How Google Could Solve the Digg Gaming Problem
Digg founder Kevin Rose said today that the site will be making changes to its ranking algorithms. The problem is that a small number of Digg users have too much influence over what makes it to the home page.
The general perception is that sites like Digg work via the wisdom of crowds, a term recently repopularized by James Surowiecki in his book of the same name. The idea behind crowd wisdom is that large numbers of people with different perspectives will be consistently better than individuals at problem solving. In Digg's case, the problem to be solved is the promotion of stories.
Most people assume that because a large number of diverse people promote stories on Digg, then the resulting top stories are the product of collective wisdom.
Unfortunately, Digg doesn't really work that way because each user has foreknowledge of other users' diggs. Diggers promote stories that have already been promoted. This is the herding instinct. If you read Surowiecki's book carefully, you'll note the wisdom of crowds relies on diversity, local knowledge and independence. It's that last quality Digg users don't have. Instead, each digg is an individually applied discrete unit of imitation, leading to an information cascade. It's one big Web 2.0 bandwagon.
One way to solve the Digg gaming problem, then, is to remove the friends system whereby diggers can tell what their friends are digging. That's a non-starter for Digg, since community is an important aspect of the site. Another way: Remove the digg rank. Also a non-starter for a site based on collective enumeration.
So how could Google solve this problem? Implement its own human-ranking system on Google News, and do it in a way that doesn't show how many people have ranked a story. Provide an alternate site -- say, Google News Community -- that displays the results of the anonymous ranking on Google News.
Another way to do this would be to allow users to anonymously rank stories from their Google Personalized Homepage. Those results would also be aggregated on Google News Community.
It's just a thought. Google's been experimenting with social search, and Google News has the traffic to make the system work. Plus, the resultant traffic to publisher sites would demonstrate Google is a boon to traffic, not a detriment. Just a thought.