-1

Right now, we have about 12 questions at -3 or lower on the homepage. This is after I closed and mod-deleted questions that are explicitly off-topic and very, very unlikely to be salvaged. It seems like the threshold for hiding questions from the homepage may be -4 or -5.

I propose two things:

For logged-in users, hide questions that are -3 or lower from the homepage. Hopefully, at least one person has also cast a vote to close if it should be closed and kicked it to the queue. If it's not already the case, I would suggest that at -3 a post gets put into the Low Quality queue for review, even if it's not on the homepage.

For logged-out users (and perhaps logged-in users under a certain reputation threshold), hid questions that are -2 or lower from the homepage. It would be nice to prevent visitors (who, hopefully, register and become members) not get the wrong idea of our community. Voting (and the rationale behind it) may not be obvious. I'd rather err on the side of caution and hide posts that are likely to not be a good fit.

I'd also encourage some network-wide data analysis to check on thresholds for this across the network. Perhaps the current thresholds require too much voting on other sites so there may be opportunities to propose changes to those communities or make default configuration changes.

| |
  • 1
    we probably better wait for a month or two to find out whether this pollution persists or goes away. What we see now may be Eternal September when folks blocked at Stack Overflow try their luck here. This year it just may hit harder than usual, because more SO askers get hit with 3CV and recent cleanup of voting fraud at SO – gnat Sep 25 at 7:24
4

Thomas, I think you are barking the wrong tree. We have a problem on our site, and that problem lies clearly in our community:

It has become almost impossible to ask a question without getting one or two downvotes at least.

It does not matter if it is a good question, a mediocre one or one of those "explicitly off-topic and very, very unlikely to be salvaged" questions - it seems there are always one or two members who downvote, in most cases without giving any sign of explanation, often not even a close vote, which could lead to an auto-generated explanation (unfortunately with a time lag which is IMHO quite too big).

In such an environment, I think the threshold for kicking of questions from the homepage must be lower than on other SE sites, not higher.

Do visitors get a "the wrong idea of our community" that way? I think not, quite the opposite, they see exactly what is going on in our community (even if we don't like it). And I am pretty convinced not the number of visible downvoted questions on the homepage scares visitors, but the unexplained downvotes for their own questions.

If you want less questions with downvotes on our homepage, try to encourage our members to reflect more about their voting behaviour, don't try to hide the downvoted questions behind a higher threshold.

Let me add, I would be very happy if those downvoters here would follow your lead what you consider to be a bad, off-topic, unsalvable question, because I think your personal threshold for this is excellent.

| |
  • Do you have some examples of good questions being downvoted? The 12+ questions on the homepage were all rightfully downvoted, IMO. Even if they didn't meet the criteria for closure and/or deletion, I'd consider them unclear or not useful. One valid concern is down voting and not monitoring the question for edits in order to undo the down vote or up vote. – Thomas Owens Sep 25 at 10:38
  • @ThomasOwens: what is "good" is surely debatable, but 1h ago this one was posted: softwareengineering.stackexchange.com/questions/416218/… - two downvotes, no comment what's wrong, not even a close vote – Doc Brown Sep 25 at 10:45
  • .. or here: softwareengineering.stackexchange.com/questions/416191/… 2 downvotes, one close vote "needs details or clarity", but no comment which details are missing (BTW, the excelllent answer by Christoph shows the question has enough details for being answerable). – Doc Brown Sep 25 at 10:48
  • or this one: softwareengineering.stackexchange.com/questions/416155/… : two downvotes, no close votes, no explantory comment (and perfectly answerable). – Doc Brown Sep 25 at 10:51
  • softwareengineering.stackexchange.com/questions/416119/… got a downvote, no close vote, some comments, but none of them explaining the downvote – Doc Brown Sep 25 at 10:54
  • softwareengineering.stackexchange.com/questions/416070/… - two downvotes, no close vote, no explanation. – Doc Brown Sep 25 at 10:57
  • 1
    ... and I could continue extending this list every day. – Doc Brown Sep 25 at 10:58
  • 1
    Of course, there also tons of crap questions which deserve downvotes, close votes and any explanation would be a waste of time. But if almost any question gets one or two downvotes by default, the differentiation between those crap questions and mediocre, answerable questions gets lost and the voting system loses some of its value. And in this situation, a limit of -4 for questions visible on the homepage is probably fine. – Doc Brown Sep 25 at 11:53
  • 1
    I wish I could see timelines of the votes to understand where votes fit in with edits. There are only daily summaries. I just posted about this on Meta Stack Overflow. I'm also slightly confused - do you think that everyone (including logged out users) should see questions at -3 and -4? Perhaps the thresholds are wrong, but I think the idea of limiting downvoted questions from visitors is a sound idea. – Thomas Owens Sep 25 at 12:11
  • 2
    @ThomasOwens: I think as long as the voting behaviour in our community does not change, seeing questions with "-3" is ok, "-4" not. I am pretty sure that is where the threshold is today. Nevertheless, I think there are too many downvotes casted indiscriminately, and that is probably part of the reason why we see so many low-score questions. – Doc Brown Sep 25 at 12:45
  • 1
    ... Counting the questions on the homepage: I see currently 46 questions there, and only 4 with no downvote (two of them with no vote at all). – Doc Brown Sep 25 at 12:58
  • ... and about "seeing the timeline" - not sure what you expect from it. That would neither change the voting behaviour of our members nor the visibility on the homepage. – Doc Brown Sep 25 at 13:21
  • 2
    @ThomasOwens: I don't have direct examples here, but it's a common occurrence for people to confuse up/downvotes to mean "I (dis)agree with this" rather than "this is (not) well-written and (not) substantiated, regardless of whether I agree with it". I think you consider downvotes the correct way, and with that in mind your suggestion is reasonable, but you've not really accounted for the other kind of downvoters who are trigger-happy, dogmatically disapprove of certain approaches, or vote harshly based on a perception of something being a "trivial" issue in their eyes. – Flater Oct 9 at 10:44
  • 1
    Sometimes I wonder if sites would be better without down-votes. I'm not talking in some hippie Disneyland sense. I've just noticed over the years that down-voting patterns are weird. For example, I see a YouTube video reporting a mass murder with 13k downvotes and 2k upvotes because, from my suspicion, people downvoted just because they didn't like the topic, not because they thought the reporting was bad (the reporting as far as I can tell was solid). They shot the messenger, you know. And I would have thought programmers were at least a smarter breed, but I've gotten doubts over the years. – DemonCode Oct 24 at 15:31
  • 1
    @DemonCode: I have no idea if the problem lies in mixing up the downvote button with a "I don't like the topic" button. And I suspect it is not a problem among software engineers in general, but only the misbehaviour of a few community members on this site - there are still a lot other SE site which don't experience such hostile voting patterns. – Doc Brown Oct 24 at 15:41

You must log in to answer this question.

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