1. 2

On HN, stories that have a lot more comments than upvotes set off a “flamewar detector”. I’m not a fan of HN’s implementation as it results in the story being taken off the FP without explanation, but the idea seems reasonable.

There have been multiple stories people have complained about staying near the top for quite some time simply because people were either arguing with each other, arguing with the story’s author, or just went off topic.

This story has 5 upvotes and 25 comments (a ratio of 0.2), and yet it’s currently at #2 in the rankings. All of the comments so far are either pointing out flaws in the article or are unrelated to its content.


  2. 3

    I don’t like that feature. Not long ago there was a great article by cperciva that was willfully misunderstood by several people. He made the mistake of trying to correct them, triggered the flame war, and the story was essentially killed. Now people who want to learn can’t.

    The flame war detector is predicated on the notion that bad stories cannot create good discussion. I disagree very much. I think we have adequate means of pushing bad content down the page.

    1. 1

      To be fair, it is undeniable that a winning strategy (if you simply want karma) is to bait people into arguments in the comments section. This is made even more profitable if you upvote/support systematic support of stories that are inherently controversial (news, non-tech, etc.). That has happened time and time again, and is not a hypothetical issue.

      I do agree that the flame war detector is not the right tool either–perhaps an automated tag for “Hey, holy shit, a lot of comments on this”.

    2. 1

      The quality of the comments is one of the reasons I like lobste.rs, and by their nature will often stray from the subject of the article.

      The quality of an article should be based on its content - which is hopefully indicated by its upvote count. Sometimes the discussion generated is better than the article itself, and I’m not convinced that changing the hotness algorithm this way will improve the working of lobste.rs.

      1. 1

        I’ve noticed a few threads recently where top comments have more points than the stories themselves. I’ve taken it to mean that folks are writing insightful responses to flawed or marginally interesting stories. I’m really happy with the current state of affairs, including that linked story, because the generation of new, intelligent content is far more valuable than holding a popularity contest for mediocre material.