For more than 16 weeks, some half a million players engaged in a kind of informational scavenger hunt, sometimes working in smaller teams, sometimes working together as a mass problem solving community.

The game required players to congregate together in both real and cyber space to collectively solve problems using information obtained online and then delivered to certain locations in the physical world.

One puzzle required them to decrypt GPS location data, send participants at various times to hundreds of phone booths scattered across the continental United States, and respond within a matter of seconds to unanticipated questions delivered by a live actor. To give proper answers, each person had to trust an ad hoc group of strangers to instantly provide the information they needed.

The game’s unlikely name is ilovebees, which was the website for a fictional honey company that served as the “rabbit hole” or jumping off point for their activities. But the real twist behind the game was that it was created as a viral marketing ploy by 4orty2wo Entertainment for the Microsoft Xbox game, Halo 2.

Elan Lee and Sean Stewart, the founders of 4orty2wo Entertainment, are veterans of “The Beast,” a similar game created to generate interest in the Steven Spielberg film, Artificial Intelligence: A.I.  The pair has pioneered a genre they describe as “search operas” but which the fans more often call “alternative reality games” (ARGs).

For hardcore players, these games can be so much more than viral marketing. These ARGs teach participants how to navigate complex information environments and how to pool their knowledge to solve problems.

Jane McGonigal, the primary community leader for ilovebees and Ph.D. candidate in Performance Studies at the University of California at Berkeley, said that ARGs are generating “players who feel more capable, more confident, more expressive, more engaged and more connected in their everyday lives.”

Alternative reality gaming could be seen as a 21st century equivalent of a much older literary form – epistolary fiction.  Many early novels, including Pamela (1740) Les Liaisons Dangereuse ( 1782 ) or The Sorrows of Young Werther (1815), consisted of fictional letters, journals, diaries, and newspaper accounts, which were presented by their authors with little acknowledgement of their fictional status. The authors often claimed to have found the materials in an old trunk or to have received them anonymously in the mail.

By comparison, consider how Stewart described the design of “The Beast” on his home page:

“Create an entire self-contained world on the (W)eb, say a thousand pages deep, and then tell a story through it, advancing the plot with weekly updates, concealing each new piece of narrative in such a way that it would take clever teamwork to dig it out.

“Create a vast array of assets – custom photos, movies, audio recordings, scripts, corporate blurbage, logos, graphic treatments, web sites, flash movies – and deploy them through a net of (untraceable) websites, phone calls, fax systems, leaks, press releases, phony newspaper ads, and so on ad infinitum.”

Like earlier epistolary stories, the ARGs refuse to acknowledge their own fictional status. One of the key rules the group uses to sustain its “reality” is that participants never look behind the curtain. The puppetmasters – those who create the game – only identify themselves once the game is completed.

The content of earlier epistolary novels turned readers into armchair detectives and amateur psychologists, piecing together the events of the story from multiple, fragmentary, and sometimes contradictory, always subjective, accounts. These ARGs take on a more public dimension, exploring conspiracies or mysteries which exploit the expansive potential of the transmedia environment.

Though read in private, these early novels became the focus of parlor room discussions as people compared notes about the characters and their situations. ARGS today offer a very similar experience of mutual debate and collaborative interpretation for a society just beginning to experiment with what cybertheorist Pierre Levy calls collective intelligence.

To be sure, there have been earlier forms of collective intelligence – people collaborating to create imaginary societies and creating relationships which extended into real world spaces. There have also been other mystery hunts – books whose hook was that they offered clues to objects hidden somewhere in the real world.

But, ARGs push it to the next level. By design, they are impossible to solve unless people put their heads together in unprecedented numbers.

Much as some early novelists – most famously, Charles Dickens – published their works in serialized form and were thus able to adjust later plot developments to the public’s initial responses, ARG designers react to the players. Sometimes, they needed to make a puzzle simpler, more often more difficult, depending on the community response. In an oft-told tale, players solved in a single day a set of puzzles puppetmasters for “The Beast” had intended to last a month.

McGonigal described a range of strategies she and the other designers used to insure that ilovebees could not be solved by individuals or isolated groups.

Specific bits of information were sent randomly to players, who would then have to compare notes to get all of the puzzle pieces. The game demanded mastery of obscure languages, forgotten codes, and other kinds of esoteric knowledge, insuring that no one player would have the expertise to solve it.

And players were asked to do “walk and talk and chew gum” tasks that required many coordinated simultaneous actions that no single player could do by themselves. The teams became smart mobs linked by cell phones and other mobile communication technologies.

But these games do more than just send people on high-tech scavenger hunts. A well designed ARG reshapes the way participants think about their real and virtual environments.

“The best pervasive games do make you more suspicious, more inquisitive, of your everyday surroundings,” McGonigal writes in an essay published online. “A good immersive game will show you game patterns in non-game places; those patterns reveal opportunities for interaction and intervention.”

Much as critics of early novels worried that readers would have difficulty separating fact from fantasy or would find themselves fixating on the affairs of imaginary people, contemporary critics of ARGs worry that players will not be able to find their way back out of the rabbit hole.

McGonigal and others see such critics as doing a disservice to the players who do what readers or theater goers have always done – engage in a voluntary suspension of disbelief in order to enjoy immersing themselves into a fictional realm.

“There are killer robots and sentient houses,” Stewart told McGonigal, “How could anyone be confused?”

A well designed ARG also changes the ways participants think about themselves, giving them a taste of what it is like to work together in massive teams. They develop an ethic based on sharing rather than hording knowledge; they learn how to decide what knowledge to trust and what to discard.

Let’s be clear. According to Levy, the collective intelligence is not governed by mob psychology. It is not a “hive mind” where everyone knows the same things and thinks in the same ways. It thrives on diversity, starting from the assumption that each individual member has something unique to share. The group seeks out different voices and perspectives to inform their discussions. There is no fixed hierarchy – people come to the foreground or fade away depending on what they can contribute at a particular moment of time.

Playing an ARG can become all-consuming for weeks at a time. Many of the most hardcore participants find themselves seeking out other ARGS or even forming teams to create their own games.

When hardcore gamers can’t find an ARG to play, they start applying their minds to other problems. As McGonigal has documented in one of her scholarly essays, team members have rallied to tackle such real world problems as locating the Washingington, D.C. sniper,  mapping governmental corruption, or even, briefly, trying to find September 11 conspiracies.

In each case, they proceeded from the assumption that the solution lie in tracking flows of information across cyberspace, looking for clues which might be overlooked by traditional law enforcement agencies and putting together the pieces in fresh ways.

One can see these efforts as distasteful in so far as they turn tragedies into games or as arrogant in so far as they assume a level of expertise they may not have earned. But they are a local consequence of the sense of empowerment players discover through participating in such robust knowledge communities. Even the most complex problems seem manageable if everyone puts their minds to them.

Levy has predicted that such knowledge cultures represent an alternative source of power that exists alongside the political authority of the nation state or the global reach of commodity capitalism. We will someday learn to use this power to change the world.

For the moment, we simply play.