I think the problem with Reddit (and Digg, and almost every other popular social news site) is algorithmic. The pattern is the same every time one of these sites becomes popular - it starts out with a small group of insightful people that submit and vote on content that they find interesting. The reason it works is that they have similar taste. Kind of like hacker news at the moment. Once the site grows, however, the initial group gets diluted and the site is swarmed with content that is not necessarily core content for the original group. And the new arrivals vote up the material that is submitted by the other new arrivals, thus diluting the site to the lowest common denominator. This is why Reddit is now filled with Ron Paul, conspiracy theories, LOLcats, and all sorts of other crap.<p>It is possible, and in my opinion not all that hard, to solve this algorithmically.<p>One way would be to make mod points weighted, so that members that have signed up early, have high karma, have the same taste (based on their upmods and submissions) as the core group have mod points that count for more than new arrivals that have yet to build any credibility.<p>Another way would be to not let new arrivals submit stories until they reached a certain karma. This would mean that submitted stories would stay more true to the original idea.<p>Another way would be to only let users have a certain amount of mod points that would be renewed after x days, spurring users to use them more sparingly, and stopping users that up/downmod everything that is not in their particular interest.<p>Yet another way would be to let the system learn from the users submissions and votes, and show stories based on this calculated using datamining. Maybe as a "my {site}"<p>There are numerous other ways of doing it, these are just off the top of my head - and yes there are probably pitfalls in some or all of these ideas.<p>The point I want to make though is that the social news sites are pursuing the wrong goals once they start to gain traction. Instead of working on and constantly tweaking their algorithms they seem to focus more on new features, design, marketing or whatever.<p>There is an excellent business opportunity in creating back-end software to drive these sites and solve the algorithmic problems that, for some reason that is unclear to me, are so hard for these sites to do properly.