4 ms·

I appreciate the idea, but I question the merit. Importantly, I think what it is really optimizing for are posts without experts on Hacker News. As the author

I appreciate the idea, but I question the merit. Importantly, I think what it is really optimizing for are posts without experts on Hacker News. As the author of https://hnprofile.com https://hnprofile.com - I've spent a long time analyzing this... For example, you post something about PostgreSQL on HN it'll have a ton of discussion. That's not just because there is squabbling, it's because people are interested. Often it is a disagreement, but it could be providing additional context. Take this comment thread on PostgeSQL 11: https://news.ycombinator.com/item?id=17962821 https://news.ycombinator.com/item?id=17962821 The comments are significantly more interesting and helpful than the article (not that the article is bad, just the collective knowledge of HN provides more value). I often upvote items because I think they are interesting, but I comment asking questions, adding context, or providing feedback. That's the value of hacker news, IMO. If you filter that out, I might as well just go to my feedly. Edit: Prior discussion about hnprofile: https://news.ycombinator.com/item?id=17942981 https://news.ycombinator.com/item?id=17942981

mirimir 8y ago
Yes. In many cases, comments help with understanding posts, and seeing value vs bullshit. Also, I sometimes find myself on HN, when researching problems, and it's comments and cited resources that are most useful.
msvan 8y ago
So even Hacker News data gets analyzed, monetized and sold to third parties. When did I opt in to you scraping my profile and determining my mood?
Tomte 8y ago
I often read the comments first, and then determine whether I should read the article. IMO, high upvote count without comments is a smell. Something is probably a little off.
- lettergram 8y ago
  Exactly, it's often super clear there are voting rings. I've seen some posts shoot up to the top, that make absolutely no sense (often they drop out relatively quickly, once they hit the front page).
mmt 8y ago
> Take this comment thread on PostgeSQL 11: That example doesn't seem to support your point. 211 upvote with only 54 comments seems like a thread with a very high votes-to-comments ratio as well as a high absolute number of upvotes, so I would expect the algorithm to rate it highly. Perhaps, enough readers looking at the thread and enjoying the article or the discussion would leave an upvote, even if they don't leave a comment. However, in the "squabbly" threads, commenters seem to overrun upvoters, and quickly at that. I've seen some where someone was creating new HN accounts to circumvent HN's comment rate-limiting, due to wanting to respond on so many sub-threads. It's difficult to imagine that kind of fervor resulting in thoughtful discourse or even conforming to HN's guideline regarding being more substantive when more divisive. As such, a high comment-count velocity would be a good "negative" measure, but, absent that, just comment count is a good enough heuristic. > If you filter that out, I might as well just go to my feedly. Personally, I don't think I'd use it as an exclusive filter, but, perhaps, as a next step after the HN front page (but before the HN page 2).
- lettergram 8y ago
  I think this post itself is a great example. At the moment, there's a post about vaccines above this post. It has had very little discussions, isn't really about "hacking". Literally the first comment is complaining how it's lacking: https://news.ycombinator.com/item?id=17997400 https://news.ycombinator.com/item?id=17997400 In the OP's link, the vaccine article appears above this post by a wide margin. It shouldn't.
  - mmt 8y ago
    > It has had very little discussions, That can be viewed as a very good thing, considering the word "vaccine" can be a lightning-rod fod flamewars, even on HN. > isn't really about "hacking" It's about science and how it's funded, which is routinely a topic of interest here, and is well within the guidelines: "That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity." > In the OP's link, the vaccine article appears above this post by a wide margin. It shouldn't. Ah, but I disagree. The discussion in this thread is very "meta" and could easily be characterized as annoying, even though I don't (yet) personally find it to be.

mirimir 8y ago

Yes. In many cases, comments help with understanding posts, and seeing value vs bullshit. Also, I sometimes find myself on HN, when researching problems, and it's comments and cited resources that are most useful.

msvan 8y ago

So even Hacker News data gets analyzed, monetized and sold to third parties. When did I opt in to you scraping my profile and determining my mood?

Tomte 8y ago

I often read the comments first, and then determine whether I should read the article. IMO, high upvote count without comments is a smell. Something is probably a little off.

lettergram 8y ago
Exactly, it's often super clear there are voting rings. I've seen some posts shoot up to the top, that make absolutely no sense (often they drop out relatively quickly, once they hit the front page).

mmt 8y ago

> Take this comment thread on PostgeSQL 11: That example doesn't seem to support your point. 211 upvote with only 54 comments seems like a thread with a very high votes-to-comments ratio as well as a high absolute number of upvotes, so I would expect the algorithm to rate it highly. Perhaps, enough readers looking at the thread and enjoying the article or the discussion would leave an upvote, even if they don't leave a comment. However, in the "squabbly" threads, commenters seem to overrun upvoters, and quickly at that. I've seen some where someone was creating new HN accounts to circumvent HN's comment rate-limiting, due to wanting to respond on so many sub-threads. It's difficult to imagine that kind of fervor resulting in thoughtful discourse or even conforming to HN's guideline regarding being more substantive when more divisive. As such, a high comment-count velocity would be a good "negative" measure, but, absent that, just comment count is a good enough heuristic. > If you filter that out, I might as well just go to my feedly. Personally, I don't think I'd use it as an exclusive filter, but, perhaps, as a next step after the HN front page (but before the HN page 2).

lettergram 8y ago
I think this post itself is a great example. At the moment, there's a post about vaccines above this post. It has had very little discussions, isn't really about "hacking". Literally the first comment is complaining how it's lacking: https://news.ycombinator.com/item?id=17997400 https://news.ycombinator.com/item?id=17997400 In the OP's link, the vaccine article appears above this post by a wide margin. It shouldn't.
- mmt 8y ago
  > It has had very little discussions, That can be viewed as a very good thing, considering the word "vaccine" can be a lightning-rod fod flamewars, even on HN. > isn't really about "hacking" It's about science and how it's funded, which is routinely a topic of interest here, and is well within the guidelines: "That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity." > In the OP's link, the vaccine article appears above this post by a wide margin. It shouldn't. Ah, but I disagree. The discussion in this thread is very "meta" and could easily be characterized as annoying, even though I don't (yet) personally find it to be.