# Why don't people vote for posts any more?

I've just done a quick bit of data collection.

I looked at the first 25 most recently posted questions.

Amongst them, they had a total of 242 views; that's an average of 9.7 per question.

Quite amazingly, these 25 questions had amassed a total of three up-votes and zero down-votes. That makes a total of three votes in total, or an average of 0.12 votes per question.

A more alarming figure is that out of the 242 views, there were three votes, i.e. a view-to-vote ratio of 80.7-to-1. In other words, for every 81 people that view a post, only one of them votes.

• Do you agree that this is a rather worrying situation?
• What do you think the causes of this are?
• How might we encourage more interaction?
• It may just be general stinginess with imaginary Internet points. I try to upvote as much as possible, but my interests (and mathematical comfort zone!) are limited. – The Chaz 2.0 Nov 13 '13 at 21:30
• I am not sure if this is a good idea, but the voters themselves could get, say, 1% of the reputation. – Librecoin Nov 13 '13 at 21:41
• Data sample too small. Take at least one day (~500q nowadays), to account for all possible time zones. – Lord_Farin Nov 13 '13 at 21:56
• The sample may be too small, but I get the same impression from the past weeks (maybe from the last two months or so) and it goes for answers just as for questions. – Git Gud Nov 13 '13 at 22:28
• I tried to make a SEDE query: here. It is supposed to bring up questions that have been voted on within a day of their inception. But it shows some strange results -- notably false positives. I'll look again tomorrow. – Lord_Farin Nov 13 '13 at 23:51
• I have a hard time voting for or against homework questions, especially the ones that go "Prove blah, blah, blah." – Jay Nov 14 '13 at 0:05
• It might be instructive to compare, if possible, the proportion of questions whose body contains the character "?" over the lifetime of MSE. – user14972 Nov 14 '13 at 2:55
• Dear FlybyNight, I have also noticed that the vote:view ratio is rather low in general. Personally, I try to upvote every single question/answer I view (unless there's a compelling reason not to do so) although I've unfortunately not been able to do that so often recently largely due to inactivity on this website. But I notice that you have 303 upvotes and 269 downvotes in a little more than a year of activity. You might be able to answer your own question by thinking about why you aren't upvoting more and why your upvote:downvote ratio is unusually low. Regards, – Amitesh Datta Nov 14 '13 at 3:23
• A year old discussion asking about the same problem: Statistics on upvoting However, (number of votes)/(number of posts) ration was discussed there. (Rather than number of views.) – Martin Sleziak Nov 14 '13 at 6:48
• Thanks @MartinSleziak – Fly by Night Nov 14 '13 at 17:49
• Your sample is too small to be very meaningful, but I, like Git Gud, also have the same impression. I try to remember to upvote questions that show some meaningful work, especially homework questions, but I often forget. I don’t actually read all that many answers: most of the time I’m busy reading questions and writing answers. When I have time to page back through the questions that have come in since the last time I was online, I do upvote some answers. – Brian M. Scott Nov 14 '13 at 18:17
• @FlybyNight I don't think a minimum reputation requirement for posting questions is a good idea. Suppose someone is studying trigonometry. How are they going to get reputation? – Jay Nov 18 '13 at 22:01
• @Jay There are hundreds of good websites and videos online that address high school and pre-university mathematics. I wouldn't be against people being unable to post high school topic questions if I'm honest. I suppose that boils down to what we think this site should be. Part of the problem is that a lot of inexperienced users are posting poorly thought-out, poorly researched and poorly formated questions. Most of the voteless questions are coming from people cut-and-pasting trivial problems that a quick Google would solve for them. – Fly by Night Nov 19 '13 at 16:26
• I have a rule of thumb. If I find a question worth answering, I upvote it. I notice a lot of answers to questions with $0$ votes, which annoys me a little bit. If it was worth it to you to answer, then you could always take the time to tick the upvote button. I also tend to upvote questions which I don't immediately see the answer to after a couple of seconds (as long as I'm somewhat within my comfort zone). – Cheerful Parsnip Nov 19 '13 at 23:41
• @GrumpyParsnip There are many users who find it more natural to keep the answering and voting processes independent. (I count myself among them.) This was discussed recently. – user103402 Nov 21 '13 at 5:57

With more and more questions it may be that questions are disappearing from the first page more quickly.

• Absolutely. If you look at questions from years ago, you see vote counts that are astonishing by today's standards. There were less people, but also very few questions, and so everybody voted each question, I guess. – Martin Argerami Nov 14 '13 at 1:23
• Not just that, but I’m pretty sure that the number of questions per day has grown faster than the number of regulars. – Brian M. Scott Nov 14 '13 at 20:47
• Also many regulars have adopted the approach of answering as many questions as possible. That is one way of using this site, but with many questions to choose from they simply then don't have the time to upvote others posts. I prefer to roam around, looking for interesting stuff, suitable problems for my courses (or contests I am co-organizing) opportunities to teach, nursing my two adopted tags, and that leaves me some time to read others answers and (possibly somewhat patronizingly) give newcomer students "passing grades" by upvoting their answers. Helps, but gives me an occasional grudge – Jyrki Lahtonen Nov 17 '13 at 12:14
• (cont'd) If many regulars adopted my approach, then we would have more votes, but less answers. Win some, lose some. – Jyrki Lahtonen Nov 17 '13 at 12:49
• I have a hard time understanding how this addresses the question. "For every 81 people that view a post only 1 upvotes." The sentence in this answer would explain why less people are viewing questions, but it does not at all answer why the people that are viewing the questions are not voting. – Matt Nov 27 '13 at 20:19

Currently, 8 SE sites receive 50+ questions per day. Below I compare their number of voters (users with the Supporter badge) to the number of questions per day. I also included MO and Physics, as they are close to 50 question mark, and are academic sites. Mathematics has the lowest ratio of all ten.

Serverfault: 40793 voters / 100 questions = 408
Superuser: 54609 / 182  = 300
MathOverflow: 6550 / 38 = 172
TeX: 12060 / 74 = 163
Physics: 6844 / 43 = 159
Ubuntu: 26498 / 223  = 119
Statistics: 5904 / 50 = 118
Wordpress: 5785 / 51 = 113
StackOverflow: 437604 / 7464 = 59
Mathematics: 23991 / 538  = 45


Notice that Math and SO are the two sites way below all others. Big city problems.

• Also, the count of voters is an overestimate for sites other than SO, because it includes many SO users who drop by once in forever to upvote a "hot" question or its answer. – user103402 Nov 17 '13 at 17:49
• +1 for interesting statistics. I suspect that many SO users only drop by SO once in forever to vote on things—many people in my office use SO every once in a while and upvote the relevant answer(s) and question(s), but don't go looking for questions to answer on a daily basis (and a good many of them need reminding to get them to upvote stuff). – Isaac Nov 17 '13 at 22:55
• turn it upside-down then Mathematics is atop of SO! – draks ... Nov 18 '13 at 21:48
• As there are many people who frequent both this site and MathOverflow, maybe comparison with MO could be interesting, too. – Martin Sleziak Nov 19 '13 at 9:09

I suspect that many—probably most—viewers cannot vote. Up-voting requires 15 rep and I'd expect that a large portion of site traffic is transient visitors who do not have any accumulated rep.

And, if you want an extreme example, this question has more than 40,000 views and only 6 13 votes.

edit: Since drawing attention to the above question increased its vote count, it is no longer the second-lowest-voted question among the dozen most viewed questions (actually, questions with 40k+ views)—it's now the third-lowest. Basically, if you look at those questions, there are two distinct kinds of questions: questions that were "hot" within the Stack Exchange network, drawing lots of viewers who also upvoted; and questions that never were hot, but drew lots of views most likely via search engine traffic, hence not so many upvotes.

• Perhaps most posters only care about their homework problem, not someone else's [inappropriate word] homework problem. – Jay Nov 14 '13 at 13:45
• @Jay: I think probably most of those posters not only don't have the rep to upvote, but aren't even looking at other questions, much less voting on them. – Isaac Nov 14 '13 at 13:47
• it gained 6+1 upvotes more...but I your point is absolute right! Do you think 15 is too much? – draks ... Nov 18 '13 at 21:53
• @draks...: good point—see my edit and look at all the questions with 40k+ views. :) – Isaac Nov 18 '13 at 22:06
• Not just 15 votes, but also requires registering to the site! – Asaf Karagila Nov 18 '13 at 22:34
• Oh, as to whether 15 rep is too high... I don't really think so. I just think we need to realize that views aren't going to translate directly into votes. (There has been some experimentation with the "Is this helpful?" yes/no prompts for non-registered/non-voting users, I think, but since that hasn't turned into a mechanism for recording upvotes, I doubt it will.) – Isaac Nov 18 '13 at 22:43

Or maybe most users are coming only for solving their own problems (homework) and this is all that matters for them.

Edit. I've also noticed that high rated users almost never upvote the answers which are not posted by other high rated users (or by people they know each other) even if these answers are not trivial at all. (Could be a few exceptions, but I can nominate only one in the commutative-algebra and abstract-algebra sections.)

• Homework or not, that's probably true and certainly a large part of the point of Stack Exchange sites. I don't think there's any reason to expect votes to follow directly from views. – Isaac Nov 17 '13 at 16:36
• @user Ad edit: How do you know that? Voting is anonymous. Did you correct for high-rep users writing better answers? – Lord_Farin Nov 27 '13 at 8:27
• You need to explain yourself. Barring "+1" comments, there is no way to recover who voted on a post. Also, given that there are over 240 users with 10k+ rep, a few incidents are not enough to warrant such conclusions. – Lord_Farin Nov 27 '13 at 8:38
• Let it be as you said, but my opinion is clearly expressed in my answer and your theoretical arguments are not enough to change it. Best regards! – user89712 Nov 27 '13 at 8:40
• But you did not phrase it as an opinion but as an observation. – Tobias Kildetoft Nov 27 '13 at 9:57
• @TobiasKildetoft Yes, this is my observation (as you can see I've started it with an I), so it is an opinion. – user89712 Nov 27 '13 at 11:02
• But calling it an observation makes it sound like it is actually based on something concrete, rather than being an opinion. – Tobias Kildetoft Nov 27 '13 at 12:50
• @TobiasKildetoft Actually it is based on many concrete observations. However, to conclude, if don't like my opinion/observation can downvote the answer (unless already did it). – user89712 Nov 27 '13 at 16:16

I will rarely (if ever) downvote. In cases where I wish to encourage the questioner to improve their question, I would start by leaving comments.

If I upvote, it is due to a question that has met several criteria for me that means it is a "good" or "upvote-worthy" question. I expect that many users have their own criteria that are similar.

In general for the Math.SE community, I would expect $(1)$ a significant percentage of questions that cannot receive appropriate attention until such time as someone versed in the field can look at it, or $(2)$ many questions are borderline or too frequent to give reasonable attention and votes to all of them.

This is my perspective as a (relatively) new participant in Math.SE.

• "I will rarely (if ever) downvote." This is a way to not contribute to this site, by leaving somehow the "downvote-not-worthy" questions at the same level with good questions. (I've noticed that some people are even proud about their not downvoting policy. I find it totally wrong as this is a form to contribute to maintain this site in a logical frame.) – user89712 Nov 18 '13 at 23:05
• I disagree with this sentiment. I mean if a question/answer has 0 upvotes then it already signifies something, why smack in the face by downvoting? Upvoting can by itself order the questions/answers by usefulness. – Adam Nov 18 '13 at 23:36
• I can agree with the sentiments expressed, but I must fall back on the principle that "I am not completely familiarized with every aspect of the possible questions posed on Math.SE" and so I cannot offer a vote (either positive or negative) on every single question that will be constructive for that question... – abiessu Nov 19 '13 at 0:45
• @Adam What does a zero score signify, really? The site has about 500K posts as of now, of them about 100K have zero score. That's one in five. Zero score can mean: bad, mediocre, never read by anyone in the know, too long, too short, posted at a bad time, etc... As of now, half of my answers have 0 score; this does not signify anything to me. On the other hand, only 6K posts have negative score. – user103402 Nov 19 '13 at 2:12
• If there is something not good about a question (gap, lack of clarity, significant "typo") I may leave a comment. – André Nicolas Nov 19 '13 at 3:21
• @AndréNicolas: that's my preferred response to questions over leaving a downvote. – abiessu Nov 19 '13 at 3:54
• @abiessu: A "silent" downvote seems kind of useless, And a downvote after a suggestion to modify is not needed. – André Nicolas Nov 19 '13 at 3:56
• @AndréNicolas: points taken – abiessu Nov 19 '13 at 3:58
• +1. I never downvote on the main site. If I see a problem, I leave a comment indicating what the problem is, and if I see that it’s been taken care of, I delete the comment. Downvotes without comments are useless — certainly to the writer of the question or answer being downvoted, and in my view largely so to future readers as well, who would be better advised to pay attention to the comments. – Brian M. Scott Nov 19 '13 at 17:24