Speculation is still hot on whether or not Google will be acquiring Digg very soon. Hotter still is speculation on why in mighty tarnation Google would want to do that. As someone involved in a Digg-clone acquisition recently, here’s my 2 cents.
I think the algorithm is definitely part of it. But the people are going to be an important part too. What Google would be buying, if they do buy, is the ability to process direct, high frequency page ranking signals on a sustainable basis.
Links = Indirect, Slow Signals = Page Rank
Currently, Google ranks search results mostly on the basis of how other web pages link to a given page. Information about this linking behaviour (which is typically agnostic to the indirect effect it has on search results) provides Google with various “signals” that they can then use to sort search results. This is a ear-to-the-ground approach to understanding what’s going on in the web. This works pretty well for traditional search results.
Votes = Direct, Fast Signals = Digg algo
Digg works with a very different type of signals. When users vote on an article, they are explicitly trying to move it up to the home page. That is a very direct signal. This is a keys-to-the-castle approach, and it works particularly well for news. After all, you want your spys and messengers to have ready access to interrupt you with important news.
Props to Digg
The problem with explicit signals is that they are easy to game, as diggers keep re-discovering. The other problem with them is that they happen a lot within short bursts of time. So not only do you have to be good at filtering out malicious votes that are not based on the merit of the article, you also have to be able to do it in record time.
While we may debate on whether or not the current Digg algorithm is gameable, we have to acknowledge that it is the most sophisticated, evolved algorithm that addresses these issues at scale. But the algorithm itself may not survive the brutal onslaught of Google users. That’s why the Digg team is important. Their experience in reacting, almost in real time, to deal with suspicious voting patterns and significant new social dynamics is extraordinary. But here’s the kicker – when they do react, they are usually not acting as moderators and manually fixing things – they fix the algorithm to account for the new patterns of usage.
This combination of algorithm, the massive Digg community and the people who are intimately familiar with both, are what would make this worthwhile to Google.