6

It seems like questions are not upvoted sufficiently. In comparison, answers are. Would it not be fair, to upvote questions automatically by counting the answers and comments, just to make clear, that a question was interesting?

3
  • By the way, comments on a question doesn't mean that it is interesting. It sometimes point to a question that is confusion and needs clarification. Similarly, questions that garner lots of answers may be soliciting primarily opinion-based responses... something that usually leads to closure. – Werner Mar 7 at 2:40
  • 1
    Also, voting on Meta is different from regular voting. – Werner Mar 7 at 3:45
  • 1
    Here is an example of a question (with a current score of -1) that asks the community to draw something for them. With 4 comments, two of which are upvoted, should this question also have a score of 4 instead, or something more (because of the +2 answer)? No, the question is poorly written, has no context in the title and only helps the individual since it won't be found by a community member easily (because of the poor structure). – Werner Mar 8 at 19:11
6

I talk for myself; others may have different opinions.

I don't upvote just-do-it-for-me questions, but if they have good answers, I upvote the answers.

I don't upvote unclear questions, but if somebodies manage to answer and the answers are good, I upvote the answers.

In general, I think those who ask for (free) help should put some effort into building good questions. If I don't see this effort, I don't upvote them.

That's why I upvote more answers than questions.

3
  • As you can see, 4 users found this question interesting, but only one did upvote. That is, because voting is opinion based and contains a judgement. I do not have a problem with that. But, a forum like this lives by questions, so questioning should be rewarded more than it is now. Even a negative comment or voting shows interest. A sum of views, comments. votings and answers divided by 10 would do, – user186178 Mar 5 at 20:01
  • @MatthiasBorck-Elsner: What do you mean by the your last calculation? – Werner Mar 6 at 3:45
  • @MatthiasBorck-Elsner I think it is a platform problem, not only of TeX.SE. – CarLaTeX Mar 6 at 7:19
5

Everyone has a different motivation for voting, and therefore an estimate of "fairness" between voting for answers and questions is difficult to assess. I vote based on the description associated with the voting buttons: It's useful/clear/shows research effort = upvote or not useful/unclear/doesn't show research effort = downvote. Others use it as a way to check that they've read the post, while others use it to correct any wrong-doing they've experienced from the community. So, voting is... complicated.

In general, though, good posts tend to receive upvotes while bad posts receive downvotes. Asking a good question is not easy, especially for new members. You need to write a good title that succinctly and adequately describes the problem. Then, you need to write a good question also...

Having been on this site for quite a while, the above seems to be a basic problem with questions on the network that people have to learn over time to achieve and may be a contributing reason for the poor voting performance on questions; users often land here with a problem... something that's very specific to them that they need to solve in a short period of time before life can return back to normal. Such questions are then poorly written, without context (only code snippets), may not include images to highlight the problematic output and possibly the expected output, asks something different than what they actually want.

The above is typically associated with new users, and we receive a bunch of those - daily! Here is a graph showing our daily new user acquisition:

enter image description here

The site seems to have stabilised at between 60-80 new users per day. This amount of new users per day is about the same as our average questions-per-day = around 60/day:

enter image description here

These statistics are also supported by the fact that new users are often just landing for a couple of questions, if that, when viewing this representation that depicts a type of user retention:

enter image description here

The lines represent new user posts by week as a percentage of answers. My interpretation is that users show up, ask their question(s), and then their question-asking drops. With that, their secondary contribution also drops - it's infrequent to see low-rep users with high voting tallies.

In short, it's not easy asking a good question, and sometimes questions take lots of time to answer, so the value may swing to be more answer-based than the other way around.

Related:

1
  • 2
    Very interesting the peak in the statistics you showed around March 2020, when the pandemic exploded! – CarLaTeX Mar 8 at 6:42

You must log in to answer this question.