I often see complaints about highly voted answers (an indicator of community respect) appearing below an accepted answer which may have far fewer votes e.g. Shouldn't the answer with more votes be above the accepted answer?
I think the meaning of the accept check mark should remain as being:
This helped me [the question asker] the most
I think it works well for the question asker to have 15 points that they can use to reward the person that they feel helped them the most, and that the accepted answer is given the advantage of being read/considered first, which if it remains useful will often lead to more upvotes too. It's the same effect as why politicians want to appear first on a voting card.
I also think it works well that the asker can change their mind about the answer they accepted, and move the accept check mark (and the 15 points) to another answer.
However, what seems to be contentious is when an accepted answer with few votes (even net negative votes) remains above one or more answers with many votes in perpetuity.
As an answer to this question points out:
There are three aspects to this proposal:
- accepted-answer stickiness
- accepted-answer reputation
- accepted-answer labeling
Stickiness I would like to see the accepted answer lose its stickiness to the top slot at some set time period after the position of the acceptance check mark was last set. I'll use 12 months purely as an example, and perhaps this could be settable per-site in the moderator tools, but my inclination is to have the same term network-wide for a consistency.
Reputation I would like to see the 15 points remain with the poster of the answer that was last accepted i.e. no change.
Labeling I would like to see the green check mark turn to amber 12 months after it was accepted as an indication that it was accepted more than 12 months ago, and that the answers are now ordered by votes alone.
Green implies more of a go ahead on trusting this answer (because it helped the question asker the most, and they said so relatively recently) than amber which implies to only proceed with caution (because acceptance was a while ago and has not been reviewed recently by the person who originally gave it).
Under this proposal:
- The asker still gets the good feeling of awarding the Accept status and 15 points to the answer they like, and even to change their mind about which answer was best, forever (see below).
- For 12 months after the accept check mark was last set, even if the community thinks there is a better answer, the accepted answer still stays at the top to give it a voting advantage, but then the stickiness disappears, and ordering becomes by votes alone.
- The community can still see which answer was the last one that it's question asker set as accepted, because it still has a coloured check mark next to it, but its colour is now amber instead of green i.e. it retains its special status, forever.
- When the asker of the question leaves the site, with a poor accepted answer occupying the top slot, the community may resent seeing the poor answer at the top, and the extra time it takes them to scroll past (or read) such answers (and sometimes long comment trails as well!), before learning that there is one or more that the community thinks to be more useful beneath it, but they will at least know that it won't be there forever.
- If the question asker is still on the site after the accept check mark has turned to amber, then perhaps they can receive a notification of that happening, so that they can (with a single click) choose to re-instate the green check mark. This could be a great opportunity for them to reassess whether they still think the answer they check marked is the best answer for their question. Either way the 12 month clock starts again, with the 15 points following the acceptance check mark, just like it always has.
In other words, the asker owns their question and gets to say and reward the answer that helped them the most forever (if they are part of the site's community for that long!).
Remembering that self-answers do not get special ordering, I think a self-answer should only ever get an amber check mark (i.e. proceed with caution because "you do realize they gave it to themselves, don't you?").
As an aside, I toyed with the idea of adding red check marks for massively downvoted accepted answers but I think that would give signals which are too mixed - besides when I see an amber check mark I'm already going to proceed with caution and looking at the voting score would be the first thing I do.
I think this solution respects the community sentiment at Do accepted answers still serve a useful purpose?, is minimally disruptive to the look and feel of the Stack Exchange network, and saves the time of users and visitors who will not as often have to read or scroll past answers in order to find the one that the community thinks is best.
It also seems to be inline with something @Shog9 mentioned in an answer when declining Can we exempt downvoted accepted answers from getting the top spot?:
And some of our devs have suggested that simply time-limiting the pinning granted by accept (say, pinned for 90 days then sorted normally) would at least prevent it from being an eyesore forever.
Would this proposal work to keep the special place that the Accepted Answer has long held and allow it to have a positional voting advantage for a (potentially long) period as a reward for being accepted, and at the same time prevent an answer that is out of favour with the majority of the community from being stuck to the top forever?