Review of The Wisdom of Crowds

Title:

The Wisdom of Crowds

Author:

James Surowiecki

Publisher:

Doubleday

Date:

2004

ISBN:

0-385-50386-5

Pages:

320

Price:

$24.95

Reviewed by Nick Christenson, npc@lvrevealed.com

August 20, 2009

Imagine there is a large jar in front of you filled with beans. By which method will you obtain a better estimate of the number of beans in that jar, by asking a single smart person or by taking the average of the guesses by a large cross-section of people? In The Wisdom of Crowds, James Surowiecki explains why it is that for many types of problems a large group of randomly selected individuals can provide a better answer than a small number of experts.

This claim contradicts many of our instincts. We are disposed to believe experts rather than the unwashed masses, but it turns out that often these instincts will lead us astray. The notion that the "crowd" may be so good at predictions explains why so few stock investors can beat the market, and why it's so hard to make a living betting sports. While I'm not a believer in the "efficient-market hypothesis", in that I do believe it's possible for a someone to consistently pick winners in many markets, the market deserves a great deal of respect as a predictive force.

These book reviews are supposed to be about gambling, so what can a gambler learn from The Wisdom of Crowds? The book is most applicable to sports betting, especially as it relates to market handicapping. Market sports bettors believe that the market generally does a better job of setting prices than individual handicappers, so (with a few exceptions), the later the line, the sharper it gets. Then they want to bet into lines that are behind the market. Surowiecki's book provides some justification for why this technique works.

I'm not sure I buy every case study that Surowiecki uses as evidence of his thesis. Counting beans by consensus seems far more plausible to me than using these methods to find lost submarines. Even after reading the book, I'd still rather have one good doctor diagnose what's wrong with me than trust the consensus of a hundred people who never went to medical school. (Of course I could easily believe that obtaining the consensus of a hundred good doctors would be better than consulting one.) The author does discuss situations in which markets can break down, but I wish he spent more time discussing the distinctions between cases that are well addressed by market-based prediction and those where such picks fail.

The crowd is not always wise, and The Wisdom of Crowds examines several factors that can cause the consensus to be led astray. Additionally, in reading the book I came up with several interesting cases that the author doesn't cover. Here's one phenomenon related to sports betting that I feel deserves further examination: Imagine using three groups of people to set lines for sporting events. The first group is random sports fans, the second group is a random sampling of sports bettors, and the third group is the same group of sports bettors, except their votes are weighted by their bet size. Who is going to make the best lines? I feel confident the sports fans will make the worst lines, and I expect that the weighted sports bettors will make the best ones. I also feel confident that any good sports bettor could crush the sports fans' consensus line.

A single expert at guessing the number of beans in a jar may not be expected to do as well as the average guess of a large crowd, and I expect that an average of skilled odds makers will set better lines than any individual linesmaker, but I don't expect sports fans to do better than an expert linesmaker. Why is this? I've got some ideas as to why this might be the case, but I don't have anything definitive or empirical, and I think it would be illuminating to examine this issue in more depth.

The Wisdom of Crowds doesn't make recommendations on playing deep-stack pot-limit hold'em, and it has no advice on avoiding unwanted attention at the blackjack table. However, I do think it's a valuable book in general, with some remote applicability for people interested in sports or pari-mutual betting. Mostly, I found the book to be interesting and worthwhile even beyond its gambling applications. It's a stretch to call it a gambling book, but I definitely recommend it as likely to be interesting to those with a gambling mind set.

Capsule:

The Wisdom of Crowds has only a fairly weak connection to gambling, but it does use sports betting markets in some of its examples, and there are some interesting connections. I can't really justify recommending it solely on its value as a gambling treatise, but there are some connections, and I do think it's an excellent book that's likely to be fascinating to those who read books on gambling.