239

Today we changed the way we sort answers on Stack Overflow. We no longer pin the accepted answer (with the green checkmark) to the top of the list of answers. By default, we now sort strictly by votes (descending order by highest score), and the accepted answer's order in the list is based on its score.

The history of feature requests to unpin the accepted answer on Stack Overflow dates back to 2013 and has been raised almost every year since. This year within the Outdated Answers project we did research that showed that we are good to proceed with the accepted answer unpinned.

Would you like to have the accepted answer unpinned on your site?

We can change the way the engine sorts answers in site settings. We would like to hear from you all if it is something you want to see on your site. (Please let me acknowledge in advance that we will not be able to run a test on each site.)

Currently we are planning to move forward with one of two scenarios, based on your feedback:

  1. Unpin the accepted answer on all SE sites by default and pin it back on a few sites that ask us to do so.
  2. Keep the accepted answer pinned on all SE sites by default and unpin it on a few sites that ask us to do so.

Please let us know what you think will work best for your site! If you can discuss this question with your community it would be awesome. We are going to collect feedback before the end of September 19th.

Site settings are per-site and child metas are also considered a "site".


Update

As Anita Taylor shared:

The Public Platform team will go through these requests and determine what to prioritize, defer and decline. We ask for your patience -- key members of the team are on vacation over the next few weeks, so we won't be able to triage these requests until mid-to-late October.


Update 2

Thanks to everyone for the feedback. Seeing that some sites do not want to unpin the accepted answer, we decided to move forward with the status quo and not to change the default behaviour on existing sites. If you think unpinning the accepted answer on your site makes sense, please do the following:

  1. Start a meta discussion on your per-site meta.
  2. When the discussion is completed and the community agrees to change the default behaviour (agrees to unpin the accepted answer), please ask one of the site moderators to add the tag to the question. It will generate a ticket and let CMs know that we need to adjust the site setting for your site.
  3. We would appreciate it if you add the outcome of the discussion to the accepted answer in this question. This will allow all of us to see the full picture.

There's no need to do this on the sites that are currently on the list; we have already adjusted the site setting for them.

44
  • 47
    My upvote on this post reflects that I think it's great to solicit per-site feedback instead of copying SO changes to the network, not that I necessarily support the change made on SO. Sep 8 at 16:47
  • 56
    This is just brilliant! Not only because I very much want to not have the accepted answer first (they're too often not the best answer), but also because of the way you are approaching this, asking for feedback before pushing network-wide. Thank you!
    – terdon
    Sep 8 at 17:07
  • 20
    I think it would be a mistake to make this different per site. I would much rather unpin it every where and have a link near the answer sort selection to navigate directly to the accepted answer. If there is an accepted answer, there's a nice visible green button or something to get there. That makes it easier to document accepted answers as well.
    – ColleenV
    Sep 8 at 17:12
  • 4
    What is the process to request this for a specific site? I'll have to check with the other mods, but this is something we specifically requested on Skeptics years ago as the concept of accepting an answer doesn't really make sense for us, and the pinning is harmful. Sep 8 at 17:58
  • 46
    @ColleenV: Speaking personally, not as a CM (I'm certainly not speaking on behalf of what the company will do): I like the idea of unpinning the accepted answer, but leaving some easy way to navigate from the question to the accepted answer (e.g. something like "This user has accepted an answer. [Click here to jump to the accepted answer.](link)."
    – V2Blast StaffMod
    Sep 8 at 17:58
  • 6
    @MadScientist: Given that it hasn't yet been decided how this will be rolled out network-wide, I think that right now, the idea is to help decide whether the accepted answer should be unpinned or remain pinned on the rest of the network besides SO - and then, once that's settled, a process will be established for having the relevant setting be enabled or disabled on specific sites.
    – V2Blast StaffMod
    Sep 8 at 18:01
  • 5
    It looks like Google lists the accepted answer first instead of highest scored Is that something SE controls? (For that particular question, that's the desirable outcome, but it may not always be)
    – ColleenV
    Sep 8 at 18:35
  • 7
    @Mari-LouA No? Votes still sorts by highest votes. Active is most recently active, meaning it's been added or edited recently, Oldest is sorted by date of post, oldest first. Neither of the other views is by score, only Votes. Making Active into Newest would just be a reversal of the Oldest sort, which, while potentially useful, is not the same as Active sort.
    – Catija StaffMod
    Sep 8 at 19:44
  • 7
    Whatever you do, please don't roll this out to all sites before we have time to discuss it among ourselves :) A week at least!
    – bobble
    Sep 8 at 20:13
  • 13
    Has this proposal been communicated to the moderators across the network?
    – Mari-Lou A
    Sep 9 at 18:46
  • 5
    Can you clarify whether this change will affect local child metas as well, and whether the setting can be different on the main site and its meta (i.e. pin on meta but not on main or vice versa)? Sep 10 at 14:14
  • 5
    @CGCampbell I think the usefulness for the asker is not relevant anymore after the accept, (his/her/or else) problem is solved, what we need to focus then is the usefulness for new visitors (so, it's best to have a "colectively approved" solution at the top, sorted via vote instead).
    – Bacco
    Sep 14 at 18:01
  • 6
    @CGCampell " we no longer care about the usefulness to the asker (?)" I think that nails it. Indeed we don't. The asker can of course kind of say that no existing answer was helpful to him, but what does it matter to anyone else? It may be a hint that people misunderstood the question or that it's still unsolved but it also may just mean that the asker moved on. In the end, the signal is not reliable. So better not put much weight on it regarding the sorting order
    – Trilarion
    Sep 14 at 22:15
  • 4
    I wonder if the simple "accepted post is pinned second, if it's not the top-scoring answer" works (?). Sep 16 at 3:33
  • 11
    Is it worth revisiting the decision to not allow us to delete accepted answers now? I have an answer that at the time was "there is no feature that supports what you are doing.", but there is now such a feature. It was accepted at the time, but there is no way for me to delete it now and the asker is unresponsive to moving the answer flag. Seems like the original limitation to prevent the accepted answer from being deleted might no longer be useful.
    – Matt
    Sep 20 at 0:34

39 Answers 39

63

Full List of Community Discussions

✅ indicates that this has been implemented on the specific site.

Site Verdict Site Discussion Link
Academia 50/50 Should the answer that appears on "top" be the OP's "accepted answer" or the answer with most votes?
Anime Does the community want the accepted answers to be pinned to the top?
Apple Do we want to unpin the accepted answer from the top?
Arqade ✅ Keep Pin https://gaming.meta.stackexchange.com/q/16087/163757
Ask Ubuntu ✅ Unpin Should we unpin the accepted answer from the top of the list?
Ask Ubuntu (Meta) ✅ Keep Pin Should we unpin the accepted answer from the top of the list on meta?
Astronomy ✅ Unpin https://astronomy.meta.stackexchange.com/questions/803
Aviation Unpin https://aviation.meta.stackexchange.com/questions/4253
Bicycles https://bicycles.meta.stackexchange.com/questions/1433/
Biology Unpin https://biology.meta.stackexchange.com/questions/4230
Board Games Unpin https://boardgames.meta.stackexchange.com/questions/2048
Chemistry ✅ Unpin https://chemistry.meta.stackexchange.com/questions/5077
Christianity Unpin Is Christianity.SE a OP's Gets it or Community Gets it community?
Code Golf ✅ Unpin https://codegolf.meta.stackexchange.com/q/23854/
Code Review ✅ Keep Pin https://codereview.meta.stackexchange.com/questions/10719
Coffee https://coffee.meta.stackexchange.com/questions/1253
Computational Science https://scicomp.meta.stackexchange.com/questions/600/
Cross Validated https://stats.meta.stackexchange.com/questions/6220/
Drupal Answers https://drupal.meta.stackexchange.com/questions/3869/
Earth Science ✅ Unpin https://earthscience.meta.stackexchange.com/q/1932/
Economics https://economics.meta.stackexchange.com/questions/2151
English Language Learners https://ell.meta.stackexchange.com/questions/5514
English Language & Usage Keep Pin https://english.meta.stackexchange.com/q/15144
French https://french.meta.stackexchange.com/questions/833
Gardening https://gardening.meta.stackexchange.com/questions/943/
German Language https://german.meta.stackexchange.com/q/1653
Genealogy & Family History https://genealogy.meta.stackexchange.com/q/3522/19
GIS ✅ Unpin https://gis.meta.stackexchange.com/questions/5303
Graphic Design https://graphicdesign.meta.stackexchange.com/q/3779/41100
History https://history.meta.stackexchange.com/q/4311/1401
Interpersonal Skills ✅ Keep pin https://interpersonal.meta.stackexchange.com/questions/4038
Judaism ✅ Unpin https://judaism.meta.stackexchange.com/q/4708
Language Learning https://languagelearning.meta.stackexchange.com/questions/720
Literature https://literature.meta.stackexchange.com/questions/1602
Math ✅ Unpin https://math.meta.stackexchange.com/questions/34076/
Matter Modeling https://mattermodeling.meta.stackexchange.com/questions/294/
Meta Stack Exchange Do we want accepted answers unpinned on Meta Stack Exchange?
Movies Unpin https://movies.meta.stackexchange.com/questions/4780
Music https://music.meta.stackexchange.com/questions/3781
Parenting Keep Pin https://parenting.meta.stackexchange.com/questions/1455/
Physical Fitness Unpin https://fitness.meta.stackexchange.com/q/865
Physics ✅ Unpin https://physics.meta.stackexchange.com/questions/13811
Politics ✅ Unpin Should accepted answers still be pinned to the top of the list of answers?
Politics Meta Should accepted answers be pinned to the top on meta?
Project Management https://pm.meta.stackexchange.com/q/896
Puzzling ✅ Keep Pin https://puzzling.meta.stackexchange.com/questions/7230/
Quantum Computing https://quantumcomputing.meta.stackexchange.com/questions/512/
RPG 50/50 https://rpg.meta.stackexchange.com/questions/11726/
Sci-fi and Fantasy Unpin https://scifi.meta.stackexchange.com/questions/13478/
Seasoned Advice https://cooking.meta.stackexchange.com/questions/3716/
Skeptics https://skeptics.meta.stackexchange.com/questions/4786
Spanish Language ✅ Keep Pin https://spanish.meta.stackexchange.com/questions/4435
Sports ✅ Unpin https://sports.meta.stackexchange.com/questions/1027/
Stack Overflow (Portuguese) ✅ Unpin https://pt.meta.stackoverflow.com/questions/8662/
Superuser ✅ Keep Pin https://meta.superuser.com/questions/14531
TeX https://tex.meta.stackexchange.com/q/8841
UNIX & Linux ✅ Unpin https://unix.meta.stackexchange.com/questions/5835/
Veganism & Vegetarianism https://vegetarianism.meta.stackexchange.com/questions/366
Webmasters ✅ Unpin https://webmasters.meta.stackexchange.com/questions/1728
Workplace Keep Pin Any thoughts on unpinning the accepted answer
Worldbuilding 50/50 https://worldbuilding.meta.stackexchange.com/questions/8441

Sites that want to remove the pin

  • Ask Ubuntu
  • Astronomy
  • Board Games
  • Chemistry
  • Code Golf
  • Earth Science
  • GIS
  • Judaism
  • Math
  • Movies
  • Physics
  • Politics
  • Science Fiction and Fantasy
  • Sports
  • Stack Overflow em Português
  • UNIX & Linux
  • Webmasters

Sites that want to keep the pin

  • Arqade
  • Ask Ubuntu Meta
  • Code Review
  • English Language & Usage
  • Interpersonal Skills
  • Parenting
  • Puzzling
  • Super User
  • The Workplace
33
  • 9
    @JeffSchaller Thank you very much for setting the table. It is perfect! We will need any way to either keep it pinned or unpin it for all sites and process the rest manually. Having a table is a great way to do so!
    – Nicolas Chabanovsky StaffMod
    Sep 16 at 14:52
  • 2
    At what point are decisions edited into this as "final"? Do we wait for a member of each site to do so, or do we say that after X days, with a margin of Y, it's decided? Sep 16 at 18:59
  • 63
    It's interesting that pinning is being used in this meta question to place a relevant meta answer at the top of the thread, showing the value of this behaviour for certain types of questions.
    – illustro
    Sep 16 at 22:05
  • 4
    Sep 19th sounded like the deadline to me, but these sound like settings in software, so... 6-8? Sep 16 at 22:29
  • 3
    The pin-to-top behaviour can be useful, but assigning that behaviour to all questions and decided by one arbitrary user is not. Bringing it back for posts e.g. tagged feature-request on meta sites only would be a nice move, but ultimately not relevant here.
    – Nij
    Sep 19 at 5:00
  • 2
    @DavidRoberts If the community wants the change it would be nice to know about it and have the site on the list.
    – Nicolas Chabanovsky StaffMod
    Sep 20 at 6:42
  • 3
    @NicolasChabanovsky as a moderator of Spanish.SE I would like to ask that both the change to "all sites" and the "per-site exceptions" are applied close to each other in time. If "unpin" wins and the sites that requested to keep the pin see that they get "unpin" anyway and have to wait 6-8 months to get back the old behavior that they never asked to be removed, they are going to be quite annoyed. This is important in our community because we have already lost more than one long-term, high-rep user due to the >2-year wait for 3-vote close, which we still haven't got.
    – wimi
    Sep 24 at 9:51
  • 4
    Maybe meta.stackexchange.com/a/108475/307535 explains the link text conversion? Sep 25 at 9:56
  • 5
    @Luuklag Yes, correct. I am going through the list and updating site settings now.
    – Nicolas Chabanovsky StaffMod
    Sep 29 at 11:18
  • 8
    @user1271772 Please remember that votes on these meta posts aren't locked. The votes on Physics were far more in favor when the change was made. You're assuming that the scores have always been what they are now and finding fault. It wasn't this close when Nicolas changed the site setting. Also, we probably should only consider upvotes, not both up and down. Editing out the current status is not OK.
    – Catija StaffMod
    Oct 4 at 13:17
  • 2
    @aepot Please feel free to start the discussion yourself - the CMs will not be starting discussions about whether to pin or unpin accepted answers on each site - the site must have the discussion themselves.
    – Catija StaffMod
    Oct 4 at 13:19
  • 6
    @user1271772 I think this situation has only two possible outcomes: We either pin or unpin, there is nothing in between. Moving forward with the most upvoted decision, even if the delta is little, seems better than doing it in the opposite direction.
    – Nicolas Chabanovsky StaffMod
    Oct 4 at 14:41
  • 3
    @NicolasChabanovsky Sites with tens of thousands of users are getting maybe 30 people voting on the change. I think the more conservative approach would be leaving things the way the vast majority of users are used to until there is a significant consensus to change them.
    – ColleenV
    Oct 7 at 19:01
  • 2
    @ColleenV But that's what they're doing now, by not changing the default behaviour unless a site specifically asks. What constitutes a consensus to change has been left to individual sites to decide, and SE just executes the decision. Any dispute regarding the decision taken by a site can be brought up on that site itself.
    – GoodDeeds
    Oct 7 at 23:18
  • 4
    @GoodDeeds A slim margin of a fraction of the users doesn’t warrant changing anything. That’s my only point. If there is a significant consensus like there has been for a number of sites, then the community has spoken and it should be changed or not according to the consensus.
    – ColleenV
    Oct 8 at 2:03
137

If the accepted post has score equal to another post please show the accepted post upper.

Suppose there are 3 answers for a post say A, B and C. Let A score be 1, B and C score be 0. Let C be the accepted answer.

enter image description here

Like you can see then if votes order is choosen then the post will be in order A>B>C. But I guess it should be made A>C>B. If 2 or most post have same score out of which one answer has been accepted then that accepted answer should be upper than the remaining answer because since OP accepted it, the accepted answer might be more helpful than the other answers at same score.

21
  • 18
    I prefer the pseudo random order, since OP is the least qualified to know which answer should be preferred, as the SO experiment showed.
    – Braiam
    Sep 8 at 16:56
  • 43
    @Braiam The experiment was not designed to test that scenario. Sep 8 at 18:32
  • 1
    @Braiam which experiment? Sep 9 at 1:46
  • 5
    @JitendraSingh meta.stackoverflow.com/q/411192/792066
    – Braiam
    Sep 9 at 3:27
  • 10
    The whole premise of this change is basically that "accepted doesn't add more value from votes" and that answers should be sorted strictly by votes. Why would that change if there are same-scored answers? Why would the accepted play a role there? It should be part of the regular random algorithm that might or might not show it first between the same-scored answers...
    – Tomerikoo
    Sep 9 at 12:45
  • 58
    @Tomerikoo Votes more accurately reflect the best answer because more people agree with it, but if the OP accepts it, that means they at least somewhat think that the answer is helpful. As the experiment shows, votes are better, but accepted should at least be a tiebreaker...
    – Anonymous
    Sep 10 at 23:47
  • 32
    The ‘‘whole premise of this change’’ is that the checkmark shouldn’t be a green vote-hammer that counts as an infinite number of votes and even disrupts the other sort orders (active and oldest).  Jitendra Singh’s request is that the checkmark should count as half a vote for sorting purposes, which seems more than reasonable, considering that it counts as 1½ votes for rep purposes.
    – Scott
    Sep 13 at 19:19
  • 17
    Acceptance as a tie-breaker seems very reasonable. After all, we have a piece of data that says this answer was useful to someone, so we may as well use it. It's not unknown to have several answers with the same low score of 0 or 1, so letting one that was some use float to the top seems right
    – Chris H
    Sep 14 at 12:01
  • 2
    @Braiam when op asks for solution, he is one who checks if it works, so for some sites for OP problem, they can be a judge. For a place like worldbuilding it is exactly as you say it - they usually do not know and they can't test and accepting is just additional cookie given to one who answered it, personally crafted by OP.
    – MolbOrg
    Sep 14 at 13:31
  • @MolbOrg you have way too much faith in OP. OP have accepted answers that don't work, and didn't work. They can accept an answer for any reason, including no reason. Acceptance is just a way that sites can bootstrap themselves, since it isn't a privilege. Nothing more.
    – Braiam
    Sep 14 at 14:06
  • 1
    But mostly good experienced users generally accepted the answer most helpful to them. In your question did you randomly accepted any answer @Braiam? Sep 14 at 14:11
  • 5
    @Braiam “They can accept an answer for any reason, including no reason.” And what’s the difference to upvotes? I have seen answer containing a non-working solution with tons of upvotes. Under normal circumstances, the OP should be the one who is motivated to actually try the posted solution while it happens very often that the other people, not having the problem, upvote an answer that just looks convincing. Granted, I’ve seen non-working answers with the checkmark too, don’t know what’s going on with those questioners, but still, SO considers question to be solved in that case.
    – Holger
    Sep 16 at 8:09
  • 6
    @Braiam Acceptance seems comparable to an upvote (especially when the asker doesn't have enough rep to upvote). Putting the accepted answer first among equally voted answers is not saying the asker knows better than anyone else, it's closer to saying acceptance is a positive signal (however big or small) and an answer with 1 additional positive signal is more likely to be better than an equivalent answer without that signal.
    – NotThatGuy
    Sep 16 at 15:37
  • 5
    @Braiam I'm not saying they're equivalent. Even if you consider acceptance to only count as 0.0001% of a vote, that would still rank an accepted answer higher than an equally voted one. The problematic cases I've seen to justify this feature is where multiple answers have 0 votes each (this was, in my possibly-more-controversial opinion, a problem even for self-answers, which is the only time it could've happened before). In that case acceptance is the only positive signal we have, so it seems very counter-intuitive to ignore that completely for ranking. [1/2]
    – NotThatGuy
    Sep 16 at 22:08
  • 5
    For 1 or 2 votes, acceptance is still a fairly strong signal. For many votes, it's unlikely for answers to have equal votes, and when they do, how we sort them seems negligible. [2/2]
    – NotThatGuy
    Sep 16 at 22:08
67

This is a modified copy of my answer to the related question on Meta Stack Overflow.


Unpinning the accepted answer is a good change for the majority of SE sites I imagine.

However, I would still very much like a clear indication (near the top of the question) that an answer has been accepted and offer a link to jump to it. As it is now, when viewing the question, there is no indication that an answer has been accepted unless you scroll through the answers (if there are other answers that have more votes).

(Something similar has already been suggested in comments by @ColleenV and @V2Blast)

Perhaps a link next to the sorting buttons, something like:

Suggested link to jump to the accepted answer

3
  • 5
    IMHO Good Idea! Very useful! Sep 15 at 19:24
  • 13
    Especially if we include the date: "Go to the answer that was accepted on DATE"
    – Reg Edit
    Sep 16 at 22:06
  • 5
    I think this could potentially be especially helpful to the asker, who may want to find the answer they accepted to review it or compare it to other answers or to get some clear signal that they actually have accepted an answer (in case they forgot whether they did).
    – NotThatGuy
    Sep 16 at 22:24
50

Stack Overflow is probably the site where the concept of accepted answers makes the most sense, and even there I think pinning the accepted answer is overall harmful in the end. For non-technical sites the accepted answer concept is often less meaningful simply because it is harder for the asker to evaluate answers, especially when the questions aren't about fixing a problem. On SO many questions can be objectively evaluated even if you have less knowledge than the answerers, e.g. now the code compiles or now I get the correct result. For more "fuzzy" subjects the asker might just accept the answer that best fits their preconceived notions, or simply the one that sounds and looks the most impressive.

For this reason I would argue that the default should be not to pin the accepted answer to the top. And any site that likes to do that different can request changing this back. If you are convinced that this is the right choice for Stack Overflow, unpinning by default is the right choice for the network as well. This is not one of the cases where SO is the outlier that needs special handling.

5
  • 1
    And any site that likes to do that different can request changing this back. Extremely unlikely that such requests would actually end up actioned in a timely manner, just saying. But I definitely see your point for SO. Sep 8 at 19:41
  • 24
    @SonictheAnonymousHedgehog Why do you think that? This should be a site setting that the CMs have access to and can change in a minute. Provided the request was status-reviewed on the child meta, I don't think this would take more than a day or two to change.
    – Catija StaffMod
    Sep 8 at 19:46
  • 1
    It makes sense, but how will you convince the don't move my cheese folks? Sep 8 at 22:19
  • 4
    @Catija: We don't have access to your internal discussions. We don't know which things are "a CM can change that in five minutes" and which things are "six to eight weeks."
    – Kevin
    Sep 9 at 5:41
  • 14
    @Kevin sure... but that doesn't excuse such an extreme statement as what Sonic said. Particularly since the question here specifically indicates that per-site configuration is possible.
    – Catija StaffMod
    Sep 9 at 5:44
43

I agree with the change, thanks for doing this.

For users who want to keep the old behaviour, I've written a small userscript, Accepted Answer Pinner (direct installation linkStack Apps post). It pins the accepted answer back on top, unless it's a self-answer that wouldn't qualify for pinning in the old situation.

4
  • 3
    As you are a moderator on ELL, could you please post something on meta asking for users' feedback?
    – Mari-Lou A
    Sep 9 at 12:12
  • 2
    @Mari-LouA great suggestion, will do once I can use my laptop again (somewhere tomorrow).
    – Glorfindel Mod
    Sep 9 at 18:05
  • 1
    Thank you! It's a pity that meta is so quiet on ELL but maybe, and hopefully, this announcement will get users excited and if it doesn't at least we can direct confused regulars to a specific post. Thanks :))))
    – Mari-Lou A
    Sep 9 at 18:08
  • +1 You should make a script that allows one to follow users because I have read through some of the posts and value your input. Basically, you make the StackExchange a well-reasoned space. Sep 14 at 23:32
40

Not every moderator or high-rep user participates on Stack Exchange Meta, and not every hi-rep or engaged user participates on their child site.

Nevertheless, I think the option to unstick the accepted answer from its pole position should be discussed by each individual meta site. Please feature this post and communicate this proposal to all the moderators across the network. I think it's possible that many moderators, especially on smaller sites, might be oblivious to this announcement.

8
  • 6
    Or at least make this post [featured]. I think that any post that says "We would like to hear from you all if it is something you want to see on your site" should most definitely be featured. But I'm not sure why the staff is a bit conservative with the [featured] tag. I've requested that similar posts be featured before and they were not.
    – 41686d6564
    Sep 8 at 19:34
  • 2
    @41686d6564 I think there's a good chance of it being featured as long as the announcement is not downvoted.
    – Mari-Lou A
    Sep 8 at 19:38
  • 2
    Some of the meta sites are pretty much dead and little to no discussion will happen regardless of how much they want it to.
    – Joe W
    Sep 9 at 0:27
  • 2
    @JoeW That maybe true but how else will you inform users from smaller communities who never ever visit Meta? At least communities will have been notified on their meta site.
    – Mari-Lou A
    Sep 9 at 3:54
  • @Mari-LouA If the meta site gets almost no traffic does asking about it there do anything? The problem is you have meta sites where question getting dozens of views is typical as almost no one out side of a very small group ever visits it.
    – Joe W
    Sep 9 at 12:03
  • 6
    If users choose to ignore their site's meta, and notices on the bulletboard i.e. featured on meta and hot meta posts, then they have only themselves to blame if changes to the system occur without their knowing or seal of approval.
    – Mari-Lou A
    Sep 9 at 12:11
  • 2
    How exactly does putting a question on a child meta that gets no traffic change anything from putting asking the question here? My point was that there are sites where users don't care about meta at all and trying to ask this question on every single meta would be a waste of time.
    – Joe W
    Sep 9 at 14:01
  • 2
    This isn't about people not seeing it because they don't use meta but the fact that if meta isn't used the post won't get any feedback (or enough feedback to be valuable) and provide no value or decision.
    – Joe W
    Sep 9 at 14:11
26

Keep the accepted answer pinned on meta sites

Oftentimes, on meta sites, the behavior of pinning the accepted answer to the top is preferred:

  • Answers can be different opinions in a discussion, and accepting the one that is eventually the community's consensus or is the one that eventually gets implemented will make that fact clear to others reading the question, especially if that one isn't the one with the highest number of votes.

  • A question may receive an official response from the team, which may not be the highest-voted answer, usually simply because other community answers were posted first, but in some cases may be due to the answer being negatively-received. However, it's still preferred to have it pinned to the top so that others can more easily see the official response.

Please don't unpin the accepted answer on meta sites, at least.

If there are concerns that the wrong answer (not representative of community consensus) would be accepted, then another, separate feature, to allow moderators and staff to change the accepted answer on meta sites, should be implemented. Personally, though, in my experience, that happens a very tiny minority of the time, and in the rare cases where it's a problem, a developer can change the accepted answer so it remains pinned even if it's later voted out.

19
  • 27
    I see what you're getting at here but I don't consistently feel like this is actually true. The thing is, with the accept still being the choice of the asker, you'll have inconsistent behavior. Some will accept the staff answer or the consensus answer while others will choose what they liked best at the time and might actually be downvoted or disliked - other times you'll accept the consensus at the time, which later changes and the asker is now MIA. For metas, honestly, I feel like we would benefit from a different system entirely but I don't think that opting to pin or not is that change.
    – Catija StaffMod
    Sep 8 at 16:45
  • 10
    What about here on MSE? I can think of quite a few accepted answers that have been staff-posted, accepted, and then downvoted into double/triple digits.
    – Ollie
    Sep 8 at 16:46
  • 3
    @Catija I've seen that happen, but in the questions I've read, that's a tiny minority. The vast majority of the time, users accept the staff answer or the consensus answer. It's an even smaller minority on the site metas. Sep 8 at 16:48
  • 1
    @Ollie I still think that such answers should be pinned. Regardless of score, they're still official responses nonetheless. It also makes it easier for later answers to argue against the reasons in the official response. Also, again, that's a tiny minority. Sep 8 at 16:49
  • 20
    My view is the exact opposite: don't pin accepted answers to the top on meta sites. It's really unfair to give one user so much say over which answer appears first, when it's an issue of community policy. I've seen cases where a meta OP accepts a low-voted answer and we have to remind people that the community consensus is the second answer in the list. See this SFF meta, referencing acceptance vs consensus and this on MSE. Sep 8 at 16:51
  • 1
    @Randal'Thor Staff have the ability to change the accepted answer; it's done on the recent feature changes post here so the current answer is pinned even though the asker is deleted. If you're running into one such case, just ask staff to do so there, so that the one that does represent consensus is pinned regardless of future voting. Sep 8 at 16:56
  • 2
    @Sonic I don't think it is (or should be) worth bothering staff for every time a trolly OP accepts the wrong answer on meta, or consensus changes after the OP disappears, or an inexperienced OP doesn't understand how acceptance works, or all the other possible scenarios. In fact I don't approve of staff meddling with acceptance at all, and I imagine they too only want to do so in extremely rare and special situations. Sep 8 at 17:01
  • 10
    @SonictheAnonymousHedgehog Yeah... If we have the ability to do that, I don't know about it and I don't think I actually can do it. I have a huge aversion to ever changing the accepted answer because that's an inappropriate choice in most cases.
    – Catija StaffMod
    Sep 8 at 17:01
  • 11
    @ThomasMarkov that's actually something we already control for since accepted self answers don't pin.
    – Catija StaffMod
    Sep 8 at 18:16
  • 2
    @Catija Well how bout that. Learn something new every day. Sep 8 at 18:17
  • 1
    Here's an example on Meta where "the" answer has been recently accepted (and thus pinned). It will never receive the most votes, both because the answer/resolution is unpopular and because it is years/decades newer than the other answers. But it is the "status completed" Answer and should probably be at the top...
    – pkamb
    Sep 8 at 19:27
  • 2
    This is an example on MSE where a staff response should be pinned at the top (and even involved some ...unpopular shenanigans to make sure that it was), but won't be once this goes live Sep 8 at 21:19
  • 1
    You probably want to separate your proposals, this will lead to murky voting.
    – Akixkisu
    Sep 9 at 13:36
  • 3
    Perhaps a "pin answer to top" power for diamond moderators on their meta would be a useful feature. Then we do not have arbitrary users steering the discussion, while still having the option to record the "accepted" proposal out of a list of proposals to some policy change. Sep 14 at 7:24
  • 1
    This may be more a case for a different feature "staff can pin specific answers in meta" to force official policy responses to be at the top, rather than going by accepted - when talking about meta you get a lot of opinion involved and there's a much less clear definition for what's "best" in some cases, you either get the consensus at the time accepted (which can change) or the answer that the asker is biased towards (regardless of group consensus or staff response)
    – Andrew
    Sep 17 at 15:31
22

Please let us know what you think will work best for your site! If you can discuss this question with your community it would be awesome.

We've already discussed this on Mi Yodeya. As of this writing, the question there is +12/−0 and the only answer is +10/−0; both ask for the accepted answer to be unpinned. So Mi Yodeya wants the accepted answer unpinned. Thanks!

2
  • Ask a moderator to set the status-review tag on the site meta question.
    – Braiam
    Sep 10 at 21:51
  • 1
    @Braiam thanks. Probably not necessary, since it's already under review, as noted in the question above.
    – msh210
    Sep 11 at 17:50
19

I'll throw in what I suggested on Politics SE: A hybrid model. Acceptance indicates that "this answer is what the asker was looking for." On Politics, that can range from "I feel informed" to "This confirms my political beliefs so I like it."

Even so, I think acceptance still has value, but shouldn't be the end-all of the matter.

Adding a site-configurable weight to the acceptance is probably a good compromise feature. Accepted answers can enjoy top billing until an answer outscores them by a significant margin (e.g. 125%+). For low-scoring answers this could get distorted but I think that, overall, it lets an asker indicate "this had particular value for me" but allows majorities of presumably knowledgeable people indicate that such an answer may not actually be superior.

0
18

We've had some discussion over on Puzzling, where there is a definite consensus on keeping accepted answers at the top (i.e., not unpinning).

One person looked at some concrete examples, which suggest to me that actually "keep the accepted answer at the top, unless its score is actually negative in which case don't" might be a better option (for us) than either "unpin completely" or "leave exactly as is".

2
  • 1
    Basically asking for this, then?
    – bobble
    Sep 12 at 23:23
  • Yup! It does seem like a pretty reasonable option; "has negative score" seems like a more reliable indicator of outright wrongness than "is outscored by another answer". (But this is a thing that will be different on different SE sites; e.g., on SO, it's possible for an answer to be excellent but out of date; it then doesn't deserve to be downvoted into oblivion, but another answer that isn't out of date may well be a better candidate for the top spot. That doesn't really happen on Puzzling where things are more self-contained.) Sep 12 at 23:27
17

To aid sites in deciding what impact this might have, is there some SEDE magic query that would list and/or count questions where this change would alter the answer order? It would be nice to have local data about the scale of the change.

2
16

I think choosing between "pinned" and "unpinned" is not good enough. I think "pinning" should be replaced network-wide with something that accomplishes the original goal of pinning the accepted answer with fewer negative side effects and possibly even a slightly improved user experience, instead of either taking away something that some of the community find useful or leaving something in place that causes a negative user experience for some of the community.

According to What does it mean to accept an answer?

Accepting an answer is important as it rewards posters for correctly answering your question, informs others that your issue is resolved, and pins the answer to the top so others reading your question read that answer first.

I interpret that to mean the goal of pinning the accepted answer is to make it more visible to future readers. There are ways to accomplish that without forcing readers to see the answer that personally worked best for the question's author before they see the best answer as voted by the community.

I would prefer an easily noticed (but not intrusive) button to bring up the answer the author chose. If there is no accepted answer, the button should not be visible. The checkmark should continue to exist so the accepted answer is still highlighted as I scroll through the answers.

Not only does this put the decision on whether to read the accepted answer first in the hands of the reader, I think it improves the accessibility experience because someone using a screen reader will be able to tell immediately if there is an accepted answer by the presence of the button. If "pinning" can change according to the site, it would be difficult for someone using a screen reader to get directly to the accepted answer (or know whether one exists) on sites that choose not to pin it. I don't have a lot of experience using a screen reader, so I'm just speculating.

I don't know if this solution is the best one, and I would support more elegant alternatives, but it's better than forcing each site to come to a consensus about whether to pin or not to pin.

5
  • 3
    You beat me to it. I was planning to write about this but was not sure about the wording. Something like "Jump to accepted answer" would work but might not be the best wording. It's somewhat similar to how other sites have a "Go to solution" link in that it gives a great level of importance to the accepted answer. I still think something like this should be implemented; ideally, while indicating (somehow) that the accepted answer is not THE correct answer.
    – 41686d6564
    Sep 8 at 19:43
  • 2
    This is actually wrong because adding any button to the current UI is more intrusive than a backstage settings solution. Otherwise you're seeing a button on all threads you might not be interested in. Regarding accessibility, in your own words: you're speculating.
    – bad_coder
    Sep 8 at 19:44
  • 4
    @41686d6564 I try to leave the exact details of the UI to people who are smarter about that stuff than I am :) I had the same qualms about exactly what the button might be labeled or documented as doing. The entire accepted answer concept is just a little bit weird. Naming it something else might make its purpose clearer, but I have no suggestions.
    – ColleenV
    Sep 8 at 19:45
  • @ColleenV I was thinking of maybe something like this where the "Show first" row is only shown when sorting by votes. Then, this could either be saved as a setting in the user's preferences or be a one-time thing (like a button) where it defaults to "Highest score" every time. I'm fine with either.
    – 41686d6564
    Sep 8 at 20:00
  • 2
    @41686d6564 I wonder if it should be somewhere else, like closer to the question, so it's clear the author chose it. Like a button under the voting buttons that has nothing to do with sort order, it just pops up the answer the author wants to highlight. One thing that confuses some is that the accepted answer can change. Switching to a "showcase" or "highlight" idea might help that. (the sort order is separated from the question by comments)
    – ColleenV
    Sep 8 at 20:05
15

Can a setting be included in the user profile to keep the accepted answer pinned to the top? My preference is to have the accepted answer shown first.

Ollie said:

(...) because it'll work for more people.

Sonic the Hedgehog said:

(...) this should be a per-site setting, because I'd want to see the accepted answer on the top on some sites but not on others.

11
  • 4
    I agree with this one, not because that's my personal choice, but because it'll work for more people.
    – Ollie
    Sep 8 at 16:32
  • 8
    If this is implemented on all sites, this should be a per-site setting, because I'd want to see the accepted answer on the top on some sites but not on others. Sep 8 at 16:35
  • 15
    I disagree with making this user configurable. This is basic functionality that should work consistently for every user so that we don't have to explain that sometimes it it's pinned and sometimes it's not. If this were an option for changing the color of the checkmark or some sort of aesthetic choice, I would feel differently.
    – ColleenV
    Sep 8 at 17:08
  • 5
    @ColleenV your argument doesn't hold by that logic we couldn't have any settings at all. Besides changing a setting from default presupposes the user knows what they're doing. E.g. I haven't seen anyone asking why they're using dark theme unaware of having chosen the setting themselves.
    – bad_coder
    Sep 8 at 17:20
  • 3
    @bad_coder Darkmode is an aesthetic choice, accepted answer behavior is not. See this article for some thoughts on why "make it user configurable" is not always the best answer: ignorethecode.net/blog/2008/05/18/… There's a difference between a preference and a configuration.
    – ColleenV
    Sep 8 at 17:24
  • 6
    What ColleenV said. My preference is for some kind of "Zoom to the Accepted Answer" button or link.
    – PM 2Ring
    Sep 8 at 17:50
  • 1
    @ColleenV the article and your argument isn't solid because it mixes business logic with presentation. The preference wouldn't overwhelm users and one sort setting doesn't qualify as the nightmare you describe it to be. (I think PM 2Ring's idea less usable that this setting.)
    – bad_coder
    Sep 8 at 17:52
  • 7
    In case this is declined, or takes too long to be implemented, I threw up a quick user script for doing this: meta.stackexchange.com/a/369579/295232
    – Glorfindel Mod
    Sep 8 at 21:13
  • 1
    Yes, it could be nice. But also it needs to not forget that se has some google traffic and the descision will have about 10% difference for that traffic. For some places, like puzzling it indeed preferable to keep accepted pinned, for other places pinning has no practical value, and google traffic most likely would preffer higher vote first.
    – MolbOrg
    Sep 14 at 14:54
  • 2
    Proposal based on the above: i) each site to decide its own default setting (pinned or unpinned); ii) each user (either at the point of joining the site or else at a specific rep level) to have their own over-ride, which can overrule the site's default setting; iii) whenever pinned setting is false, include, above the answers, a link to asker's accepted answer
    – Rounin
    Sep 14 at 20:37
  • 1
    In favour of making it user configurable: There are already some user toggles to the sorting of answers (by date of answer, by activity, or by votes). So adding a toggle show accepted answer on top should be available as well. Sep 20 at 16:52
14

Sports SE will use voting on a meta post to make the decision, or failing sufficient engagement, we can discuss it as a mod team with CMs.

If other sites need a template for this discussion, they are free to use this question and its answers.

1
13

We've discussed this on CGCC and have come to the conclusion that we'd absolutely like the accepted answer to be unpinned on our site.

2
10

A problem I see with this is the new better answers to an old question, e.g. once the Hot Network stampede has subsided. This leaves better or even the correct answers buried.

What I thought about before is vote normalization based on Q age. Of course some may want to game such system by answering later, but here's the cool thing: it works the same for downvotes. Call it the anti-necroing device (i.e. be sure to add to the topic).

Just food for thought. I'm not really familiar with posting answers on MSE, so let me know if I shouldn't have posted this tangential viewpoint.

3
  • 7
    I think it's fair to say that this is merely the first step. There's a huge amount of discussion going on on MSO around what to do with outdated answers. This was the first (and likely simplest) change. So, yes, I totally agree that in the case of active askers who accept newer, better answers, this is a bit of a loss but that's why this is such a complicated issue - there's no perfect solution that allows for the various participation levels of askers while maintaining the accepted status as an indication of the asker's opinion.
    – Catija StaffMod
    Sep 8 at 21:37
  • 10
    Alternative sorts is one of the things people have discussed, such as sorting by recent voting rather than all-time voting. So this isn't the end of the discussion, merely the beginning, I think. :)
    – Catija StaffMod
    Sep 8 at 21:37
  • 5
    From the 2021-07-21 blog post The Loop: Our Community & Public Platform Roadmap for Q3 2021 (my emphasis): "We will also be exploring the concept of a Trending sort that weights recent votes more heavily than older votes.". Corresponding MSE post." Sep 8 at 22:27
9

Right now, we are discussing this issue on Math.SE. I posted an answer there, but I think it is relevant here also.

I think the best solution would be to allow users to choose between sorting strictly by the number of votes, and putting the accepted answer at the top. Right now, there are three ways of sorting the answers: active, oldest, and votes. I would like to add a fourth option that ignores whether an answer has been accepted or not. (What this option should be called is another question.) As before, the site should "remember" the option that you used last time.

4
  • 3
    I think the best place would then be to have this as a setting on your user profile, rather then add another element to the question UI.
    – Luuklag
    Sep 9 at 13:59
  • 3
    @Luuklag: It's good that you've brought up. Judging by some of the other comments here, it seems that there is not a consensus on whether it should be a setting on your user profile, or a fourth filter on each thread. Personally, I don't have a strong preference either way—I just users to be able to choose how they sort the answers.
    – Joe
    Sep 9 at 14:04
  • 1
    I think that at least the discussion applies to the default setting, even if we have such an option. I'm assuming that the initial problem of "accept bias" affects mostly users that don't care with this type of settings, so at least changing the default behaviour can be a step forward (sorry I'm hijacking a post related to a community that I don't participate actively, but I'm thinking in general terms, not directly related to Math.SE)
    – Bacco
    Sep 15 at 23:00
  • 1
    @Bacco: Yes, I agree that we should still discuss what the default setting is (especially since there are unregistered users/new users who won't know how to change the way to sort the answers). I just think that whichever choice is made, the ability to choose between sorting strictly by votes, and putting the accepted answer at the top, is preserved.
    – Joe
    Sep 17 at 20:54
9

Accepted post is pinned second, if it's not the top-scoring answer.

My thought process:

  1. Sometimes accepted answers are underrated in terms of voting, especially in relation to Fastest Gun in the West, Hot Network Questions, and answers to old questions (slow, but well-thought-out answers => fewer upvotes).

  2. However, sometimes the most upvoted answer is indeed far better than the accepted answer (or perhaps the accepted answer is flat-out wrong).

I like this approach because it's so simple, and it has the property that "accepted answer" + "most upvoted answer" remain the top two answers (in some order).

(This was originally a comment since I didn't put much thought into it at the time, but Mari-Lou A proposed making it an answer. There's still some boundary cases that need ironing out, e.g. two equal-score top-scoring answers vs. accepted answer.)

3
  • 1
    I like this, and maybe add some labels to the two answers like "top voted answer" and "accepted answer" so users know why they are both pinned? Sep 22 at 22:55
  • 4
    I think giving the accepted answer some added visibility, but not top billing makes a lot of sense. It would be worth running an experiment on at least. If the top and accepted answer have the same score, we should do what we already do for other answers—flip a coin for the order.
    – ColleenV
    Sep 23 at 2:06
  • I had the same thought, except that up to two higher-voted answers could appear above the accepted one. Votes are more important, but the accepted answer shouldn't be completely buried. Sep 29 at 19:17
9

In Stack Overflow in Portuguese we opened a Meta post discussing the matter.

Devemos mudar a ordenação das respostas, não afixando a resposta aceita antes das demais?

Votes up are "Unpin", down are "Keep pinning".

Currently we have:

  • 12 + 1 votes to unpin 1

  • 1 to keep pinned

1 The "+ 1" vote to "unpin" is per the last post line (OP can't vote, but stated he wants the change too).

The "unpin" is active now, and the score had no recent changes.

7

The Mathematics SE community is in favor of unpinning accepted answers (judging by the response to the statement "I am in favor of ordering answers by number of votes only.")

1
  • 2
    In the immediate aftermath of the meta discussion you link, there were no compelling counter-arguments to unpinning presented. A few days ago (when I first saw the meta question), I posted such a counter-argument, and it has seen strong support since then. (Ironically, I think the post itself embodies the counter-point it describes.) I think in light of the support my answer has received, it is worth re-opening this issue and presenting the counterpoints to let the MSE community vote on, before committing to unpinning accepted answers.
    – Yly
    Sep 22 at 22:40
7

A point raised in the comments by ColleenV is that Google displays the accepted answer in the snippet. My understanding is that this is automatially done by getting the answer with acceptedAnswer set on it (see docs). Will this change for this experiment to be the top answer, or is that outside of the scope?

7
  • 2
    Google does whatever it wants so how they do things is off topic here. Sep 16 at 14:49
  • This is an excellent question! I didn't know that SE basically gave which answer was accepted to Google directly; I always assumed it was automatically parsed.
    – zcoop98
    Sep 16 at 14:50
  • 6
    @RobertLongson - Google has the ultimate say over what displays in that top box and in search snippets, but it uses Stack Overflow's "context clues" (aka Microdata) to determine if it's worth showing at all. The reason why I think it's not off-topic is because we're ultimately changing what people are going to look at the most: the first answer Sep 16 at 14:54
  • 5
    @RobertLongson That's an incorrect understanding of what's being asked here– Stack Exchange is structuring the data that describes a given QA page in a certain way so that it's easy for Google (and other engines) to parse and understand. SE is currently setting the OP accepted answer as the "right"/ accepted answer in that schema; whether or not SE will begin sending the highest scored answer in this place instead could have a real impact, and is very on-topic.
    – zcoop98
    Sep 16 at 14:56
  • 1
    @zcoop98 that's not correct. SE is doing nothing, but describing the content. Google is the one that should be able to select the highest scoring one. Sadly it seems that the score isn't included in the schema.org behavior, only total upvotes and total downvotes (I don't have a good idea of how this should work, probably a new property called totalscore).
    – Braiam
    Sep 16 at 17:48
  • 2
    Thank you for that link. I see that "accepted answer" is a property of Question, so SE doesn't have a way to suggest an order for the answer list to a search engine; even if we fiddled with the answer that was marked as "accepted", I assume the interpretation of that would be up to the site using the data. Although it does look like we could pass a list instead of a single answer. Could we send both the "accepted" and the highest voted if we wanted to? (web stuff is not my area)
    – ColleenV
    Sep 16 at 18:37
  • Sometime after reading this, I ran across the official Google tool for testing what a given page sends back as a schema: Google's Rich Results Test. Testing this page, for instance, confirms a few things: 1. SE is sending Google all answers (accepted as such, and all others under "suggested"); 2. "upvoteCount" is sent with total score, not the actual upvote count; 3. all "suggested answers" are sent with their score, meaning that Google does have a way to rank them.
    – zcoop98
    Oct 4 at 16:30
7

We've discussed this on Chemistry Stack Exchange. The only answer is by one of the highest-rep members and a moderator, with a score of +10. The general consensus is to unpin the accepted answer i.e. we support this proposed change.

Some of the reasons given there, which I think would apply equally well to some other sites too -

  1. The OP may not be the best judge of answers. This can happen if they themselves have a limited knowledge of the subject (as in the case of many students who frequent the Science sites, me included), and some of them hastily accept the first answer and never come back.

  2. The number of votes is an indicator of which answer the community, as a whole, believes to be the most useful. The opinion of the whole community is certainly more likely to be a better judge of answers than a single person's accept-tick.

7

Our vote at rpg.se What is our stance on "Unpinning the accepted answer from the top of the list of answers"? features a great variety of answers.

Until we reached the first cutoff period on September 19th, our position expressed in Doppelgreener's answer was by a significant margin in favour of unpinning the accepted answer.

Until September 24th, the voting remained in favour of unpinning accepted answers. On that day, Doppelgreener declared in an edit to align her position with the opposing views while keeping her initial answer intact. This edit was the first edit to her answer after September 9th.

In particular, her views aligned thereafter with KRyan's answer, first posted on September 14th, that steadily gained support at the same time. After that, the voting trend changed to have an influx of negative votes on Doppelgreener's answer - as of now, KRyan's answer sits at an only slightly lower score than hers.

The first answer, posted by Screamline, is against unpinning the accepted answer. As of now, it is the answer with the fourth-highest score. Our, as of now, second highest-scoring answer, posted by Thomas Markov, gives insight into various statistics.

We feature our discussion on meta until October 14th. We first featured it shortly after KRyan added their answer.

The top scores are currently about 50/50. Doppelgreener's answer has the highest participation of currently 48 votes, KRyan's answer has 35 votes (this answer had six fewer days to accumulate initial votes available to the other two), and Screamline's answer has 31 votes.

We concluded our vote with a slight majority of votes in favour of unpinning accepted answers, at 36 votes, which, however, also includes 15 downvotes bringing it to the score of 21. The minority vote in favour of pinning accepted answers ends at 31 upvotes and only 9 downvotes, scoring 22. In conclusion, we remain at a 50/50 score with no clear consensus. So we continue with the status quo, which is the old default of pinning accepted answers until we decide to revote.

2
  • 5
    Seems to be fairly strongly in favor of unpinning, though. That top answer outscores the next-highest "keep pinned" answer by 3x. And that's without considering the massive misunderstanding of how answering/the SE model works by some of the major "keep pinned" proponents evidenced in their answers and follow-up comments.
    – TylerH
    Sep 14 at 14:09
  • 6
    We've just featured this meta post yesterday, and picked up new answers since doing so, so RPG.SE's discussion is still ongoing and not something I'd call settled. Sep 15 at 14:42
6

Would you please post a link to a Data Explorer Query, which would list the Q&As (on a given site) where the accepted answer is not also the top-voted answer?

I think it would be helpful, when deciding, to see examples of which topics are affected by this decision.

3
6

Ask Ubuntu

We discussed on our meta whether to have the accepted answer unpinned or not.

Most people voted in favour of unpinning the accepted answer from the top, it seems.

A relevant concern that was strongly expressed was about obsolete answers appearing ahead of newer ones. A very popular answer proposed a solution to that instead of directly addressing the question of whether to unpin. One answer said that an advantage of pinning the accepted answer to the top was that the OP could change the accept if the old answer became obsolete. (I upvoted the answer even though I do very much want accepted answers to be unpinned). My opinion is that this rarely happens; another answer argued that it's more likely that an obsolete answer is pinned at the top while more up to date answers, often with more votes, are stuck below, and I feel the same, but it's hard to be sure (on our site, more than 9000 questions have an answer that outscores the accepted one, but I do not know, for example, how many of them are newer than the accepted answer by a margin of years).

I completely share the concern with the problem of obsolete answers, but I think that accept pinning, to the extent that it might mitigate that problem rather than exacerbating it, is a very weak mechanism, and I hear that better solutions are in the pipeline somewhere. If that is the case and we are going to have some way to deal with obsolete answers more effectively than we can with current tools, then I think all concerns about unpinning can be considered addressed and we can go ahead with it.

Ask Ubuntu Meta

We separately discussed whether or not to unpin accepted answers on meta and nobody disagreed with my opinion that we should leave them pinned. Some folks even upvoted my answer about it, so I guess they agree.

TL;DR

Unpin on main and keep pin on meta, please, and thanks a lot for asking <3

0
6

The issue with the Accepted Answer is about what Stack Overflow is and how it is intended to be used.

There are 2 possible ways to think about Stack Overflow:

  1. A help-desk where someone asks a question and gets help from the community. Accepted Answer makes a lot of sense here because you let people that OP has found a solution and no longer needs help.
  2. Wikipedia in Q/A format, of user-generated question. That is what SO is in practice. Here an Accepted Answer doesn't make sense because answers are not truly meant to help the OP in particular. The fact that OP does get help is tangential to the site's purposes and is merely useful because it entices people to generate content by asking questions for others to answer. The real goal is to help visitors who come from Google or other search engines.

When SO was new and little-known, and didn't have a critical mass of existing solutions and accumulated upvotes, it was 1) and Accepted Answer made sense, but it eventually became 2) and Accepted Answer was no longer needed.

4

Literature Stack Exchange

As one of the moderators on Literature Stack Exchange, I created a meta question asking whether the community wants accepted answers to be unpinned. Several people indicated in the chatroom that they have no opinion about the issue. The votes on the meta question suggest a slight preference for keeping accepted answers pinned at the top.

It is worth noting that answers on Literature SE don't get outdated over time, unlike answers on Stack Overflow, where validity or relevance can be influenced by technological evolution.

4

Language Learning Stack Exchange

As one of the moderators on Language Learning Stack Exchange, I created a meta question asking whether the community wants accepted answers to be unpinned. The votes on the meta question suggest a slight preference for keeping accepted answers pinned at the top.

It is worth noting that answers on Language Learning SE would rarely get outdated over time, unlike answers on Stack Overflow, where validity or relevance can be influenced by technological evolution.

4

Computational Science SE

via voting on Meta decided that we support unpinning the accepted answer from the top.

Thus,

  • if "by default" the accepted answer is unpinned for all communities, we don't need anything special.
  • otherwise, if unpinning requires explicit request, by the means of this message we express one.
2
4

French Language Stack Exchange

As one of the moderators on French Language Stack Exchange, I created a meta question asking whether the community wants accepted answers to be unpinned. The votes on the meta question suggest a preference for unpinning accepted answers.

It is worth noting that answers on French Language SE are much less likely to get outdated over time than on Stack Overflow, where validity or relevance can be influenced by technological evolution. Language does evolve, but not nearly as quickly as technology.

You must log in to answer this question.

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