what they did previously and completely disregard what hindsight has shown them. There’s a lot of studies in many areas showing just that. It’s human nature.
The problem when it comes to punting is that, as we’ve discussed many times, it’s the long-term that matters. If you rated a runner as 3-1 ($4) favourite, and it fails to win, it doesn’t mean your price was wrong. It depends on the maths.
Odds of $4 simply mean that if the race were run four times, you’d expect that the horse would win once, or 25% of the time (1/4 = 25%). So when a runner fails to win, it doesn’t necessarily mean your odds were wrong: that won’t be clear until you have a large data set of many races, enough to reduce variance to the point where you can accurately assess your own performance.
Yet hindsight bias makes it tempting to over-analyse a single or small number of races and assign too much weight (excuse the pun) when an unexpected result occurs. You may look at the result and think “I really should have known that”, when in reality, your ratings are correct in the long-run and the single result – and crucially, what led to that result – simply isn’t worth over-analysing.
It works for the other runners too: when a $26 chance (25-1) wins, some punters will look back and make false conclusions about why, and then apply that to their form analysis. They’re just using hindsight: “I should have known, and I will next time”. The fact the horse won doesn’t mean they were wrong before the race. Using the same maths, your $26 chance will win 3.8% of the time (1/26 = 3.8%). Until you have a large enough set of results to test that, you won’t know if it’s truly accurate.
The key is to resist the temptation to use hindsight to judge your predictions on a single race. It’s not easy to do – the scientists will say you’re fighting your brain’s naturals instincts – but the use of discipline and long-term thinking will only lead to better results.