The Modern Journalist

Book Review: The Wisdom of Crowds (Part 2)

Note: This is maybe the most important book I’ve read for my research. This review is going to be long and exhaustive. So, I’m breaking it up into multiple parts. You can read Part 1 here.

Surround Yourself with Overconfident Dissenters: “Overconfident people don’t do that. They tend to ignore public information and go on their gut. When they do so, they interrupt the signal that everyone else is getting. They make the public information less certain. And that encourages others to rely on themselves rather than just following everyone else.” (p 61-65)

Notes: Another really important discussion is that even if you don’t surround yourself with diverse opinions, it’s important to have people who openly dissent and question decisions. Surowiecki makes the case later in the book that even if the overconfident are wrong, they force the decision makers to clarify their thinking. He cites studies on juries.

This also encourages people to amake incorrect guesses, something that a lack of diversity in opinion discourages — because the Wisdom of Crowds says that very few people will out-guess the majority and no one person will do that every time. In other words, a group of wrong answers can lead to the right answer.

Specialization, While Bad for Decisions, is Great for Productivity: “Specialization, as we’ve known since Adam Smith, tends to make people more productive and efficient. And it increases the scope and the diversity of opinions and information in the system (even if each of the individual person’s interests becomes more narrow.” (p71)

Notes: This particular idea is fundamental to my belief that newspapers are making the wrong, collective decision about their workers. They are encourages a more skills, less specialization training methodology. They are moving away from the specialized reporter and towards a jack-of-no-trades, which will demoralize the workforce (how can I keep up with all the new technology) and take away from the reporter’s fundamental job, information gathering.

To forego specialization in favor of a multitude of diverse skills is as bad as having one specialized skill making all the decisions.

Humans Care That Rewards Seem Fair: One very important point about using markets to solve problems is that people need to have a perception that rewards are fair, which in the U.S. means that higher rewards are based upon merit — even if that isn’t the case. The perception needs to be there. Otherwise, markets crash and go away. (pp108-126)

People Fall into 3 Categories — Free Riders, Altruists and Conditional Consenters: “Twenty-fiver percent or so are selfish and always free ride…A small minority are altruists, who contribute heavily to the public pot from the get-go and continue to do so even as others free ride. The biggest group, though, are the conditional consentors.” (pp138-142)

Notes: This section really lays out what you need to do to build not only a thriving community, but also how you can make decisions. I am going to focus on the community. To create a successful community, you need to: reveal contributions from the group and set up punishments for those who don’t contribute which is meted out by the group.

This sets up a trust system: with each other, with the organization and with the enforcement system.

Open Knowledge and Sharing Promotes Commerce in Many Forms: Developing solutions — and more importantly — information that leads to solutions is more beneficial if there is a commodity market for that information. Currency is less important than credit in terms of solving problems. (pp158-172)

Notes: I loved this section about how the scientific community worked — in weeks — to discover how SARS operated. The point: if you open up your data to everyone, even when you duplicate work, you get a better set of information because people come at things from different views. You need to manage the data sharing — there needs to be a coordinating group — but ultimately, decentralizing the data allows for a better understanding of what that data means.

What continually struck me about this idea was two-fold: opening up the backend of your systems to allow people to develop on it, thus changing the way your data is distributed, is important; and the days of editors sitting in a room to determine what is news is really antiquated and doomed to failure.

Groups Led by One Figure Tend to Deliver Information the Figure Wants: “In that sense, the team succumbed to what psychologists call ‘confirmation bias,’ which causes decision makers to unconsciously seek those bits of information that confirm their underlying intuitions.” (pp177-178)

and

“That’s another reason why a popular position tends to become more popular in teh course of deliberations: it has more potential champions to begin with.” (p187)

Notes: The over-riding theme of this point is that if you don’t objectively approach a problem, you will end up with an answer you want — not because you are smarter, but because you have set up the parameters of the discussion that encourages people to — through social pressure — make sure they deliver information that won’t adversely affect their standing within the group.

Surowiecki continually emphasises that the science of crowds suggests that you must deliver as little data as possible to everyone in a factual way, then make sure they aren’t influenced through discussion and finally return their findings at near the same time.

There are so many applications to this: the one that resonates with me is how the news is determined. It makes sense that you “cast as wide a net as possible” as he continually says to make sure that you are getting diverse views. That means getting outside the newsroom, away from the group think.

Finally: “The idea of the wisdom of the crowds is not that a group will always give you the right answer but that on average it will consistently come up with a better answer than any individual will provide.” (p235)

Notes: One corollary to that: if you deliver too much information in a way that tends to slant towards an answer, you will not get an accurate reading from the crowd.