The Stupidity of Crowds
As many people know, and as James Surowiecki chronicled in his famous book, The Wisdom of Crowds, large numbers of ordinary people can often outperform experts. Lately, though, the idea has mushroomed into excessive enthusiasm for crowds.
Terms like open source and crowdsourcing have entered the lexicon and formed fanatically engaged advocates. There have also been impressive successes, like the Firefox internet browser the Apache web server project. However, much like people, crowds can be foolish as well as wise.
The ability to make distinctions between smart and dumb crowds can be the difference between a runaway success and unequivocal disaster.
Crowd Failures
Everybody knows the danger of an angry mob, but the stupidity of crowds is not limited to drunken rabble. Large numbers of highly intelligent people can exhibit behavior that is far more damaging than a few hours of vigorous rock throwing ever could.
Market Failures: From Tulip Mania to the Great Depression to the latest financial crises, markets have a way of getting out of control. During the boom, pundits find highly plausible reasons for the party to go on. New technology, globalization, whatever… Someone is always ready to give an explanation.
Asset values soar, fortunes are made and everybody is a genius. When the inevitable crash comes, pundits (often the very same people) are there again to explain how they knew it all along.
Moreover, the pundits tell us with a solemn and foreboding tone, things can be expected to worsen. Our best days are now forever behind us. Asset prices then become irrationally undervalued, at least until the next boom (which again, many will rush to retroactively predict).
The Millennium Bridge: In 2000, the Millennium Bridge opened in London with great fanfare. However, as soon as people started to walk across it started to sway wildly. What was supposed to have been a great celebration instantly turned into a nightmare.
What had happened was that that, for whatever reason, the bridge rocked a bit and people on the bridge tried to keep their balance. When they did that, they created momentum in the opposite direction. That of course, forced even more people to put their weight on the opposite foot.
On it went, from left to right, right to left. As people moved their weight to keep the balance the bridge continued to sway. The bridge was closed for two years until design modifications could be implemented that would correct for the stupidity of corwds.
The Winners Curse: Auctions often result in overvaluation. The best bid is the highest, not the closest to the actual value of the asset, so irrational behavior is encouraged. That’s the winner’s curse. It is especially prevalent in telecom spectrum auctions and high profile corporate acquisitions.
It seems that the dumbest crowds of all come with advanced degrees.
The Independence Condition
All three examples above have something important in common: interdependence. That’s what creates the feedback that makes a crowd dumb. Although we are individuals, we actually very rarely act independently. We plan our route to work to avoid traffic, try to be a team player at work, tend to watch movies that are popular, etc.
This is actually a very big problem, since virtually all of business statistics are built on the independence assumption of a bell curve. Human behavior is not like a coin flip, what one person does will influence the decision of others. George Soros calls this principle reflexivity, and he has made billions betting against it.
The essential condition of any wise crowd is that it remains truly an aggregation of independent actions. With an increasingly interconnected and codependent world, crowds are likely to get dumber, not smarter. As the stakes grow bigger, the potential repercussions become more even more daunting.
Guidelines for a Wise Crowd
Moderation: The best crowds, such as Wikipedia, Mozilla, etc. are actively moderated. Open source communities generate lots of ideas, but someone is still minding the store.
Install Circuit Breakers: After the stock market crash of 1987, stock exchanges created mechanisms to immediately stop trading if the market drops by a certain percentage. You can read more about circuit breakers here.
If you want to keep a crowd wise, protect it from its own stupidity. You need to have a plan to disrupt deleterious feedback loops.
Good Sense: If everybody else jumps off the Brooklyn Bridge, it doesn’t mean you have to. We usually know when the crowd is doing something stupid. Watch out for phases like, “It’ll be different this time” or “This is a new era” or (especially) “it’s a once in a lifetime opportunity.”
There is nothing magical about crowds, except they have the potential to absolve us of responsibility. Many who are jumping on the crowdsourcing bandwagon do so not independently, but by running with the crowd.
And the crowd is always something to bet against.
– Greg
While I think there’s definitely an issue with “crowds” and the point you raise, I wouldn’t dump the whole idea alltogether and consider reformulating “crowdsourcing” as “tribesourcing” following the idea behind social networks and how people engage in Tribes and their interconnections. Seth Godin put it very nicely in his TED talk : http://www.ted.com/talks/seth_godin_on_the_tribes_we_lead.html arguing around shared values and ideas.
Food for thought… but I personally think there’s tremendous value behind the notion of social networked tribes.
jhm
.-= jhmorin´s last blog ..iPad and DRM : here we go again, and the industry isn’t making any progress =-.
Jean Henry,
Thanks for sharing your perspective.
– Greg
Greg,
This is an insightful and realistic perspective on crowdsourcing. As with any method-based system, people tend to fall in love with the idea. Often that means forgetting that it’s just one means of many. Getting data from crowds is useful and often enlightening. But as you point out, it’s far from infallible.
Also, people tend to forget that the results are not like purely mathematical data such as mean averaging. There are many factors that can poison the well and go virtually undetected. And shouting “fire” in a crowded room does not mean there’s fire. It just means there’s someone willing to shout it.
Clients tend to love the idea of crowdsourcing because they think it gives them validation. In fact, it’s dangerous when the voice of the crowd drowns out everything else. It’s just another piece of information, not a silver bullet.
Another great post, Greg. Thanks.
John
.-= John Cavanaugh´s last blog ..Kenmore Rebranding – Risky Business =-.
John,
Thanks for your comment.
I couldn’t agree more. Crowdsourcing is an exciting concept, but like anything else, needs to be approached with good sense.
– Greg
The first thought that entered my mind was a quote from a Finance Professor I had in college (MANY YEARS AGO)…. Although unbeknown to him at the time, he was in reality, discussing crowdsourcing… Anyway, the quote he used time and again in class was “The Masses are Asses”. Now, masses regarding finance and investing are probably much different than masses using a product, providing feedback or generating code. The point is, given the circumstance, you can’t rely on the ‘wisdom’ of masses to be insightful, accurate or even meaningful. There is normal variation in any population ‘sample’. You can, however, leverage the collective tendencies from masses to provide insight into a specific problem, question or trend. Context is, as usual, paramount to effective results.
Thanks for the great perspectives… They are thought-provoking and usually lead to greater insight (at least for me).
Michael,
You make a great point about outsourcing judgement, to a crowd or anybody else. Utilizing collective wisdom can provide great insights, but in no way gives license to abandon good sense.
– Greg
Greg,
You hit on a topic that is in the forefront of my mind lately. (Actually started writing a post on this very subject yesterday.) Crowdsourcing does have a place within a social media context, but when crowdsourcing spills over into the realm of relying on it to create the tangible products that a business produces, that’s where it gets dangerous. In the business of product development, it creates a watered down mish mash of several (usually not so good ideas) to create a product that is supposed to appeal to everyone, but lacks the passion and innovation that comes from product generated by a truly creative design team. Crowdsourcing is also devaluing a lot of work going on in a variety of creative fields such not only in product design, but in graphic design, web design, writing….There are numerous sites out there now that hold “contests” for the masses to produce creative work – logos, content, websites, etc. This produces mass quantities of poor quality work, and apparently most small businesses don’t know the difference between doggy doo-doo and quality creative work. All they see is price. Low price = good. Crowdsourcing in this arena creates the Walmart-ing of business.
Cheryl
.-= Cheryl Andonian´s last blog ..The Cashier At Walgreens Is My Consultant =-.
Cheryl,
Great points! There is a similar example with cover testing in magazines. Everybody hates it and, as far as I know no one has actually tested it to see if it actually produces better sales, nevertheless it’s become a fairly standard practice.
– Greg
You your point Cheryl, I am also a writer, albeit professionally, new, however, to provide some balance to this discussion, I’ll offer a competing view. When you consider how the telephone has progressed over the years, into the internet age, and now we have smartphones, which are basically hand-held computers that make funny noises… would you say that the smartphone has devalued the telephone operator business? In reality, it has drastically changed it and automated it, however, I would not say it has devalued it; in many ways, it has enhanced the industry as a result of the newfound functionality and potential for future change.
As much as I completely agree with your points about getting a lot of noise regarding quality of work, I believe, that simply due to the enormous sample/population size with crowdsourcing, that there are undoubtedly, diamonds in the rough, that will bubble up as a result of the platform and ability for everyone to have a ‘voice’. (nice telephone analogy, huh?)
Third, I also believe that professionals (as you and others) are justifiably concerned, that your ‘turf’ has been violated and threatened by these so-called ‘masses’.. I see the exact opposite. Your job now simply changes; to 1. Recognize and find those ‘diamonds’ and 2. Use your professional experience and expertise to point out and legitimize the ‘quality’ concerns, where they exist, to your clients. There is tremendous potential to effectively and economically leverage the mindshare of millions of product testers and data analysts out there…
The industry, future applications and platforms will benefit as a result of today’s work with crowdsourcing. In the end, the ultimate customer or consumer benefits…. isn’t that a good thing and why we are all in business?
Just a few thoughts… Excellent topic and discussion… Kudos to you Greg for asking the question.
To me whether crowds are wise or stupid depends on how they are used and what for. Under certain circumstances results produced can be brilliant. Examples are successful open source projects like Firefox browser and Apache server which Greg’s article rightfully refers to.
So called agile project management methodologies (SCRUM) for software development stake on cross-functional teams. These teams achieve better results than a team of experts when working on some specific tasks that do not have a single right answer that can be deducted through analysis. Management writer Stephen Denning calls such tasks without single right answer “mistery” as oposed to “puzzles” – tasks that are better solved by experts. As software development is basically a process of discovering client requirements step by step, by closer and closer approximations, cross-functional teams work well in this area. Open source projects are also a good example and also innovative projects and new product development – when a company receives feedback from would-be customers.
All examples of good application of “crowds” I can think of though are cases of ORCHESTRATED work effort. Teams may be called self-organizing but there are in fact rules that they obey to. This is also true for open source projects and Wikipedia.
On the other hand there are examples when chaotic “crowds” produce negative results and fluctuations like market boom and bust cycles.
Using crowds is still new phenomenon and we need to learn how to benefit from it using it properly and avoiding pitfalls.
So, I agree with Greg that utilizing collective wisdom should be done carefully.
Stan,
Thanks for giving the software perspective. I would also add that the fact SCRUM stipulates cross functional teams helps preserve the independence condition.
Teams made up of people with different backgrounds and skills are more likely to bring a variety of perspectives than to fall into the kind “group think” that creates dumb crowds.
– Greg
My general thinking about crowd-sourcing is the old line “none of us is as stupid as all of us”. But I’m Italian and I exaggerate constantly. This was a great post and you made a lot of good points. The idea of moderation of the crowd and “circuit-breakers” are tempting but how does that happen exactly out in wild?
There is no stopping the momentum of turning work such as graphic design, writing, and other skills that have been the bread and butter of agency revenues into a commodity. Every product or skill set has a life cycle and this work is going to die quickly.
I think the good news is that if you really know your stuff and can consistently raise the level of your knowledge and expertise and remain relevant. Thoughts?
.-= Joanne Cirillo´s last blog ..Book Review: Putting the Public Back in Public Relations =-.
Thanks again for some great insight, Greg. We are confronted with this on a daily basis- when the mktg director needs opinion from superiors and a host of minions alike. It almost always ends up turning a great idea numb. Crowds will rarely move forward- just try to perfect “where they are”.
The wisdom of crowds is what happened to Detroit. They were moderated by the finance department.
Good points. Thanks Joanne.
– Greg
Thanks for adding your perspective, Bill.
– Greg
Hey Greg, Good topic as usual! It dates me to say this but a hootenanny ain’t a concert. Sing a-longs are fun but aren’t known for producing great music. The wisdom of crowds as I see it in action (I’m an elected member of our local government) is the central tendency of fairly diverse groups with multiple sources of competing information, because the dumb extremes tend to offset each other. But there is a famous picture of the grinning idiot faces of a lynch mob which serves as one of my reminders of the evil potential of crowds and the constant mortal and moral danger of joining a crowd. We are mistaken if we think we can’t be sucked in by stupid crowds.
Thanks, Roger.
Like most things, crowds require good sense and judgment.
– Greg
Yet another area of current fascination Greg, one that has clearly started to engage people to consider for themselves and voice their perspective – isn’t this a forum for open source disucssion and debate in an un-plugged fashion, where people (both intelligent and otherwise) get to provide their thoughts?
I couldn’t agree more with your points relating to moderation and good sense, they are logical and, well common sense. However, like a couple of others (Jean and Michael) above, I’m not convinced that all crowds are dumb and therefore that crowdsourcing should be written off as a means for generating discussion, gathering insight and occasionally coming across that gem of an idea. There may not be a single idea in any of it, but perhaps there are sufficient sparks that lead smart people/moderators down a path to product/service innovation, evolution and even revolution at some point.
I think the other intersting point about this is that technology (as per Michael’s telephone analogy) has (and continues to) raise the bar when it comes to people connecting and that mass individuals come together to discuss, debate, share, comment, create across common and diverse subjects. Personally, I think this is pretty cool.
Unlike Cheryl, I don’t believe the ‘crowd’ removes the role of innovators, creators, designers, developers, shapers. Instead, it think it adds to the roles enabling mass individual feedback, thoughts, considerations, ideas. It’s up to the intelligent ‘crowds’ to dissect what’s being offered up, taking the bits that are helpful/useful and evolving them into a brilliant idea/opportunity.
What’s more I think they are here to stay, rightly or wrongly. If we consider the growing power of Mumsnet in the UK, for example, and the influence this ‘crowd’ is having on upcoming federal elections – then is it the stupidity of the crowd or of the individuals trying to ‘manipulate’ the crowd into just accepting?
Dick,
Of course you’re right, there are smart crowds as well as dumb. The important thing is to know the difference.
– Greg
Delta’s Sky magazine had an interesting and sobering article on crowdsourcing last month:
Crowdsourcing
Welcome to the Web-based movement that harnesses the ideas and opinions of the masses. The payoff for the big idea may never be the same again.
http://msp.imirus.com/Mpowered/imirus.jsp?volume=ds10&issue=2&page=72
Frank,
Thanks for the link. Somehow, I think creative agencies will survive.
– Greg
Great article and interesting comments.
I know many companies would like to get product ideas via crowdsourcing. What’s striking to me is that in focus group work, generally the crowd doesn’t have the good ideas – because they don’t know what’s possible.
If you’re a good company at the head of the pack, your teams can see forward to things that your consumers or purchasers can’t. So crowds are potentially great for looking at current situations and backwards. But, looking ahead … as you note they tend to follow the crowd.
Another vagary of the crowd is that the crowd doesn’t have your company’s good will in mind. And that means that (a) the make-up of the crowd that gathers is out of your control and can lead you far astray and that (b) there is a crowd selfishness to guard against.
Great work. Loving your blog…
…Doug Garnett
Doeg,
Good points. Thanks.
Crowds have their place, but just as with any other method, it’s important to think things through and not just do it because it happens to be something journalists like to write about.
– Greg
Correct me if I am wrong, Scrum’s strength is to establish working teams that are NOT crossfunction in the sense that it requires a team of developers ( coders ) that have slight aspects of different disciplines ie all generalists.
If we move along the sliding scale of ‘different’ backgrounds, I would conclude that the opportunity of having QA and BSA’s is a non-starter within Scrum because these would be specialists. That being said, it would imply that the closer proximity of ‘group think’ within a Scrum teams would occur due to its like minded assembly. “That widget is awesome!”. – but is it useful?
Greg;
Great post – thanks again for making me think of things other than the daily toil.
Paraphrasing what I think has been said above: crowds can be a Massive Ass, or an amazing hive of intelligence, and it’s good to be able to tell the difference.
Reading your examples, Greg, and many of the comments, the good crowd-sourcing examples were mostly virtual crowds, while many of the not-so-happy results involved people meeting in person. It seems to make sense – alone in front of the computer, you can’t feel the heat of the angry mob or get caught up in its fear and excitement, or its stupidity. Your safe, easily switched–off virtual layer insulates you and cools your emotions. You read something one of the crowd says; lean back, stroke your chin; mutter, “you’re full of it,” or pound the desk in admiration or anger. And then, emotions vented in a way you could not do in a face-to-face meeting, you type a more measured response.
I’m not saying that one can’t have good crowds made up of real people, or that there aren’t bad ones composed of virtual meanies. Specialist task teams really do work and there are some evil virtual groups out there. It’s just that when used for “good purposes” the advantages of groupthink are there for negligible cost. All that collective brainpower without the crowd as Massive Ass.
Fred Brooks, of The Mythical Man Month fame, would be astonished. A way to run a huge software development team and not lose most of the assembled skill to overhead. If you’re like me and think technology is now our only hope for saving the planet and ourselves (along with a huge mindshift, but that’s another story), then huge groupthink projects may be one way to a cleaner future. It’s not a perpetual-motion smoke and mirrors trick, but it seems you really can add mental power to virtual teams at close to one for one efficiencies. So count me in for these virtual-think-tanks on steroids. It’s not perfection in terms of what our heads can achieve, but it is a way to inch us closer towards it.
.-= Eric Goldman´s last blog ..B2B Print Publishing: Looking for a Lifeboat =-.
Eric.
Thanks for your comment. Creating some separation does indeed help to fend off the dark side of crowds. Whether that involves circuit breakers on stock exchanges or moderators that are able to pull the plug. We have more control over the electronic world than the real one.
However it’s important to be careful, electronic markets crash the same way that real life ones do.
– Greg
Hi Michael,
I like your point of view. “The cream always rises to the top.”
Lisa
Hi Greg,
I like Michael Cerkas’s point of view along with yours.
As far as being recognized for talent verses price I do believe “the cream always rises to the top” and of couse “there is a big difference between chicken do-do and chicken salad” As an old retailer the rule was always “sell to the masses not the asses.”
In a very broad arena, I don’t think its changed much.
As always, thank you for your insight.
Lisa
Lisa,
As alway, thanks for your input.
– Greg
It is always worthwhile blogging on a trend that has got out of hand but I think that your article is selling Surowiecki short. His book is at pains to explain the factors that ensure that crowds are wise and these include the need for divergent opinions, a degree of market knowledge etc.
Furthermore I’m not sure that the Millennium Bridge example is a case of stupidity – it is more a natural instinct to stay balanced that created unforeseen ‘waves’ in the bridge structure. Of course some of the other actions that you mention (jumping off the Brooklyn Bridge) are brilliantly explained in the book by Robert Cialdini – Influence : The Psychology of Persuasion. Mass suicide as seen in the Jonestown incident when 909 members of the Jim Jones cult in Guyana quietly queued up to drink poison can be explained by the concept of Social Proof. No psychoogist would term these people stupid, they felt that they were behaving rationally because they were following the actions of others in an environment where they were unsure of their bearings and how to behave.
The Wisdom of Crowds is an inspiring book to me not so much because it identifies how markets behave but in that it has a democratic urge to value divergent opinions and voices. This should be a wake up call for all CEOs who think that only those in the C-Suite can make credible decisions. Engaging with employees may well deliver far better results…
Alan,
You are right about Surowiecki’s book. He does indeed give both sides to the story and offers good examples where crowds go wrong. I apologize to you and Mr. Surowiecki if I wasn’t clear on that point.
As for the Millennium Bridge, the people weren’t being stupid, but the crowd was and I think that is the key point. Lots of perfectly rational people can create a stupid crowd.
Thanks for the book suggestion. I’ll check it out.
And of course, thanks for a great comment.
– Greg
Good article Greg.
I always find it interesting to click through your links and learn more.
I must admit, I had not heard of the phrase “winner’s curse”. Notice in the Wikipedia article on this subject it cites Pay per Click advertising as a classic example of the ‘winner’s curse’. That would be an interesting one to flesh out.
Chris,
I’m not sure it’s such a classic example, but I can see how it might happen with some keywords. However, there is so much “long tail” arbitrage going on I doubt it’s significant.
To the best of my knowledge, the term is usually used in connection with bidding wars for companies which almost always result in the winner losing a substantial amount of money.
– Greg
Ouch, my brain hurts. Anyway I just thought this up unless you have seen it somewhere before because you have obviously read and know more than me. And that is that we seem to be a chase-after society. Rather than…er, I don’t know what the opposite would be. I’d apply this concept to disasters as well; esp. in the aeronautics industries. Chase after solutions more than preventative-type ideas/things.
Interesting idea. Thanks Robin.
– Greg