9
$\begingroup$

The topic of self-answering questions immediately is a complex one. On one hand we do not want to fill up the site with low-quality material by people hungry for raising their reputation. On the other hand, new users have a disadvantage because people have reduced their upvoting frequency (not me! I'm a site leader in overall upvotes; everyone else please keep voting with the same frequency as you did at the start, so that it's fair for the new users!), and we want users to participate, and we also want to have more material on the site that's searchable and attracts more users to visit.

If a user just asks how to do something in their favorite software, and immediately answers with how to do it, perhaps we want to consider whether or not that question is really something challenging, or if it's just regurgitating what someone already knows how to do. If a user is immediately self-answering easy questions of their own more than say, once every 6 months, I'm tempted to consider converting the question and answer to Community Wiki so the user doesn't get points (if the user is doing this to provide more searchable content on the site, then they'll be okay with this; and if they are doing this to gain quick reputation, then it's questionable whether we want to allow it). However Tristan Maxson suggested that we make just the answer a community wiki. Maybe that's a better idea. Even then, we run the risk of allowing a lot of low-quality content to fill up the site, and users will get a lot of upvotes for a question that might not actually be great, but it seems like a fair middle ground (for now) between making both the question and answer Community Wiki, and allowing the users to "run wild" with immediate self-answering for personal reputation gain. Shall we implement Tristan's suggestion? I'm in favor of it as a temporary measure for now, until we've had more time to think about it and to see the results of how this works out.

| |
$\endgroup$
  • 1
    $\begingroup$ Agree with everything listed, an official policy would be very beneficial $\endgroup$ – Tristan Maxson Oct 17 at 19:51
  • 1
    $\begingroup$ I agree too. I know that is possible to post a question (even in multiple sites/forum) and at the same time continue looking for the answer and when found it, self-answer but the ones I already see here, looks not being the case as the answer was too fast. $\endgroup$ – Camps Oct 17 at 22:38
  • $\begingroup$ @TristanMaxson I think it would be good if you posted a short answer in favor of this proposal. You don't necessarily have to include all the details of why its a good idea, but it would be helpful to have answers for each position and let users vote on them. $\endgroup$ – Tyberius Oct 19 at 14:20
  • $\begingroup$ Another point, regarding fairness. If we are all just here to provide valuable content to the community, then why should we care if the distribution of rep is fair? If rep doesn't matter, then why would it matter if some users get more of it? $\endgroup$ – Tyberius Oct 19 at 16:18
  • $\begingroup$ The issue is not the rep... It's that when people get hungry for rep, it leads to low quality content. I'm late for something, so if we can have a "town-hall" type chat here later: chat.stackexchange.com/rooms/113019/… it might be a good way to discuss! $\endgroup$ – Nike Dattani Oct 19 at 16:20
  • $\begingroup$ Oh sorry, I'm pretty new to this stack exchange methodology. Most of the questions and answers Ive made are the one's me and my collegues had while we were learning the software a month ago. Some of those incidents like the cif2cell error, the ev.x tool and mumax3 visualization issues puzzled us for days before one of us found out a suitable solution. I've googled for them, searched for them in the stack and since nothing of such sorts was found I've posted them here. And the rapid nature of questions and answers coming one after another is because theres an option to answer your own question. $\endgroup$ – Anoop A Nair Oct 21 at 15:59
  • $\begingroup$ And there's something being refered to as a community wiki, I'm entirely not sure as to what that is. But anyway If it's okay to post it there I'll be happy to do so provoded that it would show up when people search for it. $\endgroup$ – Anoop A Nair Oct 21 at 16:00
  • $\begingroup$ And regarding the quality of the answers, anyone who has started to use the software would be able to follow the procedures mentioned in my answers to resolve their issue. I've also linked the websites containing the appropriate tools and info whenever possible. $\endgroup$ – Anoop A Nair Oct 21 at 16:03
  • $\begingroup$ @AnoopANair please try not to write too many comments since it leads to the creation of new chat boxes, which becomes too hard to control: meta.stackexchange.com/q/353643/391772. The last two comments of yours, could have fit into one comment :) $\endgroup$ – Nike Dattani Oct 21 at 16:18
  • $\begingroup$ @NikeDattani Sorry for that. I'm getting used to this format of conveying ideas. So where should I post my comments for this discussion? So I have told my take on this problem so is there anything else I should keep in mind? $\endgroup$ – Anoop A Nair Oct 21 at 16:23
5
$\begingroup$

Against

Upvote this answer if you are against this proposal. To avoid double counting, don't downvote any other responses, just upvote the one you agree with.

I don't think we need a policy of converting to community wiki. Self-answers are fair game on StackExchange, so in principle there is no problem with it. I think self-answers can be a great way to address common questions in a general fashion without having to constrain yourself to fitting the parameters of a given user's question or wait for that type of question to appear.

As to the quality of these questions/answers, I think we have a simpler solution: if the q/a is useful, upvote them; if they aren't leave them alone, downvote them, or (if warranted) vote to close them. If the q/a aren't high quality, it won't be a reasonable way to hoard rep. If they do get a lot of upvotes, clearly a decent number of users felt the content was valuable. Basically, if we don't want to reward poor questions/answers, I think the easiest way is to simply not reward/upvote them. I would rather not preemptively curtail self answers, as they could be a way to provide a lot of great, tutorial-style content to the site.

One of mine own recent questions here wound up having a self-answer: Open Source PyMol Conda Package: UnsatisfiableError. I think the question is just okay, but it was pretty well received. I happen to come up with the answer a few hours after posting the question and the answer seemed to be helpful to some people as well. However, had I found the answer shortly after or even before asking the question, the implication of this post is that it would have become a bad q/a because I already knew the answer. I either should have not posted it (which would have meant other users might not have found this solution or even known that others had the same problem) or I should have made either/both the question and answer community wiki (which seems to unduly penalize me for putting in the effort beforehand to try to solve my own problem).

I would also point to another example on Chem SE. This a question I deliberately made to self-answer. I had this problem and the answers I found to similar questions on Physics SE seemed overly complicated, so when I found a different way of doing it, I thought it would be useful to Chem SE. This was again a legitimate problem I was trying to solve, I just happened to do so before ever asking a question. Instead of waiting for a new related question to come on Chem SE or adding to old questions on a different site, I decided it would be more useful to present it using a self-answered question.

My point is that I think the quality of an answer is independent of who posted it and how long after the initial question it was added. If an answer is good, we upvote it; if an answer is bad, we don't. I don't think we should be checking who posted it to see if they are deserving of upvotes.

| |
$\endgroup$
  • $\begingroup$ There ought to be limits though. You say it's "fair game on SE" but meta discussions on for ex. Space.SE point out that it's okay "as long as it's not done too much". The problem I have with allowing it, is that it can lead to quite a lot of content that's perhaps not low enough quality to downvote, but it still down significantly drop the average standards of our site. I wouldn't want the whole site to be filled with low-quality self-answers. Self-answers may be fair game, but self-answering immediately, means the user didn't really have a "question" to ask, they just tried to gain easy rep. $\endgroup$ – Nike Dattani Oct 18 at 21:48
  • $\begingroup$ Again: If they're doing it to provide the site with useful content, and not doing it just to get points, then they'll be okay with the content being there as community wiki. The only reason they wouldn't be okay with the content being converted to community wiki, is if they care more about gaining rep. If they are doing this just to gain rep, it scares me what this site might become. I could post 100 questions today and self-answer them all, but I don't do that for several reasons. $\endgroup$ – Nike Dattani Oct 18 at 21:49
  • 2
    $\begingroup$ @NikeDattani But if the question/answer isn't good, it's not easy rep. At best it gets them a little, at worst they get downvotes/flags/close votes, which could get them suspended if it happens enough. You could apply the same logic to every answer: if they care more about providing content then about rep, we can just convert every answer to wiki. If it really becomes an epidemic, with many users posting low quality self answers, then I think we could consider doing this to stem the tide. $\endgroup$ – Tyberius Oct 18 at 21:58
  • $\begingroup$ But with just 1 or 2 users, who have decent self answers mixed with occasional low hanging fruit, it's easier to take them case by case and talk to them personally if the ratio of good/bad or the rate of these self-answers gets out of hand. $\endgroup$ – Tyberius Oct 18 at 22:00
  • $\begingroup$ I don't think it's just a "little" bit of rep that gets gained this way: One case alone got 120 rep, and if the person does it many times it does add up. The question that got 120 rep displayed very little effort (the question sentence didn't even end in a question mark .. I'd edit it, but don't want to bump it up to give it more attention, especially since I just finished bumping up several good-quality unanswered questions .. maybe a future feature request would be to not bump up questions where the edit is too minor). $\endgroup$ – Nike Dattani Oct 18 at 22:03
  • $\begingroup$ Finally: out of ~1500 users, there's only been one person self-answering, and it's been done 4 times by them. all in this month. Clearly it goes against the grain of what our community does. I feel bad for the users that have the same rep but earned it all by answering other people's questions. I hope they don't get too discouraged (or resort to develop the habit of self-answering). $\endgroup$ – Nike Dattani Oct 18 at 22:04
  • 3
    $\begingroup$ @NikeDattani in regards to the 120 rep, apparently 6 users found the question/answer useful and I don't think we should be unilaterally deciding for the community what is and isn't useful enough to warrant rep. And as you said, it's one user. It would be much easier to just comment on one of this users posts and tell them to slow down the rate of such posts then to develop a whole policy for something that basically isn't happening. They are unlikely to ever see this Meta post, as metas tends to get a few orders of magnitude less traffic than main sites. $\endgroup$ – Tyberius Oct 18 at 23:30
  • $\begingroup$ His self-answered questions are getting I think an average of about 100 points in total between both the Q and the A. People upvote for all sorts of reasons, and I think the least of these reasons is that the question is useful. The issue is that if people see self-answered questions like that, and nothing is done about it, people are going to think it's acceptable to do that. I personally find it unacceptable (what goes on in other sites is a different story). Camps and Trsitan are in favor of turning answers to community wiki, so it's 3 against 1 so far. Let's see if others chime in. $\endgroup$ – Nike Dattani Oct 18 at 23:47
  • $\begingroup$ I see your point that "all answers" could be made community wiki if we all care more about the site than about rep (and trust me I do care much more about the site's survival than my own personal success on the site, which is why I'm constantly trying to get people's questions answered, and why I didn't fill up the site with self-answers, and never answered twice on the same one-topic/answer question). However someone that answers a challenging question after it's been unanswered for weeks ought to get more rep than someone that created a question just so they can immediately answer it. $\endgroup$ – Nike Dattani Oct 19 at 16:01
  • 2
    $\begingroup$ @NikeDattani I disagree. I think the time a question is left unanswered doesn't change how valuable/useful it is. A user could hypothetically have spent months answering a question, never having posted it until they arrived at an answer. I don't think they should be penalized just because they didn't leave the question sitting here for months. And again, I don't think we should be deciding for other users whether they are allowed to find an answer valuable enough to reward with rep. $\endgroup$ – Tyberius Oct 19 at 16:06
  • $\begingroup$ These things are possible in edge cases, but for the most part, a user that posts 4 questions with immediate self-answers one after the other, is engaging in behavior that is uncommon in the context of our community's etiquette. By the way, do you want to make your answer a community wiki so that you can upvote it? Because otherwise it's unfair if Tristan can upvote his answer but you can't upvote your own. Also I'd rather chat here: chat.stackexchange.com/rooms/113019/… than to create a new chat room, so I'd prefer if no one clicks the button. $\endgroup$ – Nike Dattani Oct 19 at 16:16
  • 2
    $\begingroup$ I self-answered twice in Why the Open Babel submenu is faded out in Avogadro 2 and I can't optimize geometry in Mint 20? and How to make the input generator plugin work in Avogadro 2 on GNU/Linux?. Both issues bothered me, to document it can help others, and I think to outright penalize it would be unwise. $\endgroup$ – ksousa Oct 20 at 14:53
  • $\begingroup$ @ksousa the second example was not an "immediate self-answer", and both of your self-answers were of the highest standard! Truly a remarkable amount of effort with all those screenshots! The 4 examples of "immediate" self-answers which occurred recently (all by the same user!) were super quick, with just a few lines in the Q and a few lines in the A. Finally, your use of the word "penalize" is incorrect, because we're not penalizing them, just making it so that only the Q gets reputation, not both the Q and the A. $\endgroup$ – Nike Dattani Oct 20 at 19:17
  • $\begingroup$ @NikeDattani But the first one was an immediate answer, so under this policy, despite the answer being high quality, it should have been converted to community wiki and received no rep. Thats really my whole, that this policy groups bad self answers with good ones. If the user would otherwise have earned 160 rep and because of the policy only earns 80, that seems like a penalty. $\endgroup$ – Tyberius Oct 20 at 19:22
  • $\begingroup$ Answered in chat (to prevent comment chain from getting so long that a new chat room gets created): chat.stackexchange.com/rooms/113019/… $\endgroup$ – Nike Dattani Oct 20 at 19:36
3
$\begingroup$

For

Upvote this answer if you are for this proposal. To avoid double counting, don't downvote any other responses, just upvote the one you agree with.

Tristan Maxson:

I personally think this type of answering is commonly done with the goal of gaining reputation. I think that while the upvote/downvote system should filter out bad self QA, in practice it will not. People will see a question and an answer and most people will probably upvote it regardless. This is especially dangerous when its some specific code and its hard to tell if the answer is even a good one.

This proposal does not eliminate this, but it makes it less tempting for the users seeking reputation, keeping its usefulness. We can also avoid converting it to a community wiki if it has been around for a week or so if there is a concern about people never self answering when its a long term question which they eventually solve.

(Other users can edit in details in support of this position).

$\endgroup$
  • $\begingroup$ I totally agree. If you post a question & answer it immediately, it's either to provide valuable knowledge to the community, or it's to gain rep. If it's not to gain rep, they will not mind it becoming a community wiki. If it's to provide valuable knowledge to the community, the question becomes: how hard really was this question, if the user already knew the answer? People may upvote because they like the question (or for all sorts of reasons), but how many of them are upvoting specifically because this is a difficult question which most people would not be able to figure out the answer? $\endgroup$ – Nike Dattani Oct 19 at 15:55
  • $\begingroup$ What I mean in my last sentence, is that: upvotes do not indicate whether or not this question was actually a hard one. People can upvote because the question was well constructed, or showed great effort (several paragraphs of MathJax and 10 journal references), but does that mean this question's solution was non-trivial or actually useful? Just because there's upvotes doesn't mean it's a useful question. $\endgroup$ – Nike Dattani Oct 19 at 16:07
  • $\begingroup$ @NikeDattani But thats just it. The premise of this proposal is that all self-answers are trivial or not useful. Otherwise, there is no need for a blanket policy and we can address trivial questions on a case by case basis. It seems like the proposal is to use self-answers as an indicator of bad answers and that seems like a criteria that is going to produce a lot of false positives. $\endgroup$ – Tyberius Oct 19 at 16:13
  • $\begingroup$ I agree and you raise good points. I unfortunately have to go (I'm actually quite late for something), but it sounds like all 3 of us (and others) may like to have a "town-hall" type meeting in the chat room I linked to in my last comment to your answer. We can all discuss in real time! $\endgroup$ – Nike Dattani Oct 19 at 16:18

You must log in to answer this question.

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