2
$\begingroup$

I just noticed this question, asked by a new user, which has two downvotes with no comments. I don't think this question is bad—it could use more explanation, and more demonstration of the OP's having put effort into figuring out the answer. But how can the OP know how to improve the question without getting comments?

Is there anything we can do so that we stop discouraging people from asking questions on this site by downvoting them indiscriminately?

$\endgroup$
3
$\begingroup$

There are several reasons to down-vote a post. We have discussed previously (see this and this) that we are not going to require or even demand people to explain their votes. When I see a down-vote that I don't understand I politely comment that I don't understand it, but I think demanding an explanation or stating that there is no reason for the down-vote goes too far. It is not an abuse if people are using their votes in a way different from us.

I have argued a number of times that I am against counter-voting, that is voting to null the effect of another user's vote. Everyone has their own opinion. The total reflects the opinion of the community, it is not the job of any particular user to decide what the total should be. Counter-voting encourages other users to do the same and use their votes to null votes in place of express their opinion about the question. Each vote is a manifestation of some user's opinion and if we respect their opinions then we should respect their votes even when we disagree with them.

I think it is more productive if we comment and explain why we think it is a suitable and interesting question in place of just asking others why they have down-voted the post.

In this particular case, I see several possible reasons for a down-vote.

  1. Formating: It is using an image in place of writing the expression which I think has been one of the reasons people down-vote.

  2. Content: The questions seems to simply ask how to calculate some expression for some particular values. It seems to ask how to apply a method explained in a paper in a particular case in practice. It is too localized and it doesn't seem to be a research-level question in theoretical computer science. It is also unclear what is the trouble the OP is facing in calculating the expression.

If you ask me I think we are being too open right now (I guess it is partly because people get bored that there are not as many questions as they would like) to questions which should not be here and are more suitable for CS.SE. E.g. 1, 2, 3, 4 have been in the closing review queue for sometime but have not been put on-hold. Similarly, it is not clear what this question is asking and should be put on-hold until clarified but it does not have a single close vote. And this is clearly not-research level in my opinion.

$\endgroup$
  • 5
    $\begingroup$ I believe the OP is trying to understand a particular step in this ground-breaking paper by asking how to do this calculation, which is important for the paper but I don't believe is explained very well in it ... I expect he is asking for the specific numbers because he thinks he will understand it better that if he's given an example (and I expect he's correct; seeing how an example works for one particular set of numbers is easier to understand than an explanation of the general procedure). So he's asking for an explanation of a particular step in this paper. This should be research level. $\endgroup$ – Peter Shor Sep 10 '13 at 18:31
  • 1
    $\begingroup$ @Peter, I see. I was giving possible reasons for the down-vote. Maybe you should edit the question and make your comment explicit inside the question so it is clear that this is suitable for cstheory. $\endgroup$ – Kaveh Sep 10 '13 at 18:36
  • $\begingroup$ Actually, those possible reasons for the down-vote do make sense, if you don't know the paper. I just wish people would comment when they down-vote when it's not obvious why they're doing it. $\endgroup$ – Peter Shor Sep 10 '13 at 18:41
  • 3
    $\begingroup$ @Peter, I would also prefer if people comment when they vote but simply stating this wish doesn't make it happen. It has a lot to do with the reactions users get when they comment. $\endgroup$ – Kaveh Sep 10 '13 at 18:44
  • 1
    $\begingroup$ "I have argued a number of times that I am against counter-voting, that is voting to null the effect of another user's vote. Everyone has their own opinion. The total reflects the opinion of the community, it is not the job of any particular user to decide what the total should be. " huh??? this makes no sense to me, the sentences seem contradictory. what is "countervoting"? voting in a different way than existing votes? $\endgroup$ – vzn Sep 10 '13 at 22:36
  • 1
    $\begingroup$ "I guess it is partly because people get bored that there are not as many questions as they would like"... huh, sounds like a casual, even flippant dismissal of a significant site metric to me, even used by stackexchange mgt to measure site viability/health.... ok fine, what is a reasonable number of questions per month? exactly what there is now [which is less than historical]? or is there some other target possible? ps once again, going on record again, count me for 1 in as one of the occasionally bored due to low question volume & feel it reflects poorly on site community/engagement... $\endgroup$ – vzn Sep 10 '13 at 22:37
  • 4
    $\begingroup$ @vzn, sorry but I am really busy these days and don't have any time to waste by arguing with trolls. $\endgroup$ – Kaveh Sep 10 '13 at 22:43
  • 1
    $\begingroup$ ages ago, at least one moderator did at one time care about low question volume... when it was higher than it is now...? modifying scope, a proposal $\endgroup$ – vzn Sep 10 '13 at 22:52
2
$\begingroup$

I disagree with the premise that every time you downvote, you must always leave a comment. Personally, I used to do that all the time. However, I experienced all sorts of negative reactions: people got upset, they revenge-downvoted my other answers, they flagged my comments as offensive (even when they were neutral and factual). After repeated exposure and getting burnt many times, I've learned my lesson. I know that many others have come to a similar conclusion as well.

While I appreciate others who take the time to comment, and I often do try to comment in a helpful and constructive way, I cannot agree with the suggestion that it is always rude to downvote without commenting. I think the situation is more nuanced, and there are sometimes valid reasons to downvote without commenting. (And don't forget, both upvoting and downvoting are part of the StackExchange model.) While we don't want to discourage all new questions, frankly, I suspect we do want to discourage questions that are off-topic or poorly-explained.

Personally, I think it is the poster's job to understand the culture of the site, read through the help center, read some other questions, and then post a well-posed question that makes it self-evident how the question is on-topic (e.g., in what sense it is a research-level question). While I'm often happy to try to help even new posters who don't do that, I don't agree that it is our moral obligation to do so for every single new poster.

How can the poster know to improve the question? They can read the help center. They can look at other questions that have been posted here. Or, folks (like you) who see something positive in the question can leave a comment suggesting how the question could be improved, or even edit it to improve it.

If you want to help new posters and ensure they aren't discouraged, the best things we can do are:

  1. Leave a comment explaining (constructively and helpfully) explaining the expectations of the site and how the poster can better meet those expectations.

  2. Edit the question to make it a better fit for the site, and to make it clearer why it is a good question that is on-topic.

As a bonus, you might also notice how these are things that you can do, that don't require you to convince anyone else to change their ways. That's a positive thing: that means these are actually implementable. (In contrast, trying to convince everyone to change their ways is not.)

I realize this may be an unpopular message, and I expect to be downvoted (and that's fine). But since you asked, I'm trying to share another perspective.

$\endgroup$
  • $\begingroup$ I had a good chuckle at the fact that my answer was downvoted with no comment. I couldn't ask for a more elegant example to illustrate my point that some people prefer to downvote without commenting, and there's nothing wrong with that. Dig it! (Just for fun, I am tempted to suggest that downvotes on this answer imply some sort of agreement...) $\endgroup$ – D.W. Sep 18 '13 at 3:50
0
$\begingroup$

All I've been able to do is add a comment much like what you did. I completely agree with your sentiment btw.

In the past, we had a meta question about whether we should require downvoters to explain their stance. The concern with mandating this was that people might want to downvote anonymously, which the system allows and which I think is reasonable.

But this shouldn't be abused. In general, I think it's a good idea to wait to downvote when a user clearly needs help reformulating the question, and there are people trying to help.

Maybe this is merely a case of "if you don't have anything good to say, don't say anything"

$\endgroup$
  • 6
    $\begingroup$ I am beginning to feel like I'm the only one who posts on cstheory.SE or cs.SE and knows much information theory, and that most of the community on these websites is copmpletely ignorant of it, and as a result seems somewhat hostile towards it. For evidence, see this cs.SE question which anybody who knew information theory should have been able to answer (and which still doesn't have the right answer, which I am reasonably sure can be done with arithmetic coding). $\endgroup$ – Peter Shor Sep 10 '13 at 17:59
  • 1
    $\begingroup$ @peter the votes dont really matter too much, anyone can still write comments & answer the question. what really matters is when the question is closed with no further answers possible but comments possible, or locked, disabling comments, or deleted (404). for this question, there are currently no close votes. as for stackexchange hostility, its a systemic problem that even the mgt acknowledges & cant quite resolve to date, counterintuitively to some degree the interface enables it.... $\endgroup$ – vzn Sep 10 '13 at 21:39
  • 3
    $\begingroup$ @PeterShor, Sorry but I'm not seeing the link between drive-by downvotes, ignorance/hostility towards information theory and the question you link, which is voted to +6 and has two answers also voted to +6. $\endgroup$ – David Richerby Sep 11 '13 at 3:18
  • $\begingroup$ @David: The question I linked to has only one answer voted to 6 (mine). It sat around for a month before it received more than a trivial answer, and it would be a great problem in an introductory information course. I think this shows that most people answering questions here don't really think along information theory lines—otherwise it would have received a non-trivial answer sooner. $\endgroup$ – Peter Shor Sep 11 '13 at 3:35
  • 4
    $\begingroup$ @David: For the connection of this drive-by downvote with ignorance of information theory, I suspect the down-voters didn't understand the question at all. But I also suspect that if the question had been in an area that they recognized as theory but didn't understand, they wouldn't have down-voted the question. It's the "I don't understand this at all but I'm going to down-vote it anyway without leaving a comment" mentality that upsets me here. $\endgroup$ – Peter Shor Sep 11 '13 at 3:46
  • $\begingroup$ sympathize but it seems kind of futile sometimes to complain about any particular area of TCS lacking participants.... the challenge of overspecialization is quite ubiquitous in modern theory (math/TCS). a better strategy... start a blog! write about it! its accepted promotional form & many elite researchers are doing it nowadays.... also note the TCS blog is dormant for ages, just waiting for anyone to do something with it.... the greater problem we are seeing imho in many areas/way is lack of cohesion & community interconnection here..... a lot of posts does not a community make... $\endgroup$ – vzn Sep 11 '13 at 5:21
  • $\begingroup$ last flicker of life on the TCS blog, now long comatose, basically dead... $\endgroup$ – vzn Sep 11 '13 at 5:23
  • 1
    $\begingroup$ @PeterShor, I sense your distress, but I don't see the response to that particular question as having much to do with opinions about information theory. I think any similar question -- where the poster just pasted some scanned equations and just said "How do I calculate it? Assuming x=0.987" (with no evidence of effort, no context, and no explanation of why it is in scope) -- would have gotten a similar response, no matter the area. I really don't think you can extrapolate from this question to broad statements about the community's views on information theory. $\endgroup$ – D.W. Sep 18 '13 at 19:53
  • 2
    $\begingroup$ @D.W.: this was one of the most important recent papers in information theory, and if you've read the paper, the actual intent of the question and the reason it was asked are fairly obvious. If this were a similar question about a recent important paper in complexity theory, I don't think it would have gotten the same response. But you are correct that it should be interpreted as ignorance of, and not hostility to, information theory. It does indeed look like "do this easy calculation for me", rather than "explain this important step I don't understand by showing me an example". $\endgroup$ – Peter Shor Sep 19 '13 at 12:14
  • 1
    $\begingroup$ @Peter, I don't think anyone is disputing the suitability of your interpretation of the question but that is not clear from the posted question so the down-votes were not surprising. It seems that you agree that this is not a well-written question. Your interpretation was not clear for an ordinary user from the post, and yes, it should be clear for an ordinary user of the site who might not be an expert in information theory and might not have seen the paper that the question is suitable. $\endgroup$ – Kaveh Sep 22 '13 at 22:59
  • 1
    $\begingroup$ [cont.] The person asking the question should have provided the background information and stated the question in an understandable way for other users, that is the model the site works on. How to ask a good question. $\endgroup$ – Kaveh Sep 22 '13 at 23:03

You must log in to answer this question.

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