Wisdom of Crowds

posted by Jason Kottke   Jul 14, 2004

?The problem with the global village is all the global village idiots.?
? Paul Ginsparg

?You don?t do good software design by committee.?
? Donald Norman

?There?s no justice like angry-mob justice.?
? Principal Seymour Skinner

?A person is smart. People are stupid.?
? Agent K

The wisdom of crowds you say? As Surowiecki explains, yes, but only under the right conditions. In order for a crowd to be smart, he says it needs to satisfy four conditions:

1. Diversity. A group with many different points of view will make better decisions than one where everyone knows the same information. Think multi-disciplinary teams building Web sites?programmers, designers, biz dev, QA folks, end users, and copywriters all contributing to the process, each has a unique view of what the final product should be. Contrast that with, say, the President of the US and his Cabinet.

2. Independence. ?People?s opinions are not determined by those around them.? AKA, avoiding the circular mill problem.

3. Decentralization. ?Power does not fully reside in one central location, and many of the important decisions are made by individuals based on their own local and specific knowledge rather than by an omniscient or farseeing planner.? The open source software development process is an example of effect decentralization in action.

4. Aggregation. You need some way of determining the group?s answer from the individual responses of its members. The evils of design by committee are due in part to the lack of correct aggregation of information. A better way to harness a group for the purpose of designing something would be for the group?s opinion to be aggregated by an individual who is skilled at incorporating differing viewpoints into a single shared vision and for everyone in the group to be aware of that process (good managers do this). Aggregation seems to be the most tricky of the four conditions to satisfy because there are so many different ways to aggregate opinion, not all of which are right for a given situation.

Satisfy those four conditions and you?ve hopefully cancelled out some of the error involved in all decision making:

If you ask a large enough group of diverse, independent people to make a prediciton or estimate a probability, and then everage those estimates, the errors of each of them makes in coming up with an answer will cancel themselves out. Each person?s guess, you might say, has two components: information and error. Subtract the error, and you?re left with the information.

There?s more info on the book at the Wisdom of Crowds Web site and in various tangential articles Surowiecki?s written:

- Smarter than the CEO

- Interview with Bill James

- Blame Iacocca - How the former Chrysler CEO caused the corporate scandals

- Search and Destroy (on Google bombs)

- The Pipeline Problem (drug companies)

- Hail to the Geek (government and information flow)

- Going Dutch (IPOs)

- The Coup De Grasso (fairness in business)

- Open Wide (movies and ?non-informative information cascades?)

Reader comments

RichardJul 13, 2004 at 1:59AM

“We’ve given the word mob a bad name.” Dr. Hibbert

Kabir SehgalJul 13, 2004 at 4:08AM

This sounds like a revealing read. A step away from the tomes on “independent leadership,” leadership by group is the direct democracy that some yearn for. Yet others detest.

I’ve wondered for some time now on the deterioration of groups. In other words, we know what brings people together…a common goal or interest. There is a shared interest. And in leading a business, there is shared capital, stakes, and so forth.

But what when the group breaks down and ceases to exist? Are there any trends that indicate why things don’t continue. On the surface, we could point to the opposite of formation…disagreement. But I haven’t found any research that delves into why groups breakdown and cease to operate. U.S. Steel and Standard Oil faced trustbusters in the early 20th century, thus law compelled these companies to cease and desist. The Beatles, however, and an internal squabble followed by tragic death. But is there a common thread here? Is there any commonality in demise?

Perhaps Surowiecki will look at this topic in a future book: not leadership by committee, but, dare I say it, “Death by committee?”

RemmyJul 13, 2004 at 4:20AM

Agent K, actually. Which is funny, because he was there first. But makes sense - Zed is numero uno.

Now to intelligently aggregate our feedback into article corrections…

jakeJul 13, 2004 at 9:53AM

i love these book summary posts- it’s like the kottke cliffnote series! i think this should be recurring- like we make jason read two or three books a week and post a four paragraph summary on each. but it can’t just be non-fiction books.

kowgurlJul 13, 2004 at 10:03AM

Sruowiecku has a lot of flawed arguments— one of them being his reference to the Who Wants to be a Millionaire statistics. Coreect answers for “Poll the audience” stats are better than ‘Phone a friend’. Well, duh. People ask the easier, more pop culture questions of the audience and use the phone a friend for the really difficult stuff.

Runky FunkyJul 13, 2004 at 11:01AM

Is this just another outline of an ideal (albeit one that I think worth striving for?)

Don’t the conditions he supposes seem to be things we are NOT naturally inclined towards:

1. Diversity: history and experience suggest that people tend toward like people not different people. This has good results in some cases, and bad in others. Nevertheless, people still herd around like people.

2. Independence: Really, “no man is an island,” right? So this only goes so far before it bcomes chaos. Like it or not, we’re only independent up to a point… right?

3. Decentralization: I have had my share of experience on committees and decentralized authority plays a key role in the most successfull ones. But the buck always winds up having to stop somewhere…

4. Aggregation: so someone still has to be the aggregator, or final arbiter, right? There goes independence!?

I’ve read some of his articles, not the book yet, though. I get the point and value of much of what he says, but I’m left feeling like it’s not the whole picture. It strikes me as a sort of utopian view of how things could (or ought?) to be, but without the teeth to make a difference.

My $.02

dowingbaJul 15, 2004 at 2:22AM

like we make jason read two or three books a week and post a four paragraph summary on each.

Dude, we should get him to read 52 magazines a year, too! Oh wait…

karlJul 15, 2004 at 10:18AM

There is no “collective wisdom”. There is lots of individual wisdom interacting in clever ways to achieve more than the individuals with the wisdom.

Point being, a group cannot actually hold a concept or idea -as a group-, the idea needs to be held independently by the members of the group, each in their own minds.

JoshJul 15, 2004 at 10:21AM

Not having read the book, I don’t have any right to an opinion at all — but nothing I’ve read by Surowiecki online has made much sense to me.

The key text on group-think is still probably Thomas Kuhn’s “The Origins of Scientific Revolutions,” which describes how crowds with all of the advantages Surowiecki requires still get trapped into a single gestalt, even when they think very individually. Making predictions or guessing probabilities are not the same as decision-making or critical thinking. I’d be curious to hear from someone who’s read both books and can compare.

jkottkeJul 15, 2004 at 10:43AM

To be fair and to set the record straight for all those who haven’t read the book, Surowiecki is certainly not suggesting that groups of people make wise decisions under all possible circumstances. Big chunks of the book are spent discussing when group decision making breaks down as well as situations when you’d think the group might be smarter but isn’t. WoC isn’t a condemnation of individual effort or genius, it’s just a way of thinking about some common situations where collective thinking might apply. Probably not a good idea to base too much of your impression of the book on my short summary.

Don’t the conditions he supposes seem to be things we are NOT naturally inclined towards

Perhaps, which may be why smart group thinking — aside from simple polling (i.e. guessing how many jelly beans are in a jar) — is so difficult to foster.

VictorJul 15, 2004 at 1:32PM

Interesting: “You need some way of determining the group’s answer from the individual responses of its members”

This subject alone is the branch of a type of economics called “social choice theory” developed by a berilliant man (and my honors thesis advisor, so bias beware) named Ken Arrow. Specifically he looked at voting, as it is supposed to be a way of aggregating opinion. It turns out that, given certain reasonable cireteria, it is impossible to do it mathematically perfectly, but you can try to get close. Our current voting system; basically the worst.

The reason for this is, plainly, Ralph Nader. I won’t say anything about his politics, because that is immaterial. What matters is that if he is running, he weakens the chance of Kerry. I know most peoplee voting for Nader would have preferences in this order: Nader, Kerry, a monky, bush. But by voting for Nader they lower the chances of getting anyone they want. It would be technically *more rational* (again not making value judgements, this is game theory) to vote for the man more likely to get elected, because you prefer a second choice to the 4th. The perfect voting system should take into account all of people’s preferences so the *never* have an incentive to lie about their peference (ie choose second place rather than first) in order to better their chances.

In order to learn this simple requirement of aggregating, I suggest checking out anything by, or about, Dr. Arrow. Turns out he won a Nobel Prize for this stuff.

essJul 15, 2004 at 3:03PM

Jason - originally planned to compliment your on not misrepsreseWas going to compliment our host for doing such a lovely job of not misrepresenting Surowiecki, and then I see that some commenter have jumped to the same half-assed conclusions a lot of critics and reviewers have - are we seeing a function of large discussions - dumbing toward the mean - in this small dialogue?

Kowgurl - Where’s your data on why “Who Wants to be A Millionaire” contestants choose poll the audience over phone a friend? Are you privy to exit interviews (unreliable as they may be…) Having access to Surowiecki’s work, I see no flaw in his arguments.

Jeff, having read both I’d say that Kuhn and Surowiecki both see the same problems in trying to create overarching rules that have what Runky Funky calls teeth. In my mind, Kuhn is a historian so his approach is hard to compare with Surowiecki’s -although both do want us to attend to and constantly examine our decision-making strategies.

My sense of Surowiecki’s view is this: we need to understand where we can get information and how we work as a group to do a better job of decision-making and critical thinking. He certainly doesn’t suggest that you can get the best answer to any old question by wandering around the mall and quizzing random yahoos.

(Not arguing that Kuhn should be boxed up with historians - just noting how I read him.)

In case you were wondering, Surowiecki is cute, in a quite nerdly way.

Victor, my problem with a lot of that work on voting issues is that is tends to start of with the election is question. Can you recommend anything that discusses history and context?

Fun quiz - how many consumers of fine hypertext products consider Kottke an expert on certain things but have tempered that opinion on occassion when the post is followed by a comments section?

cigarsJul 16, 2004 at 1:23AM

Hey it was good….the public as a mass tends to follow what they believe the majority to be.

kowgurlJul 16, 2004 at 11:55AM

ess: I don’t have any super-secret WWTBAM? stats, but watching the godforsaken show, you can easily see the way the questions fall in to lifeline categories. ‘Phone a Friend’ is for the hard intellectual stuff, ‘Poll the Audience’ is for when you don’t know which backstreet boy is the correct answer. Correct answers stats on the show in general reflect alot of things, not just mob vs individual intelligence.

essJul 16, 2004 at 2:37PM

“….you can easily see the way the questions fall in to lifeline categories” Like the way many contestants end up with only one lifeline left? There are many cases where a contestant knows that help is needed, but only has one option. Granted, only a contestant who has survived at least two questions can reach that pointl…

Stats usually aren’t secret. The trick is to cull a good data set from all the noise.

Surowieck does an excellent job with his real-world examples. If you would read his book I suspect you’d give him more credit.

He also address, slightly, cigars’ point about people wanting to be with the tribe cuz that’s the space were mob-think gets ugly, or at least messy.

ChrisJul 17, 2004 at 9:08PM

I’m, going to have to bookmark this discussion because I’ve got so much on right now but it’s a fascinating discussion. I’m going to put Surowieck on my wish list as well.

Just to say I guess I’ve worked with groups all my working life. I started out as a teacher and became passionate about co-operative learning. So when I came across The Social Life of Information by John Seely Brown I was engrossed. “Learning is a social phenomenom.” People work best in groups. Imagine trying to build a space shuttle by yourself. (Hey I know NASA’s a pretty imperfect oganisation - just an example.)

Now I work with project teams on their decision making processes. I will be interested to compare Surowieck with the work of Wilfred Bion (immediate post WW11 - probably the father of our modern ounderstanding of group processes.)

JoeJul 18, 2004 at 1:23PM

I will be checking out this fine book once I get done with all this school work I have to do. This seems like something I have been telling people for years. The avrage person is prety dumb I have never had a way to explane it without seeming like I am egotistical or snobey I don’t mean it as an insult but if you take a regulor person out of it’s enviorment and put it into a new one they will be stupid. If you put them into an envierment where they know what is going on then they might be the smartest person in the room. Intelegance needs context. If I where to walk into a farmers conferance I would have no idea what the hell is going on. Drop me into a internet confrence I might be one of the smarter folks in there.

sory about spelling in a hury

This thread is closed to new comments. Thanks to everyone who responded.