Mystery Behind the “Digg’s Algorithmic Mystery Tour”
Digg published a blog post titled “Digg’s Algorithmic Mystery Tour” on October, 15th. While a digg blog post is just normal, a post about the algorithm was very surprising to me. How come Digg which never bothers to blog about very visible changes, numerous bugs & issues – decided to blog about its most hidden “algorithm”? They never even verbally discussed about it in public. And since then the front page stories changed. A lot of sites not found much on the front page of digg since the v4 fiasco; started to resurface frequently. Many diggers noticed this and there has been a lot of chatter about this. Many diggers wonder, how come a small tweak to the algo is doing so much change to the front page, and so “did” I.
Finding this strange, I wanted to find an answer to this question, at least to myself. Main reason for my curiosity was that, most of these sites did not get enough diggs prior to the algo change to make it anywhere close to the “top news”. I play much with the various APIs (Digg, Twitter, Google, Facebook etc) and more of a hobby to me. What I started as a casual search for an answer has now turned to be a major revelation – big enough for me to go public about it.
Some disclaimers & notes: I am no seasoned writer and pardon my poor language usage. All data I refer to here can be availed by tweeting me. The data I am using is current (as of 11pm CST, October 23rd), however I will continue to pull and maintain the data from the API for a while. The API has a lot of limitations and I have tried my best to work around those limitations. I will be writing in the sequential order in which I was drilling the data and any inferences or opinions I make will be clearly identified to make the facts remain as facts. Some data/graphs I present may be irrelevant to the crux of the matter under discussion, but interesting nevertheless. I have separated this into several pages, not just because it is going to be very long, but because I would like your comments on the various pages – as they each show a different set of data. I would really like to hear if you agree to my view points or not.
Bottom of each page will have a "Next Page" link, which will open in a new page - as you might have to back forth to cross refer some information.