Hard-core Digg users and fans of the link-sharing site — the so-called “Digg Nation” — spend a lot of their time trying to push their favorite links to the front page, competing with each other for the number of “diggs” their links get, and debating why certain links made it and others didn’t. Now, a 17-year-old programmer has come up with an algorithm he says can predict which links will make it to the front of Digg with a high degree of accuracy. The algorithm, which developer Raj Vir is still tweaking, powers a site called Digg In The Future.
Vir, who is still in high school, says that his algorithm — which he’s been working on in his spare time for the past several months — has so far proven to be 63-percent accurate when it comes to predicting what will make the Digg front page. (He also says it will still work even with the new version of Digg, which is expected to launch soon.) So how does the software do it? Vir says he has built into the algorithm some intelligence about the site and its most popular users: both the users whose links get dugg the most, and those who digg a lot of other people’s links. As he described it in an email:
The method used takes advantage of several tendencies of Digg users. There are several “power users” on Digg who have a heavy influence on Digg’s frontpage. Two main factors taken into consideration are what I like to call “power submitters” (users who frequently submit future frontpage stories) and “power diggers” (users who frequently digg future frontpage stories). Digginthefuture [sic] keeps track of stories that have been dugg or sumbitted by successful users.
The algorithm also relies on other factors, Vir says, including the time of day (since stories submitted in the early morning hours are unlikely to reach the front page) and whether the link comes from “preferred” sites that appeal to Digg users: a list that includes Cracked, Wired, The Huffington Post, The Daily Mail and The Telegraph. One interesting element of the algorithm is that it doesn’t just look at Digg or its users: Vir says that since many of the links that make it to the front page of the site have already been shared on other social networks, the Digg In The Future software looks at frequently shared URLs from Twitter and gives those added weight.
Vir says he isn’t looking to build a business or sell his algorithm at this point (unless of course Digg or Google make him a multimillion-dollar offer, we’re assuming), although he said he is considering selling advertising on the site. But if his algorithm is good at predicting trending topics on Digg, it might be good at predicting what links and content will become popular elsewhere, and a lot of companies are very interested in doing that. Vir may wind up getting an offer he can’t refuse.
Related content from GigaOM Pro (sub req’d): Social Advertising Models Go Back to the Future