3

September 19 Update

As of the deadline to report our preferences, this question has two answers in favor of unpinning and no answers opposed. Given that the overwhelming opinion is in favor of unpinning I have made our preference known by editing it into the answer on the main meta site.

It sounds like the change will happen in mid-October.


As discussed on the main meta site:

[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.

...

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.)

This would affect about 850 questions on our site. I'd recommend looking at some of the highest and lowest scored accepted answers from that query as well as some of the highest and lowest score differences. You can sort the query output by clicking on the column headings.

Would we like accepted answers un-pinned from the top on Webmasters? We have until September 19th to give our feedback.

8
  • 2
    If the accepted answer is unpinned I would like it made clear that there is an accepted answer and to include a convenient link to the accepted answer. I've posted a "feature-request" on the main meta site: meta.stackexchange.com/a/369839/149167
    – MrWhite
    Sep 14 at 17:45
  • 2
    I really want somebody to post a contrarian answer that says that accepted answers should remain pinned to the top. I expect everybody would down-vote that answer. Then I could accept it for irony. 😈 Sep 15 at 10:29
  • 2
    @MrWhite I had the same idea. The accepted answer is what the OP decided worked for them i.e. it answers their question. However, obviously, there is more than one way to do most things and another answer may be more popular. If the highest vote is bumped to the top, there should be some way to jump straight to the accepted answer.
    – Steve
    Sep 16 at 7:18
  • 1
    @MrWhite Clean, functional and unobtrusive. I love the idea of a "go to accepted answer" link, I voted you up over there. Sep 16 at 14:51
  • I find it strange that I have yet to see anyone suggesting an additional symbol on SEO Meta when the ostensible problem is that the accepted answer isn't always the most objectively relevant/true. Sep 20 at 1:26
  • 1
    Do we need to discuss this for the Webmasters Meta site as well? It is stated in the Meta SE question that "child metas are also considered a 'site'". However, "Ask Ubuntu (Meta)" appears to be the only meta site that has expressed a preference? There is a "controversial" answer that suggests "Keep the accepted answer pinned on meta sites" but a staff mod kind of disagrees, so it's difficult to know what will happen here?
    – MrWhite
    Sep 20 at 22:14
  • @mrwhite Our meta site gets used infrequently and we have to refer back to old questions infrequently as well. I'm not really sure that the sort order here is going to affect us enough to be noticable either way. You could create a new question here to ask about the meta site if you want but personally I'm not going to be upset no matter which way things get sorted on meta. Sep 20 at 23:33
  • If I switch the SEDE query that shows the 850 question on the main site to use the meta site instead, it shows only 10 question that would be effected on meta. All but two of those have a score difference of 1 or 2. The question with the biggest score difference of 6 is webmasters.meta.stackexchange.com/questions/387 Sep 21 at 2:14
7

I think we should unpin accepted answers. An accepted answer just means the answer was useful to the question-asker. It doesn't mean it is the best answer or the answer which will more widely apply to the question. Upvotes tend to better represent both.

6

I also believe we should unpin accepted answers.

The list of affected questions you posted is really nifty! I encourage onlookers to click through and click the "score difference" column twice to sort it by the questions with the highest score difference between the accepted answer and the highest voted answer. It's easy to see that a lot of these questions would benefit significantly from having their accepted answer unpinned.

The question with the highest difference is this one, where a 136-point answer about CloudFlare written by CloudFlare's co-founder and CEO is buried under a 26-point accepted answer that has a 15-point comment saying that it doesn't really answer the question. I think that post is the clearest example that accepted answer pinning can be harmful to the site's readability.

9
  • 1
    There are also cases of down-voted accepted answers where an up-voted answer would take the top slot. Like Do search engines penalise 'Home' links and/or buttons? Sep 14 at 16:59
  • There are perhaps a few edge case(s) where the accepted answer is possibly the "better" answer, however, I think the vast majority would benefit from having the accepted answer unpinned.
    – MrWhite
    Sep 14 at 17:33
  • Yes, if we look at the question with the biggest disparity, webmasters.stackexchange.com/questions/88659/…, the highest vote is for an answer from an impeccable source, the CEO of cloudflare. Perhaps that is why it received so many votes, but even taking that "bias" out, being the definitive answer it deserves top spot and is a good case study for why unpinning is a good idea. But I still think that an indication that there is an accepted answer is useful.
    – Steve
    Sep 16 at 7:22
  • @MrWhite the edge case you found is a particularly weird one because the non-accepted, higher-voted answer is a self-answer. Sep 16 at 15:08
  • Here is a case where the accepted answer is way better than the highest voted: What is link baiting. The accepted-answer (score 2) gives a reasonable definition while the highest-voted answer (score 14) gives a definition for for another term entirely (click bait). Could be in part because the author of the accepted answer was unkind in the comments (which I just cleaned up.) Sep 16 at 15:14
  • @StephenOstermiller That's an interesting one, it may also partly stem from the general disagreement over what Linkbaiting means and how it's different from Clickbaiting. If you view the Wiktionary entry on Linkbait, one of the definitions is "Provocative headlines designed specifically to persuade people to click or share by using loaded terminology (e.g. incredible, secret, shocking, unbelievable), posing provocative questions, or tying themselves in to hot topics", and Wiktionary lists Clickbait as a synonym. You and I most definitely disagree with that - maybe it's time to edit Wiktionary? Sep 16 at 15:46
  • 1
    "general disagreement over what Linkbaiting means and how it's different from Clickbaiting" - this isn't helped by the fact that Wikipedia now redirects "linkbait" (and "link bait") - which is actually linked to in that highly voted "click bait" answer - to "Clickbait"! And "linkbaiting" just redirects to "link building" (which I suppose it is) but includes no mention of "linkbaiting" specifically. @StephenOstermiller
    – MrWhite
    Sep 17 at 15:46
  • 2
    @MrWhite I just changed Linkbait and Link bait to redirect to Link Building on Wikipedia instead of Clickbait. I also removed the erroneous definition from the Wiktionary page. Now someone just needs to write a mention of the term Linkbait in the Link Building Wikipedia article. Sep 17 at 16:38
  • I will be very curious to see how this affects our indexation (as well as other sites). Sep 19 at 23:51

You must log in to answer this question.

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