How to Harness the Crowd's Wisdom

The “wisdom of crowds” is a term first popularized in a book by James Surowiecki a few years ago. His book title was a riff off of a very popular 1841 book (yes, that's 1841, not 1941) by Charles MacKay, Extraordinary Popular Delusions and the Madness of Crowds.

Surowiecki’s main point was that a group of people with diverse opinions can often make uncannily accurate decisions and predictions – smarter in many cases than any single expert, no matter how smart. Orange-crop futures markets, for instance, predict Florida weather better than meteorologists. And just a few minutes after the 1986 explosion of the Challenger space shuttle, the stock market correctly zeroed in on Morton-Thiokol, maker of the frozen O-rings, even though it was several weeks before a team of engineers investigating the disaster figured it out.

Theoretically, at least, as social media produces “virtual crowds” of people more and more efficiently, we should be able to harness crowd wisdom to drive better decisions, and perhaps more efficient regulatory policies, as well.

But under some conditions crowds can also come to the wrong conclusions – often behaving in ways more similar to the mobs that MacKay portrayed in his own book. Stock market bubbles and lynch mobs come to mind.

Scott Page’s extremely insightful book, The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies (Princeton University Press, 2007), examines the logic behind crowd-based decisions, and is the subject of my Friday Book Share for this week.

The question Page tries to answer is: Under what specific conditions does a crowd make better decisions or more accurate predictions than individual experts do?

Using computer simulations and logical models, Page proves mathematically that the collective ability of a crowd (that is, its ability to solve a problem effectively or to make an accurate prediction) is made up of both the individual abilities of the crowd’s members and the diversity of their thinking. Stated another way, a diverse set of perspectives on how to solve a problem is just as important as having smart experts trying to solve it.

Diversity of ideas, of course, is one reason why technological progress continues to accelerate. The more ideas we share among ourselves – the more diverse perspectives have to be reconciled – the faster innovation and economic progress occur in general. Studies have shown that cities, because they tend to concentrate many people together, generate higher rates of innovation, as people run into each other, socialize, and exchange ideas. Economists estimate, in fact, that doubling a city’s size all by itself will increase average worker productivity by between 6% and 20%.

A major finding of Scott Page’s research is that when a diverse crowd solves a problem, the solution is likely to be better than whatever solution any single expert in the crowd would have come up with. In other words, in Page’s words, “diversity trumps ability.” He says this principle applies whenever four specific conditions are met:

1. The problem is difficult;
2. The perspectives and heuristics of the problem solvers are diverse;
3. The set of problem solvers is chosen from a large group; and
4. The set of problem solvers is not too small.

The reason groups tend to make wise decisions under these conditions is that the individual errors of a crowd’s members tend to cancel each other out. Errors are largely random, so some people err in one direction while others err in the opposite direction, and after summing up the pluses and minuses of a lot of mistaken ideas, what remains – the collective wisdom of the crowd – will be whatever part of the answer is most accurate. That’s why the stock market could so quickly identify the likely culprit in the Challenger disaster, and it’s why a bookie’s odds are accurate enough on a regular basis to sustain his business.

Importantly, what Page has shown is that when these four conditions apply to a relatively large group of people working collectively to solve a complex problem, a randomly selected group is likely to outperform a group made up of the best individual experts! One reason for this, Page says, is that “[t]he best problem solvers tend to be similar; therefore, a collection of the best problem solvers performs little better than any one of them individually.” When even a few non-experts are introduced into the mix, however, with completely different ideas, the experts themselves will find their own views sharpened by these new perspectives.

This doesn’t mean that putting a poet in with a bunch of chemists will improve their ability to deal with a difficult chemistry issue. But mixing a physicist into the group might.

What the idea means for a business, Page says, is that “When picking two hundred employees from a pool of thousands, provided the people are all smart, we should…not necessarily rank people by some crude ability score and pick the best. We should seek out difference.”

There is an important caveat, however. Page draws a distinction between people having diverse perspectives on how to accomplish some task or solve some problem and diverse perspectives on what the objective of the solution is. He calls these “instrumental preferences” and “fundamental preferences.” If a group of problem solvers all agree on the objective, then having diverse ideas about how to achieve the objective is very useful. But if they are so diverse that they don’t even agree on the objective, then the decision-making is not likely to be very good.

My own take on this for a business is that if you have a strong culture – that is, if the employees and managers at your business are engaged in their work and fairly united in their thinking about what the purpose or mission of your company really is – then you can benefit immensely by harnessing a diverse set of employee perspectives on how to accomplish that mission.

And Page’s research also has lessons for using crowd wisdom either as an enforcement tool against bad-behaving companies, or even as backstop or replacement for complex government regulations in a variety of areas, which is the topic we'll discuss at SXSW.

Photo: KR MEDIA Productions/Shutterstock

-----------------

If you found this Friday Book Share interesting, then you may want to review my analyses of these additional books, on previous Fridays:

The Secret Life of Pronouns, by James Pennebaker

Being Wrong, by Kathryn Schulz

The Success Equation, by Michael Mauboussin

The Signal and the Noise, by Nate Silver

Firms of Endearment, by Sisodia, Wolfe, and Sheth

Moonwalking with Einstein, by Joshua Foer

The Most Human Human, by Brian Christian

Six Pixels of Separation, by Mitch Joel