Tag Archives: aggregation

What’s in a game? Free-riders and the participation problem

Major online communities such as Reddit, Digg, and Slashdot are at the cornerstone of the social web: crowd-sourced content hubs that tackle the difficult task of collecting, evaluating, and promoting quality user-generated content. Loosely related, these platforms all perform a similar task of meta-moderating—or curating—content by tasking users to submit, up vote, or bury content based on its perceived value.

Source: Reddit.com

The only way to stratify the wealth of content they gather and distinguish it within these communities is to use intangible tiers of reputation and visibility. The idea to convert currencies of attention and reputation directly into positive feedback (ie. ‘Karma’ points) provides some tangible marker of digital success. The more popular or approved a piece of content is, the more visibility it—and its creator/poster—gains. The reputation-based Web is all about communities, participation, and peering; it is not, however, about authorial control or objectively professional qualities. Regardless, they are popular to a staggering degree and produce and promote untold amounts of content every minute of the day.

Inside and behind this content overload is a fundamental question of motivation: why does anyone contribute at all?

It seems that the mechanics at play here are some of the most interesting  qualities disseminated in economics, maths, and psychology: specifically, the ideas behind game theory.

The essential kernel of game theory is easy to grasp. Game theory is founded on a very simple but powerful way of schematising conflict. The underlying mechanics of John Von Neumann’s theory of economic games models the behaviour of perfectly logical parties interested only in winning. The dilemma of online producers and consumers alike resembles a game where players can only win by having access to quality content. Communities like Reddit and Slashdot expose this content through necessary participation, but those who contribute nothing still benefit from the results even if they put in no work to achieve them—often known as the free-rider dilemma.

The free-rider game is originally based on public transport, but remains true for modelling online communities. It posits that those who are willing to pay for the use of the railway system subsidise those who do not, because a single non-paying commuter is unlikely to bankrupt the whole system or take up an unreasonable amount of room on the train. Reddit users can easily rationalise lurking or browsing to their heart’s content, because their non-participation has a minuscule effect on the health of the system and the entire systems keeps chugging along.

And here’s the thing: the free-riders are right. They can reap all the benefits of vibrant, online communities without contributing any energy themselves. It is perfectly rational and provides the best possible outcome for the end-user. There is nothing wrong with thinking this way, that is, until everybody starts thinking ‘rationally’. Individually rational action—letting others do the participatory work—leads to a collectively irrational outcome: no one has any access to content—the trains stop running. The bargain of leeching off the system is a fiction; and ultimately, the only bargain offered by non-participation is a Faustian one.

But, since these repositories of user-generated content represent a kind of online collectivism that mired in an anonymous tide of mass mediocrity it is almost impossible to rely on anything other than the altruism of enthusiastic amateurs to produce meaningful work. Web 2.0 has demonstrated a frightening degree of indifference to participation as beautifully outlined by Jakob Nielsen’s treatise on participation inequality, and the intermittent contributions of idle users cannot be relied upon. However, the reputation-based currencies (such as ‘karma’) that these content-aggregators employ might as well be labelled as the bribes they are, because without an overt reward system, these communities would languish.

Source: Nielsen Group

With that in mind, these communities are resoundingly successful at encouraging contributions. Even if their scoring systems can be criticised as self-aggrandising, self-indulgent, and insular, the economics of reputation do directly encourage users to actively work within the rules of the system: it is effectively bribing them to participate. And in doing so, the system becomes universally more valuable to everyone. So long as users participate actively in posting and rating content they are organising the platform and allowing content to move properly through the system. Sure their internet points may be relevant only to their digital peers, but they are a powerful motivator so long as they derive some meaning from their boosted reputation.

Remembering that most-users don’t participate—and that they’re perfectly rational to do so—I’d love to hear why some of you do choose to get involved!

 

Huxley’s Nightmare: Twitch plays Pokemon

Famed science-fiction author Aldous Huxley dreamed a of dystopia in his novel Brave New World whereby the onslaught of trivial media had transformed the character of society into a burlesque. Years earlier, his grandfather—and noted evolutionary biologist—T.H Huxley had first supposed that a million monkeys on a million typewriters might one day collectively compose Shakespeare.

Neither Huxley could reasonably have anticipated that their speculations might converge and manifest themselves through the lens of Nintendo’s Pokemon franchise.

Source: imgur user Rhycool

In February 2014, a user identified as ‘twitchplayspokemon’ launched a channel on Twitch.tv—an online platform for live-streaming video games—running an emulated version of Pokemon Red, an 18 year old Gameboy game. Twitch.tv boasts a lively community of gaming enthusiasts streaming, watching, and interacting around a shared interest in playing and experiencing games. The community dynamics hinge on chat features embedded in the Twitch platform, and the interactions between streamers and viewers transforms typically solitary gaming pursuits into social experiences.

Twitch plays Pokemon takes this community interaction one step further, and by bootstrapping a chat-parsing bot onto the emulator viewers can input commands into the chat stream to dictate the inputs of the game. Typing in simple “up”, “down”, “left”, “right”, “A”, “B”, or “start” will feed the corresponding action into inputs for the game.  Predictably, thousands of people clamouring in chat and spitting out hundreds of commands simultaneously creates unmitigated chaos. The avatar gets stuck on terrain, he constantly flits between menu screens, spins around in circles, and ultimately behaves like a man possessed. Without pattern or purpose, the result is a bizarre mix of the trivial and captivating. For every hour spent in inaction and anarchy, brief flurries of cooperation incrementally nudge the game forward and demonstrate a complexity of collective action that verges on impossible.

This communal watch-and-play experience showcases some of the best and worst qualities of user-driven content; Web 2.0’s greatest strength also represents a terrible vulnerability, to aggregate the wisdom of the masses you have to invite their madness in too. The beating heart of Twitch.tv lies in its users coming together to share and interact, but their collective activity does little to build a community. Far from it, as TwitchplaysPokemon has demonstrated time and time again the contributions of the many are not necessarily helpful. In fact, the disruptive—often maliciously antagonistic—anarchy of the users creates something wholly unique and unanticipated from what should be relatively mundane, boring elements.

Source: imgur user Rhycool

Counterpoint to Tim O’Reilly’s position that harnessing collective intelligence has an effect liken to that of the sum being greater than the whole, the Twitch community thrives not on inclusiveness and cooperation like so many other Web 2.0 hubs, but rather on the toxic levels of animosity and trolling behaviour between users. The raw energy and interest that Twitch produces is not generating a collectively ‘intelligent’ behaviour across the sum of interactions. While there is something to be said for the collective joy of achieving the most basic goals when 100 000 voices are all arguing at once, the most compelling part of Twitch—the thing that binds the community together—is how devastatingly broken it all is.

After more than 2 weeks of continuous play, Pokemon was finally beaten. This is either an incredible feat of crowd-sourced collaboration, or a tacit endorsement that monkeys might well compose Shakespeare given time. If nothing else it opens the way for a new approach to how games and gaming are consumed and integrated on social platforms. The potential for participatory collaborations on Twitch need no longer be constrained to a simple viewer/streamer relationship. The possibilities for building a community—nay, a society—around even the most trivial of media, when experienced cooperatively, is limitless should Twitch—or other platforms—choose to leverage this position.

Looking forward, it certainly appears that Twitch plays Pokemon has ushered in a brave new world for Twitch.tv.

 

A Delicious Mess

Following on from my soap-boxing about the state of content curation online, I’ve setup a Delicious feed available here and embedded in this blog under my ‘What’s Interesting?’ tab at the top.

I’ll be collating and curating interesting articles about disruptive innovation, information futures, and just plain good writing in this feed, so stay tuned for more of that!

Can you Digg it?

Curation, Aggregation, and Web 2.0

Tools for curating, sorting, and managing web content usually take the form of social aggregators such as Digg or Reddit. The act of curating is not one of careful selection by a trained expert, but rather the weighted consensus of the masses promoting or up-voting content they find notable.

Web 2.0—nay, the entire information profession—has a problem; the barriers to information creation and storage have fallen in recent years. This has resulted in the amount of information on the Web proliferating beyond all expectations. Finding the right information among the endless supply of trivial and irrelevant data has become almost impossible. The rational response would be to trust our curation to trained professionals, able to disseminate and sort through this wealth of information and categorise it based on merits of accuracy and quality.

Instead, popular aggregators and the wisdom of crowds have emerged as the determining values of qualitative merit on the Web.

There is a very real risk that the Web—the most powerful source of knowledge available—is mislabelling, misrepresenting, and misplacing important data, and being unable to distinguish it from the unfiltered noise of the masses. We have trusted the most important resource in human history to the collective rule of enthusiastic amateurs.

This pollution of data poses a threat of eroding and fragmenting any real information stored on the Web. Users have come to rely on the anonymous and amorphous ‘rest of the Web’ as their authoritative filter. Content aggregators remix information drawn from multiple sources and republish them free of context or editorial control. These aggregated opinions of the masses are vulnerable to misinformation as users have too much control and too little accountability. The risk of aggregating information is the risk of privileging the inaccurate, banal, and trivial over the truth.

Digg.com, founded in 2004, was the first notable aggregator of Web 2.0 content. Voting content up or down is the core of the site: respectively ‘digging’ and ‘burying’ material based on contributors input. This supposedly democratic system allows content of merit to be promoted and displayed. But, this assumes that all opinions and user-generated regulations are equally valuable and relevant in determining merit.

The collective judgements of a group—the clichéd ‘wisdom of the crowds’—can be an effective measure of certain types of quantitative data. Called upon to guess at the number of jellybeans in a jar, the aggregated guesses of a thousand contributors would provide a relatively accurate figure. However, if that same group was called upon to disseminate the value of a news story their opinions would not represent a collective truth about the value or merits of the piece. The voting process of Digg or Reddit is transparent and instant, and causes contributors to cluster around popular opinions—promoting sensationalism and misinformation. Content that grabs the attention of users will quickly be promoted and rise to be seen by more users, regardless of its accuracy.

The momentum of a popular story is exponential: the more users see something, the more popular it becomes—exposing it to even more users. The infinite shelf-space and shelf-life of the Web means that once a piece of information has seen any exposure it almost impossible to control. Instantly a lie can spread across the Web by the zeal of its promoters, and be cross-referenced by a dozen news aggregators. Lies become widespread and pollute enough aggregation sites that they become the valid—supposedly authoritative—result of any Google search on the topic. The wisdom of the crowds is fickle and closer to a mob mentality; it is impossible to aggregate their wisdom without aggregating their madness as well. After all, there is a fine line between the wisdom of crowds and the ignorance of mobs

However, non-trivial and important content is still being created, promoted, and viewed on the Web; aggregated information services do capture these notable pieces of data in their trawling. In practice an old problem remains: time and effort must be manually expended to sort out the real information from the useless noise. Exactly the sort of time and effort that professional curators, librarians, and information professionals were traditionally employed to expend.

Digital media theorist Andrew Keen, in his book The Cult of the Amateur (2007) likens the community of Web 2.0 to evolutionary biologist T.H Huxley’s humorous theory that infinite monkeys on infinite typewriters would eventually create a masterpiece such as Shakespeare. Keen sees this infinite community of empowered amateurs as undermining expertise and destroying content control on the Web. He argues that their questionable knowledge, credentials, biases, and agendas means they are incapable of guiding the public discourse of the Web with any authority at all.

Another perspective on this comes from the 1986 book Amusing Ourselves to Death, wherein television commentator Neil Postman theorised about the erosion of the public discourse by the onslaught of the media. He frames the media in terms of the dystopian scenarios offered by Huxley’s grandson—science fiction author Aldous Huxley—in the novel Brave New World, and compares them to the similar dystopia of George Orwell’s 1984:

 

‘There are two ways by which the spirit of a culture may be shrivelled. In the first—the Orwellian—culture becomes a prison. In the second—the Huxleyan—culture becomes a burlesque’ (Postman, 1986, p.155).

 

In one dystopia, Orwell feared those who would deliberately deprive us of information; in another, Huxley feared those who would give us so much information that the truth would be drowned in a sea of irrelevance.

And, the culture of Web 2.0 is essentially realising Huxley’s dystopia. It is cannibalising the content it was designed to promote, and making expert opinions indistinguishable from that of amateurs.

User-generated content is creating an endless digital wasteland of mediocrity: uninformed political commentary; trivial home videos; indistinguishable amateur music; and unreadable poems, essays, and novels. This unchecked explosion of poor content is devaluing the work of librarians, knowledge managers, professional editors and content gatekeepers. As Keen suggests ‘What is free is actually costing us a fortune. By stealing away our eyeballs, the blogs and wikis are decimating the publishing, music, and news-gathering industries that created the original content these Websites ‘aggregate’ (Keen, 2007, p.32).

In a world with fewer and fewer professional editors or curators, knowing what and whom to believe is impossible. Because much of the user-generated content of the Web is posted anonymously—or under pseudonyms—nobody knows who the real author of much of this self-generated content is.

No one is being paid to check their credentials or evaluate their material on Wiki’s, aggregators, and collaboratively edited websites. The equal voice afforded to amateurs and experts alike has devalued the role of experts in controlling the quality and merit of information. So long as information is aggregated and recompiled anonymously then everyone is afforded an equal voice. As Keen dramatically states, ‘the words of wise men count for no more than the mutterings of a fool (2007, p.36)’.

We need professional curation of the internet now more than ever. We need libraries and information organisations to embrace the idea of developing collections that include carefully evaluated and selected web resources that have been subject to rigorous investigation. Once upon a time we relied on publishers, booksellers, and news editors to do the sorting for us. Now, we leave it to anonymous users who could be a marketing agency hired to plant corporate promotions; it could be an intellectual kleptomaniac, copy-pasting other’s work together and claiming it as their own; or it could be, as Keen fears, a monkey.

Without professional intervention, the future is a digital library where all the great works of human history sit side-by-side with the trivial and banal under a single, aggregated category labelled ‘things’. And we would have no-one to blame but ourselves.