10

Those of you frequenting Stack Overflow might have noticed: accepted answers are no longer pinned on top. To clarify: currently, it does not matter if the accepted answer has a lower score than others; it is always shown above it. In the past, many people complained about this, often stating that it's more important what the community thinks (and expresses in the form of upvotes) than a single user, even if though they are the author of the question. (Of course, in the old and new situation, if the accepted answer has the highest score of all answers, it will be shown on top.)

This change will be rolled out to other sites in the network as well, including English Language Learners. The company has offered us the choice there to keep the current behaviour if we decide it is beneficial for our community. @MariLouA asked me to post a question here, that sounds like a good idea and it will be interesting to hear your opinions.

I'm going to do some kind of poll with two answers: one to keep the current behaviour and one to have it changed. Please:

  1. upvote the answer representing your choice (downvotes will not count)
  2. edit additional reasons to choose one or the other option

Comments on the answers can be used to improve the reasons mentioned in them (in line with their normal intended usage).

4
  • 2
    Thank you so much for bringing this new development to everybody's attention.
    – Mari-Lou A
    Sep 11 at 9:54
  • Looking at the votes on the answers : Poppa bear says yes. Baby bear says no. Mama bear says meh. That's not reaching a consensus for which I'd decide to make site-wide changes for everyone whether they want them or not. That would require an overwhelming majority of +90% IMO.
    – Mazura
    Sep 16 at 18:07
  • Thanks for the script that lets us opt out if this does come to pass. But that's not how it should have to (not) work if I don't want it to. Best comment on the Meta : "how will you convince the don't move my cheese folks?" That's the neat part; you don't.
    – Mazura
    Sep 16 at 18:12
  • There is a SEDE query listing all of the questions where the accepted answer has a lower score than the top scored answer and it includes the difference in score. I took a quick look and in most of the cases where there was a big difference, neither answer was clearly "wrong". There were many instances where the accepted answer had a more comprehensive explanation than the succinct answers the voters preferred. In many cases the "short" answer was posted first.
    – ColleenV
    Sep 17 at 13:26
10

We should unpin the accepted answer

Possible reasons for this:

  • The community as a whole is better suited to determine what the best answer to a question is than the vast majority of question authors.
6

We should keep the accepted answer on top

Possible reasons for this:

  • When a new answer, better than the ones before, is posted late, the earlier answers may already have gathered so many upvotes (for example from Hot Network Question activity) that it's difficult for the best answer to reach that score as well. If the author of the question accepts the answer and it's pinned, it will get its deserved share of attention.
  • Our active voter base might be skewed towards native or near-native speakers, and Stack Exchange veterans, and vote more often for comprehensive answers. Question authors might be more likely to go for simpler answers, which might also be more useful for the majority of the visitors (anonymous learners of English who arrive here via search engines)
8
  • 2
    That would not be my reason for keeping accepted answers at the top. Many times an accepted answer on ELL is also the answer with the most upvotes. And sometimes a better, more nuanced answer can be posted weeks or even years later but not earn many upvotes.
    – Mari-Lou A
    Sep 11 at 7:39
  • 2
    As a result the OP may change their mind and accept the new and better answer, this decision deserves to be highlighted.
    – Mari-Lou A
    Sep 11 at 7:47
  • 2
    Moreover an accepted answer says that it was the most helpful to the OP regardless of the number of votes it attracted. It tells the community that to all intents and purposes the problem has been solved.
    – Mari-Lou A
    Sep 11 at 7:50
  • 1
    I find the arguments of @Mari-LouA compelling. It does not often happen that anyone switches their vote to a better late answer but it does happen.
    – mdewey
    Sep 11 at 13:24
  • @Mari-LouA the late answer case is a really good one (and even applicable to other sites in the network - it might be worth mentioning on Meta.SE as well).
    – Glorfindel Mod
    Sep 11 at 16:21
  • 5
    An accepted answer should indicate the one most helpful to the OP, but let's not forget the countless cases where somebody just accepted the first answer thrown at them and never came back to revisit that decision.
    – Glorfindel Mod
    Sep 11 at 16:22
  • @Glorfindel totally agree, it's very true that askers often accept the first answer posted within an hour. And sometimes the OP accepts an answer that happens to agree with their own personal interpretation.
    – Mari-Lou A
    Sep 11 at 18:36
  • I think that the last bullet point here isn't true - it seems like early, correct but short answers are preferred by voters, but the author of the question often accepts the answer with more explanation, or where the author of the answer has responded to comments.
    – ColleenV
    Sep 17 at 13:28
6

No matter which way we do it, there will be potential problems and pitfalls.

When a mediocre answer gets hastily accepted, and better answers come along later on, it's nice to see that mediocre answer get pushed downward.

However, when a well-thought-out answer comes late but is ultimately accepted, it will be sad if that answer sits at the bottom because the question is now in the rear-view mirror for most of the people who voted on it, so the newly-accepted answer has trouble amassing many more upvotes, even though it might be deserving of getting them.

In the end, sometimes the new system will be an improvement, but sometimes it will be inferior to the old way of doing things. C'est la vie.

1
  • Which is why I argued on Meta that "pinning" should be replaced with something better instead of making us choose pinned or not pinned.
    – ColleenV
    Sep 13 at 17:30
2

There are several advantages and disadvantages to pinning the accepted answer at the top and until recently I was undecided.

Contrary to many users who fervently sustain that an incorrect or low scoring answer should never be shown at the top, I think the current system works admirably well on ELL for two reasons.

  1. Users who read an accepted answer that is inaccurate or misleading will downvote it. Once an incorrect answer has been accepted, the number of downvotes increases exponentially as soon as a correct answer (regardless of its length) is posted. This voting pattern makes it clear to visitors and users alike which answer is considered best by the community.

  2. An accepted answer says that it was the most helpful to the author of the question regardless of the number of votes it accrued. It tells the community that to all intents and purposes the problem has been solved. In 2015 Taryn (Staff ♦ Mod) explained why their suggestions for unpinning the accepted answer were overruled

…because the pinning of accepted answers is part of what makes our sites different from others - it's the indication that an answer worked for the OP - we don't want to lose the signal for those answers or hide them way down on a list of a ton of answers.

  • Despite the differences in scores, one author may have understood the OP's dilemma better than a higher scored answer. And sometimes a better, more nuanced answer can be posted days, weeks or even years later but earn very few upvotes. If I am that OP I would want to thank the user by accepting their answer and seeing it pinned at the top of the page. This shows my gratitude and it has the given advantage of being read first, and if it remains useful it will often lead to more upvotes.

Why accepted answers have never really been a problem on ELL

  • Language sites such as ELL and EL&U are non-technical, so there is no such thing as an obsolete answer. A nontechnical question may ask about obsolete practices or archaic terms but a well-supported answer will not become outdated next year or in ten years.

In an era where advancements in technology are being continually made, it makes sense that an answer accepted in 2010 on Stack Overflow, but now obsolete, should not be displayed at the top position. This is not true for social science and the humanities where any changes that do occur happen more slowly and over a longer period of time.

2
  • I think you make a good case for why pinning the accepted answer should be replaced with a better implementation of increasing its visibility. Ideally the top-voted and accepted answers should be side by side, but I wouldn't want that when I'm browsing on my phone. Maybe there is an "accepted" tab that no matter where you are in the answer list, you can flip to the accepted answer without losing your place?
    – ColleenV
    Sep 20 at 17:38
  • 1
    Side by side could work if the left navigation panel never existed. The day that feature is turned on permanently is the day I say "Goodbye". The easiest solution I found was in a comment, it suggested placing the accepted answer immediately below the highest scored one or it could be placed second place whatever the tab selected. I've been thinking this for a bit, and I've come to the conclusion the accepted answer is not a problem here. Besides on ELL there are more questions w/o accepted answers.
    – Mari-Lou A
    Sep 20 at 17:49

You must log in to answer this question.

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