How was the book?

You should read this. You might not agree with all the arguments by James Surowiecki, but he will give you new perspectives.

Experts vs. groups is the fundamental question of this book. Great virtue of collective decision-making is that ”errors people make effectively cancel themselves out.” The Wisdom of Crowds is an argument against ”our excessive faith in the single individual decision maker”, because it’s hard to identify a true expert and he also has humans error – biases and blindspots.

James Surowiecki is ”merely” a staff writer at the New Yorker. My expectations were fairly low not because of the writers profession, but the topic seemed so self-evident. I was wrong; the book has intelligence and deep understanding of human nature.

What are the key learnings?

Learnings are:

• On average people get things right without any communications.
• Some people will do better, but on average they get it close enough.
• The simplest way to get reliably good answers is just to ask the group each time.

One of the key lessons is that ”we don’t always know where the good information is.” The key to wisdom of crowds is diversity, independence and decentralization or private judgement.

Surowiecki uses a lot of examples to demonstrate the power of crowds. Examples of accuracy of crowds:

• Francis Galton noticed that groups are remarkably intelligent on averaging things. For example the ox on the market which weighed 1198 pounds. And the guestimate was 1197 pounds.
• Who Wants to Be a Millionaire audience got the answers right 91 % of the time.
• On Hazel Knights tests the accuracy was on room temperature 72,4 degrees when the correct was 72 degrees.
• In case Challenger the crowd got the company responsible of the accident right – which was Thiokol.
• Google in built on wisdom of crowds.

Collective intelligence has three problems:

• Cognition
• Coordination
• Cooperation

Conditions that make good group intelligence are:

• Diversity
• Independence
• Private judgement

About diversity. James G. March and Scott Page have studied diversity and the learnings from their studies indicate that too smart teams doesn’t work well compare to teams that have also less knowledgeable team members. In fact, people who know less will improve the group’s performance. Alternatively, the ”homogeneous groups are great doing what they do well, but they become progressively less able to investigate alternatives. They spend too much time exploiting and not enough time exploring.” The gains comes from diversity. The value of expertise is overrated. 

Forecasting is difficult even for experts. J. Scott Armstrong has noted ”expertise and accuracy are unrelated.” Therefore, expert thinking and forecasting have little to do compared to crowd thinking. Experts are as likely to disagree as to agree. Experts are also surprisingly bad at calibrating their judgements, because they routinely overestimate the likelihood that they are right. The problem is that experts do not recognize that they are wrong and especially how wrong they got. So chasing a guru who knows everything is a waste of time. Pool of gurus is much better – wisdom of crowds. This doesn’t mean that there would not be people who can outperform group of gurus. For example Warren Buffet has outperformed S&P 500 Index since the 1960s. Try to find smart people, but not the smartest, because he might lead you astray.

Cognitive diversity is important for wise decision-making. You need it to conceptualize problems in a novel way. Homogenous groups tend to do groupthinking. Groupthinking ”shares an illusion of invulnerability, a willingness to rationalize easy possible counterarguments.” Bay of Pigs is a classic example of groupthinking. In addition, Salomon Asch has demonstrated in his experiments of three same size lines that people tend give up on their opinion if the group has different opinion. 70% of test subjects fell pray, because of peer pressure. They didn’t want to stand out. Diversity is important also, because it helps individuals to express more freely their opinion. 

About independence. ”Independence of opinion is both a crucial ingredient in collectively wise decisions and one of the hardest things to keep intact.” Independence means relative freedom from the influence of others and it is important to intelligent decision making. Why is it so important? ”First, it keeps the mistakes that people make from becoming correlated. Second, ”independent individuals are more likely to have new information” which gives diverse perspective and you won’t be making the group any dumber. Being independent is difficult, because we are social beings and we want to learn which is social process. Surowiecki sees that the more people tend to socialize with each other the more there will be personal contact and the likelihood that group’s decision will be less wise. A controversial observation that should be considered in team building. Surowiecki asks that ”Can people make collectively intelligent decisions even when they are in constant interaction with each other?”

Social proof is a concept that was tested by Stanley Milgram, Leonard Bickman and Lawrence Berkowitz in 1968. The put people on a corner of a street looking up in the sky. The more there were people staring in the sky the more other people stopped and started looking also. I.e. crowd becomes more influential as it becomes bigger. The governing assumption is that ”the best thing to do is just to follow along.” 

Hearding should be noticed, because it explains that why people are following the same strategy although there would an incentive to follow alternative strategy. Somehow, people see that following the heard is a safe bet. For example during the Bowling bubble in the 1950s. The stocks of AMF and Brunswick were bought with out the understanding of the limitations of the bowling market. Or as John Maynard Keynes has written ”Wordly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally.” 

Information cascade plays a vital role in decision-making. It is a process how the decisions are made. The groups are better on making decisions than individuals, but individuals are better on coming up with solutions than groups. A successful recipe for making decision is that one should have a lot private information and pay less attention to what everyone else is saying. ”The more important the decision, the more likely it is that the group’s collective verdict will be right.”

Decentralization is a key ingredient in wise decision-making. Weakness of decentralization is that no one guarantees that information available somewhere will be available elsewhere. Strength of decentralization is that is ”encourages independence and specialization on the one hand while still allowing people to coordinate their activities and solve difficult problems on the other.” Linux is an example of decentralization or any other crowdsourced system. 

Norms and convention play a vital role in complex world. And the most successful norms are internalized. The Milgrams experiment of asking people to give their seat in subway is a test of internalized norms. The test person who asks the sitting subject are both playing their internalized norms and conventions. Test person doesn’t have the guts to ask the seat and the subject doesn’t have the guts to say no (internalized norm) or she feels that she has earned the seat, because she was there earlier (convention).

People are odd and the ”ultimatum game” underlines that. When two persons are given a chance to share 10 USD, typically less than 2 USD offers are rejected not because that’s more than nothing, but because the receiver feels that the other party is getting too much by having 8 USD. Receiver wants to punish the greedy counterparty. Moreover, that’s why typically offers tend to yield around 5 USD. Expect in a case where the offering party has earned the position on making offers for example via test. People want to see a relationship between accomplishment and reward.

Virtues of decentralization are:

• The more people have responsibility of their own environment the more they are engaged
• Decentralization makes easier to coordinate. Hard to believe, but the writer suggests that Zara for example does this brilliantly.

How should we change according to the book?

Vote is a great way to aggregating the opinion of the members of the group. Do not emphasize consensus over dissent. Second, ”group decisions are no inherently inefficient.” Third, the groups should really have power to make decisions, take responsibility.

Talkativeness is petrol for the group, because it helps the people to think and make wise decisions.

Making decisions have two separate ways – evidence-based or verdict-based. The decision situation should not start by making conclusions. Share all the information and talk a lot about the topic every day. 

Shadow of the future is a great way to cooperate. When everybody knows that there will be work done in the future it will build durability of the relationship.

What should I personally do?

Follow the shadow.

Summary

The book in six words – ”Who the hell wants to hear actors talk?” Harry Warner of Warner Bros. (1927)