Would you find it more useful if I taught you how to buy your next car for the cheapest possible price or if I explained how to predict the foreign policy of South Korea? I thought so. I’ll tell you about the car.

The traditional way of purchasing a car puts all the weapons in the hands of the car salesman. You visit a showroom, get acquainted with the guy on the sales floor and learn a little about the options you have. Then, when you have been warmed up, the salesman asks you, “What will it take to put you in this car today?”

Sadly, the salesman has all the information you really need to set the price. He knows the real costs of different optional extras, the demand for different colours and so on, but you are on the spot now. So you give him a figure. Perhaps you give him one that is ridiculously low. He looks hurt, maybe he is even a little rude. Embarrassed, you go higher, coming pretty close to your own maximum price. He goes away, consults his manager, pushes you up a little further. And the deal is done.

You wouldn’t have agreed to the deal if you weren’t willing to take it, of course. But is it the best you could have done? Almost certainly not. He knew what he was doing and you didn’t.

So here’s what you should do. Research the car carefully on the internet and decide exactly what you want. Determine the colour, the extras, everything. Then, call every dealer within, say, a 20-mile radius. When they answer, tell them exactly the car that you want. Then inform them that you are calling all the dealers in the area and asking about the same car. You are going to buy the car at 5pm from the dealership offering you the best deal. You will ring back soon and seek a price — the full price, with nothing at all left to be added on later.

The dealer may object that if he gives you a quote over the phone, the next dealer will just come in £50 lower. You simply tell him that, yes, this might indeed happen. That is why, you explain, he has to give you the very lowest price he humanly can, so as to avoid anyone underbidding with a price the dealer would have been willing to accept.

When the witching hour arrives, you go to the dealer with the best offer, cheque in hand, and pick up your car. If there is any change in the terms, you go to the second-best showroom, although this shouldn’t be necessary.

What has happened here? You have forced the salesman to provide you, in the form of his lowest price, all the information he has about the real cost of the car. The advantage has moved from the dealer to you. Try it, and you will save thousands of pounds. I have further good news. The insights that produced this method of buying a car can also help you with South Korean foreign policy. Now there’s an additional advantage I bet you hadn’t anticipated.

In a fascinating new book, Predictioneer, Bruce Bueno de Mesquita, the Julius Silver Professor of Politics at New York University, explains why he thinks that experts so often get their predictions wrong. His argument is that while they may understand their field, they do not understand how decisions are made. The professor provides his car purchase method as an illustration of how outcomes differ depending on the way decisions are structured. A different price emerges when the salesman has the upper hand.

The professor then points out that, armed with this knowledge, there are many outcomes that you can shift in your own direction. Take the appointment of a new chief executive. The order in which candidates are considered and voted on by the board will alter the choice. A shrewd board member will realise that. The professor has been hired by people as a consultant to help them stop candidates they don’t want.

The tool he uses is game theory. Since the Second World War this field of study has investigated and modelled the way decisions are made. How should you behave, given that your behaviour produces a reaction from others? The outcome results from an interaction between your choice and the response of others. Using computers you can simulate the interaction.

Game theory is fearsomely hard to follow. Or at least I find it so. But I can see that, in principle, computer modelling allows us far more insight into, say, South Korea’s future foreign policy than we think that we have.

There is still a big role for subject experts — they provide the inputs into the model — but if the professor is right, the future belongs to a new group. Those who are experts not in particular areas but in how decisions are made.

daniel.finkelstein@thetimes.co.uk

Times Recommends

• Bank charges hit us for being human
• Sorry, Father, thou still shalt not steal
• The shed where Santa’s failures go to die