Risk taking is an inherent part of the story of human development. From the first cave-dweller who
ventured tentatively into the next valley to find better food resources and spread his/her genetic
code more widely, much of human progress has depended on the propensity of people to accept
risk. Natural selection has shown that risk taking is generally, for all its dangers, a successful strategy
and our genetic make-up includes programming which hard-wires our brains to both consciously and
sub-consciously seek out and manage risk and the opportunities that often accompany it.
One of the features of the hard-wiring of the human brain is that, complex and amazing though it
undoubtedly is, it can’t quite cope with processing information at the speed required to deal with
the complexity of the information it has to deal with and it also has to manage the fact that, more
often than not, we are in possession of limited knowledge and information regarding the decisions
we have to make. Our capacity to manage the risk and uncertainty that we are pre-disposed to seek
out is limited.
Things happen too fast and are too complex to allow our brains the luxury of processing the
information methodically. Exploring every option, assessing and examining every facet and arriving
at a perfectly optimal conclusion about what to do next is impractical. When that cave-dweller set
out towards the next valley and a sabre-toothed tiger appeared in his path the cave-dwellers who
survived (and passed on their genetic code) were the ones who had the innate and instinctive
capacity to size up the situation, make instantaneous decisions and act extremely quickly. To run, to
hide, to grab a weapon, etc. Their brains short-cut cognitive processes allowing them to act quickly
on the basis of only the most important information, ‘there’s a tiger, I need to run or hide’.
Those that stood and deliberated on the possible alternatives, ‘what are all my options here and
what is the probability of a successful outcome for me by adopting each one of these?’ Well, you can
guess what happened to them.
Not all the quick thinking cave-dwellers survived of course. It didn’t matter, what mattered was that
the slow-thinkers ALL perished in those types of situations.
The ability to short-cut certain cognitive processes is adaptive and natural selection has ensured that
this ability has become an established feature of the human mind. Over the millennia, our brains
have developed many ways to short-cut and speed up our cognitive processes. We have developed a
variety of techniques, mainly sub-conscious, which allow us to short-cut the time it takes to arrive at
decisions and actions and to compensate for our information or knowledge gaps. Without these
techniques, these short-cuts, it is hard to imagine how we would function at all.
All very interesting, I hope you agree, but how does it relate to horse racing and betting?
The problem with these mental short-cuts is that they may be adaptive for life in general but they
can cause us to make errors in certain situations. As life, particularly in the last few centuries, has
become increasingly complex our adaptive short-cuts, useful for surviving sabre-toothed tiger
attacks, sometimes work against us. They play situational tricks on us. Certain of the mental short-
cuts work in such a way that our thinking can produce biases, known as cognitive biases. The way we
think at speed is useful for many situations but for other situations or problems this type of thinking

can lead to errors in judgement and decision making. Some of the biases our brains have a tendency
to fall foul of can particularly work against us when it comes to betting on and analysing horse
racing.
Let’s look at just a few of these in a bit of detail and I’ll outline some strategies that you might want
to employ to counteract your natural propensity towards these biases and the resultant errors of
judgement they can produce.
Anchoring Bias
Let’s start with the Anchoring Bias.
The fundamental principle of the anchoring bias is that people will have a tendency to rely on an
initial piece of information too heavily when making judgements related to it. That initial piece of
information becomes an ‘anchor’ which influences the individual’s assessment or judgement. For
example, if we look at a house and I ask you if it is worth more or less than £100k and then ask you
to estimate its value you will be very likely to place a lower valuation on it than you would had I
initially asked you if you thought the same house was worth more or less than £200k. You will
anchor that initial information (the £100k or £200k figure) and, regardless of its accuracy or
relevance, you’ll take it into account when arriving at your own valuation.
The anchoring effect has been illustrated in many studies and is used widely by skilled negotiators.
The opening price in any negotiation, the initial starting point offered up by one of the parties, has
been shown to be unduly influential in terms of the final price agreed. So, in day to day life, if you
are negotiating to buy, don’t wait for the seller to suggest a price, get in first with a very low offer
(create a low anchor) and if you are selling get in first and anchor a high price.
There are several ways in which the anchoring bias can affect your betting. Firstly, if you look at a set
of prices or odds offered on a race, anchoring bias theory would suggest you are likely to be sub-
consciously influenced by those prices when arriving at your own estimate of the true probability. If
you see a horse shown as a 6/4, your own assessment of its chance is likely to be unduly influenced
by that figure. Your assessment will drift towards that anchor (upwards or downwards).
Looking for value bets means trying to identify horses whose actual probabilities of winning are not
in line with the odds on offer. By looking at the odds on offer before arriving at your own judgement
the anchoring bias is influencing you to adjust your own assessment into line with the odds on offer,
thus negating your original purpose of objectively comparing your own, independent assessment
against that of the market.
One professional odds compiler who works for a large betting organisation recently tweeted about
being influenced by odds (good or bad) he had seen prior to arriving at his own decision, illustrating
the difficulty we face when trying to control our minds and resist the effect of these extremely
powerful biases.
The only way to effectively counteract this and avoid being influenced by another’s assessment is by
compiling your own probabilities (odds/prices) BEFORE you look at anyone else’s because, once
you’ve seen them the anchoring bias is very difficult, if not impossible, for us to consciously
overcome.

There are other ways you can be anchored or unduly influenced by others opinions. Imagine
someone mentions a horse is a 130 rated horse in their opinion. Now, imagine they said it was rated
90. With no other information on the table you would be more likely to subsequently rate it lower if
you’d heard them say it was rated 90 than if you’d heard them say it was rated 130. Regardless of
how clever you are or how objective you may be and whether you consciously decide to ignore that
opinion bear in mind that you are still likely to be influenced in your subconscious by the information,
whether you like it or not.
A solution in that example is to have your method of rating horses relying on as little subjective
evaluation from your own opinion as possible, not always easy as our objective data is often
incomplete in terms of a horse’s actual ability, there is almost always a subjective element. But try to
limit it. You might also try to limit your exposure to other’s opinions as far as possible until you’ve
arrived at your own.
Confirmation Bias
The confirmation bias is the name given to our tendency to assimilate and interpret information in
a biased manner based on pre-existing beliefs or entrenched positions we may already have. When
conducting analysis or data gathering, we may subconsciously ignore information contrary to our
view and actively seek or give undue importance or weighting to information supporting or
confirming it.
It isn’t hard to see the confirmation bias at work in our daily lives. In politics, where views and ideas
on issues with significant degrees of complexity are often simplified into yes/no polarities, there is a
strong tendency for political supporters to selectively interpret information that supports their
ideology or viewpoint and to ignore evidence that undermines their position.
I am sure you’ll also have seen in your own lives how difficult it can be to get someone to change
their mind on a topic, particularly if they have an emotional attachment to their point of view,
regardless of how compelling the information or factual data you may present to them.
Racing, it is often said, is a game of opinions. One of the reasons it is an opinion based sport is that
we don’t often have a great deal of information to go on. The interpretation of the limited data is
compromised by its scarcity and we have to develop opinions and draw conclusions extrapolated
from that limited quantity of data or information. The information we do have may also be
compromised by being of poor quality, its accuracy may be questionable (think official going reports
for example).
It is important then that we are as objective as possible with the data we do have but the danger
posed by our tendency toward confirmation bias is that we are unable to be as objective as we
would like to be. There is a likelihood that we’ll seek out or give weight to only the data that
confirms our (subjectively arrived at) opinion and we’ll ignore data that does not support our view.
We may have an opinion, for example, that a horse has lots of potential to improve because of some
factor or combination of factors. Once we sow the seed of that opinion in our brains we’ll tend to
subconsciously start to look for information to support that opinion and we’ll also tend to overlook
information or data that would throw doubt on or fail to support that opinion.

We may start to find excuses for a horse, looking for data that supports our view that it is a horse of
significant potential. I’m sure we’ve all had our ‘cliff’ horses, horses that we follow time and again on
the basis of some opinion we’ve arrived at regarding its merit and nothing that happens, no matter
when it is even staring us in the face, can change our view. And the important thing to remember is
that this will happen subconsciously, you won’t be in control of how your subconscious is influencing
your thinking.
We must be wary not to seek out and notice only the facts that support our pre-existing opinions
and we must be wary about failing to give due weight to the information that contradicts our
opinion.
In practise the confirmation bias may affect our analysis of a race by our noticing some bad luck ‘our’
horse may have had in a race while failing to notice the bad luck others may have had. Or we simply
over-emphasise the bad luck and its effect on the result or performance. We then classify our horse
as having been ‘unlucky’ and mark him up, wrongly, for his effort when we should actually be doing
nothing of the sort.
Or we may seek to blame other factors such as jockey, trainer, ground conditions and suchlike as our
brains work subconsciously to find information to support our beliefs concerning the horse and its
ability.
That is not to say we shouldn’t be looking at such factors, just that we should be aware that we may
have a tendency to over-emphasise the effect or result of them depending on whether they support
or fail to support our pre-existing belief.
Another feature of confirmation bias is that we can have selective memory when recalling
information based on whether it supports our position or otherwise. We tend to ‘remember’ things
selectively, for example we may remember the good rides of a jockey we admire while forgetting the
howlers he may have had. And we may do the converse for a jockey we have an adverse opinion of
and in so-doing we’ll arrive at an inaccurate representation of that jockey’s true ability.
Again, you can seek to counter the confirmation bias by maintaining an awareness of its existence.
Challenge yourself by actively questioning your judgement and the basis on which you’ve arrived at
it. Consciously task yourself with finding evidence AGAINST your belief or judgement in order to
ensure you haven’t subconsciously dismissed it.
Availability heuristic
A bias that can be powerful in our context is known as the availability heuristic. As we saw in the
introduction the ability to process information and act quickly is adaptive. One technique our brains
employ in order to short-cut our information processing/action timescales is to utilise the
information that is most readily available to it. Such information tends to be information that is
either more recent or more vivid. It may be information associated with some strong emotional
connection, some novel event or some trauma associated incident.
The availability heuristic is useful to humans in a variety of ways. But the Nobel prize winning
researchers Tversky and Kahnemann suggested, on the basis of experimental work, that there are
predictable patterns of error when this heuristic (or way of thinking) comes into play as people

attempt to assess probabilities. As assessing probabilities accurately lies at the core of successful
betting, this heuristic would seem to present an obvious barrier to us.
In racing terms the manifestation of this tendency is that we may place too much emphasis on the
most recent information. Imagine two horses of equal ability with no factual evidence to
differentiate the likelihood of either finishing in front of the other. One of them won, on television, a
week ago at a major meeting. The other won a month earlier in a midweek race at an unheralded
track. The two performances were of equal merit.
One performance is recent, likely to have been more well-publicised and more heavily imprinted on
our memory (assuming we watched it on television). Our subconscious tendency will be to recall that
information more readily and to give it more weight than the information that is more difficult to
recall relating to the other horse.
Be conscious of that tendency. One way to counter it is to try to absorb information uniformly. For
example, plenty successful bettors will tell you they watch as much racing as possible. While I
suppose they don’t do it for this precise reason, one of the effects of that will be to even out the
availability of recall of the different performances.
You should try to review past performances in a methodological, structured way. For example, if
betting in a race, take time to watch the last two performances, or at least the most recent, of EVERY
horse in the race (if you have time to do so most races are available via the online versions of the
two dedicated racing channels). This ensures that you are planting a recent memory of the
performance in a relatively uniform way for each horse and aren’t relying wholly on your existing
memories of the performances, some of which you may not have witnessed or some of which may
have occurred further in the past rendering them less likely to be recalled and less likely to be given
the appropriate weight or importance by your subconscious.
The Gambler’s Fallacy
This bias is, as you can guess from its name, particularly apposite to our subject matter. The basis of
the concept of the Gambler’s Fallacy is that people have a tendency to believe that if something has
happened more(or less) frequently than would be expected in the past that it will happen less (or
more) frequently in the future.
For example, if when tossing a coin five heads in a row come up people will, in general, give more
weight or probability to the next toss rendering a tail, ignoring the fact that the probabilities for the
next coin toss are unchanged by the events that have already happened (i.e. following the five heads
the next coin toss is just as likely to be another head as it is to be a tail).
There is also a bias known as the Reverse Gambler’s Fallacy which would lead to a person, in the
above example, giving undue weight to the next coin toss being heads, on the basis they will have
concluded that there is some physical advantage (an unbalanced coin for example) rendering a head
more likely. Of course, this thinking may not represent a fallacy as they may be correct, and in racing
terms this is an important point to note.
So how does this apply to our assessment and decision making in betting on racing?

The thing to look out for and be wary of is that we all have an inbuilt, subconscious tendency to look
for patterns and there exists an ever-present danger that we will give an inaccurate weighting to the
likelihood or probability of future events based on events that have occurred in the past that have
no bearing at all on probabilities of for future events.
Often the patterns we observe and base our reasoning on are statistically irrelevant because they
are either derived from small samples or are statistically insignificant. Our subconscious mind
doesn’t care about that, it’ll go right on ahead and start influencing our thinking, no matter what the
statistical nuances are.
There are lots of potential examples. Say, the first four favourites win at a race meeting. Does that
affect the probability of the next favourite winning? The Gambler’s Fallacy would suggest a general
disposition, in this case, towards people thinking that the next favourite is more likely to lose than
the actual probability simply because the first four have won. ‘They can’t keep winning’ is how you
might hear it expressed or ‘the next one is bound to lose’ and so the assessment of the chance of the
next favourite is incorrectly arrived at.
I am sure all the readers of this article would be wary of the obvious trap outlined in that situation
but what about the effect of the draw on a given day?
Say at Ascot the first two straight course, big field handicaps are won by horses drawn on the stands
rail, winning at 10/1 and 8/1. Should this influence the probabilities of a win for the horse drawn on
the stands rail in the next similar race later in the day? How would you arrive at a meaningful,
statistically sound method for assessing whether, and by how much, it would affect that probability?
How would you know you weren’t simply a victim of the reverse gambler’s fallacy ‘the first two
drawn on the stands side have won so the probability of the next winning is higher’? You may be
right, of course, there may be some track bias but, equally, those first two results may be simply
down to randomness and chance. What would your method be to understanding whether it was a
track bias or a random result? How would you quantify it?
Frankie Dettori’s magnificent seven at Ascot in 1996 is a good example of both sides of the fallacy at
play. Having won the first six races punters started to bet Dettori’s horse in the final race at odds
much, much shorter than those generally accepted to represent the horse’s true chance (he was
being bet at 2/1 having been 12/1 before the earlier events had transpired). They were falling prey
to the Reverse Gambler’s Fallacy, believing the fact that he’d won the first six races rendered the
chance of him winning the last to be disproportionately higher. (They were also likely to have been
falling foul of several other cognitive biases or errors of judgement).
The bookmakers, on the other hand, couldn’t believe their luck before the last. Here they were with
the opportunity to lay (take bets on) a horse at 2/1 when his true winning chance was probably six
times those odds. While they were correct in assessing that the Reverse Gamblers Fallacy effect had
shortened the odds to a point where they didn’t reflect the true probabilities and were right to stand
the horse at the prices, some over-stretched themselves on the basis of the Gambler’s Fallacy.
Here is what Barry Dennis was quoted as saying afterwards. "Don't panic, something is coming on
the outside – this is going to beat him. I knew he couldn't ride all seven, it isn't possible. Half a

furlong out – I do not know the challenger but he simply must get up. Twenty yards to go. Reality – it
is not going to get up. Frankie's done it. I stood on my stool, staring, not hearing a thing, in a trance."
And Dennis wasn’t alone. Bookmaker Gary Wiltshire faced financial ruin. It was the day, according to
some, that the mug punters got their revenge. The day falling prey to the Reverse Gambler’s Fallacy
paid off and when the Gambler’s Fallacy proved costly for some.
One area that the Gambler’s Fallacy can lead us down the wrong track is when assessing trends.
Trends are a popular method of analysing races. Historical information is analysed to provide
pointers to future runners. ‘Ten Year Trends’ is a popular feature of the tipping columns in the
Racing Post on big race days.
The data set is so small however that a ten year trend often has no statistical significance or meaning
at all. That favourites have won the last five renewals of a particular race has no bearing on what the
price of the favourite should be the next time it is run. But because it is a ‘good race for favourites’
people may overestimate the probability of the next one winning (the Reverse Gambler’s Fallacy) or
underestimate it (the Gamber’s Fallacy – ‘the favourite has win this the last five times, it is due to get
beat’)
I’m not saying ignore trends. They can potentially provide some useful pointers (for example if
particular trainers are known to target a race or if it has proven a race that horses with certain
profiles tend to do well in) but be wary of giving a disproportionate weighting to the information.
Summary
Alan Potts, a respected professional gambler once said, “As with golf, it’s the six inches between
your ears that will ultimately decide whether you win or lose.”
He was right, of course, and managing what happens between the ears, in the brain, is vitally
important. Understanding how you think can help you to avoid mistakes and costly errors of
judgement. I’d encourage anyone serious about their betting to think about the irrational biases that
our subconscious tendency to short-cut our thinking can lead us to encounter. Only a fraction of
them have been outlined above, and only a fraction of the different situations where they might
lead us astray have been explored.
To understand more you can google ‘cognitive bias’ and find lots of information on the internet. For
those looking to explore the topic more deeply there is an excellent book by the eminent
psychologist Daniel Kahnemann called ‘Thinking, Fast and Slow’ which looks at how our brains make
decisions in the way we do and which explores in depth some of the biases mentioned and explains
the research behind our exploration of them.
Happy thinking!