Here's an example of a not-so-good question: Domain and sub domain in different hosting providers

It's about system management, so it doesn't belong on Stack Overflow. Googling for the question title brings up a bunch of responses, and the top response is a tutorial from the AWS side, so clearly the OP (who has been on Stack Overflow for 4+ years) hasn't indicated that any research. On the other hand, it is clearly written and is answerable by a person who knows the GoDaddy side.

As of this morning it's been viewed 13 times, downvoted twice, and has no close votes or explanatory comments. This doesn't seem very useful: it doesn't help the OP in any way, and doesn't leave a good question to help the next user.

If close votes are to be useful, it seems that they must be less "expensive" in relationship to downvotes. One approach would be to dramatically lower the number of rep points required: from 3,000 to, say, 500. That would still limit close votes to people who had been on the site long enough to have a vested interest.

Another possibility is that when a user of sufficient rep clicks the downvote button, s/he is asked if a close vote would be more appropriate.

Or, as a completely different alternative, accept that closed questions aren't adding anything to the site and eliminate closure altogether.


Before asking this question I tried this and this. The closest I found to a duplicate was More effective closing / downvoting of junk questions to help with the signal-noise ratio?.

  • 1
    Note that users with 15 or more rep can flag a post, and put it in the close votes review queue. Then >3K users can cast their votes. The difference between actually casting a vote and doing that is just one close vote (the queue is unfortunately very full, though). – Erik von Asmuth Aug 10 at 11:08
  • How many high-rep users spend their time reviewing close votes? I've done it a couple of times, but it's certainly not the first thing that I think of when I go to the site. That's why I'm suggesting lowering the initial friction (or removing the need). – kdgregory Aug 10 at 11:21
  • 2
    @kdgregory not enough, that is for sure, given the size of the CVQ being at 10K. – rene Aug 10 at 11:28
  • You're on place 3620 (with 79 reviews) of the 7879 users that I know of that have reviewed at least once in the CVQ on SO. – rene Aug 10 at 11:38
  • @rene - do you have any way of finding out how many people with < 3k rep flag questions for closure? – kdgregory Aug 11 at 10:40
  • @kdgregory no, not really. The API doesn't have review data and in SEDE the tables related to reviews and flags are anonymized. Only SE staff can query that info on their internal SEDE instance – rene Aug 11 at 10:42

One approach would be to dramatically lower the number of rep points required: from 3,000 to, say, 500. That would still limit close votes to people who had been on the site long enough to have a vested interest.

The downside of lowering this privilege (so much) is that it will be available to people who are lucky to place their first and only answer on a Hot Network Question. I do not expect those users to be able to determine what is on-topic and off-topic for Stack Overflow. In fact, the question you mentioned is an excellent example of why we shouldn't do this; that user has 538 reputation and apparently doesn't know such a question is off-topic.

Another possibility is that when a user of sufficient rep clicks the downvote button, s/he is asked if a close vote would be more appropriate.

Well, downvotes and close votes serve different purposes. Downvotes are for questions which are low quality and/or lack proper research; close votes are for questions that don't belong on the site at all. Of course, some questions deserve both, but they're by no means one-on-one.

  • 1
    To add to your point: the max amount of rep you can earn by just editing is 1K, to prevent users from gaining certain privileges just by editing. Setting the limit at 500 would allow users that have never posted anything this privilege. – Erik von Asmuth Aug 10 at 11:31
  • Re "Downvotes are for questions which are low quality and/or lack proper research," we also have close options that seem to cover those cases: either "unclear what you're asking" or "debugging help needs an example". Perhaps the reasons for closure should be scaled back to off-topic or duplicate? – kdgregory Aug 11 at 10:43

You must log in to answer this question.

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