7

For the full background, please see this MSE question.

In short, currently, if the question-asker “accepts” one of the answers, that answer is “pinned” to the top position, shown to all visitors at the top of the page, the rest of the posts appearing in order of score, so that the highest voted post, if different, appears directly beneath it.

It was noticed on StackOverflow this caused problems for that site. The most pressing problem for SO, which is technically oriented, is that old, obsolete, or dangerously outdated answers could and would be shown before more updated or correct ones. SO posts often get thousands of views by voting users long after the original question is posted, and so new answers often rapidly accrue votes and over-take highly upvoted, previous, out-of-date answers. Therefore now SO shows answers (if using the default sort) strictly in descending score order.

This turned out to be popular for that particular site, and so now the Powers That Be are asking their other sites for their preferences.

So, do we want to take the option? Would you prefer the status quo, where the OP-accepted answer is shown first (except where the OP themself posted the answer, of course), or should we switch to strictly score-based sorting

To inform your decision, you can view the complete list of which answers would be affected on Main (6,888 answers; average score difference between accepted and top-voted = 4.6 pts) and Meta (137 answers; average score difference between accepted and top-voted = 4.25 pts). Note that the settings for Main and Meta are independent; we can unpin on Main and keep the pin on Meta, for example.


It would be best if answers to this question took a concrete, binary stance — keep pinned or unpin — and optionally, but encouraged, a rationale.

Mod note (Andrew Leach ♦): Because there is a binary stance, only vote up the answer you support. Only the upvotes on each stance will be counted. If you vote up one answer and vote down the other, you effectively double your vote, which is undesirable.

46
  • 1
    This discussion about a highly upvoted but wrong answer may offer some food for thought. I'm not saying it supports pinning or unpinning, but it was a real controversial question on ELU where pinning the accepted answer had an effect.
    – ColleenV
    Sep 21 at 17:59
  • 1
    @ColleenV BTW, feel free to post an answer taking the “keep the status quo” position if you’re still inclined. We need both positions for the Metazens to vote on.
    – Dan Bron
    Sep 21 at 18:18
  • 1
    @ColleenV As you can see, I am too. I had wanted it for years, and nothing to do with SO’s (ill defined, as you say) problem with obsolescent technology. I’ve only ever seen the accepted answer be meaningfully outscored by competing answers when the OP churlishly accepted it because everyone else disagreed with him.
    – Dan Bron
    Sep 21 at 18:20
  • 2
    @Araucaria-Nothereanymore. I know the SWR-hugger type you mean. It’s part of why I consider rep meaningless (though not the biggest part). I’ve read all the comments you’ve left in this thread, and I’ve ended up at: I agree with jw that you point out real deficiencies in the voting system, no question. But I also agree with jw that the fix isn’t pinning accepted answers; it puts too much power in one person’s hand. In fact, that’s my main complaint with it: more often than serving as a positive check against mob voting, I see accept-pin abused by petulant OPs who don’t want to see the truth.
    – Dan Bron
    Sep 24 at 19:30
  • 1
    @AndrewLeach Ok, it’s been a week, and the votes have been stable for a few days. We have about 11 for unpin (the 10 upvotes on my unpin answer, plus myself who can’t vote on it), and 6 for keep pinned (the 6 upvotes on the keep pinned answer I posted). I’m not sure how to account for the votes on the other answers, but I suspect some keep-pinned users voted on both (and anyway even if we count them independently the total is still < 11). Thus I intend to edit the MSE post to say EL&U wants to unpin. If you disagree or think it premature, feel free to roll back my edit.
    – Dan Bron
    Sep 27 at 9:36
  • 1
    @AndrewLeach Fair enough, but I’m not clear on how we will distinguish votes on his answer from votes on the earlier keep-pinned answer in terms of deciding whether to keep pinned or unpin. I’m happy to hold off fir any amount of time. I’ll check back in in a week.
    – Dan Bron
    Sep 27 at 9:54
  • 1
    @AndrewLeach They’ve already been directed to this post in a table summarizing all child meta Qs on this topic. That brake also reports the conclusions in a single column which indicates either “keep pin” or “unpin”. I edited a link to this meta-Q into that table when I posted it. Now all the remains is editing in our conclusion. That’s what I was describing above.
    – Dan Bron
    Sep 27 at 10:02
  • 1
    -1 for not editing your post (why do I have to search through the comments?) and not announcing the decision to unpin and for not saying you have edited the aforementioned MSE post.
    – Mari-Lou A
    Oct 5 at 10:34
  • 1
    @Mari-LouA I said in an earlier comment I’d give it another week and I gave it another week. Another user rolled back my edit dude to the post scores being now equal (disregarding the upvotes-only rule), so I will wait yet another week, and I will edit the MSE post again according to the upvotes on each position. We need to make a decision.m: keep pinned or unpin. If keep pinned has more upvoted in a week I’ll edit the MSE post to say keep the pin, and vice versa.
    – Dan Bron
    Oct 5 at 11:48
  • 1
    @NikeDattani, the deadline in the title only made explicit what was already in one of Mr. Bron's comments, and that deadline was already an extension. The hope was that making the deadline more prominent would motivate those who haven't already voted to do so. (Note, incidentally, that expecting the community to agree on the deadline leads to an infinite regress: how would the deadline for agreeing on the deadline be agreed upon?) Given the reasonable arguments that have been thoughtfully presented on both sides of the issue, consensus seems unlikely.
    – jsw29
    Oct 7 at 20:42
  • 3
    @NikeDattani We need some deadline. It doesn't have to be the 12th but it has to be something. Not making a decision is, as I explained to others, making a decision. I've happily pushed it back a couple times, and I'll happily do it again if activity on the thread warrants it. As for determining what that decision is: all the other sites which have updated the MSE post have done it on the strength of votes, so far as I can tell. That's the point of making arguments & assessing their persuasiveness to the community. Having said that, \new method is for a mod to add "status review" for a CM
    – Dan Bron
    Oct 7 at 20:52
  • 3
    @NikeDattani Because not making a decision is making a decision. The deadline has been extended a couple times because of new activity in the thread. So long as the votes are changing, it’s worth re-extending. But we can’t extend indefinitely because not making a decision is making a decision. As for why me, or jsw, or anyone else: because this is how meta sites are run: it’s on us, the regulars of the community to drive meta and make community decisions. It’s not on the mods, it’s not on the CMs, it’s on us. It’s our duty, and it’s our call.
    – Dan Bron
    Oct 7 at 21:09
  • 3
    @NikeDattani I’m not going to rehash the same topics. This will be my last comment. We need a deadline, so we have a deadline, though we can and will be flexible about it. We will use votes because there is no other way to determine what the consensus actually is; you might find A’s 14 pts compelling, I might not, and vice versa. How do we know? Voting. If voting fraud is rampant, we have bigger problems to worry about than pinning the accepted answer. If you need comfort there will be a sanity check, the CM status-review should provide that. Thanks, nice talking to you.
    – Dan Bron
    Oct 7 at 21:26
  • 2
    If the voting is close, why choose? We can already sort by date or votes, so why not have a check-box so users can decide for ourselves whether to pin accepted answers?
    – Lawrence
    Oct 8 at 6:42
  • 3
    @Araucaria, as I have already said in my comment on your earlier post, I agree that acceptances sometimes work in the way you describe and that, when they do, they may produce better results than voting, but I am not convinced that this happens as often as you believe (Dan Bron, Mitch, and those who upvoted Dan Bron's answer are presumably not convinced either). There probably isn't much to add to the arguments, on both sides, that have already been stated, very clearly, on this page, so the only thing left is to wait and see how the voting turns out.
    – jsw29
    Oct 9 at 16:23
17

The accepted answer should appear on top (this is the status quo).


We have the opposite problem to Stack Overflow. Answers on SO often get thousands of views and votes by active voting members long after a question is posted. New posts can therefore easily overtake an early accepted answer. Here, however, the number of views by active voting members bombs after about 48 hours. The only way of promoting a later, superior, more helpful answer and making it easily visible to readers is through the 'accepted answer' mechanism.

We are on the verge of having many of the best answer posts on EL&U suddenly consigned to obscurity under a pile of mediocre early answers, reducing the value of this site to readers.

An accepted answer says that it was the most helpful to the author of the question regardless of the number of votes it may have accrued. It tells the community that to all intents and purposes the problem has been solved.

Despite the differences in scores, the author of the accepted answer may have understood the OP's dilemma better than a higher scored answer. And sometimes the OP may change their mind and accept a better, more nuanced answer that was posted days, weeks or even years later. If I am the author of the question I would like 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 it will lead to more upvotes if it is seen at the top of the page rather than lost in a sea of answers.

Finally, the top answer is not necessarily the "best". HNQ visitors are not always the most objective or knowledgeable in the field of language. Native speakers have an instinct that is matchless to any artificial intelligence software or language learner but that doesn't mean they are able to explain or understand why one answer is superior to another.

For instance, this question "Closet" vs. "Wardrobe" Why is the first more common in the US? posted on December 14 2017 at 12:49 has three visible answers. The “best” answer, with 87 upvotes, was posted on the same day at 13:39. But the author of the question accepted an answer which was posted six days later. Was the author of the question wrong to accept an answer with only 13 upvotes? Does that answer not deserve the top position? You be the judge.

Another reason for pinning the accepted answer at the top is that it keeps the two most significant answers at the top, namely, the accepted answer and the most popular of the other answers. And it does so without additional programming, compared to sorting by votes and then pinning the accepted answer below the top-voted answer (if they are different answers).

10
  • 3
    Arranging answers strictly by vote count would make more sense if all answers appeared at approximately the same time and were seen by approximately the same number of prospective up- and downvoters. But in reality, answers may appear years after a question is posed—and if a late-arriving answer is good enough to win the question poster's check mark after such a delay, it seems a shame to leave it at or near the bottom of the stack just because it wasn't around when the earlier answers started attracting upvotes.
    – Sven Yargs
    Sep 22 at 19:55
  • @SvenYargs This is a CW answer. Feel free to edit in that argument if you’re so inclined. (Also if you want to talk to the author of it, you need to explicitly @-ping ML, otherwise I’ll be the only one notified. CWs get messy that way.)
    – Dan Bron
    Sep 22 at 20:18
  • 7
    This may be a good argument for the existence of the institution of accepting, in addition to voting, but it doesn't really support displaying the accepted answers above those with the highest net votes. It may well be that 'the author of the accepted answer . . . understood the OP's dilemma better than a higher scored answer' and that the answer was 'the most helpful to the author of the question', but that doesn't mean the answer is likely to be more helpful to the future visitors to the page, and it is their interests that should govern how the answers are displayed.
    – jsw29
    Sep 22 at 20:46
  • @SvenYargsSven please edit the post in any way you please. I don't mind if some lines are deleted or even its entire content. Start from scratch if that makes it easier. I already knew this would be an uphill battle.
    – Mari-Lou A
    Sep 23 at 4:42
  • @jsw29 and it is their (future visitors) interests that should govern how the answers are displayed I disagree, the author of the question has a higher status than a visitor or a user who casts their vote. Without questions there would be no answers. Without questions there would be no Q&A SE sites. By diminishing the value of the questioner, we could see fewer answers being accepted in the future because it will become unclear what its purpose was.
    – Mari-Lou A
    Sep 23 at 8:15
  • 2
    Why should an accepted answer come at or near the bottom of a pile because of the number of upvotes it's received? We have all seen answers that attracted ridiculous number of upvotes thanks to HNQ. A huge crowd of visitors is great way for posters to earn rep but it is not analogous to people who are experts or passionate about the subject. Sometimes I agree with the most upvoted answer but not always, and if I am the author of the Q my opinion should count. And if the person asking is an expert in the field, then even more so their opinion should be valued.
    – Mari-Lou A
    Sep 23 at 8:19
  • 1
    This fear that visitors will learn something incorrect or be mislead by an accepted answer with lower votes is a false one. Why? Because whenever an accepted answer is objectively bad or incorrect it attracts mass downvoting, as well it should. Remove the accepted answer from its top position that same bad answer might be shrugged off or completely ignored. Even LQ accepted answers serve a purpose. See tchrist's fascinating answer posted in 2018 to a 2010 question. Were it not for the old accepted answer, that answer would not exist.
    – Mari-Lou A
    Sep 23 at 8:37
  • 2
    @Mari-LouA I'm voting here based on the general expectation. Yes, there are lots of exceptions (late good answers, voters voting for non-substantive reasons, multiple good answers, what -I- want from -my- question vs what others want). There's also an issue of tyranny of the majority, high voted questions tending to get more votes not (or not necessarily) because they are good but because they are highly voted already.
    – Mitch
    Sep 23 at 13:20
  • @Mari-LouA Sorry for being so telegraphic...only so many letters can fit in a box. I am choosing my vote based on those criteria and those criteria led me to vote up on unpinning. Also, I think I may have waffled on which box to put this in, thus misleading you. I just wanted to give an explanation in addition to voting.
    – Mitch
    Sep 23 at 18:06
  • 1
    Wow! Six upvotes in less than two days. What happened?!
    – Mari-Lou A
    Oct 9 at 11:27
7

There is an interesting idea on Meta suggesting pinning the accepted answer below the top voted answer. That system might give an advantage to the accepted answer over other highly scored answers, but never over the top voted answer.

I'd like to know if anyone thinks that would be worth testing. We may not be able to choose that option for this effort, but if there's interest we might be able to get it on SE's radar.

5
  • 3
    I’m in the camp that acceptance really doesn’t need any special attention. I’d buy your feature to put a little direct link to the accepted answer in the footer of the Q, or sort the accepted answer higher than other answers with exactly equal score, but that’s about the extent of it. I think the acceptance mechanism was introduced to solve the main problem with forums in 2008: quickly getting to the answer in a sea of irrelevant comments, but I think by now we’ve learned that forcing all answers to be answers is more than adequate for that need. OP’s opinion shouldn’t carry more weight.
    – Dan Bron
    Sep 23 at 14:21
  • @DanBron I tend to agree, but I thought the idea was interesting enough to try to get it in front of more people.
    – ColleenV
    Sep 23 at 14:25
  • @DanBron I'm in the camp where SWR's and similar should not be taken too seriously when considering how the site should be run. for non SWR-type questions, acceptance is a good indicator of several things, though not perfect. But it's there to counteract another non-perfect system, the so-called 'voting' one. That's why it's desperately needed. Sep 24 at 1:25
  • To see why I downvoted this proposal, please see the end of my post :) Sep 24 at 1:32
  • 1
    I like the idea (+1), but I’d put the accepted answer on top. If it isn’t the top-voted answer, then that answer will only be one position away from the top. You get both answers (top + accepted) at / near the top without any additional programming.
    – Lawrence
    Oct 8 at 6:35
7

Answers should appear only in order of score (this is the new behavior on Stack Overflow).

The mandate of SE in general is to create a “library of reusable knowledge”, and each site attempts to do that in its subject area by attracting and retaining domain experts.

Those experts become “the community”, and the community expresses its will — and that very domain expertise — principally through voting.

I do not think it advances our agenda to allow one question asker — who, by definition, does not know the correct answer to his question — to outweigh the aggregate opinion of this expert community.

The “accept” button should and does express “this is the answer which most helped the OP”, and it still will, even after this change. The difference will be the first-listed answer will be more likely to help anyone else who has the same question down the road.

On a more practical level, I really only see situations where the accepted answer having a lower score than some or most other answers when the OP came in with an axe to grind and accepted the answer which best flattered his preconceptions, despite everyone else disputing it. There are even cases where a diamond moderator had to delete the accepted answer (10K+ only link) because it was highly controversial and didn’t meet the standards for an “answer” on EL&U.

33
  • 1
    But this answer in comparison to the accepted one, supports keeping things the way they are. Sep 21 at 23:32
  • 1
    While I agree that an answer that has >30 upvotes is likely to be very good, when the difference between the answer accepted and the most popular is <10, who is to say that the OP does not recognize a right answer when it is posted? It is up the answerer to convince not only the community but also the OP. And sometimes the author of the question does know the answer and they post the question for the greater benefit of the community and its "library". In those cases they do know which is the best answer.
    – Mari-Lou A
    Sep 22 at 9:04
  • Lastly, HNQ visitors are not always the most objective or knowledgeable in the field of language. Native speakers have an instinct that is unmatchable to any artificial intelligence or language learner but that doesn't mean they are able to explain or understand why one answer is superior to another.
    – Mari-Lou A
    Sep 22 at 9:07
  • @Mari-LouA HNQ is a good point, but those questions are anomalies on many dimensions anyway. I’m not stressed about them. I’ll add links to this Meta-Q later today so you can see exactly which answers on Main would be effected if we adopted this change.
    – Dan Bron
    Sep 22 at 10:52
  • I see only one option to vote for! It would be nice a Keep it pinned box for anyone brave enough to cast their vote against the current flow. :)
    – Mari-Lou A
    Sep 22 at 11:11
  • 3
    @Mari-LouA (a) There's nothing stopping you creating that post. (b) Only upvotes should be counted in a two-post contest, in order that people cannot vote up one post and vote down the other and effectively double their vote.
    – Andrew Leach Mod
    Sep 22 at 12:57
  • 1
    @Mari-LouA As you like. Please see the footnote.
    – Dan Bron
    Sep 22 at 15:03
  • 3
    A check mark is a visible "thank you" and may influence future readers in search of an answer, but poser is equal to all other voters and gets only one up/downvote in "community judgement". I vote to keep answer order fully dynamic. // Just as there is a green square on the main page to indicate that question has an accepted answer, I suggest that if the decision is to unpin, we request a similar small symbol on the answer page to reminder readers that there is an accepted answer below (if it doesn't happen to be at the top.)
    – DjinTonic
    Sep 23 at 9:54
  • 4
    @Mari-LouA My reasoning for not wanting the accepted answer pinned to to the top has two assumptions. 1) the question poser -doesn't know the answer-, that's why they're asking, and further they are likely not to have the knowledge to judge the correctness. An accept by the OP doesn't mean an answer is right, but that it was most helpful to the OP. 2) Though the question is motivated by a single user, the answers are read and judged by everybody, so the utility of the accepted answer to all the other users is questionable, that is I see votes as a better marker of quality than the OP's check.
    – Mitch
    Sep 23 at 13:09
  • 3
    @Mitch Voting order is most often an indication of when the answer was posted. Voters don't get notified when new answers, which may be definitive, provide real evidence etc appear. There is only the remotest semblance of peer review here because far more users have upvote privileges than downvote ones, and also only the remotest semblance of democracy because early posted answers have thousands more opportunities to be upvoted than later ones, especially those posted months or years after the original question. It's misleading to pretend early answers aren't artificially promoted. Sep 24 at 1:57
  • 2
    @Mitch And thus misleading to pretend that highly upvoted answers would be more highly regarded by the community. The community just didn't see the later ones. The accepted answer feature is one check in the face of this problem. Sep 24 at 1:59
  • 3
    @Araucaria-Nothereanymore. Yes, the OP is the only one who sees all the activity. I think then that it is a matter of opinion whether you think the OP is a good assessor of what the best answer is -and- if you think that is more important that voting. You have pointed out a very important problem with any voting on things that are presented (like answers) at different times or where everybody sees the voting counts before voting. But I am not convinced that a user 'accepted' mark, or in addition such a mark leading to it being put at the top comes anywhere near helping.
    – Mitch
    Sep 24 at 3:17
  • 1
    @NikeDattani (1) All I can say is that after I posted my comment pointing out the narrow margin, the next day three more votes were posted supporting the change. The post is on meta, it was featured on the bulletin for weeks, the lack of response and debate also counts. Low rep users are probably oblivious of the feature and the majority of veterans are absent and largely indifferent.
    – Mari-Lou A
    Oct 8 at 0:09
  • 1
    (2) I see that the current system works. Never ever did anyone post on EL&U complaining specifically about the accepted answer. When it was mentioned it was in connection to a bad or demonstrably incorrect answer, which shows that the system works. A bad answers was highlighted when it was accepted by the OP, but good answers were upvoted even more so. The vast majority of users posting questions are not idiots, they have usually invested energy and time in their question and are happy when two or more answers are posted.
    – Mari-Lou A
    Oct 8 at 0:13
  • 1
    @Mari-LouA it’s surprising to me too, I saw that a couple of them were un-upvotes for unpinning; presumably these then went to keep-pinning. That suggests to me an appeal was made in chat, maybe. Regardless, the tally’s the tally. If the community wants the pin, we’ll keep the pin.
    – Dan Bron
    Oct 12 at 10:11
5

The accepted answer should appear on top!

I urge readers to vote to pin the accepted answer at the top of the page. This feature does a lot of silent good work whilst the very few annoying instances where an OP picks a (seemingly) obviously wrong answer are very scarce indeed, however memorable these are (see Cag51's post on Academia and the information therein).

Despite the terms voting and upvoted, there is a severe problem in terms of how democratic the voting system is. Early posts get voted on a lot, later posts much, much less so. Around 48+ hours after a question is first posted, the number of views by active voting members drops off significantly.

Pinning the accepted answer allows relatively 'late'-arriving but excellent and helpful answers to be recognised and pinned to the top of the page where they currently benefit readers. Without this feature some of the best and most helpful information on the site will languish unseen underneath a list of earlier mediocre answer posts. This current system still retains the benefit of having the highest-voted answer directly beneath the selected one. The new one will see many existing good posts vanish into obscurity.

Another benefit of the current system is that the Original Poster is the only member who is routinely alerted to new answers, especially those that arrive weeks, months or years after the question is originally posted. Because of this, they are in by far the best position to curate their own question page, and, if appropriate, accept a late answer. Certainly, the slew of voters on the original few answers will not be notified and will not get the chance to vote anew on the full range of answers.

Lastly, the current system affords some respect and agency to people who ask questions on the site. Whilst there are always vaguely annoying members in every aspect of the daily life of every SE site, we don't allow this to destroy or make us abandon useful and helpful features of the site. Where the odd muddle-headed OP might select the wrong answer, this is rarely anything more than an annoyance, and a rare one. In contrast if we in essence lose the selected answer feature, users will lose the benefit of many excellent posts and the helpful information that they provide. The vast majority of people asking questions here are sensible adults fully capable of making appropriate decisions regarding selected answers.

Some SE sites, for instance SO, get thousands and thousands of views by active voting members. So, for example, the highest voted answer on SO has over 33,000 votes. On these sites a very high number of votes over a quite sustained period may be the best indicator of the accuracy and helpfulness of an answer. Here, however, this is not the case.

The voting system on English Language & Usage is a good thing. However, it is not perfect for many reasons, including those detailed above. The 'accepted answer' feature helps provide checks and balances within the system. In particular it defends against the unintended and unwelcome tyranny of the early upvoted answer. Just like a healthy democracy, where second chambers and the separation of the legislature, the judiciary and the police provide safety in the form of checks and balances, the same is true of the accepted answer feature in its current form. Vote to keep it!

10
  • 1
    Regarding point 2, I found 3383 questions where later answers outscore the accepted answer and 4353 questions where a lower scoring accepted answer was posted after the highest scoring answer. Browsing the results, I'd say it's also common that lower quality answers are pinned. It'd be interesting to know how often the most researched answers get snowed under now vs. under the proposal, but I think it requires a qualitative analysis. Perhaps it's doable by taking a small sample?
    – JJJ
    Sep 24 at 5:13
  • Point 10. Truth be told when a new answer is posted, it bumps the question up on the home page. Edited posts and new answers are rushed to the front of the queue. Nevertheless the newer better answer will always attract fewer views. An exception can be made for old questions that are visited tens of thousands of times. A new better answer will slowly earn upvotes every time that question visited and when it is bumped to the front.
    – Mari-Lou A
    Sep 24 at 5:38
  • BUT if the newer better answer is accepted AND pinned at the top of the page it stands a better chance of getting read by future visitors especially if there are dozens of other answers.
    – Mari-Lou A
    Sep 24 at 5:45
  • All regular contributors to this site would probably agree that the voting system is far from perfect and that the answers with the highest score are often not really the best ones. The question here, however, is not whether that net votes are the perfect indicator of which answers are the best, but whether they are better indicators than acceptances. Yes, it is possible that a very thoughtful questioner's acceptance will counter the unfairness of the votes, but does that happen more often than the acceptance being awarded to an answer that is worse than the one with the highest score?
    – jsw29
    Sep 24 at 16:28
  • @jsw29 Yes, I believe it does. And most often on questions where it actually matters. At any rate, under the current system the top-voted answer is pinned directly under the accepted one, so there's no great loss there. However, there will be when the accepted answer disappears out of view. I'd say it's routine for answerers such as Prof Lawler to come late to questions, get accepted and then garner the upvotes from that. I used to answer a lot of old questions knowing that I wasn't going to get much rep, but wanting to get the accepted answer so that future readers would be helped by my post. Sep 24 at 17:13
  • @jsw29 This post on Academia SE shows that, even on rather unfortunately uninformative scales used, only a very, very small number of posts seemed to have egregiously disfavoured answers accepted. However, there were a very significant proportion of posts where the accepted answer appeared much later than the highest voted one. This suggests that far from being fickle, obstinate and dimwitted OP's are being extremely careful, considered and judicious. Sep 26 at 1:36
  • +1 from me. The decision in favour of changing, unofficially announced, has been made and it will be a matter of days, I imagine, before it is implemented. meta.stackexchange.com/revisions/369914/57
    – Mari-Lou A
    Oct 5 at 10:49
  • 1
    @Mari-LouA Unless we can marshal enough votes from peeps unaware of the current poll, maybe. Oct 5 at 12:56
  • 1
    @Mari-LouA Btw, in relation to your observations re point 10, that only happens if you have the question sort ordered by activity. Us old-timers know that that's what provides the most interesting viewing. But the default ordering is by time of asking and that's what most users see! :) Oct 7 at 23:57
  • 1
    @Mari-LouA We are slowly and surely turning the tide! Oct 9 at 0:46
3

I'm just posting this as an answer to make an argument for unpinning. Please upvote Dan Bron's answer to practically indicate support for unpinning according to the rules Andrew Leach laid out.

I'm in favor of unpinning because the current system has no good way for anyone but the original poster to adjust the position of the pinned answer, and aside from situations where the original poster is mistaken or biased (which Dan Bron mentions, and Araucaria argues are rare), I think it's not uncommon for the original poster of a question to just not be active on the site any more.

Araucaria's answer talks about situations where there is a not-so-good (or even outright incorrect) early answer and a late better answer. I agree that the voting system isn't set up to handle this well, but there is always at least some chance for the newer answer to catch up by votes over time, especially if people use tools like bounties to draw attention to it. However, if the not-so-good early answer was accepted and the OP is no longer around, there's absolutely no way for the late answer to reach the top spot under the current system of pinning accepted answers (short of the drastic step of special diamond moderator intervention). So I don't think pinning accepted answers is a good solution to the problem of early answers having an advantage relative to late answers.

2
  • This completely ignores the fact that where there is a slew of early upvotes, sometimes by non-expert readers from the HNQ, there may be literally thousands of views per day, but this will drop to tens per day, or only a fraction of that, a week later. It may take decades to catch up. It also ignores the relative risk. The highest voted answer is currently highly visible, appearing second. The pinned answer, in those rare cases where wrong, is just an annoyance. If unpinned hundreds/thousands of useful posts will effectively vanish into obscurity buried under mediocre or worse posts. Oct 5 at 12:13
  • It also ignores proportionality and numbers, There are very few of those posts where annoyingly, an egregiously wrong answer has been accepted, whereas under the new system there will be a mountain of useful posts effectively buried Oct 5 at 12:13
-1

Pin the accepted answer. Point by point:

  1. The accepted answer feature constitutes one of the checks and balances to the voting system.

  2. Highly upvoted answers are often just early answers.

  3. Highly upvoted answers which later prove to be wrong will no longer be able to be superseded by demonstrably correct well-evidenced and well-researched answers.

  4. Late answers, however good, however correct, do not get highly upvoted

  5. Question posters are not thickos who are unable to recognise a later well-argued and well-evidenced answer to be better than an early one.

  6. The current proposal is a disincentive—especially for experts in that area—for members who wish to post late answers. They may not expect to accrue lots of upvotes, but they can at least hope to be more widely read if their answer is accepted.

  7. Given that early answers, even those which are highly upvoted, are often subsequently proved wrong, the proposal will condemn later correct answers to obscurity and promote misinformation.

  8. Many of the, correctly, highest voted answers to important, popular and useful questions are so precisely because the OP was able to accept the answer and promote it to a more prominent position where it was read more frequently and thus upvoted. This will no longer happen.

  9. Users are unable to downvote answers which they previously thought to be useful after a better, more well-informed, well-researched answer comes along (unless it is edited).

  10. Question posters are the only users who are routinely alerted when a new answer is posted. The thousands of users who already saw the question and voted on the smaller number of answers do not routinely get the opportunity to revote when a later answer appears.

  11. Because of the point above, question posters are the only users who are in a good position to curate their question pages. No one else will do so and then revote as new answers appear, often years later. They are in an ideal position to decide which answer should be at the top of the page.

  12. Many good answers are the result of detailed or diligent research. See, for example, answers by Sven Yargs, StoneyB, Tchrist, Mari-Lou A, JEL, Janus Bahs Jacquet. However, research takes time. If these folks give a stellar answer after a few days research, one of the ways future readers can benefit from it is if the OP selects it over an older, not well-researched but highly upvoted earlier answer. The current proposal disincentivises research.

  13. SE has done a lot of work to promote the role of question posters on the SE network. For example, they recently increased the rep for votes on questions so that it matches the rep for votes on answers. The current proposal is a step backwards, demoting the importance and the agency of people who ask questions.

  14. Some SE sites, for instance SO, get thousands and thousands of views by active members. So, for example, the highest voted answer on SO has over 33,000 votes. On these sites a very high number of votes over a sustained period may be the best indicator of the accuracy and helpfulness of an answer. Here, however, this is not the case.

  15. There are many types of question on the EL&U network. A prominent one is Single Word Requests. The current proposal would be very appropriate for a single-word-request site. However, this is not what EL&U is meant to be. The current proposal will not be supported by serious and dedicated contributors to EL&U (in terms of content), apart from answerers of single-word-requests and similar type questions. Serious SWR askers, who research their questions, such as Mari Lou A, and serious askers and respondents of more linguistics-based, etymology-based questions will not be in favour of this proposal. Indeed it will reduce their contributions. If you wish to see less from Sven, JL, Mari Lou, or you wish to see less future contributors like Stoney, etc then vote for this proposal. Also do this if you want to see more people misinformed.

  16. The voting system of SE and EL&U is a GOOD THING. However, it is not perfect for many reasons, including those detailed above. The 'accepted answer' feature provides a balance to this system. Just like a healthy democracy, where second chambers and the separation of the legislature, the judiciary and the police provide safety in the form of checks and balances, the same is true of the accepted answer feature in its current form. Don't vote to remove this! Keep true democracy alive!

You must log in to answer this question.

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