Site   Web

March 6, 2009

Webstock 09 – The Wisdom of Communities

Live blogging the Wisdom of Communities presentation at Webstock 09 by Derek Powazek, author of Design for Community: The Art of Connecting Real People in Virtual Places.

Derek’s presentation is based on the Wisdom of Crowds. What is the wisdom of crowds and how should it be applied to the web?

Elements of Wise Crowds include:

– Diversity

– Independence

– Decentralization

– Aggregation

Take the Wisdom of Crowds mentality online by:

1) Giving people small simple tasks

One Wisdom of Crowds technique example is the site Hot or Not where users rate singles based on appearance. Another example is Threadless – a crowdsourced tshirt store where people uploads tee designs and the crowd votes to decide whether or not they get printed. Voting data isn’t displayed until the poll ends, which is fairer. An example of Wisdom of Crowds that didn’t work is Assignment Zero. The idea was to ask people to write news stories but it was too hard, not a simple task and it resulted in collective freakout.

2) Giving the tasks to large diverse groups

Large diverse groups are best for Wisdom of Crowds . Failures occur though, for example the 2007 Chevrolet Tahoe was featured on Donald Trump’s TV series, The Apprentice, where the show sponsored an online contest in which anyone could create a commercial for the new Tahoe by entering text captions into the provided video clips. The prize was for the winner’s ad to air on national television. What could possibly go wrong? EVERYTHING.

The viral marketing campaign backfired when hundreds of environmentally conscious parodies flooded YouTube critiquing the vehicle for its low gas mileage.

3) Design for selfishness

It’s the self-centeredness of our individual thinking that makes a group diverse and… smart.

4) Aggregate the results

Result aggregation is important. The game outcomes change commitment in people. Games are the result of clever web apps. The way to win the game is to do something that benefits the community. Points attributed to individuals create karma whores e.g. the Slashdot (“me too” posts).

Favrd makes a game out of Twitter. It uses favorites on Twitter to aggregate results. Derek’s site is It uses crowd sourcing for content but is basically a storytelling site.¬† The site has morphed into a series of live storytelling events and several books.

The Flickr Top 500 photo thing created a game. Therefore it attracted “karma whores” and bad behavior including group spamming etc. Therefore, Flickr turned it into Photos of Interestingness over Last 7 Days

In your web applications, by all means use voting algorithms. BUT popularity does NOT have to rule.

What About the Trolls? Trolls derail the game. Ignore them and take away their power.

Seeing Things Рwhen the brain fails to have context, it fills in the blanks. Use this knowledge with your web sites and apps. For example the Cloud Patterns  experiment used an In Control group vs an Out of Control group and showed them a series of identical cloud patterns. The Out of Control group were consistently told their answers were wrong while the In Control group were given positive feedback on their suggestions. The Out of Control group began to see patterns in the clouds as a way to take control back.

Use such science to make your web apps and interfaces awesome – give people ownership and make them feel in control. Let them share their personal stories.