27
\$\begingroup\$

Sometimes, not that rarely, rules change in RPGs. Answers that were good become objectively, definitely wrong overnight, through no fault of their writer.
The accepted course of action is to do nothing, hoping that the previous voters realize the change in the facts, and modify their votes accordingly.

However, they do not. Either they do not know there was an errata, and even if they do, you can not expect people to follow all their votes, for ever. And if they do, they might not change their votes, because of the time constraint ("Your vote is locked unless the answer is edited"). The right answer for this question has only 4 votes, 10 less than the wrong one. Half a year has passed since the errata, so our approach of waiting for the problem to fix itself clearly does not work.
Someone who is unaware of the errata gets the wrong answer, which is against the intentions of this site.

Update after a week: The votes are now 11 to the right answer, 9 to the old one, very far from convincing.

How can we fix this?

I see 6 options:

  1. Downvote the old answer, upvote the new one, comment on old one
    • Pros: Does not offend anyone, moves things in the right direction, takes small effort
    • Cons: I just demonstrated that it failed to give the expected result even after 7 months
  2. Create a meta question about it
    • Pros: Does not offend anyone, moves things in the right direction
    • Cons: I just demonstrated that it failed to give the expected result even after a week, also quite high effort (I planned doing this question for a month)
    • Cons: Invisible for a casual user, who just wanted to find the right answer for the original question
  3. Completely rewrite the old answer to reflect the new ruling
    • Pros: Now the highest rated answer is also the best one
    • Cons: Against current rules, more importantly could start editing wars
  4. Edit the old answer, add "obsoleted by rules change" as first line
    • Pros: Low effort, highly visible
    • Cons: The writer of the answer could delete it
  5. Use the flag "in need of moderator intervention"
    • Pros: Low effort, highly visible
    • Cons: Not immediatel clear why it was flagged
  6. Create new flag "obsoleted by rules change"
    • Pros: Low effort, highly visible, and clear
    • Cons: Might not be technically feasible

I would suggest option 4 or 6.

\$\endgroup\$
  • 2
    \$\begingroup\$ I suspect much of it is votes that are locked... \$\endgroup\$ – Shalvenay Feb 2 '16 at 23:14
  • \$\begingroup\$ Data: 10 upvotes are locked, 5 are not locked. Of the 10 locked upvotes, 3 of them were cast after the comment pointing out the errata. (Note that all existing votes will unlock if someone does an edit on the post to fix typos, etc.) \$\endgroup\$ – SevenSidedDie Feb 2 '16 at 23:27
  • 7
    \$\begingroup\$ Calling meta’s attention to the problem seems like a reasonable solution for these cases when they come up. \$\endgroup\$ – KRyan Feb 3 '16 at 16:57
  • 5
    \$\begingroup\$ I don't think it's particularly reasonable to call someone "lazy" for not keeping up with hundreds or thousands of maybe-updated, maybe-not votes, their current correctness, and so forth. Now, if someone is using my script to check for retractable votes (shameless plug!) and doesn't catch this, then they might be lazy. \$\endgroup\$ – user17995 Feb 4 '16 at 2:20
  • \$\begingroup\$ should the second half be an answer? \$\endgroup\$ – András Feb 9 '16 at 22:29
  • 3
    \$\begingroup\$ Numbers 5 and 6 would be altering stack norms in a pretty significant way and I'd certainly be loathe as a moderator to begin enforcing any kind of "correctness" standard wrt answers. Four is also fairly borderline. It's much much much better to simply let voting take care of this. \$\endgroup\$ – wax eagle Feb 9 '16 at 22:31
13
\$\begingroup\$

Downvote and (optionally) comment the old answer, upvote the new answer.

This actually has a twofold effect. It notifies the OP of the old answer that there may be an issue that needs to be addressed, and it also serves to begin the process of floating the new answer to the top and dropping the old answer to the bottom.

If this is still insufficient, you could (and I've done this myself), offer a bounty to award the new answer and thus draw attention to the question.

If the old answer is accepted, there may be issues there, but a note to the OP of the question that the ruling has changed is not out of line. In rare cases the answerer of the accepted answer will realize it's incorrect and that there is now an established correct answer and want to delete their answer. In this case the OP should flag it and ask the moderators to delete their answer since they cannot do it themselves. Moderators are typically quite responsive to these flags (though in rare cases they may see value in retaining the original answer if it's providing other value outside of the incorrect answer).

\$\endgroup\$
  • 2
    \$\begingroup\$ If you initially upvoted the answer before it became inaccurate, do you encourage making a trivial edit to the answer to allow you to downvote? \$\endgroup\$ – PotatoEngineer Feb 8 '16 at 23:58
  • \$\begingroup\$ Someone commented about the rules change 7 months ago on the old answer \$\endgroup\$ – András Feb 9 '16 at 22:03
  • \$\begingroup\$ @András just edited to suggest adding a bounty, that's another option. \$\endgroup\$ – wax eagle Feb 9 '16 at 22:27
  • \$\begingroup\$ This sounds fine, but there's no mention of a meta question to call attention to it if it's not moving along, which seems like the next step. \$\endgroup\$ – DCShannon Feb 9 '16 at 23:02
  • \$\begingroup\$ @DCShannon that's not a typical route for this. Not every incorrect post needs a meta post about it. Normally we expect voting to take over and resolve this without additional intervention. When that fails, comments and bounties are the next step. \$\endgroup\$ – wax eagle Feb 9 '16 at 23:05
  • \$\begingroup\$ @waxeagle Of course not every incorrect post needs a meta. It would be the next step if it's not moving along. A week is probably not enough time, but calling attention to a question that needs work on meta is a pretty normal thing to do. \$\endgroup\$ – DCShannon Feb 9 '16 at 23:07
11
\$\begingroup\$

Here are the options available to you if you disagree with the top voted and/or accepted answer:

Your options #1 and #2 are acceptable but other answers have covered why #3-6 are not. Instead I will cover the range of options you do have within the SE system to deal with cases like these. In general, they can be used in any combination.

Comment on the top/accepted/incorrect answer

Leaving a comment for the person who wrote the answer that is now outdated gives them a chance to review the facts and determine whether they want to change their answer accordingly or not. If changes can be made to the answer to accommodate the new information they can integrate that info into their answer and the answer should now be (more) correct. If the new information would reverse their answer then they might agree to delete it. In the rare case that an author realizes that their accepted answer is incorrect they can request a mod to delete it by flagging it.

If the author is no longer active, this has a low chance of working.

Upvote better answers and/or downvote incorrect ones

Obviously the standard way to indicate an answer's quality is to vote, one way or the other. Obviously you have only one vote but you can try to attract more...

Draw attention to the post to attract other voters

There are several ways you can attract eyes onto the post:

  • Start a bounty (including a custom message with the bounty pointing people to flaws in the answer may help inform voting)
  • Point people to it in chat (you can even explain your issues with it there and people may look into it and vote according to their whims)
  • Point people to it here on Meta.se (if there is a meta-worthy issue to address)
  • Post your own answer (see below) which bumps the question into the queue of active questions

Post your own answer

If there are no correct answers to upvote, write your own! This also has a couple of non-obvious advantages: it bumps the question in the active queue (and maybe even in the Late Answers queue) which means it might get more eyes, and it notifies the author of the question of the new answer which might get them to look at the answers again as well.

Comment on the question

Especially when the issue is with an accepted answer, leaving a polite comment to the author of the question may let them know that you believe there is new information that may change which answer they want to accept. Of course they are free to not act on it at all if they do not want to for whatever reason.

Be patient

Sometimes the Stack moves slowly, so don't expect that your actions will necessarily have immediate results. But the idea is that, over time, the correct answer will win out, so just do what you can and then leave the rest to the system. An important idea on Stack is that no one person determines what is correct or incorrect, instead depending on the wisdom of crowds. It is just that sometimes crowds take a while to work.

\$\endgroup\$
  • 2
    \$\begingroup\$ If there are no correct answers to upvote, write your own! I think you ought to make that point number 1 in this answer. \$\endgroup\$ – KorvinStarmast Nov 9 '18 at 20:16
5
\$\begingroup\$

@waxeagle's answer is the correct one, but if you're unsatisfied with the results of that course of action, and you really want to mark a particular answer as being the correct one in a visible way, you could use a bounty. Obviously, it's up to you to decide if it is important enough to you to be worth spending your rep on.

Pros:

  • Draws attention to the question by placing it in the Featured tab.
  • Using a customised bounty message, you can explain why you're bountying so that the attention you draw can be directed the way you want it to.
  • Once the bounty is awarded, the answer it was awarded to has a thing on it to draw people's attention to it.

Cons:

  • Works much better while the bounty is active than once it's awarded.
  • May cause additional, unnecessary answers that don't really add anything.
  • Obviously, it costs rep.

Note: There is a conflict of interest here, given the context of this meta, so I'm suggesting this for the general problem rather than your specific example.

\$\endgroup\$
0
\$\begingroup\$

Part of the virtue of the Stack format is that no one person can enforce their view of which answers are correct.

That includes mods. That also includes you (whoever the "you" is that is reading this at the moment.)

Of the methods of dealing with an "incorrect" answer suggested in the answer:

  1. This has already been done.
  2. This has already been done.
  3. This is not permitted because it alters the intent of the author of the post. It also invisibly hijacks the votes on a post to falsely endorse the lone editor's idea of what is correct.
  4. This is not permitted because it alters the intent of the author of the post. It has a similar hijacking problem, though to a lesser degree because it is visible instead of invisible and therefore less likely to deceive readers.
  5. Moderators aren't allowed to make that decision. We even have a specific, hard-coded flag rejection reason that says so:

    flags should not be used to indicate technical inaccuracies, or an altogether wrong answer

    Don't send us flags about incorrect answers. They'll only be rejected, which would only (with a lot of rejections) result in losing flagging privileges entirely.

  6. Not technically feasible (since flag reasons are network-wide, and hardcoded, they cannot be customised per-site), and would be rejected even if it was possible (for the same reasons as #5).

Basically: chill. Fixing this is not your job. You've already done plenty, and it's on the path to voters correcting this, as is proper on a Stack. If you're so impatient for Truth To Win Out that a mere week is insufficient for the normal site mechanisms to work and you feel the need to immediately make it correct according to your own personal standard, you gotta seriously chill out. A year from now the votes will likely have a wider division.

If that's not good enough: sorry, no one user gets to decide which answers are correct and which aren't, so you just have to deal with being unhappy about the votes.

\$\endgroup\$
  • 2
    \$\begingroup\$ This sounds fine, but I don't see an issue with #4. The original author can always roll back the edit if it offends them. \$\endgroup\$ – DCShannon Feb 9 '16 at 23:04
  • 4
    \$\begingroup\$ @DCShannon Even authors are discouraged from complete reversing their answers, because it breaks voting. Non-authors are even more strongly unwelcome to do so. \$\endgroup\$ – SevenSidedDie Feb 9 '16 at 23:07
  • 4
    \$\begingroup\$ #4 isn't rewriting the answer, it's adding a small note. #3 was rewriting the answer. \$\endgroup\$ – DCShannon Feb 9 '16 at 23:08
  • 2
    \$\begingroup\$ @DCShannon A note at the top is as much an attempt to reverse what the answer is stating as #3. \$\endgroup\$ – SevenSidedDie Feb 9 '16 at 23:08
  • 5
    \$\begingroup\$ Alright, that's gonna be one of the rare cases where I disagree with your judgement. \$\endgroup\$ – DCShannon Feb 9 '16 at 23:09
  • \$\begingroup\$ @DCShannon I don't see why it would be controversial; top and bottom of the post affirms the fundamental design principle of Stack Exchange that no-one gets to decide alone what is true. Not editing other users' answers to reverse them follows directly from that. \$\endgroup\$ – SevenSidedDie Feb 9 '16 at 23:11
  • 4
    \$\begingroup\$ I don't agree that it's reversing or destroying the answer. All the content and reason is still there, and it's still correct for the version for which it was written. It's just a note about a rules change, which is an objective fact. If the author wants to rewrite the note, they can do so. If they want it gone completely, they can do that. \$\endgroup\$ – DCShannon Feb 9 '16 at 23:13
  • \$\begingroup\$ @DCShannon A note about a rule change that the note-writer asserts makes the answer incorrect. That's not for an editor to decide ever: that's the voters' job. \$\endgroup\$ – SevenSidedDie Feb 9 '16 at 23:18
  • 1
    \$\begingroup\$ You are advocating democracy, which is nice, but democracy is famously bad at handling facts: geekologie.com/2015/12/… \$\endgroup\$ – András Feb 10 '16 at 8:27
  • \$\begingroup\$ @András I'm not advocating democracy, I'm explaining the basic functioning of the site. You can argue with that until blue in the face but it will not change. \$\endgroup\$ – SevenSidedDie Feb 10 '16 at 15:30
  • \$\begingroup\$ As of this date the vote split is 14 to 5 in favour of the newer answer. It took less than a year. \$\endgroup\$ – SevenSidedDie Oct 21 '16 at 23:04
  • 1
    \$\begingroup\$ @SevenSidedDie you make it sound like "less than a year" were a good enough speed \$\endgroup\$ – András Nov 9 '18 at 20:02
  • \$\begingroup\$ @András If you can see an alternative that doesn’t violate a core principle of how the site works, I’m all ears. \$\endgroup\$ – SevenSidedDie Nov 9 '18 at 21:01
  • \$\begingroup\$ @SevenSidedDie, with the current rules, it took nearly a year to correct something obvious. So maybe the rules are wrong. \$\endgroup\$ – András Nov 9 '18 at 22:57
  • \$\begingroup\$ @András The only rule involved is “everyone gets one vote per post”, so no single person gets to decide what’s right and wrong. That’s not a rule we can change. Even mods are not allowed to decide what answers are right and wrong. \$\endgroup\$ – SevenSidedDie Nov 9 '18 at 23:01
0
\$\begingroup\$

Comment on the old obsolete answer and the question. If there is no reaction to the comments, then edit a note on the obsolete answer, as per suggestion 4 in the original post.

This gives the original asker and answerer an invitation and opportunity to consider the change (to edit their answer, to select another answer as the best one, to change their votes). If there is no action, then editing the obsolete note into the answer unlocks votes and is a usual case of editing useful content from comments to answers and questions.

The errata note might look something like:

After this answer was originally posted there has been errata / rules changes [link]


[the previous answer here]


Justification of adding in the errata

From the help center, "When should I edit posts?":

  1. To fix grammar and spelling mistakes
  2. To clarify the meaning of the post (without changing that meaning)
  3. To include additional information only found in comments, so all of the information relevant to the post is contained in one place
  4. To correct minor mistakes or add updates as the post ages
  5. To add related resources or hyperlinks

Adding a mention of new errata fits the criteria 3-5. It somewhat conflicts the parenthetical clarification of criterion 2, but even this is arguable.

Some other quotes from the same web page, out of order, out of context:

All contributions are licensed under Creative Commons, and this site is collaboratively edited, like Wikipedia. If you see something that needs improvement, click edit!

Editing is important for keeping questions and answers clear, relevant, and up-to-date. If you are not comfortable with the idea of your contributions being collaboratively edited by other trusted users, this may not be the site for you.

Any time you see a post that needs improvement and are inclined to suggest an edit, you are welcome to do so.

\$\endgroup\$
  • 2
    \$\begingroup\$ Actually, on second thought, I will downvote, because this is also altering the intent of the post's author. \$\endgroup\$ – Miniman Feb 14 '16 at 11:37
  • 1
    \$\begingroup\$ @Miniman I added in some justifications for editing in the errata, and removed the "EDIT"-word from the suggestion. \$\endgroup\$ – Thanuir Feb 14 '16 at 12:00
  • 2
    \$\begingroup\$ collaborative editing is one thing. completely changing the intent on an answer is another \$\endgroup\$ – Wibbs Feb 14 '16 at 14:52
  • 2
    \$\begingroup\$ @Wibbs I see a vast difference between completely changing the intent of the post and adding a notice that the post is no longer up to date. An example of the former would be editing and rewriting the post according to the new information. \$\endgroup\$ – Thanuir Feb 14 '16 at 15:48
-1
\$\begingroup\$

I think the solution would be not make things more complex. Maybe just to give more relevance to how old questions and answers are on the site would help people to realize that the once right answer seven years ago could not be correct today.

To add functionality to solve things is not efficient. Rules can change... not only forward but backward. Some changes are reversed or again superseeded by future rules. Then, what you fix now, could be again wrong in a year or more. No one is going to track that madness.

Dates are small things on these sites, we can show them a little bit more and maybe add something that says "5 months old" or "posted three years ago".

\$\endgroup\$
  • 2
    \$\begingroup\$ We don't have the power to change the software for just this site. If this is a suggestion you feel strongly about, you might want to pair it with a [feature-request] on Meta Stack Exchange and link to that request in this post here. \$\endgroup\$ – SevenSidedDie Feb 15 '16 at 18:43

You must log in to answer this question.

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