reddit and digg – first impressions

Thanks to Walton’s mention, I’ve been spending quite a bit of the weekend looking at reddit, and comparing it to digg. Both sites are a step forward from the old classics, Slashdot and Kuro5shin, as well as the much newer de.licio.us (which as I’m sure everyone who follows it knows, has just been bought by Yahoo).

digg and reddit both allow their users to directly determine the content that appears on the site. However, they have a significantly different voting system. digg is older (being launched in November 2004), and has a very simple voting system. A registered user submits an article, and it immediately appears in a queue of new posts. Users can then dig the article (a positive vote). If it receives enough positive votes within a certain time period (I’m not sure exactly how the algorithm works), it will be promoted to the main site. The more positive votes in the shortest space of time, the higher it appears. It works well, and is an innovative democratic advance over a site such as Slashdot, where posts are approved by the editors. I stopped visiting Slashdot a while ago. They do have interesting posts, and I still visit occassionally, but I get sucked into reading inane comments such as fisrt post, or ya gotta be kiding, M$ sucks, and getting depressed about the state of humanity. I’m sure those exact comments exist multiple times 🙂 . With the limited number of eyes approving article s, there’s a significant delay, and the number of posts appearing on the site is relatively small. I revisited it, just for the comparison, and found I spent a lot more time on digg.

digg has its problems, in particular with repeat posts, such as when a new wave of users resubmit an article that’s already dropped off. This is probably one of the reasons for the important difference in reddit’s voting system. reddit launched in July 2005, and allows both negative and positive votes. So the top ranked articles are not just those that are popular amongst a significant portion of the users, but those that achieve some sort of consensus. Repeat posts can therefore be voted down by older users, and are less likely to be ranked highly. A user also accumulates karma, which is dependant on the success of stories they submit and vote upon. As far as I know karmas only rewards it accumulators with bragging rights, although I have a vague, probably mistaken memory of reading somewhere that it also affects the weighting of the vote. Its an interesting concept – the secrets of the algorithm are kept secret, and one downside is the phenomonen of so-called karma whores, who vote for stories they know will be popular in order to improve their own karma, doing nothing for the overall variety. If digg tends to suffer from repeats, reddit tends to suffer from a lack of variety, and potentially divisive articles may not rate highly. Anything by Paul Graham, one of the founders (and who presumably has lots of karma according to the system) seems to shoot up the charts, and there’s a few legacy topics (such as Lisp) that achieve strange success. Perhaps this is what lead me to think that karma plays a role in the weighting of a vote.

I found some excellent articles on both sites, and will be using them both extensively in future. I can relate to the sentiments of one reddit commentator, who said that he will never visit another site as he feels that anything worth reading will appear on it. Both also have a great interfaces, with reddit’s being fantastically simple, and digg not far behind. It warms my heart to see more sites with such great design, cottoning on to the fact that complicated intefaces with multiple choices puts people off – compare your first impression of digg or reddit to say CNN.

Perhaps I haven’t yet made enough use of the friends or customisation features, but I found myself getting ideas about implementing a similar concept with a focus on local topics. reddit had a large number of articles about the Bush/Cheney horror show, and while it’s heartening that more Americans are waking up to how they’ve been duped, the American demographic bias does result in a limited variety. Give me well-written articles about Mbeki/Zuma (they must exist somewhere), or an in-depth analysis into the Western Cape ANC’s infighting (being even more wishful here). Let me see the pick of Tectonic, Commentary, the Mail and Guardian, IOL, News24 and ITWeb. Introduce me to those niche bloggers with the monthly pearls of wisdom I may never otherwise come across.

I’m sure we’re going to see even more variations of the idea appearing, and I wait with interest to see how well these ones will age as their communities change, and what else will appear.

2 comments

  1. Thanks for the write-up, I just wanted to clear up some things (perhaps this is a sign that we need some better documentation).

    Karma presently has no effect on the weight of user votes, it also can only be gained from submissions, not voting on them. [from our FAQ] http://reddit.com/help/faq

    How do I accumulate karma?

    The formula was and continues to be constantly evolving (or, intelligently redesigned, if you’d prefer). Presently, the easiest (and most effective) way is by making good submissions: items that other people want to read and will subsequently become popular. Those redditors who best understand what people want to be reading about will be the best rewarded.

    Also, you’re correct about Paul Graham’s essays being popular reddit submissions, but I can assure you that he is not a founder. However, he is a partner in Y Combinator, which provided us with our seed funding, but he didn’t get any extra karma for that.

    All that being said, you make an interesting point about localization. Our “recommended” section is a step in that direction, provide you with more links you’ll be interested in, but your desired level of specificity is going to require a lot more South Africans using our site :).

    On a totally unrelated note: I had the good fortune to visit your beautiful country last April and I can’t wait until I can get back and finally get some color back into my pale, pale skin.

  2. Reddit also has a different focus from Digg. Digg is specifically a tech news site, whereas Reddit aims to deliver whatever’s new on the web. A lot of the stories on the front page of Reddit are tech stories at the moment, but that’s just because early adopters of new technology tend to be techies. (In 1985 most email was about computers; that’s not true now.)

Comments are closed.