4 ms·

I also would appreciate knowing if the mods see this. I'm worried that flagging is possibly automated and vulnerable to campaigning.

dang 8mo ago
We eventually saw it, but only randomly. In case you didn't see them yet, here are some of my other comments in this thread: https://news.ycombinator.com/item?id=47141443 https://news.ycombinator.com/item?id=47141443 https://news.ycombinator.com/item?id=47141678 https://news.ycombinator.com/item?id=47141678 https://news.ycombinator.com/item?id=47141517 https://news.ycombinator.com/item?id=47141517 Re the concern about flagging, the situation is much as I've described in these past threads: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=by%3Adang%20coalition%20flags&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que.... Specifically, when I looked through who had flagged the current post, I saw the usual coalition between users who appear to be consistently flagging for political reasons, and other users who have quite different flagging patterns than that. In any case, virtually all of the accounts that flagged the thread were established HN users. Sometimes when people bring this concern up, I go through and make a list of other stories that the same accounts had flagged, to illustrate the point that their flags are not exclusively targeting one specific topic or vector. I've done that here in a collapsed reply, if anyone wants to take a look. I hope this explanation helps - your posts in this thread seemed to me to be in good faith so I wanted to respond in kind. If you still have a question that my comments and links to past explanations haven't answered, I'd be happy to take a crack at it.
- dang 8mo ago
  Here are some stories that flaggers of this submission also flagged. I have no idea why, except for the handful of obvious spam, but it illustrates the point I made in the parent comment. The rise and fall of peer review - https://news.ycombinator.com/item?id=47123133 https://news.ycombinator.com/item?id=47123133 - Feb 2026 (0 comments) Ladybird adopts Rust, with help from AI - https://news.ycombinator.com/item?id=47120899 https://news.ycombinator.com/item?id=47120899 - Feb 2026 (692 comments) Pope tells priests to use their brains, not AI, to write homilies - https://news.ycombinator.com/item?id=47119210 https://news.ycombinator.com/item?id=47119210 - Feb 2026 (440 comments) Music Discovery - https://news.ycombinator.com/item?id=47114672 https://news.ycombinator.com/item?id=47114672 - Feb 2026 (56 comments) The 7-Year Bug That Took 3 Minutes to Fix - https://news.ycombinator.com/item?id=47090261 https://news.ycombinator.com/item?id=47090261 - Feb 2026 (1 comment) AI made coding more enjoyable - https://news.ycombinator.com/item?id=47075400 https://news.ycombinator.com/item?id=47075400 - Feb 2026 (97 comments) RFC 3092 – Etymology of “Foo” (2001) - https://news.ycombinator.com/item?id=46934499 https://news.ycombinator.com/item?id=46934499 - Feb 2026 (52 comments) Launching My Side Project as a Solo Dev: The Walkthrough - https://news.ycombinator.com/item?id=46845567 https://news.ycombinator.com/item?id=46845567 - Feb 2026 (9 comments) There is an AI code review bubble - https://news.ycombinator.com/item?id=46766961 https://news.ycombinator.com/item?id=46766961 - Jan 2026 (249 comments) Proof of Corn - https://news.ycombinator.com/item?id=46735511 https://news.ycombinator.com/item?id=46735511 - Jan 2026 (307 comments) XLibre XServer 25.1 Changes - https://news.ycombinator.com/item?id=46474846 https://news.ycombinator.com/item?id=46474846 - Jan 2026 (4 comments) Python Data Science Handbook - https://news.ycombinator.com/item?id=46120611 https://news.ycombinator.com/item?id=46120611 - Dec 2025 (61 comments) NTSB Preliminary Report – UPS Boeing MD-11F Crash [pdf] - https://news.ycombinator.com/item?id=45995834 https://news.ycombinator.com/item?id=45995834 - Nov 2025 (228 comments) Best shipping logistic aggregator in India - https://news.ycombinator.com/item?id=45924139 https://news.ycombinator.com/item?id=45924139 - Nov 2025 (0 comments) WebDAV isn't dead yet - https://news.ycombinator.com/item?id=45698070 https://news.ycombinator.com/item?id=45698070 - Oct 2025 (128 comments) Unicode Footguns in Python - https://news.ycombinator.com/item?id=45689443 https://news.ycombinator.com/item?id=45689443 - Oct 2025 (20 comments) AGI is not imminent, and LLMs are not the royal road to getting there - https://news.ycombinator.com/item?id=45627171 https://news.ycombinator.com/item?id=45627171 - Oct 2025 (124 comments) Super Ace: Your PH Home for Jili Slots and a 300% Welcome Bonus - https://news.ycombinator.com/item?id=45624939 https://news.ycombinator.com/item?id=45624939 - Oct 2025 (0 comments) Pkgbase Removes FreeBSD Base System Feature - https://news.ycombinator.com/item?id=44730021 https://news.ycombinator.com/item?id=44730021 - July 2025 (42 comments)
  - ryandrake 8mo ago
    As always, thanks for the transparency. It seems that people, even "established HN users" will flag literally anything. Do you feel that there is any remaining article quality signal that can be obtained from the current flagging mechanism?
    - dang 8mo ago
      If the above list gives the mistaken impression that flagging is basically random, that's an artifact of the way I cherry-picked the list. The flagging system has problems, for sure, but it's a vital part of how HN's system functions. If you squint and look closely, though, I think you can detect this in the above list. The weirdest "wtf?" cases of flagging are ones where the threads had a lot of comments and were on the frontpage. That means upvotes won the tug-of-war with flags, as they should have in most of those cases. Conversely, it you look at the submissions in the list which had 0 comments or very few, it looks to me like most were either spam, low-quality articles, or dupes. Remember, also, that some flags are just mistakes - the link is easy to fat-finger or misclick, and the UI doesn't provide feedback about that. That's likely to change soon as part of work that tomhow and I are planning.
      - ryandrake 8mo ago
        I know I'm changing the goal posts here, but out of that list, while the articles' quality does vary, I don't see much (or any) rulebreaking, besides https://news.ycombinator.com/item?id=45924139 https://news.ycombinator.com/item?id=45924139 and https://news.ycombinator.com/item?id=45624939 https://news.ycombinator.com/item?id=45624939
        dang 8mo ago
        Are you saying that flaggers flag things that shouldn't be flagged? If so, I agree, of course. That sort of flag is usually outweighed by other variables though.

dang 8mo ago

We eventually saw it, but only randomly. In case you didn't see them yet, here are some of my other comments in this thread: https://news.ycombinator.com/item?id=47141443 https://news.ycombinator.com/item?id=47141443 https://news.ycombinator.com/item?id=47141678 https://news.ycombinator.com/item?id=47141678 https://news.ycombinator.com/item?id=47141517 https://news.ycombinator.com/item?id=47141517 Re the concern about flagging, the situation is much as I've described in these past threads: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=by%3Adang%20coalition%20flags&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que.... Specifically, when I looked through who had flagged the current post, I saw the usual coalition between users who appear to be consistently flagging for political reasons, and other users who have quite different flagging patterns than that. In any case, virtually all of the accounts that flagged the thread were established HN users. Sometimes when people bring this concern up, I go through and make a list of other stories that the same accounts had flagged, to illustrate the point that their flags are not exclusively targeting one specific topic or vector. I've done that here in a collapsed reply, if anyone wants to take a look. I hope this explanation helps - your posts in this thread seemed to me to be in good faith so I wanted to respond in kind. If you still have a question that my comments and links to past explanations haven't answered, I'd be happy to take a crack at it.

dang 8mo ago
Here are some stories that flaggers of this submission also flagged. I have no idea why, except for the handful of obvious spam, but it illustrates the point I made in the parent comment. The rise and fall of peer review - https://news.ycombinator.com/item?id=47123133 https://news.ycombinator.com/item?id=47123133 - Feb 2026 (0 comments) Ladybird adopts Rust, with help from AI - https://news.ycombinator.com/item?id=47120899 https://news.ycombinator.com/item?id=47120899 - Feb 2026 (692 comments) Pope tells priests to use their brains, not AI, to write homilies - https://news.ycombinator.com/item?id=47119210 https://news.ycombinator.com/item?id=47119210 - Feb 2026 (440 comments) Music Discovery - https://news.ycombinator.com/item?id=47114672 https://news.ycombinator.com/item?id=47114672 - Feb 2026 (56 comments) The 7-Year Bug That Took 3 Minutes to Fix - https://news.ycombinator.com/item?id=47090261 https://news.ycombinator.com/item?id=47090261 - Feb 2026 (1 comment) AI made coding more enjoyable - https://news.ycombinator.com/item?id=47075400 https://news.ycombinator.com/item?id=47075400 - Feb 2026 (97 comments) RFC 3092 – Etymology of “Foo” (2001) - https://news.ycombinator.com/item?id=46934499 https://news.ycombinator.com/item?id=46934499 - Feb 2026 (52 comments) Launching My Side Project as a Solo Dev: The Walkthrough - https://news.ycombinator.com/item?id=46845567 https://news.ycombinator.com/item?id=46845567 - Feb 2026 (9 comments) There is an AI code review bubble - https://news.ycombinator.com/item?id=46766961 https://news.ycombinator.com/item?id=46766961 - Jan 2026 (249 comments) Proof of Corn - https://news.ycombinator.com/item?id=46735511 https://news.ycombinator.com/item?id=46735511 - Jan 2026 (307 comments) XLibre XServer 25.1 Changes - https://news.ycombinator.com/item?id=46474846 https://news.ycombinator.com/item?id=46474846 - Jan 2026 (4 comments) Python Data Science Handbook - https://news.ycombinator.com/item?id=46120611 https://news.ycombinator.com/item?id=46120611 - Dec 2025 (61 comments) NTSB Preliminary Report – UPS Boeing MD-11F Crash [pdf] - https://news.ycombinator.com/item?id=45995834 https://news.ycombinator.com/item?id=45995834 - Nov 2025 (228 comments) Best shipping logistic aggregator in India - https://news.ycombinator.com/item?id=45924139 https://news.ycombinator.com/item?id=45924139 - Nov 2025 (0 comments) WebDAV isn't dead yet - https://news.ycombinator.com/item?id=45698070 https://news.ycombinator.com/item?id=45698070 - Oct 2025 (128 comments) Unicode Footguns in Python - https://news.ycombinator.com/item?id=45689443 https://news.ycombinator.com/item?id=45689443 - Oct 2025 (20 comments) AGI is not imminent, and LLMs are not the royal road to getting there - https://news.ycombinator.com/item?id=45627171 https://news.ycombinator.com/item?id=45627171 - Oct 2025 (124 comments) Super Ace: Your PH Home for Jili Slots and a 300% Welcome Bonus - https://news.ycombinator.com/item?id=45624939 https://news.ycombinator.com/item?id=45624939 - Oct 2025 (0 comments) Pkgbase Removes FreeBSD Base System Feature - https://news.ycombinator.com/item?id=44730021 https://news.ycombinator.com/item?id=44730021 - July 2025 (42 comments)
- ryandrake 8mo ago
  As always, thanks for the transparency. It seems that people, even "established HN users" will flag literally anything. Do you feel that there is any remaining article quality signal that can be obtained from the current flagging mechanism?
  - dang 8mo ago
    If the above list gives the mistaken impression that flagging is basically random, that's an artifact of the way I cherry-picked the list. The flagging system has problems, for sure, but it's a vital part of how HN's system functions. If you squint and look closely, though, I think you can detect this in the above list. The weirdest "wtf?" cases of flagging are ones where the threads had a lot of comments and were on the frontpage. That means upvotes won the tug-of-war with flags, as they should have in most of those cases. Conversely, it you look at the submissions in the list which had 0 comments or very few, it looks to me like most were either spam, low-quality articles, or dupes. Remember, also, that some flags are just mistakes - the link is easy to fat-finger or misclick, and the UI doesn't provide feedback about that. That's likely to change soon as part of work that tomhow and I are planning.
    - ryandrake 8mo ago
      I know I'm changing the goal posts here, but out of that list, while the articles' quality does vary, I don't see much (or any) rulebreaking, besides https://news.ycombinator.com/item?id=45924139 https://news.ycombinator.com/item?id=45924139 and https://news.ycombinator.com/item?id=45624939 https://news.ycombinator.com/item?id=45624939
      - dang 8mo ago
        Are you saying that flaggers flag things that shouldn't be flagged? If so, I agree, of course. That sort of flag is usually outweighed by other variables though.