Numerous examples from everyday life seem to convince us of the inability of the collective mind to make the right decisions. Meanwhile, under certain conditions, a group, even consisting in the majority of people who do not shine with intelligence, often turns out to be closer to the truth than its smartest members.
On a cool autumn day in 1906, the English scientist Francis Galton left his home in Plymouth and went to the annual exhibition of animal husbandry.
Perhaps wandering between the stalls, staring at prize stallions, sows and dairy cows, is odd entertainment for an 84-year-old gentleman. But Galton was known for the breadth of his interests. This, it seems, was the last encyclopedist of modern times, a physician by education, made a contribution to meteorology - he discovered anticyclones, did a lot for forensics (he was one of the founders of fingerprinting), for genetics, psychology and anthropology, invented an ultrasound generator ("Galton's whistle"), developed the first psychological tests, new methods of mathematical statistics, traveled across Africa …
While wandering through the exhibition, Galton came across a crowd in front of one of the pavilions. An unusual game was offered to the visitors: a well-fed bull would be brought out onto the lawn, and those gathered had to guess the weight of the meat that could be obtained from it. For six pence, anyone can buy a ticket with a number, on which they must indicate their grade, as well as their name and address. The most accurate guessers will receive prizes. There were eight hundred people who wanted to, among them were farmers and butchers, but there were also quite a few onlookers who were completely ignorant of animal husbandry and came just to gawk.
When the contest was over and the prizes were distributed, Galton asked the organizers to give him the "ballot papers." He, like many intellectuals of his time, had a low opinion of the mental qualities of the average person and wanted to prove with the help of an unusual fair competition that the average English voter could not correctly assess even the weight of a bull, let alone political programs and statesmen, voting "for" or "against" - even more so.
By the way, a contemporary of Galton, the French writer Gustave Le Bon, in his book Psychology of the Crowd (1895; reprinted several times, there is also a Russian translation) sharply criticized the behavior of any crowd. He was annoyed by the growth of democracy in the late 19th century and was very worried that the ordinary citizens of France could begin to determine the policy of France. “When the crowd acts,” Le Bon said, “they always act stupidly. The crowd can be brave or cowardly, it can be cruel, but it is not capable of being smart. He believed that a panel of jurors often passed sentences that would never be approved by each of them; that parliaments pass laws that each member, if asked personally, would reject.
Out of 800 tickets, Galton rejected 13 - they were illegible, and for the remaining 787 he calculated the average value of the estimated weight of beef after the bull was slaughtered and skinned. He expected this meaning to be far from the truth. But he was wrong. The average crowd opinion was £ 1,197 and the real value was £ 1,198. At the conclusion of an article he published in the scientific journal Nature, Galton admitted: "The result is more evidence in favor of the reasonableness of democratic voting."
Since the time of Galton, numerous examples have accumulated that, under certain conditions, a group turns out to be smarter than each of its members, and often even smarter than the smartest. Even if the majority in the group are not very informed and not very smart people, even if it is led by a person who is not bright in mind, she can work out the right decision.
Promotional video:
Psychologists have repeatedly experimented with collective intelligence In the early 1920s, sociologist Hazel Knight asked a group of Columbia University students (USA) to estimate the temperature in the classroom. The average group decision was 22.5 ° C, while in reality it was 22.2 ° in the hall. In the end, this is not surprising: it is clear that the audience should be at about room temperature, at least 20. But later more complex experiments were carried out. A group of 200 students were asked to rate the weight of different subjects. The average group estimates were 94% correct, which was more accurate than almost all individual results.
In another experiment, a group of 56 students was shown a jar filled with multi-colored pill candies and asked to write on a piece of paper the number of pills in the jar. The average rating of the group was 871. In fact, there were 850 pills in the bank. Only one of the group gave a figure closer to the real one. In all these cases, the students did not discuss the assignment with each other and made estimates strictly individually, just like competitors for a prize at a livestock exhibition.
But here is a much more complex and responsible case than weighing beef or counting sweets by eye.
In May 1968, the American nuclear submarine Scorpion disappeared en route from duty in the North Atlantic to the base. The data on the place of the last radio contact with the boat only allowed us to assume that it should be looked for in an area with a diameter of 20 miles and a depth of thousands of meters. The reasons for the death of the boat were completely unclear.
Scientist John Craven, a civilian Navy officer tasked with investigating the disaster, took an unusual path. He gathered a group of people of different specialties - from submariners to mathematicians, and asked each of them to answer questions, to which, in fact, no one had answers: what happened to the boat? at what speed was she going at that moment? how steeply did she sink to the bottom when she sank? To spur the imagination of the participants, a bottle of the best whiskey was offered for each answer closest to the truth (the truth should have been revealed when the boat was found).
After processing the results using probability theory, Craven obtained a collective assessment of the location of the lost boat. Five months after the disappearance of the "Scorpion" he was found at the bottom 200 meters from the place indicated by the collective mind. Moreover, this place was revealed only after mathematical processing and averaging of the answers; none of the experts specifically named this point. Although none of them knew the speed of the boat, nor the depth at which it went, nor the steepness of its fall inland, the group as a whole, as it turned out, knew this. History, unfortunately, has not preserved information about who got the bottle of whiskey.
Another tragic incident occurred on January 28, 1986. The space shuttle Challenger, after taking off from the launch site at Cape Canaveral, exploded 74 seconds after launch. Eight minutes later, a message about it appeared on the tape of the stock exchange financial news agency.
There is no time for a minute of silence on the American stock exchanges. Within minutes, investors began to dump shares of four major companies involved in the launch: Rockwell (this company built the shuttle itself and its main engines), Lockheed (creators of the launch complex), Martin-Marietta (manufacturers of the external tank for fuel) and "Morton-Thiokol" (the creators of a solid-propellant rocket that accelerates a spacecraft in the first seconds of launch).
Twenty-one minutes after the explosion, shares of Lockheed dropped five percent, Martin Marietta three percent, and Rockwell’s six percent.
But the shares of Morton-Thiokol fell the most. So many bidders tried to sell these securities, and there were so few people willing to buy that trading in Tiokol had to be stopped for almost an hour. An hour later, the value of his shares fell by six percent, and by the end of the day - by almost twelve. Meanwhile, the shares of the rest of the firms participating in the creation of the "Challenger" gradually went up, and by the end of the exchange day, the financial damage to them turned out to be small.
In fact, this means that the collective intelligence of the exchange market has decided that "Tiokol" is to blame for the tragedy. Meanwhile, on the day of the disaster, there was no indication of this. Not in the press, not on television. And the next day, the newspapers did not report any signs of Thiokol's guilt.
Only six months later, a specially created commission, which included reputable engineers and scientists (among them was the famous physicist, Nobel laureate Richard Feynman), discovered the causes of the spacecraft disaster. The rubber seal rings on the Tiokol upper stage froze on a cold January morning, became brittle and allowed hot gases to escape, which should only be emitted through the rocket nozzle. The gases burned through the wall of the fuel tank, a powerful explosion occurred.
And the market, half an hour after the accident, having no information, decided that "Thiokol" was guilty.
How could this happen?
The choice was small (only four firms), and it could be purely random. Or maybe the owners of the shares thought that if the construction and shuttle flights were canceled, the Tiokol would suffer the most (the other three firms do a lot more than missiles). Or the halt in trading caused by a purely random decision of some investors to dump the shares of this company caused panic among the rest of the stock exchange traders. All this could be, and yet the fact is amazing.
Two economics professors tried to figure it out. First of all, they looked to see if the Tiokol employees had sold their shares on January 28, who could immediately understand that the problem was in the rubber rings. No, they didn't. Didn't the employees of its competitors get rid of the Tiokol shares, who also knew the subject and could quickly guess what was the cause of the explosion? No, it wasn't. Didn't someone buy up the shares of the other three firms involved in Challenger while dumping the shares of Thiokol? To do so would be logical for an informed person who knew that the other companies had nothing to do with it, and their shares would soon grow, and Tiokol was to blame. No, there were no such market participants.
The two professors did not come to any convincing conclusion.
What exactly happened on that January day? A large group of people (shareholders of four aerospace firms, potential shareholders and owners of shares in their competitors) were asked the question: how much do you think the shares of these firms are worth after the death of the Challenger? And this group of many thousands, in which, most likely, there were no Nobel laureates, answered correctly. It is possible that there were several people who immediately understood what had happened. But even if there were no such people, some fragmentary information about the explosion and about the structure of the space shuttle, which was in the minds of the market participants, formed a picture that turned out to be close to the truth. As was the case with the "Scorpio" and with the determination of the weight of the bull, as well as in experiments with students.
Another, less dramatic episode repeats on the same New York Stock Exchange every spring. It offers advance pricing (called futures) for Florida orange juice. The harvest of oranges, from which the juice is made, will appear in Florida in a few months. Nevertheless, prices worked out by a large collective of stock exchange traders predict summer weather in Florida more accurately than long-term forecasts of meteorologists. Prices are high - there will be few oranges, the weather is bad, and if the pre-set prices are low, then the summer will be great and there will be a lot of oranges …
So what does all this mean? American economist and psychologist James Surovetsky, who studied the problem, came to the conclusion that averaging eliminates the mistakes made by each member of the group. If a sufficiently large group of different and independent people is asked to make a prediction or estimate the probability of an event, the mistakes of different individuals will mutually annihilate, leaving the truth or something close to it. Of course, for this to happen, the members of the group must have some kind of truth.
Surovetsky sets forth four conditions for a group decision to be correct. The opinion of the members of the group should be diverse (everyone should have some kind of information of their own, even if it is a wrong interpretation of the actual facts). They should be independent (everyone's opinion should not depend on the opinion of their neighbors). The group should be decentralized (there is no “boss”, a recognized authority in it, for the opinion of which others could follow). Finally, a mechanism is needed to identify a common solution. For example, in the case of the bull, these are the organizers of the competition, who collected all the marks, and Galton, who calculated the average.
However, it is enough to look through the daily newspapers to find examples of how the collective mind, which seems to meet all these conditions, can be wrong. This is most clearly seen in the example of mass opinion polls. For example, sociologists at the University of Maryland recently asked Americans what, in their opinion, the percentage of the annual national budget the United States spends on aid to other countries. The arithmetic mean was 24 percent. In fact, this share is less than one percent. The reasons for this distortion, in general, are quite understandable: it is flattering for the people to think that we are, they say, selflessly feeding the whole world …
Another poll, conducted at the height of the Cold War, found that nearly half of Americans considered the Soviet Union a NATO member. Perhaps the fact is that the American press, in pursuit of sensations, inflated disagreements between the members of this defensive alliance so that it was already becoming unclear who was friend and who was enemy.