I think the concept is interesting - but needs a bit of intellectual hacking so to speak. The point that if a story is on digg or reddit it doesn't belong on H

I think the concept is interesting - but needs a bit of intellectual hacking so to speak. The point that if a story is on digg or reddit it doesn't belong on HN is a symptom of something else, namely that digg or reddit is associated with a low common denominator, and thus an article that appears on these sites first shouldn't appear here. The real problem to be solved here is twofold: 1) Quality - An article doesn't become bad simply by appearing on reddit, it only seems so because it appears in the colorful company of lolcats and conspiracy theories. It might also have something to do with the psychology of having a "better" site over here at HN. 2) Categorisation - not all quality articles should have high HN quotient. Political articles for instance should have a markedly harder time getting a good HN rating. I think that the HN quotient should involve the data you propose, but it should also go beyond that and consider the points I made above. If the HN quotient is a measure of being uniquely hacker the problem is coming up with this definition. It's probably impossible to come up with a hard definition, but since we have ratings of articles, users with ratings and longevity data for these (how long has the user been a member), ratings for comments to articles (including the users that commented) for this site it should be possible to get an idea of what is uniquely hacker news. This could be done by using some statistical kung-fu. Example: Each article gets a number of points that distinguishes it as more or less uniquely hacker news - the more points the better: - articles points/3 - submitted users longevity (days since user was created)/5 - average of (commenters karma + commenters longevity) for top level comments to the story. (second level comments and below often have less to do with the submitted article than the comment it responds to) When these are added together it gives a number which is the "uniquely hacker news" quotient. Note that there are many ways of getting this number, the above is just a simple example that favors submissions that old users that are in the conversation find interesting and comment on. It could easily be made smarter. Now we have a basic definition of what constitutes a definition. The ineresting bit of course would be to see whether submissions with a high score correllate with articles that are submitted/not submitted to other sites, and the timing of this. Oh well - just rambling along here. Feel free to tear down my algorithmic proposals all you want :-)