22

Explanation whether duplicate or not --> see end of post.

Imagine you have a question about a game. You first search for it and find it answered a long time ago. Back then the answer was correct, but since then the game mechanics changed several times, so that it is now outdated and obviously wrong for the current version of the game.

How is it possible to get the very same question answered for the current version of the game?

In my case I just asked the same question again (What is the rule of how many new quests you get at one day in Hearthstone 4.X?). But the moderator said this does not comply with the policy (Outdated Answers Due to Patches), which in this case says, that I am supposed to update the answer of the old question.

But I have just the same question and not a new answer for that! Besides a bounty, I don't see another possibility to direct the attention of people to the old question again. And if I haven't enough reputation, I would not be able to put a bounty anyway. It cannot be the solution to wait for a new answer to the old question, can it? So if somebody was new here, it seems to be impossible for her/him to get an up to date answer to the old question. Is that right?

If yes, wouldn't it be reasonable to change something? Maybe it should be possible to mark outdated answers and move questions to the top again. Of course after some approval. Just an idea...

There is a very similar question (How should we deal with out of date questions/answers?) as suggested in the comments, but the answer is not applicable to my case. The accepted answer in this thread just suggests to edit the question and write new or edit existing answers. But I have exactly the same question, but no answer for it. So I cannot take the actions suggested in this thread.

EDIT 1

Here I would like to explain why this post might be a duplicate, but is necessary anyway (Please excuse my poor English, I am not a native speaker)

This question might be a duplicate, yes, but this is what the topic of the post is about. There is currently no way to get answers to a very same question which was asked before, but the answers became wrong and outdated in the meantime, due to game patches etc. The policy suggests here and here (same links as above) that the old answers and/or questions are supposed to be edited. To edit the question doesn't make sence, because I have the very same question. And in my opinion there is no point of editing the answers either, because on the one hand I don't know the answer, because I am searching for it, and on the other hand who will notice if I put a warning that the answer is wrong?

Who might notice:

  • The original authors, who might not know the answer.
  • Users searching for the answer to the question, who don't know it, either.

I think there is a lack of the possibility to direct all users attention to an old question which needs a new answer. It must be the attention of all users, because it is much more likely that one of them knows the answer compared to just the few people which were involved in the thread back then. Of course an expensive bounty is one solution, but not everybody can afford a bounty. Besides the bounty, the only possibility is to ask the very same question again, which is not allowed by the policy. We need something else!

Ironically, the very same thing is currently happening to this meta post, since there were already answers to quite similar questions. But the answers of those cannot be applied to the case I described. That is, why I think it is important to discuss it again and why I think, the existing threads are not enough (because almost nobody cares about / notices old threads = the problem what this post is about).

  • The moderator? You don't mean me, do you? – Studoku Nov 25 '15 at 17:31
  • @Studoku Yes, I do! Are you not? I'm sorry... – Rob Nov 25 '15 at 17:59
  • Apology accepted. In future, you can recognise mods by the diamond ♦ next to their username. – Studoku Nov 25 '15 at 18:33
  • I think the duplicate I raised is applicable still because essentially the recommendation is to either edit the out of date info with a disclaimer and/or add comments so that others (e.g. original author) can make corrections. Such actions will also bring questions "to the top" so to speak, so others will also become aware of the changes. – user101016 Nov 25 '15 at 19:54
  • Also, I have seen others with high rep place bounties on such questions in the past. – user101016 Nov 25 '15 at 19:56
  • 9
    @CamelCase That is the problem: If I edit an answer in the old thread nobody except maybe the author of the answer and others who are searching for an answer to the question will notice. Neither does the original author nor I know the answer to the question for the current version of the game. So it remains unanswered and outdated and nobody is allowed to ask the question again or to do anything else about it. The ironical thing is, that the same problem now happens to this meta post. It kind of is a duplicate, yes, but in my opinion there is still no good solution. – Rob Nov 26 '15 at 15:32
  • @Rob - There's no easy solution to this issue. I've thrown a bounty on the old question, as that really is the best we can do. – au revoir Nov 26 '15 at 22:59
  • @JasonBerkan Thanks, but somehow the question about the game became a minor matter. Good mechanics became more important, now. ;-) – Rob Nov 27 '15 at 21:56
  • @Rob - where are we with the original question? It looks like changes have been made to reflect the current state of the game (although this looks like it will change again in the next patch if I am reading it right!). – user101016 Nov 30 '15 at 16:39
  • @CamelCase Sorean posted a very satisfactory answer in the original thread. But yes, it will change again, and then we will have 2 outdated, high rated answers in the thread. By the way, I don't know what to do with this meta thread? Am I supposed to accept an answer? They have 0, 0 and -1 votes... I think it's not reasonable for one to be chosen, especially not by me. – Rob Dec 1 '15 at 13:40
  • @Rob - High rated because people found them useful/worthy. I wouldn't associate highest answer or accepted answer as the correct answer as a rule of thumb, more a guideline. Otherwise we would only have questions with 1 answer. The post has had a lot of publicity in the past week so the increase of votes is expected (plus it has a bounty). As for accepting a meta answer, you don't have to accept any. Is there an eta on the bug fix? May be best to wait until then. – user101016 Dec 1 '15 at 14:07
  • @Rob - also, you may only be able to see the net affect of voting on answers. There have been 42 votes (at this point) across all 3 answers. Pretty even split of agree/disagree for each answer. – user101016 Dec 1 '15 at 14:10
  • @CamelCase You mean the comment votes? Or where can I see the number of votes an answer received? After I read all the answers, comments and linked threads, I think I understand both views on the subject. I will do it later when I have more time... – Rob Dec 1 '15 at 14:22
  • @Rob - when you get a high enough rep you can click on the answer votes. It gives you a break down on the up and down votes. – user101016 Dec 1 '15 at 14:38
  • @Rob - In particular, it takes 1000 rep to see the vote breakdown. gaming.stackexchange.com/help/privileges/established-user Unfortunately, as shown by the rep league chart at stackexchange.com/leagues/7/year/gaming/…, it takes most people over a year to gain that much. – Mike R Dec 1 '15 at 18:35
5

I think it's pretty clear from the comment votes that the only thing we all agree on is there is a problem here.

In a perfect world, a new question is asked, marked as a dupe, asker either a) comments on the original answer to say it's out of date or b) complains on their dupe loudly enough that someone does it for them (if their rep is too low, for example). The author of that answer gets a notification and updates the answer. We close the dupe, everybody wins.

Equally perfectly, the author of the original question notices the comments on the answer, and removes the checkmark. When a correct answer is posted, they accept the new answer.

I don't think I'm risking too much by saying we can all agree the best case is either of those two (or something very close to them).

That's not always going to be the case, and we need a policy for when this stops working. Let's briefly go over the rationale for my proposal.

I don't like editing the accepted answer because the votes go to the wrong place. If the editor is wrong, they have nothing to lose. Clearly if the answer is abandoned, the author doesn't care what happens to their rep, so why should a third party edit net them more or dock them? And if we edit in a disclaimer, why let the incorrect information stand? That doesn't seem to be in line with the spirit of our edit policies, at least not to me.

Bounties are great, but it puts too much onus on to few users. I just hit 3K, I'm not placing any bounties that will lose my voting privileges. It seems to be a trend that folks with extremely high rep forget how much of what you can do is based on relatively high rep scores. 200 rep is a lot when you have 3K.

Finally, down-voting simply doesn't work, if the answer has been abandoned, that original author doesn't care. All the reasons bounties are subpar apply here as well.

If it becomes clear to reasonable members of the community that the accepted answer to an abandoned question is both out of date and abandoned by it's author, we should delete the answer. What flag to raise could be debated, but I say we use "moderator attention." We normally shy away from one user making categorical statements about accuracy, but that's because we rely on the voting system to make sure the good stuff rises to the top. The system falls apart on mature questions, for a two simple reasons. First, a question with an accepted answers will draw less view from the community, which gives less chance for those down-votes or a new, accurate answer. Secondly, rather than starting at zero like normal, the answer "starts" with up-votes.

We have a responsibility to prune old information from the site, especially when that old information is inaccurate. We've yet to have an actionable policy about this, and we truly need one with the games of today. So please, if this proposal is unacceptable to you, let's work on hammering one together. Our band-aid fixes won't keep working forever.

(I'm aware the question that spawned this has been taken care of, and if you think this meta had absolutely nothing to do with that, well, I admire your faith in humanity. Me, I think it played a major role, so pointing to the question in question as an example of the system working doesn't hold much water with me.)

EDIT: I should really have slept on this before posting. I thought what I was saying was obvious, but it l clearly isn't.

I'm not proposing we immediately delete a question as soon as it's outdated. We'd first behave normally, commenting on the question and answer to give everyone a chance to correct it normally. If neither the author of the question nor the answer makes any change within some agreed upon grace period (7 days? 10? 14? I don't know, please suggest something, I'm here everyday, my perspective of reasonable is probably skewed), then and only then do we take steps to delete the accepted answer.

  • How do we determine a question is abandoned? How about if the answer is out of date? Why should mods be tasked with deleting this? They're not going to be experts in every single game. Does that mean we close outdated questions so that 10K+ can vote to delete? That runs headlong into another meta, where we leave old questions alone. At the end of the day, all you're doing here is creating more problems. – Frank Dec 4 '15 at 0:18
  • 3
    There's no reason to have spoiler markup in a meta answer. I've removed it. – Ash Dec 4 '15 at 0:21
  • Nice point on obsolete answers starting with upvotes, especially if they were very well received. Also not to forget is that questions can't be voted on again unless edited after a period of time, so a +21 question would have to be first edited before any of those voters could even change their vote. I see deletion as a better option, although I don't like the idea of taking away the answerer's rep just because the answer is outdated. – GnomeSlice Dec 4 '15 at 0:29
  • Wait your stance is to delete all out of date questions and answers? How is that fair, people will lose thousands of rep. – Aequitas Dec 4 '15 at 0:36
  • 2
    Sorry, I just re-read your proposal; I had originally thought you were proposing deletion of questions and answers. If it's just about deleting out of date answers, I would support that, but only for games where there is effectively only one version playable. So for MMO's, and other online only games, generally. – Frank Dec 4 '15 at 0:39
  • 2
    @Aequitas 1st, I said nothing about questions. 2nd, rep is permanent after 60 days. 3rd, even if the rep WAS lost, if someone can't be bothered to edit their own question after its brought up that it's wrong, they don't care about their rep. – Dallium Dec 4 '15 at 0:40
  • 2
    @Frank these days, it's most games. You have to actively try to keep a game out of date on PC, and if your (modern) console is connected to the Internet, you CANT play previous versions. I guess I don't see the benefit of keeping an answer out of date for a tiny fraction of users who both play the game on that exact version and just so happen to have the same question. Minecraft is really the only example I can think of off hand that has a decent chuck of users who play old patches. That doesn't mean they don't exist, though. – Dallium Dec 4 '15 at 1:17
  • While it isn't desirable to have inaccurate answers hanging around with positive scores and checkmarks, it's not the mods' job to determine the accuracy of an answer. In fact, there's a specific flag rejection reason that indicates that flags should not be used to point out technical inaccuracies in an answer. – Unionhawk Dec 4 '15 at 1:18
  • @Frank as to how we determine whether a question is abandoned, the first step would be to check the original authors profile. If they are still active, they should be given more or less unlimited time to make whatever edits are necessary. If they haven't logged in for years... I don't have any concept of a fair grace period, I've been here everyday since I joined (I'm assuming that's not typical). I would think a week would be too little time. It not like I think we should burn questions the instant a patch drops as a first step. – Dallium Dec 4 '15 at 1:21
  • @Unionhawk because we rely on the voting system to determine accuracy. I mentioned that in the answer. Am I wrong in thinking there's no extant method for a user to mark an answer for deletion? – Dallium Dec 4 '15 at 1:22
  • 2
    @Unionhawk right now they aren't, yeah, because the current practice is to just shrug and let out of date info stand. I don't think we'd have much success with a feature request asking for a mark for deletion flag for answers, but it might be worth a shot. The subject matter in other stacks isn't as changeable as it is here. The bald fact is this is going to happen more and more as games continue to be released bigger and buggier and get major patches well into their lifetimes, even single player games. We need to be able to do SOMETHING. – Dallium Dec 4 '15 at 1:46
  • 1
    At the moment, I can't support this proposal. Answers are useful forever, unless they're for an online game. I'd support a mod delete for outdated answers, with those caveats, but we'd definitely have to put bounds and process on it. – Frank Dec 4 '15 at 2:27
  • 1
    @Frank Which people? Updates are automatic in the majority of cases. – nukeforum Dec 15 '15 at 23:50
  • 1
    @nukeforum Well, I had more meant to use libraries to research their issues, not play Skyrim. Just to clarify, as that wasn't what I had meant. :) – Frank Dec 16 '15 at 1:19
  • 1
    Either way, though, adding additional curation efforts isn't really feasible. The only ones who can delete upvoted answers are mods, and we're not going to be putting this all on their shoulders; they have enough to do. You're not going to get enough downvotes for old answers for even 20k'ers to help. I know; I've tried. – Frank Dec 16 '15 at 4:44
2

Our duplicate policy shouldn't apply to outdated questions

Stack Exchange works around votes and accepted answers, allowing good answers to rise to the top. This does not work as intended if answers earn votes then become incorrect.

In many cases, the asker of the old question is no longer an active user. This means that even if the updated answer is posted, and it gets enough attention and votes to rise above any other answers, it will always be below the accepted, incorrect answer.

Editing the correct answer into the accepted answer, which is often suggested, goes against our edit policies and means we allow users to attribute answers to other users. It also sets a dangerous precedent- if we edit outdated answers, what about incorrect answers?

Instead, users should be able to re-ask outdated questions and get the answers they need- that's what we're for, right? The old question should be closed or edited and linked to the new one if it gets a good answer.

  • 1
    What? No! How do we determine if an answer is outdated? It's a bit more effort for us to try to keep old questions up to date, but that's why we rely on established users to do what is needful; it's still a dupe, so one of us can bounty the original. Done and done. – Frank Nov 27 '15 at 20:32
  • 9
    @Frank Aside from the fact that that almost never happens, whereas answers becoming outdated are getting more and more common. – Studoku Nov 27 '15 at 20:52
  • 1
    Entropy kind of dictates that; it happens on SO quite often. If answers are outdated, fix it! Don't gunk up the process by allowing dupe to exist. – Frank Nov 27 '15 at 21:13
  • 7
    @Frank the gunk is the outdated questions and answers, not the new relevant ones. The outdated content serves no useful purpose and may actively confuse users. – Dallium Nov 28 '15 at 2:06
  • 2
    @Dallium No, the gunk is multiple questions all asking the same thing. Outdated answers can still serve a purpose, if those versions can still be played. If it's an online thing, and you can't play anything but the latest version, downvote them! That's what it's for! – Frank Nov 28 '15 at 2:41
  • 1
    The whole point of duplicates is to gather all the information into a single spot. We're trying to get away from the forum format, where people have to dig through multiple threads to find their answers. – Frank Nov 28 '15 at 2:45
  • 6
    @Frank for the question to be outdated, by definition, the version can't be playable (or at least, not playable in an official capacity). Honestly, from what (little) I know about how questions age, it seems clear that the system was designed from the ground up to deleted outdated questions. Don't they not lose rep after a certain amount of time has past if the question is deleted? And no, downvoting an answer that is incorrect NOW, but was correct back then is NOT what downvoting is for. Why would you want to hang onto questions that are terrible because they're old? – Dallium Nov 28 '15 at 2:46
  • 6
    We have a responsibility to prune old information, and pointing new users at questions that are multiple patches out of date doesn't serve our mission at all. – Dallium Nov 28 '15 at 2:47
  • 5
    @Frank So now we're punishing people who answer questions that happen to become outdated.? – Studoku Nov 28 '15 at 2:48
  • 1
    No, we're judging content that is no longer useful. We don't judge users here, @Studoku. You know that. Besides, by that point, they've already gained a ton more rep from the answer than they're losing. If an answer is no longer useful, that's what downvotes are for. That's literally in their description! – Frank Nov 28 '15 at 3:41
  • 3
    We downvote non useful answers. If your answer is wrong, but it was right back then, it is still not helpful, now. If the author of the answer is still around, and is concerned about losing their rep, they should equally be ensuring the answer is up to date. If you post a comment to explain your downvoting due to the information being invalidated, the author will be alerted. I will personally convert my answers to wikis if they concern online games I do not play, anymore. – user106385 Nov 28 '15 at 8:01
  • 5
    I feel like @Frank 's solution is great, in a world where no one ever leaves a website and is always dedicated to it's goals. But here in the real world, the original asker either isn't a memeber here anymore, or doesn't play the game anymore, or both, and so gives exactly zero fucks as to the accuracy of the answer to their question. So that totally inaccurate accepted answer stays at the top, regardless of disclaimers. And it's up to US, the community members who DO GAF, to make sure ACCURATE information rises to the top. Close old, upvote new. – Dallium Nov 28 '15 at 9:59
  • 3
    @Arperum what second question? We deleted the old one. It wasn't relevant so we deleted it. – Dallium Nov 28 '15 at 22:25
  • 4
    @Dallium Wait, what? We're deleting the old questions now? Who's proposing that? I'm not seeing any proposal for that. – Frank Nov 29 '15 at 4:31
  • 3
    @Aequitas and if the original author edits them or if the question author revokes the check mark, they don't have to be. Our mission is not historical preservation, it's to provide answers to questions. If that answer is wrong, we have a duty to ensure it gets corrected or removed. – Dallium Dec 4 '15 at 2:15
-3

I believe this meta is a duplicate of the following:

How should we deal with out of date questions/answers?

The concerns and answers provided cover the concerns highlighted by the OP of this meta.

If the question is still relevant, but in some way no longer accurate, it should be edited to reflect the current reality. Answers can then be edited, or a new answer written as needed.

This will allow for amendments, where appropriate, to be made to the question. More importantly, it allows for incorrect answers (due to a patch) to be given a disclaimer.

If the question is abandoned, and the accepted answer is inaccurate and it is unlikely for the accept to be moved, editing it to add a disclaimer notifying readers about that fact and pointing them to a newer answer is considered the Right Thing To Do.

As above, but addresses concerns where the accepted answer does not change.

The OP is concerned moreso (in my opinion) with the possibility that no one will answer the question without a new question being added. However, any edits will bring the question to the top of the active questions list. Edits will notify the original asker, and to people who answered the original question. They may also be the ones with the expertise to provide an updated answer.

A bounty can be used to promote a question, and an option exists to request for updated answer.

I think this still holds true:

I personally see no problem with this method, and don't see any reason that we need any sort of special mechanics to revoke acceptance or any such beyond the tools we already have.

  • The current mechanics is, to edit the question? Although,I want to know the answer to the very same question? Editing will bring it to the top. So to discuss again,I would have to rewrite the question of the thread you are citing? Does this "bringing to the top" also apply when editing an accepted but outdated answer? This is what actually needs editing. Otherwise someone has to edit the question again,after a correct answer was given. And if somebody cannot edit answers and gives a new and correct answer,who accepts the answer if the asker is not active anymore? That is still an open problem. – Rob Nov 27 '15 at 21:53
  • 3
    Are you sure people will notice an edited answered question? 99% of the time it's at the top of the list because some spammer or NAA post bumped it. I don't think people are genuinely looking for subtle signs that answers are out of date. – Troyen Nov 28 '15 at 4:42
  • 2
    "Oh look, an answered question is at the top of the site. Ignore" – Dallium Nov 28 '15 at 6:06
  • If you don't have enough reputation to use a bounty, then for sure your edit will go into a review queue. This gets the attention of at least 2 high rep members. Also, as I said previously, it gets the attention of people involved in the original question. Plus people may be watching the game's tag, and of course people watching old questions rise to the top of the active list. Surely, most of these actions will help get an updated answer for the question? It did for the question that was the source of this meta. – user101016 Nov 29 '15 at 22:58
-3

If we're not closing new questions as duplicates because our existing answers are outdated, we're failing one of our core tenants: To cut the noise to signal ratio. By allowing them to exist, we will start versioning questions; how does this mechanic work in version 1? Version 2? Version 3? Are we really making the internet a better place by deliberately allowing duplicates because the current answers are outdated? Should we be duping the old ones to the new ones? We'll eventually end up with long dupe chains, and that's even worse for user experience and SEO. Now they have to follow a long chain to actually get their answer. Do we not dupe them at all? Now users have to troll through several versions of the same question, trying to find the correct one that applies to them. That's one of the exact things SE was created to counter, not encourage.

I subscribe to the late @MarkTrapp's answer: We expect more from established users, so that new users can get answers to their duplicate questions. If it's a dupe, dupe it. Then bounty the original. We're here to help curate the site; it's not the best solution, but it's much better than allowing duplicates to exist. It keeps all the answers in a single place, whether or not you can play the old versions.

If it's an online game, and an answer is no longer relevant because it's for an older version, use your downvotes. That's what they're for, after all; denoting that an answer is no longer useful. That will hopefully spur the answerer to update or delete their answer, and then everybody wins. In extreme cases, if an outdated answer is accepted, and an update or delete isn't happening, we can flag for mod removal. If not, the downvotes will indicate that an answer is no longer useful, and make it less likely for users to read it.

That's using the tools at our disposal to their maximum effect. We don't need to allow duplicates. We just need to expect more of ourselves.

  • 3
    Are we really making the internet a better place by deliberately allowing duplicates because the current answers are outdated? Yes. Unequivocally, absolutely, yes. – Dallium Nov 28 '15 at 10:02
  • 5
    @Dallium No. No, we're not. We're creating additional cruft, which is one of the very things SE was created to prevent. Having to search through multiple threads to try to find an answer is why SE is so focused on questions and answers. We're killing that process by allowing duplicates. – Frank Nov 28 '15 at 14:54
  • 2
    You're essentially putting the new user experience above that of site curation. At the end of the day, we're here to help many users. Not just the one that asked the question. – Frank Nov 28 '15 at 15:23
  • 4
    @Dallium as I also said on the other answer: your duplicate question can and will confuse people. People will find the old question that doesn't contain any correct information anymore, but not find the new and correct information. By updating the old question with new answers they do find the new and correct information. – Arperum Nov 28 '15 at 16:19
  • 5
    Essentially what you are proposing is to let old, outdated, inaccurate answers stand because you can't be bothered to curate the site. – Dallium Nov 28 '15 at 22:24
  • 1
    @Dallium, I think it is better to assume the user can scroll down and check the next answer, if the previous does not work, then to go against one of the basic tenants of stack exchange. We dont do duplicates. Period. That is why "duplicate" is its own close reason. – user106385 Nov 29 '15 at 0:56
  • 3
    The problem with this approach is that if nobody wants to bounty the question to get it updated, we have a question with an incorrect answer that will appear to others as correct. There should be another solution. – Mike R Nov 30 '15 at 19:18
  • @MikeR You're welcome to come up with one. So far, this seems to be the best we have. Allowing duplicates is a non-starter. If no one is willing to bounty the original, then Arqade is no longer fulfilling its function, and we might as well shut the site down – Frank Nov 30 '15 at 21:57
  • 3
    The duplicate question is only a symptom. The root problem is that there is an accepted answer which was previously correct but no longer is. What do we do in those cases? Do that here as well. ( I couldn't tell what the policy regarding regarding out-of-date answers is. There are multiple meta questions on that topic and the accepted answers for those don't seem to agree with each other or with the policy that edits should not change the intent of the answer.) – Mike R Nov 30 '15 at 22:35
  • 3
    I'm not sure we have a set policy, but I believe a case could be made for flagging for deletion. Only for games where there's effectively one version,though; MMO's and the like. For offline games, we'll just need to keep accruing different versions answers; they're all useful. – Frank Nov 30 '15 at 23:03
  • 1
    @Frank I agree with the flag for deletion idea, and that a distinction needs to be made between online games with forced updates and other games. – Mike R Dec 1 '15 at 15:37

You must log in to answer this question.

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