Yes, I started with “reduce the reputation”, so I'm going to be downvoted into oblivion.

Ever since the early days, it was apparent that many questions got answers but no upvotes, and the general response was that answerers should upvote the questions they answer most of the time.

I can think of only a few reasons not to upvote a question I've answered:

  1. I've run out of votes.
  2. A new user lacking the 15 rep to upvote.
  3. The question is a good one, but so badly asked it doesn't deserve upvoting as is.
  4. The question is technically valid, but so trite it doesn't deserve upvoting.

Well, looks like the team is prepared to help with #1, by providing more votes for questions.
For #2, let's just say that new users (<15rep) get a dispensation.
For #3, now everyone has access to the edit button to improve the question.
For #4, well, if the question is so trite, the answer can't be that difficult.

So: if you answer a question and don't upvote it, I propose that you earn less reputation from answer votes. Perhaps only +5 rather than +10.

I'm aware that this discloses whether you've voted on a question if you've answered it. I don't think this is a problem. Remember that you can vote up or down.

Before proposing this, I wanted to see how many people would be affected. The answer is, a lot. (Hoping these queries mean what I think they mean since this is my first foray in the data explorer.)

  • Zero-score questions: On SO, 5.5% of questions have a score of 0 and no answer. 19.4% have a score of 0 and one answer. 11.6% have a score of 0 and two answers.
  • Question score distribution: On SO, 46% of all questions have a score of 0. About 1% have a negative score (all these queries look at score, not upvotes, but this 1% figure shows that downvotes are second-order).
  • Question upvote excedent: Only 13.2% of questions have more upvotes than they have answers. 15.7% have as many. 32.3% have one less. 37.9% have at least two more answers than upvotes. More breakdown.
  • I thought people might be refraining from upvoting questions because they were afraid of running out of votes. So here's a question vote deficit breakdown by question time. So, yes, time has an effect, but it's small: if you ask between 0:00 and 1:00 UTC, and get at least one answer, you have a 31.8% chance of getting more upvotes than answers. Between 6:00 and 7:00, the chance drops to 26.8%. That's barely statistically significant.
  • 20
    I spend a lot of time trying to make sure that bad questions get answered. That will stop happening if you reduce my rep gain. I already lose points when I downvote the question. The site isn't going to be better off by upsetting users that spend time trying to get even the bad, poorly-asked, sometimes unsalvageable questions answered.
    – Cody Gray
    May 9 '11 at 1:04
  • 9
    I upvote questions which I find interesting. I see no point upvoting it otherwise, just to game the rep.
    – vartec
    May 9 '11 at 1:08
  • 3
    Is it fair to rework bad questions using assumed knowledge? I can't say that all questions can be edited to be made better.
    – jcolebrand
    May 9 '11 at 1:09
  • @Cody: I spend quite a bit of time on bad questions as well. If they're salvageable, I salvage them (by editing). And upvote them. May 9 '11 at 1:10
  • 4
    @Gilles: Yes, I'm one of the top editors for the year on Stack Overflow. So I indeed edit as many questions to improve them as I can. Doesn't change the fact that lots of questions are just bad and don't deserve an upvote. The only question is whether you/the community think they deserve an answer. Or, indeed, any attention at all.
    – Cody Gray
    May 9 '11 at 1:16
  • 5
    @Cody why would you spend any time trying to answer bad questions? It just teaches people asking bad questions that it doesn't matter if you can't ask a good question. Answering bad questions is part of the problem.
    – user149432
    May 9 '11 at 1:51
  • @Mark - I don't think elitism is a value that's advantageous to the community. I personally regard the question as poor. I don't think it merits an upvote. But I'm not sure that's the same thing as making the call that it doesn't deserve an answer. You say, "Answering bad questions is part of the problem". But I didn't realize we had such a problem.
    – Cody Gray
    May 9 '11 at 6:54

Horse Feathers. Why should I upvote the questions I answer? Sometimes I answer poor questions. Frequently I answer middle-of-the-road questions that are perfectly accurately described with a total vote score of 0. Other times I upvote well-done questions that I'm not qualified to answer.

I don't owe the question-asker a cookie, and they don't owe me one either.

Am I encouraging bad answerers? Well, I don't do this all the time, and I have some criteria that attempt to avoid reinforcing serial bad-question-depositors. Really, the 'middle-of-the-road' case is what I want to emphasize.

  • 3
    You realize you're giving bad question-askers a cookie by answering their questions, right?
    – user149432
    May 9 '11 at 1:56
  • expansion to follow ...
    – Rosinante
    May 9 '11 at 1:57
  • 7
    @Mark Some questions are bad because their askers are idiots. Others are bad because their askers don't know better. The latter can be improved and safely answered. I think we should remain mindful of the "kinda meh, but not atrocious" kind of users that still deserve an answer and can grow into being better community members.
    – Adam Lear StaffMod
    May 9 '11 at 3:16
  • 3
    You had me at Horse Feathers. May 9 '11 at 3:25
  • If you feel so strongly against those questions, downvote them! May 9 '11 at 16:07

As a question asker, I don't particularly agree with this idea.

I don't think of voting as a binary option, and I'm certain that I'm not alone. When it comes to questions, upvoting means that I find the question useful, downvoting means that I find the question not useful, and abstaining means that I think it's a plain old question. It doesn't have to be a bad question in order to not get upvoted, and it doesn't have to be a good question in order to avoid downvotes. It may just be a question.

While I'm obviously invested in getting answers to my own questions, I wouldn't upvote a few of my own questions if I were just a passerby. This is a particular one I cite often, it's a very basic question. I know it's not worth all that much, after all it has 0 votes. But is it really so worthless, so "trite", that people should earn less reputation for answering it?

Answer votes are tooltip'd to suggest voting on what's useful. Which is generally what is helpful and correct. And if there's one thing the answers were, it was useful, helpful, and correct. The fact that I may have been somewhat silly in asking the question doesn't change the fact that those answers can still be found to be useful. And in the end, votes indicate that a bunch of people found it useful. If one believes that the answer is covering such simple information that it's not worth the full effort of +10 reputation, then that person probably believes the answer to not be useful, and should appropriately abstain from upvoting answers that are not useful.

If someone has the kindness to address basic, plain old questions that I may ask, I don't need a pity upvote just so that the user can get the full reputation they deserve based on however many users find their answers useful.

  • It's got a sympathy upvote now. Not because I care about your reputation, but because I've been struggling against Outlook recently myself. May 9 '11 at 16:05

Gilles, I agree that it doesn't make sense to answer a question if you're not willing to upvote it, but I don't think giving half rep is the answer.

It would seem if we wanted people to upvote questions, we should be providing positive reinforcement for the behavior we want.

I suggest giving +5 or +10 for every day you hit your question vote max (as of today, that is 40 votes). I am rather sure we would see an increase in question voting if this policy was implemented.

  • I don't like giving reputation purely for voting, because that's completely uncontrolled. Reputation should come from visible (hence auditable) participation. May 9 '11 at 16:07
  • @Gilles, responding late because someone upvoted the answer again... voting is auditable... go to the votes tab on your user page Aug 5 '13 at 18:09
  • Auditable requires that the information is visible by someone other than the person being audited. Aug 5 '13 at 18:18
  • @Gilles, stack exchange employees can do such an audit as required... Aug 5 '13 at 18:33
  • A handful of SE employees cannot audit millions of users. Aug 5 '13 at 18:37
  • When are you proposing that an audit would be required by non-SE mods? Aug 5 '13 at 18:37
  • When you get reputation from someone voting on your post, the decision that grants you reputation (the vote) comes from an external source (the voter). When you get reputation for a suggested edit, the decision (acceptance) comes from an external source (the reviewer). If you got reputation from casting votes, there would be no other user involved, and even no way for anyone other than a handful of employees to see why you got that reputation. How can anyone control that you did something desirable, as opposed to, say, randomly upvoting 40 questions? Aug 5 '13 at 18:41
  • if the goal is to encourage more question voting, one could consider it an acceptable outcome if a person randomly upvotes to get their daily "vote points". If we truly desire to limit random voting, optionally implement a vote rate-limiter (perhaps no more than 30 votes in as many minutes), to keep people from gaming the system just for points. The rate-limiter should not apply to the ability to cast a vote, just the ability to get points from voting. Aug 5 '13 at 18:50
  • The goal is not purely to encourage more question voting, otherwise the software could just add N votes per question automatically. The votes still have to rate questions according to a subjective but nonetheless broadly agreed-upon metric. Aug 5 '13 at 18:53
  • Your question is not asking about a subjective-but-nonetheless broadly agreed-upon metric. Your question suggests that those who answer questions should upvote (presumably so good questions get more votes). My suggestion accomplishes the same Aug 5 '13 at 19:23

Answering a question, and voting the questions are two separated actions. Some questions are worth to be answered, but they do not deserve a vote. For example, I imagine it is difficult that a question ending with "please help me; it's urgent," will be voted; yet, if the question is not too generic and it's clear what the OP is asking, the question is worth of an answer.

What is the effect of what proposed if not to punish who answered, after his answer has been up-voted? The more the answer is voted, the more the answerer is "punished;" they would lose 5 points for a vote, 50 points for 10 votes, 100 points for 20 votes.
What is worse is the answerers would be "punished" for a possible problem with the question, not the provided answer, for which they are not responsible.

If you think the question has some kind of problems, then simply down vote the question, not a possibly excellent question; after all, if X users (with X greater than 2) have voted the answer, it cannot be that bad.

  • It's all in how you frame it: think of it as gaining more points if they upvote the question. May 9 '11 at 16:06
  • @Gilles The problem is yet the same: who answers would get more reputation for something they didn't write (the question). I would rather prefer users get more reputation for something they did: giving a useful answer, asking a good question, etc.
    – apaderno
    May 9 '11 at 19:14

You must log in to answer this question.

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