It was suggested that requiring a category for a down-vote would create an asymmetry between up- and down-votes, so I am amending my suggestion to include up-votes along with down-votes.

This may be a partial duplicate, but the two ideas seem to belong together, so I am repeating the "comment" suggestion, (but adding to it the suggestion that an anonymous comment feature be added for down-votes.)

When I invest considerable time and effort in formulating and composing a question which I post to math.se, and have it down-voted, that is discouraging. It reduces my motivation to contribute, and rarely communicates anything helpful. I often suspect some kind of bigotry, bias or other petty motive is behind the down-vote, rather than an effort to improve quality of content.

As I understand it, the reason down-voting is anonymous is that it encourages people to down-vote because, for some reason down-votes are considered valuable. IMO this encourages quantity over quality of down-votes.

I will suggest that down-votes be categorized. In other words, down-voting would require the selection of a reason (poorly formulated|off-topic|insufficient effort|is open-ended|don't like user|that's homework you should do on your own). I also recommend including the suggestion that the down-voter leave a comment (possibly anonymously) explaining the reason.

I understand that anonymity of down-voting is a way of avoiding "down-vote wars", but it has its drawbacks as well.

I believe the suggested features would also encourage participants to be more thoughtful in their voting practices. That is, it would remind them of what a vote is intended to communicate.

  • 5
    $\begingroup$ How do you propose to force people to be honest about their down-vote reasons? $\endgroup$
    – postmortes
    Jun 3, 2019 at 17:10
  • 4
    $\begingroup$ Variants on this idea have been discussed many times before here. Some of the main problems are 1) lack of anonymity leading to harassment, 2) a failure of duality (in the sense that upvotes aren't treated in the same way), 3) a bunch more noise that's not necessarily helpful, especially since people have no reason to explain their votes, and 4) that requiring vote explanations is a fundamental change in how things work here and decreases the usefulness of the feedback, because users can be targeted for making quality-control efforts. $\endgroup$
    – user296602
    Jun 3, 2019 at 17:14
  • 1
    $\begingroup$ We really can't protect against dishonesty. But the value of honest down-votes would be increased. $\endgroup$ Jun 3, 2019 at 17:17
  • 2
    $\begingroup$ How to tell if an upvote/downvote is honest or dishonest? $\endgroup$ Jun 3, 2019 at 17:24
  • 1
    $\begingroup$ @ArcticChar isn't that already a problem? $\endgroup$ Jun 3, 2019 at 17:25
  • 3
    $\begingroup$ You said the value of honest downvote should be increased. I am asking how to locate such votes. $\endgroup$ Jun 3, 2019 at 17:27
  • 1
    $\begingroup$ @ArcticChar the good-faith assumption is that most votes are honest. We should be concerned with the quality of honest participation. We can't protect against dishonesty. But, even if a vote is not properly motivated, having it categorized would be more constructive than the simple implication that "I don't like the questions (or the person asking it)" $\endgroup$ Jun 3, 2019 at 17:34
  • 1
    $\begingroup$ Why do you propose this for downvotes? What about upvotes? What about "I've upvoted this because it was downvoted, and I don't know what kind of quality this answer has, but I just dislike the fact it has a negative score" reason for upvote? Wouldn't that be great too? $\endgroup$
    – Asaf Karagila Mod
    Jun 3, 2019 at 17:41
  • 1
    $\begingroup$ @AsafKaragila I amended my suggestion to include up-votes. One of the categories could simply be "I like/dislike the question". At least that would communicate to the other person what motivated the vote. $\endgroup$ Jun 3, 2019 at 17:44
  • 3
    $\begingroup$ @AsafKaragila There is only one choice: allow voting on voting. Recursively probably. $\endgroup$
    – rschwieb
    Jun 3, 2019 at 17:45
  • 1
    $\begingroup$ @rschwieb: Recursive implies that there is a well-founded relation to define the halting condition. I don't see that here. $\endgroup$
    – Asaf Karagila Mod
    Jun 3, 2019 at 17:46
  • 2
    $\begingroup$ @rschwieb I don't understand how that is the "only" solution; or even an effective solution. My primary objective is to improve the quality of feedback communicated by votes, by including more information than a simple Boolean value. $\endgroup$ Jun 3, 2019 at 17:58
  • 2
    $\begingroup$ @StevenThomasHatton Sure, you can gauge the quality of the votes by their meta-votes. But yeah seriously if you didn't detect it already this was tongue-in-cheek. I did not think that aspect would go unnoticed. $\endgroup$
    – rschwieb
    Jun 3, 2019 at 18:00

1 Answer 1


This pops up if you flag a post:

On first flag

Something similar to this, pops up if you select should be closed: Should be closed image

Do you see how, a lot of these are similar to the suggestions you make for the downvotes ? You'll Also find that you can get 5 downvotes on an answer, before it cancels the reputation points from a single upvote. On a question the downvotes mean more. You'll also find a downvote asks you to leave a comment, it's not generally enforced. People who are spammy or crackpots, don't generally give up unless challenged. They are subject to the Dunning Kruger effect, as are some of your own postings.

Past postings by you, state that you being off-topic is on the community. Not entirely true, your being off-topic could come down to the level of postings, in a given math community. More advanced, adequately researched and on topic questions, are likely to warrant moving them to MathOverflow. Off-topic rants, could have been simply deleted, instead of downvoted.

  • $\begingroup$ "They are subject to the Dunning Kruger effect, as are some of your own postings." This sounds borderline like a personal attack. This is not a bad answer, but try to keep your comments restricted to the topic at hand, not the user themselves, please. $\endgroup$
    – YiFan Tey
    Aug 10, 2019 at 6:14
  • $\begingroup$ topic included the persons postings in the body of the original post. $\endgroup$
    – user645636
    Aug 10, 2019 at 9:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .