I am overviewing my thoughts related to the two concepts ‘probability’ and ‘prediction’. The main reason why people want predictions, is to avoid risk materialising. Yes – I know some people would like a sound prediction for winning numbers on the next Euromillions jackpot. If you’re that type, move on swiftly.

But look – life is too full of risky and uncomfortable situations. People desperately want to avoid adverse outcomes, which they see as coming at them from that nebulous place they call ‘the future’. Prediction has to be about the future. Well, it’s obvious that nobody (I know) cares about predicting the past.

I have covered some of the concepts in various ways as below:

1. May 2012 – I covered how humans assess risk.
2. 2013 – my thoughts on prediction were rudimentary. I conflated probability with prediction.
3. Dec 2020 – I touched on so-called predictability and preventability.
4. July 2021 – I explored the implications of predictions when the future is not real.
5. April 2022 – I dealt with risk in the context of what is gambling and what isn’t.

Probability, risk, future and, and prediction  – must all hang together as concepts. But wait – what is a concept? You can head off to the nearest dictionary if you want. For this post I take concept to be an idea in the mind of a person. I fully accept that it can mean a range of different things.

What is probability?

Probability is the chance of something (an event) happening or not. The bookies (in England) recently had up up for bets odds of 6/4 that Boris Johnson would not be the Prime Minister by end of 2022. That meant that if you wanted to bet that he would be PM by end of 2022, your £50 bet (say) would bring you approx £125. Those winning a bet in their favour would punch the air, probably saying, “I knew it. I predicted it!“. Those who lost would sulk or cry in their tea, saying “My prediction was wrong.” They are all confused.

On separate matters, getting it right on the 1 in 100 chance is not a prediction. You were right some would say, “Because you were lucky.”

Probability statements don’t aim to be right. People aim to be right. If a person is crossing a busy road with their eyes closed, the probability is overwhelmingly perceived to be not in their favour of surviving unscathed. So let’s say that you think the probability of they being run over i90%. What happens if they are not run over? What does that mean? Is the probability statement wrong? It isn’t. How? The statement left a 10% probability that they would not be run over.

Probability statements can be right and wrong at the same time. How? If the event happens, the statement was right to the greater probability. If the event doesn’t happen it was also right to the lesser probability – and vice versa if you substitute the word ‘wrong’ in the latter. Of course, I expect people to say ‘That’s not fair. You can’t be right and wrong at the same time!‘ – which means that people like conviction which is what predictions supply. This is exactly how prediction and probability are confused and conflated.

Use of probabilities

How can probability statements be used? Depending on context, action should be taken to avoid the greater probability of an adverse event. It gets very tricky when probabilities are about random events (a pandemic or asteroid impact), with no discernible previous pattern. Nobody should be overly occupied about the probability of winter happening every year between Dec and February, as there is a well established pattern about that. One should be interested if a probability estimate of an individual committing suicide or homicide is estimated at *51% or more. It would be stupid to wait until a such probability estimates rose to say, greater than 70%. There are cost implications (see below), that get tangled in all this.[*Caution: this does not mean that such probabilities less than 51% are unimportant].

Asteroid impact

Everybody knows that the risk of an asteroid – one large enough – striking the earth and wiping out all or most of humanity, is pretty small. What ordinary people might want to know is the chance of such a catastrophic event. If they hear something like (for example) ‘1 in 100 million over the next 30 years‘, that’s reassuring as they can move on swiftly. But unordinary people – such as leaders of major economic powers – might look deeper. For them 1 in 100 million over 30 years would NOT mean that such an asteroid cannot fall in the next 5 or 10 years.

What this means is that the value of a risk assessment is different based on ‘who you are’ and the nature of impact you might conceptualise. Joe Bloggs walking down the street is simply concerned that an asteroid won’t fall on his head any time soon. If you’re the leader of a Nation, the life of your Nation and its people is a broader matter that is considered over a longer time frame than ‘any time soon’. Bloggs is mainly concerned that his house won’t be flattened. Nation Leader is concerned that his country’s economy and people are not flattened at any time over the next few years. Individuals with lower levels of responsibility engage with risk differently compared to those with much higher levels of responsibility.

Prediction

A prediction is inherently binary. What does that mean? Binary means one way or the other. If someone was to predict that the sun would rise in the West tomorrow morning – and it doesn’t – then the prediction is wrong. A prediction cannot be right and wrong at the same time – in contrast to probability statements.

People use the word prediction loosely – and the research is yet to arrive, showing how dangerous that can be. Prediction means that a person can see into the future. That’s total nonsense and I’m not going to be occupied debating it. I’ve actually heard highly trained professionals using the term ‘risk prediction‘. Of course, I have to button it cuz if I were to tell them “Sorry, you’re confused!“, I’d get myself into big trouble. But they are confused. How? Risk is not an event – it is a probability. Logically you can’t predict a probability. You have to estimate it. Jeez – it’s causing me distress, thinking about this nonsense.

Luck

The probability of the ordinary person striking a hard ball moving at 100mph with a bat [38 in (965 mm) long and width no more than 4.25 in (108 mm)] is pretty small. Amusingly people don’t say that a good batsman who achieves that most times is lucky. The situations in baseball , table tennis and badminton are even more demanding. Demanding of what? It’s called ‘skill and experience’. Skill and experience can change the odds in favour of an expert.

So the seasoned batsman who achieves against seemingly impossible odds, is not simply ‘lucky’. The non-seasoned novice who strikes the fast ball is seen as lucky.

Luck is about throwing dice, with no true application of skill.

Cost

Taking action to attempt to avoid an adverse outcome means expenditure on manpower and resources. There is the usual dilemma: “What if we spend all this money and nothing happens? If we spend money something might happen anyway. Whatever we do it could be a waste of money.” Think about it. The whole point of spending money is so that the thing doesn’t happen! But what ‘people’ do not balance is ‘What if you don’t spend the money and something happens – what’s the cost of that?‘

But readers may have overlooked the word ‘happen(s)’. It brings me right back to ‘people’ needing a guarantee about the non-existent future. There ain’t!

With asteroid impact, there are serious consequences for human and other life on earth.

The questions for decision-makers who aim for risk control, ought to be as follows:

1. What would it mean if a low probability event has catastrophic consequences if it happened?
2. Can we in some way limit our spending to the extent that we are proved wrong by history?
3. What is the impact of no advance spending or under-investment in aiming to avoid risk materialisation? [Think New Orleans disaster, Fukoshima and the COVID pandemic).

In the above – taking asteroid impact as an example, they spend some money – not endless money – to be prepared. An asteroid is most unlikely to come out of ‘nowhere’ within a couple days. So if they started building defences years ago, they already have a template to work with. That could then be enhanced, because they will have time. That’s better than starting from scratch.

I can’t go into each scenario imaginable. Suicide and homicide reduction require a separate set of thinking. However, I think my three questions are still good in those scenarios.

Risk among the mentally disordered

This is a complex area of analysis, which I won’t delve deeply into. Why? Because lay persons have a habit of misinterpreting complex dynamic issues, and then misapplying knowledge not baked by experience and training. The mindmap below gives a bare bones outline only.

Conclusions

1. Prediction and probability are distinct concepts.
2. Risk management and control should avoid predictions.
3. Probabilities can be used in a cost-effective manner to avoid risk-materialisation. [I did not say ‘perfectly’].
4. Depending on luck or happenstance is not a great way to manage risk.