Kevin Rose states that they’re doing a hard job at battling the “so-called” bury brigade. Kev says:

For the same reason that we don’t expose all of our back-end methodologies for the Digg promotional algorithm, we also don’t expose the details of how the burying algorithm works. We spend a lot of time analyzing our data and understanding how people Digg and bury content. We have spent the last 2.5 yrs building systems that ensure a diverse group of users promote or bury stories.

The way Kevin answers the issue of bury brigade makes it sound like he’s dealing with it the same way that it deals with the gaming of Digg to promote stories. Algorithmically. The problem, however, is that they are two largely different processes.

Gaming Digg to promote a story involves a few red flags that are relatively easy to track.

a) A set group of users who vote in similar patterns
b) The promotion of a stories from a particular URL
c) The velocity of how many votes a particular story accumulates

There are probably other factors that I’m not accounting for, but using these factors alone you could see how an algorithm could flag stories. For example stories that:

a) are promoted by a group of users who tend to vote in a very similar pattern — all the time
b) accumulate diggs unusually fast and
c) are promoted from a URL which is known to be suspicious (for being labeled as “spam”)
… could be flagged, ranked, and possibly pulled for gaming Digg.

The problem with this is that it doesn’t really address the problem with the Bury Brigade, which I think DOES exist. And the way that they influence the availability of stories on Digg has very little in common with the patterns around gaming. The Bury Brigade, as I have called them, is no specific group of people.

Rather, they are naturally congregating groups of users who have probably never “friended” each other, but have similar interests, and more importantly *grudges* in common. For example, some groups don’t like blogs. Others don’t like anti-Digg posts. Others yet don’t like Microsoft. Or Sony stories. But what many of them do not have a grudge against are specific URLs.

And its precisely because there are no specific set of users and these are actions that are not against a specific set of URLS, I think its impossible that any algorithm will be able to identify these behaviours, and therefore, counter the “bury brigade”. Of course, while it would be possible to find people who bury stories certain types of stories consistently — through certain keywords, for example — the fact is that flagging those accounts and invalidating their buries, would only eliminate a tiny portion of those users who are burying naturally in groups.

The promotion of stories has a specific set of features that can allow an algorithm to identify certain patterns that would enable it to flag suspicious stories — either while on the frontpage, or while it is achieving a certain velocity of votes on the way to the frontpage. The identification of buries by “bury brigades” would be far more difficult — particularly because the way that people bury is probably in the same people submit and vote stories. And while it continues to be difficult, bury brigades will continue to exist, and they will be influencing digg news in a way that will be hard to fight.