Translator: Ivana Krivokuća

Reviewer: Carlos Arturo Morales How to make good decisions? If you open a book on rational choice, you will likely read

the following message: look before you leap,

analyze before you act. List all alternatives,

all the consequences, and estimate the utilities

and do the calculation. This is a beautiful mathematical scheme, but it doesn’t describe

how most people actually make decisions. And not even how those who write

these books make decisions, as the following story illustrates. A professor from Columbia University

had an offer from a rival university, and he could not make up his mind

whether to accept, reject, go or stay. A colleague took him aside and said,

“What’s the problem? Just maximize your expected utility!

You always write about doing this!” Exasperated, the professor responded,

“Come on, this is serious.” (Laughter) The method of listing pros and cons

and doing the calculation is an old one. Benjamin Franklin once in a letter

to his nephew recommended exactly that. Listing all pros and cons,

and then weighting and adding. And at the end of his letter,

he wrote the following: “If you do not learn it, I apprehend

you will never get married.” (Laughter) Did you choose your partner

by a calculation? I asked my friends who teach this method

as the only method of rational choice how they chose their partner

if they had any choice at all. (Laughter) All of them said, “No, no, no.” There was one exception. And he told me that he applied

his own theory. He explained to me

that he listed all the alternatives, all the consequences. For instance, will she still talk to me

after being married? Will she take care of the kids

and let me work in peace? And then he estimated

for each of these women the probability

that it will actually occur. And multiplied it by the utilities

and made the calculation. Then he proposed to the woman

with the highest expected utility. She accepted. He never told her how he had chosen her. (Laughter) I’ve met him recently.

Now they are divorced. I will talk today about two ways

of making decisions. One is the one that is taught

at the academia; it has many names: Benjamin Franklin’s bookkeeping method

or expected utility theory. And this is a method

that works in a world of risk, that is, when the probabilities, the consequences

and alternatives are known. Here, statistical thinking is enough. A key example is lotteries

or if you play the casino. Then you can calculate

how much you will lose. In a different world,

the world of uncertainty, calculation is not enough. You need smart rules of thumb,

that are technically called heuristics, and good intuitions, which are also based

on smart rules of thumb. I will talk today about decision making under uncertainty and also about the dangers of using systems that work for known risk and applying them

to the world of uncertainty. Choices between two jobs

or between partners are all in the world of uncertainty. You can’t calculate everything,

you don’t know the consequences, and there will be surprises. I will make four points and then illustrate them

with two examples. First, the best decision under risk is not the best decision

under uncertainty. Second, heuristics that you need in order to make good decisions

in the uncertainty are indispensable

for good decision making. They are not, as it’s often claimed, a sign of a kind of mental retardation or of just mental laziness. Third, complex problems do not always require complex solutions, and that’s again

in the world of uncertainty. And finally, more information, more calculation, more time

is not always better. Less can be more. Let’s go to the first example. This is sports. How does an outfielder

catch a flying ball? In baseball, in cricket, or maybe in soccer,

where the goalie has to get it. How does he or she know where to run? There are two theories about that. One is it’s a complex problem,

you need complex mental processes, and the other one is

it’s a complex problem under uncertainty, and you need to find

a simple method for that. Let’s look for the first one. Richard Dawkins, in his famous book

“The Selfish Genes,” proposed the complex method. So what does the outfielder do?

He or she calculates the trajectory. Have you ever calculated a trajectory? (Laughter) Okay, that’s what you do? (Laughter) And this formula doesn’t even have

wind in it or spin, so it’s not enough. But what else could it be? What you see here is the idea to apply a theory that works

if you know everything, like under known risk,

to the world of uncertainty. The key idea is that you say,

“Oh, he behaves ‘as if’ -” and Dawkins puts in the “as if” – the player would calculate that. What’s the alternative?

How do real players catch a ball? That’s my question. And a number of experiments show that real players use

a number of simple heuristics. I’ll show you one. This one works when the ball

is already high up in the air. It’s called the gaze heuristic. It has three steps. First, fixate your eye

on the ball, start running, and finally, adjust the running speed so that the angle of gaze

remains constant. This player here does exactly that. He runs so that the angle of gaze

remains constant, and that brings him there

where the ball will land. Do you want to see it again? Here it is. Importantly: the player

can ignore to estimate or calculate every variable that’s necessary

to estimate the trajectory. Every one. It’s a heuristic that belongs

to a family of heuristics that just looks at one good reason. And then you get there. You’ll find the same heuristics

in evolutionary history, so birds and fish, when they hunt a prey or a mate – which is sometimes not so different – they just keep the optical angle constant

in three-dimensional space, and that’s enough. No trajectory correlations. This heuristic is used

by players unconsciously. If you have ever interviewed a player and asked him how is he doing this

so well, then you get “intuition.” And it’s intuition, meaning the person knows what to do,

but doesn’t know why. But how does this intuition function? As you see here,

it functions by simple rules. The same rule can be used deliberately. Every rule that we studied

can be used deliberately, and that’s very different

from what you might hear in some claims about decision making that think that heuristics

are unconscious, and statistical thinking is conscious. Don’t believe that. Here’s an example. Remember the miracle of the Hudson River? What had happened? A plane hit shortly after takeoff a formation of Canadian geese. They flew in both engines

and silenced both engines. The pilots turned around to see whether they could get back

to La Guardia Airport, or they would have to do something

more risky, like the Hudson River. How did they make this decision? Did they do calculations?

They didn’t have much time. They used the same heuristic,

now deliberately, the gaze heuristic. How does it work in this case? You fixate the tower

through your windshield – that’s what the pilots did – and if the tower is slowly

moving upwards, you won’t make it. And that’s exactly what Jeffrey Skiles,

the co-pilot, is saying, in other words. Here is another instance where heuristics

can help us to make a safer world, and the decisions are done very fast. My second illustration

is the world of finance. We know now that the theory of finance is part of the problem, or was part of the problem

of the financial crisis, not its solution. Why? Because it’s a theory

about known risk, and it’s applied

to the world of uncertainty, and suggests certainties

that are illusiory. Calculations of value, of risk,

and that sort of things. I’ll give you one example. Assume you want to invest money, and you don’t want

to put everything in one basket, but you want to diversify. But how? Harry Markowitz,

from the University of Chicago, got his Nobel Prize

for finding the solution. When Harry Markowitz

made his own investments for the time after his retirement, he used his Nobel Prize-winning

optimization method. So we might think. No, he did not. He relied on a simple heuristic

that we call 1/N. You divide your assets equally. For instance, if you have

just two alternatives, you divide your money 50-50, and so on. The interesting question

is how good is this simple heuristic that doesn’t need much calculation

in the real world of investment, as opposed to the theory of investment? A study by DeMiguel et al. looked at that and gave the complex

Nobel Prize-winning method 10 years of data

to estimate its parameters, and then to estimate what’s happening

the following months. The window wasn’t shifted

until there was no data left. What was the result? According to common measures,

1/N made more money than the Nobel Prize-winning

mean-variance model. The interesting question is now not just to show that something simpler

does something better, but the real question is,

can we identify the world where simplicity pays

or where the complex calculation pays? I’ll show you here

three features of this world where 1/N outperforms mean-variance, or at least very likely

outperforms mean-variance. One feature is predictive uncertainty

is large – that’s the case with stocks. The second one is the number

of alternatives is large, and you can see this

because the complex methods need to estimate more parameters, 1/N not, and then generate more errors. And finally, the learning sample

is small; it was 10 years. Now one can ask the following question: if I would have 50 alternatives,

how many years of data do I need so that mean-variance

actually gets better than 1/N? What do you think? Ten years is too little. Eleven? Twelve? The best estimate is 500 years. So in the year 2500, we can switch from our intuitions, 1/N, to doing the calculations, provided that the same stocks

are still around in the stock market in the first place. Do our banks understand that? No. I recently got a letter

from my internet bank which said: “With Nobel Prize-winning strategy

to success in investment.” And then I read, “Do you know Harry Markowitz?

No? You should know him.” And then a story was told that he won the Nobel Prize

for solving the problem and the bank has now adopted his method, and there was a warning

about people’s intuitions. What this bank has not understood is that they sent the letter

500 years too early. (Laughter) This is my second illustration about the power of simple rules

in an uncertain world, and also about the damage that can happen when you rely on methods

that work very well in risk, but apply them blindly to real world. Ask your own bank what they use. That this is not just a one-shot

is shown by this slide. There are 20 studies,

and what you see here – we have an optimization model

that’s widely used multiple regression, and we have three heuristics. The minimalist is too simple.

It just picks something randomly. The other two heuristics

have a different philosophy. You already know 1/N, you just throw

the weights away and do it equally. One good reason is a heuristic that goes with

the first good reason that it can find, and then it’s there. I’m not going into details,

they are all mathematically studied, and what you see here

is something important. When you know already all the data, that’s called fitting, then the complex model is the best one. So you can, you’re flexible enough,

and you explain the hindsight. When you have to predict,

then something interesting happens; it’s a crossover. Every one of the simplifications

is more accurate, not just more frugal and faster. This condition is like hindsight. For instance, I hear often on the radio a financial adviser being asked why did Microsoft go down yesterday, and he has always an answer. This is hindsight. If you would be asked if Microsoft

is going up or down tomorrow, that would be prediction. And prediction is hard. That’s an illustration. Let me get the general picture. What we’re studying

at the Max Planck Institute is how do people and should people

make decisions under uncertainty, and the first question

is a descriptive one: what’s in the adaptive toolbox? There are many more heuristics

that I can’t tell you today, and lots of social heuristics. So, people trust their doctor, and the study of ecological

rationality asks the question: in what situation is this

a good idea and when not? If your doctor knows the medical evidence, has no conflicts of interest,

and doesn’t do defensive decision making – he has fear that you

might turn into a plaintiff – that’s a good idea. But that’s not the case in most countries. Most doctors – we have studies –

don’t know the evidence, they have conflicts of interest,

and they do defensive decision making. And finally, how to create situations,

environments, and also strategies that are intuitive and that help people

make better decisions? Let me finish. Decision making

under uncertainty is different from decision making under risk, and heuristics are not the second

best strategies, that we often hear. They can do better

than even optimization strategies in a world of uncertainty,

not in a world of risk. And finally, more information,

more time, and more computation is not always better. Less can be more. Thank you for your attention. (Applause)