Ask HN: Is this normal HN ranking?
I came back looking for a TechCrunch story about SV undermining democracy [0], and was surprised to find it had already dropped off the front page.
Found it on page 2, but it's ranking was lower (44) than articles that it should seemingly outrank—based on comments (211), points (247) and time posted (12 hours ago). See screenshot [1].
For instance, at 38 is an older article with only 3 comments and 29 points.
At 41 was a significantly older article with only 38 comments and 125 points.
So, this article is ranking lower than even a much older article with roughly half the votes and ~5.5x fewer comments.
Is there some other "bonus" or penalty for articles (beyond those that are overtly problematic per HN guidelines)?
[0] https://news.ycombinator.com/item?id=49232221
[1] https://imgur.com/a/t4Cn2Fi
Comments
The HN algo is intentionally obscure to prevent people from gaming the system. For example, if there's a bunch of traffic going directly to the discussion link, it will heavily punish the weighting of the post as it's assumed the OP is sharing the link to farm upvotes. There's a bunch of other black magic involved in the HN ranking system so to try and understand it based on numbers alone is a waste of time.
I've seen this cited around here frequently, and I don't doubt that there were good intentions for implementing it that way. However, in my anecdotal experience, I've noticed that otherwise high-quality and thought-provoking posts that happen to be critical of the tech industry (or other VC/SV-aligned interests) seem disproportionately impacted by algorithmic deranking.
I acknowledge that it may be pure coincidence, but the problem with an opaque system is that deliberate antagonism can't be categorically excluded as a possibility. And considering the financial interests of the platform's ownership, I'm also not convinced we should extend the benefit of the doubt very far, either.
Comments reduce a post's ranking.
Huh? Why?
I've always suspected it's an attempt to defuse flame wars, with the expectation that a lot of comments means a post is highly controversial and stirring up arguments in the comments section.
Seems unlikely. Using comment count as an indicator of poor quality discussion would treat topics with intense interest and engaging discussion like flame wars.
So, I find this explanation hard to believe, especially given that the mods here are active, and that there are other mechanisms for exposing flame wars (including flagging, down votes, etc).
High-quality posts will usually be upvoted by lurkers, ensuring a favorable upvotes-to-comments ratio. But yes, it's a blunt instrument that sometimes has false positives where the mods intervene to turn off the penalty: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
Makes far more sense than comment count alone.
Wondering about this as well.