Before I ever asked anything on SO itself, I was active on TeX.SX . My experience there is that valid questions, which are clear enough and not missing relevant data, most often get at least 1 up-vote; and that answer-providers often up-vote the question if it's a 0-vote or sometime even if it's a 1-vote question. You could say the logic is "if it's worthy of my reply, it's worthy of not being a zero-voter".

Now, TeX.SX is a much smaller site/community of users than SO itself. And yet, here, it seems that questions have a harder time getting any upvotes at all. Is this just my impression or is there more of an upvote stinginess in the SO culture? And, more importantly, is it not worthwhile to encourage voting over 0 as a means for 'vetting' a question?

Edit: See also this answer to the related question on How come so few SO questions have an answer accepted.

  • 5
    SO gets a lot of questions, and we only have 40 upvotes each (fewer if we upvote some answers). Commented Jul 8, 2013 at 8:03
  • 6
    Less specialisation, lower influx of questions per topic == more focus per question. Stack Overflow is a veritable firehose, and we have only so many votes per day. Commented Jul 8, 2013 at 8:03
  • 6
    Questions move by quickly on SO. Questions on smaller sites stay on the first couple pages much longer than on SO.
    – user206222
    Commented Jul 8, 2013 at 8:04
  • 2
    @MartijnPieters: But isn't that balanced by the high number of users (both experienced/high-reputation and inexperienced newbies)?
    – einpoklum
    Commented Jul 8, 2013 at 8:09
  • 2
    the quality of most questions is poor, unclear, and most of the time gets flagged and edited a couple of times before it becomes clear and understandable.
    – user221081
    Commented Jul 8, 2013 at 8:09
  • 2
    @einpoklum: Try to follow the C#, PHP, Java or Python tags. The questions accumulate fast. You generally do not come back later and scan through the new questions accumulated so far, you only look at new ones coming in. Commented Jul 8, 2013 at 8:13
  • 6
    FWIW: I downvote a lot :-)
    – PeeHaa
    Commented Jul 8, 2013 at 9:36
  • 1
    Questions should never get upvoted. The only consideration is how many downvotes it deserves. Commented Jul 10, 2013 at 19:51
  • 3
    @user414076: Ok, why should questions never be upvoted? (The designers of the site apparently didn't share your opinion.) Commented Jul 10, 2013 at 20:17
  • 1
    Every single question is a tax on those that must answer or moderate it. Most every question is a duplicate of something already answered. Or it's so heavily localized that nobody will find value in it beyond the asker. But the other strong opinion I have is that people ask too many questions, get too much rewards for asking them, when the only reward should be an answer. Give reputation and trust to the people who provide answers. The best that askers should hope for is to not have trust taken away. And yes, I'm serious. Commented Jul 10, 2013 at 20:32
  • 1
    @Keith I agree with the nameless user, in fact, I think the only vote you should be able to cast is a delete vote, albeit with a good motivation.
    – CodeCaster
    Commented Jul 10, 2013 at 20:33
  • 2
    Consider the end result of the "there are no more good questions" approach: If it is our goal to moderate away almost every new question, eventually askers get tired and/or frustrated, new issues don't get presented, and the good answerers have no reason to stick around. The site dies. We should encourage new, good questions, and moderate to remove the noise. Commented Jul 10, 2013 at 20:34
  • 2
    @GeorgeCummins, I realize I'm mostly yelling at clouds and fighting a losing battle, I don't expect that many people agree with me, I know better than to expect anything to actually change. But I'm sick of seeing the same crap day after day, and the fact it's frankly easier and more rewarding to continue to spoon-fed these people rather than moderate and keep the site clean. So yall keep at it, enjoy yourselves, and I'll just continue to dial back participation (for my own sanity). Commented Jul 10, 2013 at 20:45
  • 2
    @user414076: This sounds a lot like "Everything that can be invented has been invented." In fact, every time that Apple (or really anybody else for that matter) releases a new version of their OS, we get all kinds of new problems and questions that nobody has encountered before....
    – lnafziger
    Commented Jul 10, 2013 at 20:59
  • 3
    @lnafziger, I understand new tech will prompt new and incredibly useful questions. But I also understand that the C# tag has 80,000 questions titled "object reference not set to instance of object." Commented Jul 10, 2013 at 21:27

3 Answers 3


The number of questions asked every day makes a very big difference.

There are some active users browsing tags they are interested in. On smaller site there's a great chance they will eventually visit all questions asked in particular tags. On StackOverflow they will see only a tip of the iceberg.

Such active users are usually the most active voters.

Another factor is, that on each sites you have 30/40 votes a day. On smaller sites it's much more that the number of questions asked a day, making in theoretically possible to upvote (almost) all questions. On StackOverflow it's quite typical to be out of votes.

  • But shouldn't the number of questions be balanced out by the number of users? On SO there are a lot more knowledgeable users on all levels (1000+ rep, 10,000+ rep, 100,000+ rep...) ; seems to me everything should just scale.
    – einpoklum
    Commented Jul 11, 2013 at 7:15
  • 1
    But the mass makes it harder to browser good content, you're drown in the mass of new questions... Interesting phenomenon, it's just like there are nuch more people in big cities, but you have usually less friends than people living in small towns... Commented Jul 11, 2013 at 8:46
  • But that's partly due to cities being organized in an atomized way which doesn't encourage forming closer interpersonal bonds. The question is whether SO has a "refraing from voting on questions" culture.
    – einpoklum
    Commented Jul 11, 2013 at 13:56

Content is King(TM).

On average, good1 questions will receive upvotes and bad2 questions will receive downvotes.

While SO has a huge number of users, this does not necessarily mean that every good question will receive an upvote and that every bad question will receive a downvote.

However, this also means it is unlikely that "on average" good questions are not being upvoted.

1. defined as at least one standard deviation in the right direction from the mean3.
2. defined as at least one standard deviation in the wrong direction from the mean.
3. defined per tag cluster, you'll notice some tags have wildly different ideas of 'average'.

  • I wonder if a skilled SEDE user could determine how many highly-upvoted questions (say, 10+) SO gets per day compared to TeX. Commented Jul 10, 2013 at 20:23
  • SO probably gets more bike sheds than TeX does. Beyond +5ish you no longer can tell if the question is actually any good from votes alone.
    – user7116
    Commented Jul 10, 2013 at 20:27
  • True, but many of the easy questions already have answers, and we're getting pretty good about whacking new dupes. Commented Jul 10, 2013 at 20:30
  • @GeorgeCummins Different tags are a bit different with respect to dups, some are more aggressive about closing, some end up answering the question repeatedly (or at least have more people answer the dup before it can be closed).
    – Servy
    Commented Jul 10, 2013 at 20:35
  • @user7116: So what you're saying, is that 68% (Normal distribution function integrated over -1 to +1) of questions should not / do not get any votes - despite being average questions. Is this a good thing?
    – einpoklum
    Commented Jul 11, 2013 at 7:14
  • @einpoklum: why do you think the average should be >0?
    – user7116
    Commented Jul 11, 2013 at 13:10
  • @user7116: I think that should not be a principle, just a consequence of other principles. First, many/most questions with negative votes should be corrected or closed. Second, questions with no votes at all should be voted for by someone who has 'vetted' them as reasonable, worthwhile questions.
    – einpoklum
    Commented Jul 11, 2013 at 13:55
  • @einpoklum: why? All you're doing is diluting the usage of the question "score". An average question should have a 0 (you can have an average question which should not be closed AND should not have up or down votes). An above average question should be >0, and a below average question should be <0. (re: "why not make 10 louder?")
    – user7116
    Commented Jul 11, 2013 at 14:08
  • 1
    @user7116: An average question should not, in my opinion, have a score of 0. That makes it indistinguishable from a question nobody's bothered to read through and consider. I'd say an average question, after a short while, should have a low score, greater than 0.
    – einpoklum
    Commented Jul 13, 2013 at 16:08

I think if a question is helpful for other users, it will get its share of upvotes. I've come across many questions that have. I don't know about the TeX site, and I've never seen it come up in a Google search but when I first became a user here it was because it came up near the top of every search I did. I would suspect there are more "inexperienced" users here because of that phenomenon, and most newer users don't really understand the voting system. Similarly, I've seen many questions get great answers, the "asker" posts a comment that says, "WOW! Awesome answer, this helped so much!", yet they never upvote or checkmark the answer.

  • Yeah, that's frustrating when your reputation is not that high... and you feel awkward adding a comment saying "so please upvote me then" because it makes you sound greedy.
    – einpoklum
    Commented Jul 11, 2013 at 7:18
  • But... I don't think the evidence supports your assumption that worthwhile questions get their share of upvotes. What's more, questions themselves are not supposed to be helpful, just, well, "poignant" would be the adjective I guess.
    – einpoklum
    Commented Jul 13, 2013 at 20:19
  • Wrong, on both counts. First, you shouldn't ask for upvotes. In fact, I think that's actually against the rules. Second, questions CAN be helpful. If I ask "How do I format a date in Access?" and you also need to know the answer, then the question is helpful to you and you should upvote it. Commented Jul 14, 2013 at 22:14
  • So, what you're saying is that in your view, question votes should essentially mean "I'm also interested in knowing the answer to this question", but not "this question is well-put/thought-provoking/etc."?
    – einpoklum
    Commented Jul 15, 2013 at 7:22
  • I think they can mean both. I tend to only upvote questions that I can also gain knowledge from. I don't go looking through Android questions trying to find well-thought-out questions, because I really could care less about Android. But I do go through C# questions trying to find ones that I also have questions about, and I'll upvote those. Commented Jul 15, 2013 at 12:22
  • What knowledge would you gain from reading a question, rather than an answer?
    – einpoklum
    Commented Jul 15, 2013 at 12:25
  • If it gets an answer, I gain knowledge. Commented Jul 15, 2013 at 13:55

You must log in to answer this question.

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