12

This question is inspired by a recent interaction, with one of the most helpful and prolific users of this site, as outlined in comments to an old answer, here.

Is it really this community's intention to require comments, prior to down-voting? I was unaware of this. Is it documented in any official pages (e.g., any side panels or the site's Tour page)? (This prior answer is relevant, but not definitive, nor sufficient in scope to address this inquiry.)

Does a majority (or perhaps majority of Moderators) really subscribe to this notion? I can see the argument and perhaps it might indeed make people feel better (which I naturally do consider important). I never personally viewed a down-vote as particularly negative nor rude, however—simply an efficient mechanism to express one's preferences regarding a question or an answer, and to effectively motivate improvements or (mostly) signal extrinsic accuracy.

How does this putative policy differ (if at all) for questions, vs. answers? How does it differ concerning old questions, for which newer information or additional research might be relevant? Are users really expected to have to engage substantively, prior to their being able to express a hopefully constructive preference? In particular, is it not desirable to allow quick expression of preferences, for users willing to express them (and take the Reputation penalty), especially in situations in which an answer is old and may be improved? Is the onus to suggest edits or to comment really to be delegated to the user, as opposed to merely being preferred?

In this particular situation, I had admittedly not initially noted that I was down-voting such an active user. So naturally, more constructive engagement could have been preferable. Nonetheless, I am surprised that this appeared to be interpreted negatively and might be contrary to policy.

This seems to run contrary to the usual practice across Stack Exchange or at least how the model is generally outlined and presented. If this is to be the determination of the relevant quorum, perhaps this ought be more substantively outlined and documented?

6 Answers 6

Reset to default
15

There's never been a requirement to provide a reason to downvote, more a sense that downvoting is sufficiently unusual that where it's used, giving an explanation is probably useful. (At a technical level requiring comment is not enforceable locally in any case.)

As questions and answers start at zero votes, most of the time one can show approval by upvoting and simply leave other content entirely alone. The exception is spam or similar, where a rapid application of downvotes can trigger automatic removal.

7
  • 1
    @CobyViner On assessing answers: That doesn't depend on downvotes. Over time, content will attract upvotes, and in a situation where there are alternative answers, it's likely that the vote distribution will not be even amongst all of them.
    – Joseph Wright Mod
    Jan 23 at 8:37
  • 4
    @CobyViner On the local consensus: Pretty early on in the life of the site, the 'regulars' came to the conclusion that downvotes are a dis-intensive to new(er) users, the same as closing questions. So there was a move toward voting up good content and leaving other content simply unvoted. Of course, there are users who downvote, and they are entitled to (I have no issue provided it's not targetted), but most of the 'regulars' tend toward just using upvotes.
    – Joseph Wright Mod
    Jan 23 at 8:40
  • 3
    @CobyViner On the idea that a comment is useful: A downvote without feedback doesn't really allow the poster to know what to do. So the idea that something that is sufficiently bad to need a downvote would also benefit from a supporting comment seems to flow naturally.
    – Joseph Wright Mod
    Jan 23 at 8:42
  • Interesting. Not sure I quite agree (especially on answer assessment, for which down-votes could certainly substantively contribute, if they were more popular or to perhaps valuably disagree with a populist answer). Notwithstanding, I really appreciate this clear perspective on matters—thanks!
    – Coby Viner
    Jan 23 at 8:43
  • 2
    @CobyViner Re "Why is there a sense that it's unusual, however?". At least in the past, it is definitely unusual. See tex.meta.stackexchange.com/questions/4190/… , physics.meta.stackexchange.com/questions/6754/…. E.g., in 2015 only 0.15% of posts on tex had a negative score. (I'd be interested in the new data, but I do not have time at this moment.)
    – Chris Chudzicki
    Jan 23 at 14:12
  • 1
    This won't be popular, but there should definitely be a written sitewide policy prohibiting users from publicly "outing" downvoters especially when voting is supposed to be "anonymous" (which it certainly is not). It is tantamount to bullying. There are some users here who feel they are beyond reproach and they can't stand being disagreed with in matters of opinion. That's why I've stopped voting and interacting. Reporting to mods does nothing as such behavior is part of the SE culture unfortunately. I also think rudeness should warrant downvoting.
    – LaTeXereXeTaL
    Jan 30 at 2:12
  • 3
    @LaTeXereXeTaL I think it's worth noting that whilst one can ask about a vote, there's no way of forcing that person to reveal it. (Voting is anonymous in the sense that no-one here can see individual votes.) On site-wide policy, perhaps you could open a separate meta question?
    – Joseph Wright Mod
    Jan 30 at 6:54
24

Voting is, by design, anonymous, so you should not be required to comment on a downvote. However since a downvote with no explanation will be percieved as (and in fact, is) unfriendly and unconstructive, it's better to simply leave it. If you have no time or desire to comment or edit the post or post a new answer, simply leave the post without voting.

It's true that other sites on the network use downvoting more freely but that isn't something that makes those sites welcoming and it doesn't really affect the way downvotes will be perceived here. If you post a badly formed question on stackoverflow you may get to -10 within minutes, it's just the way it is. But here as it is obvious that very few posts get downvotes, a person who receives one is likely to feel it as personal criticism and you should consider that before downvoting. It does not require any site policy statement.

9
  • 3
    +1 for “[…] downvoting … isn't something that makes those sites welcoming […]” and “a person who receives one is likely to feel it as personal criticism”. If I could, I would give more upvotes.
    – enthu
    Jan 23 at 10:54
  • Interesting, thanks. While I take your and others' points, it might be helpful to clarify in a visible manner, as this does differ markedly from other SE sites and does seem to run contrary to the site's inherent design.
    – Coby Viner
    Jan 23 at 13:28
  • 1
    @CobyViner the (diamond) mods on the site have very limited control over the information that is presented in the help pages, so it is not really feasible to communicate the downvoting culture that we have to new community members. However, most new people quickly understand how this works, in most cases before they have the privilege to downvote and/or before they encounter a post that they would have downvoted if it was on another SE site.
    – Marijn
    Jan 23 at 16:04
  • @Marijn I'm not sure how you expect people to intuit that. Perhaps I was particularly clueless in those regards, but I hardly think that's entirely fair… it is the prevailing culture on many other SE sites, and indeed officially advocated by SE, writ large. Admittedly for reasons that are not entirely applicable here and these answers have given strong reasons to prefer otherwise. That's precisely the sort of thing that users could be informed of, somehow.
    – Coby Viner
    Jan 23 at 16:24
  • 3
    @CobyViner perhaps surprisingly, even though I agree that this is unexpected/un-expectable, it still happens and many people do in fact grasp the culture quickly. When I started myself I once downvoted a high-rep regular, he got angry, I asked clarification in the general TeX.SE chat room and that was that :)
    – Marijn
    Jan 23 at 16:36
  • 1
    That said, I agree with you that it would be preferred to make this explicit for new users, but that is not something that we are able to do unfortunately. But again, it works itself out somehow in general.
    – Marijn
    Jan 23 at 16:37
  • Fair enough. Thanks for your perspective.
    – Coby Viner
    Jan 23 at 16:37
  • 3
    @CobyViner perhaps this meta question will encourage long-time users to explain our preferences a bit more gently when such a situation presents itself - thank you for posting it in any case, I think it is good to bring attention to the topic.
    – Marijn
    Jan 23 at 16:46
  • @Marijn I hope so. Thanks—that's much appreciated!
    – Coby Viner
    Jan 23 at 16:47
14

Many of the people posting to this site are (relative) newbies. Many are also not fluent English speakers. Furthermore, (La)TeX is often not welcomed in many environments, with a preference for Word clearly expressed by management.

Questions: A question that is poorly or not clearly phrased or doesn't contain precise details usually won't get a usable or correct answer, and wastes the time of someone trying to help. Not providing information on why a question can't be answered in its original form simply discourages an OP and is unlikely to win an enthusiastic user; that's what an unexplained downvote does.

Answers: Sometimes an answer is simply wrong or will have unintended consequences. A downvote without explanation is unhelpful, especially if the answer looks plausible or attractive, which it often does. Of course, if an answer is clearly spam. a downvote is justified; but in that case, there's also the option to flag it.

Personal experience: On another site, I once posted, in answer to a request for information, an announcement that a manual, published by my employer, previously available only for sale at a nontrivial price had been released as a free PDF file. This was very quickly downvoted, with the reason given that the posting was for personal aggrandizement. It was my first posting on that site, but not intended personally. A poster on tex.sx, experienced on the other site, intervened to rectify the situation, but that was my last posting, and nearly my last visit to the other site.

Downvotes can be taken as personal attacks. I strongly prefer to avoid that possibility.

0
10

I never personally viewed a down-vote as particularly negative nor rude, however—simply an efficient mechanism to express one's preferences regarding a question or an answer, and to effectively motivate improvements or (mostly) signal extrinsic accuracy.

Well you own example shows that starting an interaction with a down-vote is not really efficient.

If you don't explain why you downvote it seldom improves a question or answer as people typically don't understand the reason and ignore the downvote (or are slightly hurt).

If you explain why you downvote you not only have to explain what is wrong with the answer but additionally why you thought it fit to downvote. And you have to do some extra work to motivate people as downvotes don't motivate: they either depress or make angry or are ignored. Personally I don't like to do something for people who think that they can push me by first giving me a slight slap on the head.

It would have been much more efficient to leave a comment which clarifies your open point and ask friendly if the answer could be improved.

6
  • 1
    I don't think that's true… as a matter of fact, I had commented prior, just not explicitly providing the reason for the down-vote. It was correctly inferred. I'm not sure how you drew these conclusions.
    – Coby Viner
    Jan 23 at 8:51
  • 1
    sorry but it really doesn't matter what SE thinks of downvotes, it matters what the people you interact with think about it. And it is not efficient: You now produced lots of comments and a meta question only because of one downvote. So obviously something is wrong with your idea that it is efficient (and you are not the first).
    – Ulrike Fischer
    Jan 23 at 9:01
  • 1
    the times in the comments show clearly that you did downvote at the begin of the interaction, egreg had no time to reject your comment nor sensible time (e.g. a day) to address it. And sorry but community norms can't change human behaviour: the majority of people don't react with joy if someone takes something from them. Sites with lots of downvotes aren't described as efficient but as unfriendly.
    – Ulrike Fischer
    Jan 23 at 9:14
  • 2
    how in your view can a "negligible reputational loss" "effectively motivate improvements"? Either you believe that downvotes have a noticable effect on people, then this can only be an unfriendly effect as you expect people to do something to avoid it, or you believe it is "negligible", then why doing it at all?
    – Ulrike Fischer
    Jan 24 at 9:35
  • 1
    I doubt that many apart from the op ever notice a downvote on an answer if the overall sum is positive. I certainly never would check that. Do you really click to see the numbers when you read an answer? And consider to add another answer if you see a single downvote in the middle of 20 upvotes?
    – Ulrike Fischer
    Jan 25 at 10:10
  • On occasion. Not from a single down-vote, but from a few, yes. I think we can agree to disagree here. Appreciate your perspective on matters and certainly acknowledge yours is the majority here—thanks for your input!
    – Coby Viner
    Jan 25 at 19:26
4

My opinion:

In theory, no.

If everyone doesn't take downvotes personally, would the site be a better and more efficient place? Probably yes.

Unfortunately, in real life there are many people who don't _want_ their posts to be downvoted. And there isn't much can be done about it.

Regarding the efficiency, I think this site is small enough (at least for now) that it can afford to take every user to look through every questions without them getting downvoted (we I is the real purpose of downvoting questions).

On the other hand, if you leave a comment for improvement and downvote at the same time (which is what you did in this case), it would probably be a good idea to set a timer to check the answer later for improvement to retract the downvote. But if you do that, you might as well set a timer and downvote _after_ there's no improvement for X days if you still want to.

0

There is a saying, along the lines of: "A written job reference tells you nothing about the applicant. It only tells you something about the person who wrote the reference."

I believe it is the same with downvotes. Unless a post is clearly spam or utterly wrong, a downvote tells use something about the voter, rather than about the post.

That can also be said about upvotes, but good answers deserve attention.

3
  • 2
    ...so bad answers also deserve attention? If attention = votes.
    – Werner Mod
    Feb 2 at 0:28
  • ... Not what I meant. Hard to explain.
    – rallg
    Feb 2 at 20:53
  • 3
    Your answer, interpreted as written, implies that upvotes also need a comment, otherwise it only says something about the voter, not the content/answer. Regardless, it goes to show how good answers aren't easy to write up. The same goes for good questions.
    – Werner Mod
    Feb 2 at 20:59

You must log in to answer this question.

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