LESS IS (SOMETIMES) MORE - UNCONSCIOUS INTELLIGENCE - Gut Feelings: The Intelligence of the Unconscious - Gerd Gigerenzer

Gut Feelings: The Intelligence of the Unconscious - Gerd Gigerenzer (2007)


Everything should be made as simple as possible, but not more so.

—Albert Einstein1


The pediatric staff of a leading American teaching hospital is one of the best in the country. Years ago, the hospital admitted a twenty-one-month-old boy; let us call him Kevin.2 Nearly everything was wrong with him: pale and withdrawn, he was drastically underweight for his age, refused to eat, and had constant ear infections. When Kevin was seven months old, his father moved out of the house, and his mother, who was often out “partying,” sometimes missed feeding him altogether or tried to force-feed him jarred baby food and potato chips. A young doctor took charge of the case; he felt uncomfortable having to draw blood from this emaciated child and noticed that Kevin refused to eat after being poked with needles. Intuitively, he limited any invasive testing to the minimum, and instead tried to provide the child with a caring environment. The boy began to eat, and his condition improved.

The young doctor’s superiors, however, did not encourage him in his unconventional efforts. Eventually, the young doctor could no longer impede the diagnostic machinery, and responsibility for Kevin was divided among a coterie of specialists, each interested in applying a particular diagnostic technology. According to their conception of medicine, the doctor’s responsibility was to find the cause of the tiny boy’s illness. They felt that they couldn’t take chances: “If he dies without a diagnosis, then we have failed.” Over the next nine weeks, Kevin was subjected to batteries of tests: CT scan, barium swallow, numerous biopsies and cultures of blood, six lumbar punctures, ultrasounds, and dozens of other clinical tests. What did the tests reveal? Nothing decisive. But under the bombardment of tests, Kevin stopped eating again. The specialists then countered the combined effects of infection, starvation, and testing with intravenous nutrition lines and blood infusions. Kevin died before his next scheduled test, a biopsy of the thymus gland. The physicians continued testing at the autopsy, hoping to find the hidden cause. One resident doctor commented after the boy died: “Why, at one time he had three IV drips going at once! He was spared no test to find out what was really going on. He died in spite of everything we did!”


One day in the 1920s, the editor of a Russian newspaper met with his staff for their regular morning meeting. He read out the assignments for the day—lengthy lists of events and places to be covered, addresses, and instructions. While talking, he spotted a newly hired reporter who did not take notes. The editor was about to reproach him for not paying attention, when, to his surprise, the man repeated the entire assignment word for word. The reporter’s name was Shereshevsky. Shortly after this event, the Russian psychologist A. R. Luria began to investigate Shereshevsky’s fantastic memory. Luria read to him as many as thirty words, numbers, or letters at a time, and asked him to repeat these. Whereas ordinary humans can correctly repeat about seven (plus or minus two), the reporter recalled all thirty. Luria increased the elements to fifty, then to seventy, but the reporter recalled all correctly, and could even repeat them in reverse order. Luria studied him for three decades without being able to find any limits to his memory. Some fifteen years after their first meeting, Luria asked Shereshevsky to reproduce the series of words, numbers, or letters from that meeting. Shereshevsky paused, his eyes closed, and recalled the situation. They had been in Luria’s apartment; Luria was wearing a gray suit and was sitting in a rocking chair and reading the series to him. Then, after all those years, Shereshevsky recited the series precisely. This was most extraordinary at the time, given that Shereshevsky had become a famous mnemonist who performed on stage and had been exposed to massive amounts of information to recall in each performance, which should have buried his old memories. Why did Mother Nature give perfect memory to him, and not to you and me?

There is a downside to such unlimited memory. Shereshevsky could recollect in detail virtually everything that had happened to him—both the important and the trivial. There was only one thing his brilliant memory failed to do. It could not forget. It was flooded by images of his childhood, for example, which could cause him acute malaise and chagrin. With a memory that was all details, he was unable to think on an abstract level. He complained of having a poor ability to recognize faces. “People’s faces are constantly changing,” he said, “it’s the different shades of expression that confuse me and make it so hard to remember faces.”3 When reading a story, he could recite it word for word, but when asked to summarize the gist of the same story, he had to struggle. In general, when a task required going beyond the information given, such as understanding metaphors, poems, synonyms, and homonyms, Shereshevsky was more or less lost. Details that other people would forget occupied his mind and made it hard to move from this flow of images and sensations to some higher level of awareness about what was happening in life—gist, abstraction, or meaning.

More memory is not generally better. Ever since Luria, prominent memory researchers have argued that the “sins” of our memory are necessary byproducts of a system adapted to the demands of our environments.4 In this view, forgetting prevents the sheer mass of life’s detail from critically slowing down the retrieval of relevant experience and so impairing the mind’s ability to abstract, infer, and learn. Freud was an early advocate of adaptive forgetting. By repressing memories that include adverse emotional attributes or induce negative emotions when recalled, one can gain some immediate psychological advantage, he argued, even when the long-term costs of repression are harmful. The psychologist William James held a similar view when he said: If we remembered everything, we should on most occasions be as ill off as if we remembered nothing.”5 A good memory is functional, and makes bets on what needs to be remembered next. The same functional principle is used in the file menu of many computer programs, such as Microsoft Word, where only the most recent items are listed. Word bets on the hypothesis that what users looked up last is likely what they will look up next (Figure 2-1).

Yet we needn’t conclude that less memory is always better than perfect memory, or vice versa. The question is, what environmental structures make less than perfect memory desirable, and what structures favor perfect memory? I call this an ecological question, because it is about how cognition is adapted to its environment. What would a world look like in which a perfect memory is advantageous? One such world is that of the professional mnemonist into which Shereshevsky moved, where no abstraction is required. The philosophical world in which perfect memory would flourish is a completely predictable world, with no uncertainty.


Figure 2-1: The Word program keeps only recently opened files in memory and ``forgets’’ the rest. This usually speeds up finding what one is looking for.


The world in which forgetting is adaptive is larger than we think. For people with painful and traumatic experiences, the ability to forget can produce relief. For children, the ability to forget seems to be essential for language learning. When Jeffrey Elman, a cognitive scientist, tried to get a large artificial neural network with extensive memory to learn the grammatical relationships in a set of several thousand sentences, the network faltered.6 Instead of taking the obvious step of adding more memory to solve this problem, Elman restricted its memory, making the network forget after every three or four words—to mimic the memory restrictions of young children learning their first language. The network with the restricted memory could not possibly make sense of long, complicated sentences. But its restrictions forced it to focus on the short, simple sentences, which it did learn correctly, enabling it to master the small set of grammatical relationships in this subset. Elman then increased the network’s effective memory to five or six words, and so on. By starting small, the network ultimately learned the entire corpus of sentences, which the full network with full memory had never been able to do alone. If parents read the Wall Street Journal to their babies and only talked to them using its sophisticated vocabulary, the babies’ language development would probably be impeded. Mothers and fathers know this intuitively; they communicate with their infants in “baby talk” rather than by using elaborate grammatical structures. Limited memory can act like a filter, and parents unconsciously support this adaptive immaturity by providing limited input in the first place.

Starting small can be useful in areas beyond language development. For instance, a new business may develop more steadily with fewer people and less money than with a large group and a $10 million investment. Similarly, if a firm asks someone to do something spectacular, and pays him large sums of money, it may actually doom the project to failure. The rule “create scarcity and develop systematically” is a viable alternative in human as well as organizational development.

Cognitive limitations can help and hinder. It is easy to imagine situations where it is an advantage to start big. But cognitive limitations are not bad per se; they are only good or bad relative to the task at hand. The more complex a species, the longer the period of infancy tends to be. Humans are an extreme case, where a substantial fraction of their lifetime is spent in an immature state—physically, sexually, and mentally. One of our greatest minds, Albert Einstein, attributed his discovery of the theory of relativity to being a slow starter: “But my intellectual development was retarded, as a result of which I began to wonder about space and time only when I had already grown up. Naturally, I could go deeper into the problem than a child with normal abilities.”7


In 1990, Harry Markowitz received a Noble Prize in Economics for his pathbreaking work on optimal asset allocation. He addressed a vital investment problem that everyone faces in some form or another when saving for retirement or striving to earn money on the stock market. Say you are considering a number of investment funds. In order to reduce the risk, you don’t want to put all your eggs in one basket. But how should you distribute the money over the various assets? Markowitz showed that there is an optimal portfolio that maximizes return and minimizes risk. When he made his own retirement investments, he surely relied on his Nobel Prize technique—or so one might think. But no, he didn’t. He used a simple heuristic, the 1/N rule:

Allocate your money equally to each of N funds.

Ordinary folks rely on the same rule intuitively—invest equally. In face, about half of the people in studies follow it; those who consider only two alternatives invest on a fifty-fifty basis, whereas most consider three or four funds and also distribute their money equally.8 Isn’t that intuition naive and financially foolish? Turning the question around, how much better is optimizing than 1/N? A recent study compared a dozen optimal asset allocation policies, including that of Markowitz, with the 1/N rule in seven allocation problems.9 The funds were mostly portfolios of stocks. One problem consisted of allocating one’s money to the ten portfolios tracking the sectors comprising the Standard & Poor’s 500 index, and another one to ten American industry portfolios. Not a single one of the optimal theories could outperform the simple 1/N rule, which typically made higher gains than the complex policies did.

To understand why less information and computation can be more, it is important to know that the complex policies base their estimates on existing data, such as the past performance of industry portfolios. The data fall into two categories, information that is useful to predicting the future, and arbitrary information or error that is not. Since the future is unknown, it is impossible to distinguish between these, and the complex strategies end up including arbitrary information. The formula 1/N would not be better than optimal policies in all possible worlds, however. These policies do best if they have data over a long time period. For instance, with fifty assets to allocate one’s wealth, the complex policies would need a window of five hundred years to eventually outperform the 1/N rule. The simple rule, in contrast, ignores all previous information, which makes it immune to errors in the data. It bets on the wisdom of diversification by equal allocation.


Does it pay to hire a renowned investment adviser for deciding which stocks to buy? Or is it better to save the consulting and managing fees and do it yourself, as long as you diversify? A forceful chorus of professional advisers warns that John Q. Public should not be left to his mere intuition: he cannot pick the stocks for himself, but needs their insider knowledge and sophisticated statistical computer programs to make money on the stock market. True?

In the year 2000, the investment magazine Capital announced a stock-picking contest. More than 10,000 participants, including the editor-in-chief, submitted portfolios. The editor laid down the rules: he chose fifty international Internet equities and set out a period of six weeks in which everyone could buy, hold, or sell any of these stocks in order to make profit. Many tried to gain as much information and insider knowledge about the stocks as possible, while others used high-speed computers to pick the right portfolio. But one portfolio stood out from all others.

This portfolio was based on collective ignorance rather than on expert knowledge and fancy software, and it was submitted by economist Andreas Ortmann and myself. We had looked for semi-ignorant people who knew so little about stocks that they had not even heard of many of them. We asked a hundred pedestrians in Berlin, fifty men and fifty women, which of the fifty stocks they recognized. Taking the ten stocks whose names were most often recognized, we created a portfolio. We submitted it to the contest in a buy-and-hold pattern; that is, we did not change the composition of the portfolio once it was purchased.

We hit a down market, which was not good news. Nevertheless, our portfolio based on collective recognition gained 2.5 percent. The benchmark proposed by Capital was its editor-in-chief, who knew more than all the hundred pedestrians together. His portfolio lost 18.5 percent. The recognition portfolio also had higher gains than 88 percent of all portfolios submitted, and beat various Capital indices. As a control, we had submitted a low-recognition portfolio with the ten stocks least recognized by the pedestrians, and it performed almost as badly as the editor-in-chief ’s. Results were similar in a second study, where we also analyzed gender differences. Interestingly, women recognized fewer stocks, yet the portfolio based on their recognition made more money than those based on men’s recognition. This finding is consistent with earlier studies suggesting that women are less confident about their financial savvy, yet intuitively perform better.10

In these two studies, partial ignorance rather than extensive knowledge paid. Was that a one-off stroke of luck, as financial advisers are quick to suggest? As there is no foolproof investment strategy, name recognition will not always be a winner. We conducted a series of experiments that together suggest, however, that mere brand-name recognition matches financial experts, blue-chip mutual funds, and the market.11 You may be asking whether I myself trust the collective wisdom enough to put my money where my mouth is. In one case I did and invested some fifty thousand dollars in a portfolio created by the name recognition of the most ignorant pedestrian group. After six months, the portfolio had gained 47 percent, better than the market and mutual funds managed by financial experts fared.

How can Jane and John Q. Public’s collective ignorance be equal to renowned experts’ knowledge? Peter Lynch, the legendary money manager of Fidelity’s Magellan fund, gave exactly this advice to laypersons: invest in what you know. People tend to rely on the simple rule “Buy products whose brand name you recognize.” This rule only helps if you are partially ignorant, that is, if you have heard of some and not all of the stocks. An expert such as Capital’s editor-in-chief who has heard of all the stocks cannot use it. In the United States alone, investment consultants earn about $100 billion annually in advising others how to play the market. Yet there is little evidence that advisers can predict much better than chance. On the contrary, some 70 percent of mutual funds perform below the market in any year, and none of the remaining 30 percent that happen to beat the market do so consistently.12 Nevertheless, ordinary folks, companies, and governments pay the Wall Street clergy billions of dollars to tell them the answer to the big question “What is the market going to do?” As Warren Buffett, the billionaire financier, put it, the only value of stock forecasters is to make fortune-tellers look good.


A few years ago, at Kansas State University, I gave a lecture on fast and frugal decision making. After a lively discussion, my kind host invited me out for dinner. He did not say where. The ride was long, too long, I thought. I guessed that he was taking me to a special restaurant, perhaps with a Michelin star or two. But in Kansas? Indeed, we were headed to a very special restaurant, albeit of a different kind. The Brookville Hotel was packed with people eager to dine, and when I sat down and looked at the menu, I knew why my host had taken me here. There was nothing to choose from. The menu listed exactly one item, the same one every day: half a skillet of fried chicken with mashed potatoes, cream-style corn, baking powder biscuits, and homestyle ice cream. The people around me came from all over the place for the pleasure of not having to make a choice. And you can bet that the hotel knew how to prepare their only dinner; it was delicious!

The Brookville Hotel features a radical version of less-is-more—the zero-choice dinner. It embodies the reverse of the New York City ideal that more choice is always better, with menus that resemble encyclopedias rather than helpful guides. This idea that more choice is better flourishes beyond menus, fueling the lifeblood of much bureaucracy and commerce. At the beginning of the 1970s, Stanford University had two retirement plans that invested in stocks or in bonds. Around 1980, a third option was added, and a few years later, there were 5. By 2001, there were 157 options.13 Are 157 better than 5? Choice is good, and more choice is better, says the global business credo. According to rational choice theory, people weigh the costs and benefits of each alternative and pick the one they prefer. The more alternatives people have the higher the chance that the best one will be included and the more satisfied customers will be. But this is not how the human mind works. There is a limit to the information a human mind can digest, a limit that often corresponds to the magical number seven, plus or minus two, the capacity of short-term memory.14


Figure 2-2: Do customers buy more when there is more choice?

If more choice is not always better, can it also hurt? Consider Draeger’s Supermarket in Menlo Park, California, an upscale grocery shop known for its wide selection of foods. Draeger’s features roughly seventy-five varieties of olive oil, two hundred and fifty varieties of mustard, and over three hundred varieties of jam. Psychologists set up a tasting booth inside the grocery store.15 On the table were either six or twenty-four different jars of exotic jams. When did more customers stop? Sixty percent of the customers stopped for the wider selection, compared to 40 percent when fewer alternatives were offered. But when did more customers actually purchase any of the jams on offer? With twenty-four choices, only 3 percent of all shoppers bought one or more jars. However, when there were only six alternatives, 30 percent bought something. Thus, overall, ten times as many customers purchased the products when the selection was limited. Shoppers were more attracted by more alternatives, but many more bought the products when there was less choice.

Smaller selections can pay. Procter and Gamble reduced the number of versions of Head and Shoulders shampoo from twenty-six to fifteen, and sales increased by 10 percent. Unlike Draeger’s, the global supermarket chain Aldi bets on simplicity: a small number of products bought in bulk, which allows for low prices, and an absolute minimum of service. Yet the quality of its products has a good reputation and is constantly monitored, much easier to accomplish with a small selection. Forbes estimated the fortune of the owners, the two Albrecht brothers, immediately behind that of Microsoft founder Bill Gates and the aforementioned Warren Buffett.16 Is less choice also better for matters of the heart? An experiment with a group of young singles who were given online-dating profiles got the same pattern of results. These young people said that they would rather choose among twenty potential partners than among four. After having gone through this process, however, those given more choice found the situation less enjoyable and said that it did not increase satisfaction or reduce the feeling of having missed out on a better match.17


A golfer has to go through a long list of steps when putting: judge the line of the ball, the grain of the turf, and the distance and angle to the hole; position the ball; align shoulders, hips, and feet to the left of the target; prepare for backswing; and so on. What advice should a coach give to a golfer? How about “Take your time, concentrate on what you are doing, and don’t get distracted by anything around you.” That seems like clever consulting to some, and patently obvious to others; in any case, it is backed by research on the so-called speed-accuracy trade-off: the faster a task is performed, the less accurate it becomes. And in fact, when novice golfers are advised to take their time, concentrate on their movements, and focus their attention, they perform better. Should we give expert golfers the same advice?

In an experiment, novices and expert golfers were studied under two conditions: they had either only up to three seconds for each putt or all the time they wanted.18 Under time pressure, as mentioned, novices performed worse and had fewer target hits. Yet surprisingly, experts hit the target more often when they had less time than when they had no time limit. In a second experiment, players were either instructed to pay attention to their swing or distracted by an unrelated, second task (counting tape-recorded tones). When they were asked to pay attention to their swing, as one might expect, novices did better than when they were distracted. Yet with experts, it was again the opposite. When experts concentrated on their swings, their performance decreased; when experts’ attention was distracted, their performance actually improved.

How can we account for this apparent paradox? Expert motor skills are executed by unconscious parts of our brains, and conscious thinking about the sequence of behaviors interferes and becomes detrimental to performance. Setting a time limit is one method to make thinking about the swing difficult; providing a distracting task is another. Since our conscious attention can focus on only one thing at a time, it is fixed on the distracting task and cannot interfere with the swing.

Golf is not the only sport where taking time can hurt an expert. Indoor handball is a team sport in which players face a constant stream of quick decisions about what to do with the ball. Pass, shoot, lob, or fake? Allocate the ball to the left wing player, or to the right? Players have to make these decisions in an instant. Would they make better decisions if they had more time and could analyze the situation in depth? In an experiment with eighty-five young, skilled handball players, each stood in front of a screen, dressed in his uniform with a ball in his hand. On the screen, video scenes of high-level games were shown.19 Each scene was ten seconds long, ending in a freeze-frame. The players were asked to imagine that they were the player with the ball, and when the scene was frozen, to name as quickly as possible the best action that came to mind. After their intuitive judgments, the players were given more time to inspect the frozen scene carefully, and asked to name as many additional options as they could. For instance, some discovered a player to the left or right they had overlooked, or noticed other details they were not aware of under time pressure. Finally, after forty-five seconds, they were asked to conclude what the best action would be. This final judgment was in about 40 percent of all cases different from the first choice. How did their intuitive first choices compare with the final decision upon reflection? To measure the quality of actions, professional-league coaches evaluated all proposed actions for each video. The hypothesis of a speed-accuracy trade-off suggests that when players have more time, they choose better actions because they have more information. As with the expert golfers, however, the opposite was the case. Taking time and analyzing did not generate better choices. In contrast, the gut reaction was, on average, better than the action chosen after reflection.

Why was the gut feeling so successful? Figure 2-3 shows the answer. The order in which possible actions came to players’ minds directly mirrored their quality: the first action was substantially better than the second, which in turn was better than the third, and so on. Thus, having more time to generate options opens the door for inferior ones. This ability to generate the best options first is characteristic of an experienced player. Inexperienced players, by contrast, will not automatically generate the best actions first, and for them, more time and reflection may help. That the best option tends to pop up first has been reported for experts in various fields, such as firefighters and pilots.20


Figure 2-3: Are skilled ballplayers better off with more time to think before acting? The first spontaneous option that came to mind was the best one; the others were inferior (based on Johnson and Raab, 2003). Therefore, experienced players are well advised to follow their first gut feeling.

The speed-accuracy trade-off is one of psychology’s well-established more-is-better principles. This earlier research was generally done with naive students rather than experts, however, and as we have seen, more (time, thought, attention) is better doesn’t apply to expertly mastered skills. In such cases, thinking too much about the processes involved can slow down and disrupt performance (just think about how you tie your shoes). These processes run best outside of conscious awareness.

Stop thinking when you are skilled—this lesson can be applied deliberately. The famous pianist Glenn Gould was scheduled to perform Beethoven’s opus 109 in Kingston, Ontario. As usual, he started to read through the music first and then play it. Three days before the concert, however, he had a total mental block and was unable to play through a certain passage without seizing up. In desperation, he used an even more intense distraction technique than the experiments with the golfers had. He turned on a vacuum cleaner, a radio, and a television simultaneously, producing so much noise he could no longer hear his own playing. The block vanished.

In competitive sports, the same insight can be deliberately used to undermine your opponent psychologically. For instance, while switching courts, ask your tennis opponent what he is doing to make his forehand so brilliant today. You have a good chance of making him think about his swing and weakening his forehand.21 In sports, emergency units, and military actions, decisions need to be made fast, and striving for perfection by prolonged deliberation can lose the game or somebody’s life. An ad for a computer game featuring U.S. covert operations in the Pacific theater in 1942 once caught my eye. It showed a picture of two marines on a road, facing a misty landscape with trees, bushes, and a wooden bridge across the road. Four locations were marked, and the question was asked: “What is the enemy hiding behind?” Having carefully inspected all the locations, I suddenly noticed the solution printed upside down beneath the picture: “You took too long to answer. You’re dead.”


Gut feelings are based on surprisingly little information. That makes them look untrustworthy in the eyes of our superego, which has internalized the credo that more is always better. Yet experiments demonstrate the amazing fact that less time and information can improve decisions. Less is more means there is some range of information, time, or alternatives where a smaller amount is better. It does not mean that less is necessarily more over the total range. For instance, if one does not recognize any alternative, the recognition heuristic cannot be used. The same holds for choices between alternatives. If more people buy jam when there are six as opposed to twenty-four varieties, that does not imply that even more will buy when there are only one or two alternatives. Typically, there is some intermediate level where things work best. Less is more contradicts two core beliefs held in our culture:

More information is always better.

More choice is always better.

These beliefs exist in various forms and seem so self-evident that they are rarely stated explicitly.22 Economists make an exception when information is not free: more information is always better unless the costs of acquiring further information surpass the expected gains. My point, however, is stronger. Even when information is free, situations exist where more information is detrimental. More memory is not always better. More time is not always better. More insider knowledge may help to explain yesterday’s market by hindsight, but not to predict the market of tomorrow. Less is truly more under the following conditions:

A beneficial degree of ignorance. As illustrated by the recognition heuristic, the gut feeling can outperform a considerable amount of knowledge and information.

Unconscious motor skills. The gut feelings of trained experts are based on unconscious skills whose execution can be impeded by overdeliberation.

Cognitive limitations. Our brains seem to have built-in mechanisms, such as forgetting and starting small, that protect us from some of the dangers of possessing too much information. Without cognitive limitations, we would not function as intelligently as we do.

The freedom-of-choice paradox. The more options one has, the more possibilities for experiencing conflict arise, and the more difficult it becomes to compare the options. There is a point where more options, products, and choices hurt both seller and consumer.

The benefits of simplicity. In an uncertain world, simple rules of thumb can predict complex phenomena as well as or better than complex rules do.

Information costs. As in the case of the pediatric staff at the teaching hospital, extracting too much information can harm a patient. Similarly, at the workplace or in relationships, being overly curious can destroy trust.

Note that the first five items are genuine cases of less is more. Even if the layperson gained more information or the expert more time, or our memory retained all sensory information, or the company produced more varieties, all at no extra cost, they would still be worse off across the board. The last case is a trade-off in which it is the costs of further search that make less information the better choice. The little boy was hurt by the continuing diagnostic procedures, that is, by the physical and mental costs of search, not by the resulting information.

Good intuitions ignore information. Gut feelings spring from rules of thumb that extract only a few pieces of information from a complex environment, such as a recognized name or whether the angle of gaze is constant, and ignore the rest. How does this work, exactly? The next chapter provides a more detailed look at the mechanisms that allow us to focus on these few important pieces of information and ignore the rest.