Last week’s report from Compete.com that the social aggregator site Digg.com had lost a staggering one-third of its visitors in a single month prompted many a dirge to be sung to the site’s imminent death. If this proves true, it’s a big damn deal, because sites like Digg are supposed to be going the other direction. As social content proliferates, we desperately need social aggregation/curation sites to keep our collective heads from exploding, and Digg has arguably been the top player in that space for half a decade. So what went wrong?

What we have here is a failure to cooperate. In my book on game theory and social media marketing, which I will shamelessly plug on this blog until someone eventually buys a copy, I suggest that the success of social media ventures like Digg is entirely dependent on delicate social contracts that demand cooperation from ALL participants. It takes very few bad apples to spoil the social media bunch. Digg’s value from the start has been predicated on the belief that stories are being promoted fairly on the basis of actual popularity, but that value has been compromised nearly from the start.

If you did a Google search for stories about Digg’s demise, you would turn up results stretching back to 2006. Only Castro has survived a longer deathwatch. But the incidents that drove those early predictions also sowed the seeds for the site’s present predicament. In short, popularity is the site’s currency, but there is no way to prevent counterfeiting. A small number of extremely active users have been able to hijack the site’s top content by relentlessly promoting their favorite stories, making a mockery of the site’s emphasis on the wisdom of the crowd.

Once users come to distrust the fairness of the voting system, they’re either less inclined to vote themselves, or they abandon the site altogether, which heaps even more unearned power on the bad apples. At that point, the site is no longer a curator for truly popular content, but merely a tyranny of geekdom. In game theory, we call this cycle of mutual defection a death spiral, because it is self-reinforcing and, well, because it leads to death. Digg’s death, in this case.

Digg claims that it has addressed its bad-apple problem by “tweaking the algorithm” that decides the weight of individual diggs. But no game theorist would be satisfied with this solution, because it fails to deliver any penalty for the bad apples gaming the system; it merely invites them to figure out the new algorithm and how to beat it.

So what could Digg have done differently? They could have recognized that they’re dealing with what behaviorial scientist John Platt called a “social trap.” In Platt’s simple formulation, social traps occur when a given behavior produces positive results for the individual—the content owner trying to promote their own stuff on Digg—and negative results for the group, i.e, the rest of us poor saps looking for good content. As long as the individual is only accountable to themselves, the negative behavior is self-reinforcing, resulting in “locked-in behavior,” even though the individual’s long-term interests are imperiled by the behavior.

Platt offers several ways out of the social trap; the most important of these is the notion of “counterreinforcers.” Since destructive behavior is self-reinforcing in the social trap, counterreinforcers discourage this behavior by offer some negative consequence that the player must evaluate before taking the action. This may not seem like a terribly radical notion to you, because, in fact, social media models that evolved after Digg had counterreinforcers built into their DNA. On Twitter, blowhards masquerading as worthy information sources are subject to immediate counterreinforcers—they get unfollowed, and their stuff doesn’t get retweeted. Friending and de-friending on Facebook work much the same way. These are self-perpetuating cooperative systems that can’t be easily gamed.

It’s no surprise, then, that Digg’s turnaround plan, announced by CEO Kevin Rose last week, involves making the site more like Twitter or Facebook by allowing the user to select which news sources to follow. Under such a system, Digg’s content bullies would have to earn their followings by digging good content, and the social trap would effectively be eliminated.

But it may be a case of too little, too late. Death spirals are nearly impossible to reverse, and old life forms must make way for new ones in the evolutionary chain. Liz Gannes thoughtfully observes that Digg wants to be the “Twitter of News.” The problem is that we already have a Twitter of News. It’s called “Twitter.”