Reality Is Broken: Why Games Make Us Better and How They Can Change the World (33 page)

Just three days into the game, it was clear that the crowdsourcing effort was an unprecedented success. More than 20,000 players had already analyzed more than 170,000 electronic documents. Michael Andersen, a member of the Nieman Journalism Lab at Harvard University and an expert on Internet journalism, reported at the time: “Journalism has been crowdsourced before, but it’s the scale of the
Guardian
’s project—170,000 documents reviewed in the first 80 hours, thanks to a visitor participation rate of 56 percent—that’s breathtaking.”
⁴

A visitor participation rate measures the percentage of visitors who sign up and make a contribution to a network. A rate of 56 percent for
any
crowdsourced project was unheard of previously. (By comparison, roughly 4.6 percent of visitors to Wikipedia make a contribution to the online encyclopedia.)
⁵
It’s
especially
breathtaking considering the mind-numbingly tedious nature of the actual accounting work being performed.

So what accounted for this unprecedented participation in a citizen journalism project? According to Willison, it all boiled down to rewarding participants in the right way: with the emotional rewards of a good game.

“The number one lesson from this project: Make it feel like a game,” Willison said in an interview with the Nieman Journalism Lab. “Any time that you’re trying to get people to give you stuff, to do stuff for you, the most important thing is that people know that what they’re doing is having an effect. If you’re not giving people the ‘I rock’ vibe, you’re not getting people to stick around.”

The “I rock” vibe is another way of talking about classic game rewards, such as having a clear sense of purpose, making an obvious impact, making continuous progress, enjoying a good chance of success, and experiencing plenty of fiero moments. The Investigate Your MP’s Expenses project featured all of these emotional rewards, in droves.

The game interface made it easy to take action and see your impact right away. When you examined a document, you had a panel of bright, shiny buttons to press depending on what you’d found. First, you’d decide what kind of document you were looking at: a claim form, proof (a receipt, invoice, or purchase order), a blank page, or “something we haven’t thought of.” Then you’d determine the level of interest of the document: “Interesting,” “Not interesting,” or “Investigate this! I want to know more.” When you’d made your selection, the button lit up, giving you a satisfying feeling of productivity, even if all you’d found was a blank page that wasn’t very interesting. And there was always a real hope of success: the promise of finding the next “duck pond” to keep you working quickly through the flow of documents.

A real-time activity feed showed the names of players logged in recently and the actions they’d taken in the game. This feed made the site feel social. Even though you were not directly interacting with other players, you were copresent with them on the site and sharing the same experience. There was also a series of top contributor lists, for the previous forty-eight hours as well as for all time, to motivate both short-term and long-term participation. And to celebrate successful participation, as well as sheer volume of participation, there was also a “best individual discoveries” page that identified key findings from individual players. Some of these discoveries were over-the-top luxuries offensive to one’s sense of propriety: a £240 giraffe print or a £225 fountain pen, for example. Others were mathematical errors or inconsistencies suggesting individuals were reimbursed more than they were owed. As one player noted, “Bad math on page 29 of an invoice from MP Denis MacShane, who claimed £1,730 worth of reimbursement, when the sum of those items listed was only £1,480.”

But perhaps most importantly, the website also featured a section labeled “Data: What we’ve learned from your work so far.” This page put the individual players’ efforts into a much bigger context—and guaranteed that contributors would see the real results of their efforts. Some of the key results of the game included these findings:

And the game detailed:

Bringing these numbers to light helped clarify the true extent of the crisis: a far more pervasive culture of extravagant personal reimbursement than originally suspected.

So what did the players accomplish? Real political results. At least twenty-eight MPs resigned or declared their intention to do so at the end of their term, and by early 2010 criminal proceedings against four MPs investigated by the players were under way. New expense codes are being written, and old codes are being enforced more vigorously. Most concretely, hundreds of MPs were ordered to repay a total of £1.12 million.
⁶

It’s not all the doing of the
Guardian
’s gamers, of course. But without a doubt, the game played a crucial role. The citizen journalists helped put significant political pressure on the British government by keeping the scandal in the news. The longer the game continued, the more public momentum built to force major policy reform.

Investigate Your MP’s Expenses enabled tens of thousands of citizens to participate directly in a new kind of political reform movement. Instead of just clamoring for change, they put their time and effort into creating evidence that change was needed. Crucially, the crowd of gamers did all of this important work faster than any individual organization could have, and they did it for free—lowering the costs of investigative journalism and speeding up the democratic reform process.

Not all crowdsourcing projects are so successful. Working together on extreme scales is easier said than done. You can’t crowdsource without a crowd—and it turns out that actively engaged crowds can be hard to come by.

In 2008, New York University professor and Internet researcher Clay Shirky sat down with IBM researcher Martin Wattenberg and tried to work out exactly how much human effort has gone into making Wikipedia. They looked at the total number of articles and edits, as well as the average article length and average time per edit. They factored in all of the reading time required to find knowledge gaps and spot errors, and all of the hours of programming and ongoing community management required to make those edits hang together coherently. After a lot of clever math, they worked out the following estimate:

On one hand, that’s no trivial effort. It’s the equivalent of rounding up a million people and convincing them to spend a hundred hours each contributing to Wikipedia, for free. Put another way: it’s like persuading ten thousand people to dedicate five full-time work years to the Wikipedia project. That’s a
lot
of effort to ask a
lot
of people to make, for no extrinsic reward, on behalf of someone else’s vision.

On the other hand, given that there are 1.7 billion Internet users on the planet and twenty-four hours in a day, it really shouldn’t be that hard to successfully pull off lots of projects on the scale of Wikipedia.
⁸
Hypothetically, if we could provide the right motivation, we should be able to complete one hundred Wikipedia-size projects
every single day—
if we could convince all 1.7 billion Internet users to spend most of their free time voluntarily contributing to crowdsourced projects.

Maybe that’s unrealistic. More reasonably, if we could convince every Internet user to volunteer just one single hour a week, we could accomplish a great deal. Collectively, we would be able to complete nearly
twenty
Wikipedia-size projects
every single week
.

Which really makes you wonder: with so much potential, why aren’t there even more Wikipedia-scale projects out there?

The truth is, the Internet is littered with underperforming, barely populated, or completely abandoned collaboration spaces: wikis that have no contributors, discussion forums with no comments, open-source projects with no active users, social networks with barely a few members, and Facebook groups with plenty of members but few who ever do anything after joining. According to Shirky, more than half of all collaborative projects online fail to achieve the minimum number of participants necessary to even begin working on their goal, let alone achieve it.

It’s not for a lack of time spent on the Internet. It’s just incredibly difficult to achieve the necessary critical mass of participation on any given serious project.

For one thing, some participatory networks are more rewarding than others—and the most readily rewarding networks aren’t, as a rule, the ones doing serious work. Online games and “fun” social networks like Facebook provide the steadiest stream of intrinsic rewards. They’re
autotelic
spaces—spaces we visit for the pure enjoyment of it. Their primary purpose is to be rewarding, not to solve a problem or get work done. Unlike serious projects, they are engineered first and foremost to engage and satisfy our emotional cravings. And as a result,
they
are the projects that are absorbing the vast majority of our online
participation bandwidth
—our individual and collective capacity to contribute to one or more participatory networks.

A second and more pressing problem is the fact that, across serious crowd projects, our participation resources are increasingly being spread too thin.

In the past month, I’ve been invited to join exactly forty-three Facebook groups. I’ve been asked to help edit fifteen wikis and contribute to nearly twenty Google Docs. And I’ve been (unsuccessfully) recruited for nearly twenty other assorted collective intelligence projects, each one requesting me to spend valuable online time voting, ranking, judging, editing, sorting, labeling, approving, commenting, translating, predicting, contributing, or otherwise participating in someone else’s idea of a worthy mission. I may be an extreme example—I’m a highly networked individual with many personal contacts doing interesting work online. But I’m certainly not alone in feeling overwhelmed by participation requests. Increasingly, I hear the same complaint from friends, colleagues, and clients: there are simply too many demands, from too many people, on our online engagement.

I call it “participation spam.” It’s the increasingly unsolicited requests we receive on a daily basis to participate in someone else’s group. If you’re not getting participation-spammed yet, you will—and soon.

By my own back-of-the-envelope estimate, there are currently more than 200 million public requests for crowd participation on the Internet, across thousands of different networks, ranging from citizen journalism, citizen science, and open government to peer-to-peer advice, social networking, and open innovation. This estimate factors in, for example, more than 1 million public social networks created on Ning, more than 100,000 wikis on Wikia, more than 100,000 crowdsourcing projects on Amazon’s Mechanical Turk, at least 20,000 videos awaiting transcription and translation on DotSUB, as well as myriad smaller clusters of open collaboration, such as the more than 3,300 public “idea spaces” for proposing and developing innovative ideas on IBM Lotus’ IdeaJam and more than 14,000 on Dell’s IdeaStorm.

With 1.7 billion people on the Internet, that works out to about 8.5 people per crowd.

That’s a very small crowd.

It’s certainly not a big enough crowd to build a resource on the scale of Wikipedia.

This problem is likely going to get worse before it gets better. As it becomes easier and cheaper to launch a participation network, it will likely become equally difficult to sustain it. There are only so many potential participants on the Internet. And as long as participation is designed as an active process requiring some mental effort, there are only so many units of engagement, or mental hours, each participant can reasonably expend in a given hour, day, week, or month.