I'm curious about the factors that influence voting.

One of my motivations is the number of '0-score' answers there are. Indeed, even @egreg has one page of '0-score' answers- I find this really interesting.

I've been an active member on the site for over a year- I love the community that we have here. In that time I've seen some truly heroic answers that have received very little attention and less than 5 votes. And conversely I've seen significantly less heroic answers go platinum with over 25 votes very quickly.

We have a really diverse and interesting world-wide community, and I'm interested in what factors the community perceives as the main influences when voting.

To spark the discussion, I perceive the following factors as primary when it comes to voting:

  • timing: Monday morning (from my perspective, GMT-8hrs) seems to see a lot of votes, Friday night, not so much
  • tags: tikz questions seem to be the most popular and accessible; questions on (for example) expansion (which I find much more interesting) seem to receive a lot less.

6 Answers 6


Maybe it's my personality projecting on it but I value the votes very very very little (including downvotes on my questions/answers though I always want to see what the mistake was that I've possibly made). Somewhat stupidly, I consider downvotes on others' question and answers a lot because I want the spirit to stay as it is (relaxed, quite kind, and willing to help further even though the votes are saturated and very unlikely to get more) and unfortunately not everyone takes these votes lightly. Hence my concern.

I'm kind of among the actively voting users. Since I don't value the votes much, I'm clicking those arrows much easier, I might even write down the steps because I've just did on a 5-answer question.

  1. What's this question? Oh, a MWE. That's handy. About (random example replaced here) table of contents, section misalignment. Man, why would you do that? Anyway MWE is nice. voted.
  2. Ah it has an answer, is there any @ characters in it? (just joking!). Well what did they use ? I didn't know that. But there is a package for this. Anyway, voted. Next.
  3. Yep, that's what I would sort of do. Voted.
  4. This can not work. Let me compile it. Dang, it's working how come? Let me tweak a little. Still works. But this is not suitable for such and such. Comment and voted for fooling me.
  5. Nope got the question wrong.
  6. No idea what this is about or how to use it (happens mostly on bidi, Emacs, LyX questions)

I think I'm using the votes as mark as read button. And I stop voting when I see the yellow warning signs You have 5 votes left. But honestly I keep reading the questions in the same manner. However the system is not sorting the questions on technical proficiency. So many questions goes away unvoted. But if you happen to receive votes for an ancient question that's probably me searching for something else :).

Also you might want to consider what tag is being followed the most to get the pattern normalized. I'm following the TikZ questions among other things but I'm mostly spending a lot of time on core questions (mostly asked by Bruno) that received answers that are rock hard and some of them are at the How to draw Christmas tree with TikZ? level questions. However, I can only vote once.

So 5-10 votes in a TikZ question are like 2-3 votes on a normal question and I'm also not comfortable with it. We had some time ago a discussion to keep the high-rep users back a bit to let new users rep-charge themselves. And I think the system is working in the face of such imperfections. Lastly answers also have a lot of pictures but it's not getting that much of attention so putting images are not the only factor for the votes.

Because we have the experts having fun here too, votes seem to be a measure of participation instead of proficiency on TeX.SE. So I guess the votes are showing the fun factor not anything else. That's why we can never have Careers 2.0 here because everyone is too nice to take this as a challenge and compete for some weird virtual trophy.


Anything that has a pretty picture gets more votes. So TikZ answers tend to get lots of votes. They are also easy for non-specialists to vote on, since you can look at the picture and say "Cool, get's my vote." But really technical answers are not so easy for non-specialists to vote on, nor are they likely to even bother reading them, so the questions themselves get fewer views and therefore fewer votes. Now you can vote on the basis that "If X wrote this answer it must be good.", but if you do that too much (even with the best of intentions) the voting fraud bots will catch you and much hilarity will ensue.


Ok, here's my take on this: Confessions of an Upvoter

As a kind of disclaimer, I'd like to stress that at this level I consider myself a decent "TeXnician", nothing more. Since most of the answers posted up to the time of typing this come from really experienced users I am looking up to, I thought an answer from a less experienced user could be fun. More so, because I happen to have an Electorate badge.

As other users pointed it out, TikZ questions will always get more votes for the fun factor involved, especially the Christmas tree kind. I must confess, I also tend to upvote them, although to a lesser extent lately.

Recently my participation has somewhat slackened, I am bogged down with other tasks for the coming months, but I still read questions on a daily basis and also try to find an answer (where no @s are required ;)). Or at least check the MWE to see if I can reproduce and trace the error, especially if I think it's non-trivial. If I already took the pains of copy-pasting the code and compiling it, I upvote the question. I see it as an encouragement for new users to ask away, a question with 0 upvotes can be quite a turn-off.

Now that I think of it, I often vote to close duplicates. Sometimes it is just a matter of doing a more thorough search, but at other times the link to a previous question is less obvious. I gladly vote to close questions that are blatantly of RTFM type. I feel that if we don't curb them, we tacitly encourage them, which is in no way constructive. (And I get slightly annoyed by users posting answers when a comment already suggests it's been answered and 3 or 4 closing votes confirm that...) Nevertheless, I very seldom downvote anything. It needs to be serious business for me to do so.

I upvote mainly on answers that either are very simple and elegant or from which I learned something new: a new feature or a new way of addressing an issue. Also, I like detailed answers with good explanations about how and why the code posted actually works. I have learned a lot from such answers and they definitely worth being rewarded. Good "competing" answers are definitely part of this category. And I upvote sensible comments too.

And there are some other factors involved in my voting process, with which some of the users may not (fully) agree. For instance, I tend to upvote a question/answer with 9 votes, to secure a badge for the user. At the point where it earned 9 votes, I think it also deserves a 10th one. I know there are lots of users who avidly collect badges and nobody can deny the fun factor involved. (In the early days I was one of those hunters myself. :)) And this leads to the second (somewhat controversial) reason for voting: I practically was preying on interesting questions (and answers) to upvote when I had some 50 votes missing to get the Electorate badge... All right - now I said it all. :)

  • 1
    Your answer had 9 votes, thus: +1 ;-)
    – Stephen
    Dec 3, 2012 at 18:53
  • @Stephen: Thanks! :D I didn't think it would backfire on me! :D
    – Count Zero
    Dec 3, 2012 at 22:08
  • 1
    @Stephen: Hmmm, this has 11 votes now, so I perhaps you back to 10. :-) Dec 5, 2012 at 4:37

i can speak only for myself. i tend to ignore anything that i probably can't use, unless it's asked or answered by a few "core" participants. even then, i'll consider the content of the question and answer before voting (or not).

unlike most people here, i have a day job that is tex-based, so anything that's tagged with an [ams*] i have to look at, although i will up-vote only questions or answers i consider valid. i will comment on a lot more of them though, and if the comment results in an edit that improves the item sufficiently, then i will up-vote. if an item on an [ams*] topic is really good, i will up-vote and consider it for inclusion in the ams documentation (though getting that updated is a slow process). if a real or plausible bug is reported, i will pick it up for consideration, up-vote, and add either a comment or an answer with a workaround (if one is known). i will also pick up a really great suggestion for enhancement, up-vote and comment. (real-world users are our best debuggers, as demonstrated by knuth.) if i make a blunder in an answer or comment, i'll usually delete it, even if it's been up-voted, unless i think an edit can make it useful over the long term.

i tend not to down-vote, but if i recognize something that i think has already been covered, i'll look for the earlier material if i have time, and vote to close if others agree. (i don't think i've ever posted the first vote to close.) i will definitely flag spam for moderator attention.


There were other people poiting out that the technical answers generally get small number of votes, because people cannot understand it, and testing it works takes a lot of time and effort.

This is something that other sites suffer from as well. For instance on math.SE posts get much less upvotes than here. For comparison:

                      MATH    TEX
Number of questions   87k     30k    (almost 3x larger)
Nice question badges  3.1k    3.8k   (20% smaller)
Nice answers badges   6.6k    6.5k   (almost the same)

The reason is that most of the things in maths are understandable to a small part of the site visitors.

  • 1
    Also the competition is mind-bogglingly fierce. I know because I was active until last year. Also smartness and correctness is the currency there, conversely here people just lift each other up in case of small mistakes. So I would also include that the appreciation is relatively lower. And they have a very high unanswered ratio. So garbage collection is quite effective on the users' perception that if it's alive on TeX.SE it's (almost always) a valid question. Homework questions are contaminating Math.SE a lot though the intention is purely innocent.
    – percusse
    Dec 2, 2012 at 22:27

As a relative newcomer to TeX.SX I can only provide my two cents worth of emerging experience. At first I was looking for insights into areas where I needed to widen my perspective. I thus searched for tag and provided votes on entries that asked for me pertinent questions or similarly relevant replies or both. This way my voting was governed by my own "needs". Fairly soon I started to follow the entries in "real time" and of course found much that I was not really aware I was interested in. Thus the voting has widened in scope but is still bsed on question that in some way interests me and for answers, answers that provides me with new and interesting information (particulary if they are succinct and well written). What I thus might still miss are posts that come in under my radar either because my own understanding of TeX/LaTeX is lacking or because the heading is poorly written in the sense that the true content is not clear. I guess I have not come to the point where I simply must try to understand everything. I can imagine visitors may remain in any of these steps which will then govern the voting.

You must log in to answer this question.

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