10

Recently, @Toscho gave me a gentle nudge to this SO.meta discussion regarding commenting on OP's accept rates. While the accepted answer to that question was "no", the competing answer of "yes" was not especially far behind in overall votes, so I don't view the accepted answer as representing a consensus of the community. Further, what may be appropriate - or even consensus - for SO may not be appropriate for other network sites. So, I thought it would make for a worthwhile discussion topic for WPSE.meta.

So: let us discuss among ourselves!

| |
  • 2
    Perhaps this is a separate discussion, but it applies to the question/OP at hand: what do we do with OPs who leave comments indicating "that worked!", fail to accept the answer, are told via comment to accept the answer, still fail to accept it - and then keep posting questions for which they persist in failing to accept correct answers? – Chip Bennett May 1 '12 at 15:34
  • Just live with it? If an answer is correct and the question not too localized … everybody wins anyway. – fuxia♦ May 1 '12 at 20:28
  • 1
    So, the acceptable response to users who habitually refuse to adhere to site/network mechanics is, "just live with it"? I have to ask: what, then, is the point of having mechanics/conventions/rules? – Chip Bennett May 1 '12 at 21:52
  • That's not what I said. But we can live with good answers even if they are not accepted answers. – fuxia♦ May 1 '12 at 22:03
  • Chip & @toscho, wouldn't this suggestion help to minimize the "Accept Rate" & "Abandoned Questions" issues? – brasofilo May 30 '12 at 22:40
6

On the WPSE question referenced, the OP had (and still has) a zero percent accept rate, out of ten questions asked.

(Edited to add: of those ten questions, multiple questions have answers that the OP has indicated, via comment, to be correct, and to which WPSE users have replied with further comments instructing the OP to accept that answer - and which the OP still refuses to accept.)

Currently, WPSE has several such users. The problem would be more isolated if the site were more mature, but as it is, a small number of such users has a disproportionate impact on the overall quality of the site. To me, these numbers are indicative of:

  1. A user not caring enough to understand site mechanics to accept answers, or,
  2. A user who asks so many poor-quality questions that cannot be answered reasonably

Clearly, accept rate is intended to be used for some purpose, or else it wouldn't have a place of such prominence in the UI.

Taking the course of simply ignoring such a user doesn't help anyone. The user is potentially frustrated that none of his questions get answered. Community members who make a good-faith effort to answer his questions are frustrated when their correct answers are not accepted. The overall quality of Questions and Answers on the WPSE site is diluted by unanswered and/or poor-quality questions.

Thus, the only other course of action is for the community to attempt to encourage/instruct users in proper site mechanics, including accepting answers and asking high-quality questions. One means of such encouragement/instruction is commenting on users' accept rates.

Note: I generally only leave such a comment if the user has asked at least 10 questions.

Also, I generally try to follow up such a comment by perusing the user's previous questions, to see if any can be improved, or if any can be easily answered. Quite often, such perusal reveals several questions that are off-topic, not constructive, too localized, or not a real question. Occasionally, a good question will remain unanswered. I do my best to answer such questions, and then if/when the OP engages with the answer, I leave a comment regarding how to accept an answer if the OP considers it to be correct.

| |
  • Well said I'm with you on this 100% – Bainternet May 1 '12 at 14:03
  • Very well said! – Adam May 1 '12 at 16:33
  • I agree 100%, I will continue to do so if it is clear the user has a history of not accepting good answers, regardless of the logic here meta.stackexchange.com/questions/23321/…. – Wyck May 1 '12 at 21:26
4

I know, the boilerplate is hot…

I´ve already been through this some month ago. After talking to our kind & polite mods, I was told to not do this … and I did it anyway just to know what will happen.

…and touching it hurts!

So I saddled my ~┌┐⌐ and went out to chase them down, those ignorant bastards. I left some comments on user questions to »fix« their accept rate as they had ~10+ open questions with different sort of "solutions": »fixed it«, "nvm, error is gone", "blabla", and worse. Here´s what happend: Those who reacted did it in the most aweful way, that you can imagine. They marked a solution on every question!. Even if it wasn´t fixed and even if there was only a part of a solution. First I didn´t get it, but then there was one user who marked on of my answers as solution that was definitively not the solution.

"Sorry/Ludo/Mensch ärgere Dich nicht"

Do you know the game "Sorry/Ludo/Mensch ärgere Dich nicht"? If yes, then you know where nagging about the accept %-rate brought me:

…and it hurts even years later!

Here comes the really sad thing (in short): If you´re encouraging users to "work" on their accept rate, then we´ll end up with questions that got the wrong answers marked as solution. And fixing this is close to impossible without reading ~20k questions incl. approximately 80k answers.

What do we learn from it?

Don´t tell them to "work on the accept rate". Never ever do this.

Sidenote

But we should maybe write something somewhere up, that we can link in case, which tells those users how the site works and when & how to start to accept answers as solution.

| |
3

While I totally understand the urge to nag about low accept rates … I see some problems:

  • It is noise. We have already developed a habit of creating overly long comment discussions. Reading through these comments to get facts about the question should be as easy as possible.
  • A good (on topic) question deserves a good answer. I answer questions sometimes which have an accepted answer already – we are here to share knowledge after all. A low probability to get the ✔ mark should not stop anyone to provide a useful answer.
  • Unnecessary comments may break the garbage collector. Use comments to make the question better, not the asker.
| |
  • 1
    But, in our case, all too often the questions themselves are noise, a non-trivial amount of such questions aren't actually high-quality, and making questions better without engagement from the OP is often impossible. – Chip Bennett May 1 '12 at 14:10
  • 2
    There are many better ways to improve (or remove) a question than nagging about accept rates. – fuxia♦ May 1 '12 at 14:11
  • 1
    My issue is with the presumption of "nagging." The manner in which an OP is informed/encouraged regarding answer acceptance may or may not constitute "nagging." I agree that, in all cases, community members should remain polite and respectful. I do not, however, believe that politeness and respect are mutually exclusive from encouraging/informing users about answer acceptance. – Chip Bennett May 1 '12 at 14:22
  • 2
    @ChipBennett I guess my problem with this practice as it currently exists is that it publicly reprimands a user (generally by a non-moderator, which may seem more cruel and self-serving), calling out the "offender" while implying that marking an answer is and should be linked with others answering future questions (which I feel it shouldn't). Maybe educating the user on this should be left to SE itself (the system more gently prompting the user when they create a new question to mark past answers if their score is below X or something) but not made publicly. Just my 2 cents on the matter. – William May 8 '12 at 19:52
1

I have a lot of sympathy for Chip's answer.

First of all - users shouldn't refrain from answering questions from those with low accept rates. The answer is to the community, not just the OP. However, this is an ideal - and I suspect one that is not adhered to. This means low-accept rate users have a negative impact on the community. Regardless, users who do not accept an answer pollute the 'unanswered' pile.

The occasional gentle and polite reminder on the other hand can have a positive effect. Not only can it prompt a user to clean up their answered questions, and those in the future - it also makes others aware of the mechanism to accept answers (an unawareness which I suspects accounts for a large proportion of 0%s).

I accept that such comments do not specifically address the question, and as the discussion on SO Meta points out, are technically 'off topic'. However, I do not feel that the comments (I've seen) constitute nagging or in anyway abrasive (and if it is, flag it).

If we take the stance that such comments should be flagged (and removed) because they are off topic then this presumably applies to other such comments: So any comment that addresses how a user interacts with the community (e.g. prompting to post solutions as an answer, not in the question/comments) should also be avoided. While moderators have the power to fix such things, this presents its own problems:

  • (Unnecessarily) takes up the moderator's time
  • A comment can encourage participation (and not just from the OP).

Personally, it seems that the bigger issue facing WPSE is large proportion of 'unanswered' questions (answered or otherwise) than a flood of comments prompting answers to be accepted. But to be clear, I'm not advocating a witch hunt - but the odd (very occasional) comment could help the community a lot.

Chip also mentioned poor quality questions. I don't think that's particularly relevant here. If the low accept rate is the result of poor answers to poor or off-topic questions - then the issue is probably more of posting 'on-topic' questions.

| |
  • What I don't understand is the assertion that using the comment system to carry out the site's conventions and mechanics is somehow - suddenly - "off topic". How many times have we been told in other meta discussions that we are supposed to use questions to instruct community members regarding site mechanics? But now, when we do, we're told that such comments are off-topic. Well: which is it? – Chip Bennett May 1 '12 at 15:15
  • 2
    Also: I do think that poor-quality questions are on topic for this question. IMX, there is a strong correlation between accept rate and quality of questions asked. Encouraging such OPs to improve their accept rate often leads to a response of, "but I don't have any good answers." I believe that every good question can be answered, and that unanswered questions are far more likely to be a function of question quality, rather than answer quality. And if we fix one (question quality), we are more likely to fix the other (accept rate). – Chip Bennett May 1 '12 at 15:18
  • This is my point - I think comments should be about interacting with both user and question (within reason). And site mechanics , I would include in that. Regarding your second point, I would consider 'I don't have any good answers' a valid reason not to accept an answer. The action here would be to improve the question / ask for clarification / flag the question. – Stephen Harris May 1 '12 at 15:21
  • "The action here would be to improve the question / ask for clarification / flag the question." - I agree completely, and always try to do both. My main point is that this is a both/and proposition, not either/or. – Chip Bennett May 1 '12 at 15:32
1

Here are some ideas I came up with in another thread. They definitely need some discussion and refinement, but it's a starting point:

  • Communicate more clearly to new users when they register and ask their first few questions. Make it clear that accepting answers is an integral part of the process and something they should be doing.
  • When a question has at least one upvoted answer that hasn't been accepted after 3 or 4 days, send a nag email to the OP. Let them know that there are N upvoted answers and that they're expected to accept one if it solves their problem.
  • Show a warning on questions where the asker has low accept rate. The goal isn't so much to discourage people from answering, but to encourage the user to improve their rate so that it will be removed. Kind of like a scarlet letter ;)
  • If a user has asked ~4 questions without accepting an answer to any of them, then the system should prevent the user from asking new questions until they accept answers on at least ~2 of them.
  • Give moderators the ability to mark an answer as accepted on questions that have been abandoned. Maybe setup a process where users can submit questions for review.
  • Give moderators the ability to mark a question as "abandoned," which would remove some reputation points from the asker. If the asker gets 2 questions marked abandoned, they'd be prevented from asking any new questions until they accept answers to the abandoned questions.
  • Give a badge to users who answer a question with ~4+ upvotes on an abandoned question.
  • Add a private messaging component, so that users can send the asker a note reminding them that they should accept good answers. This would probably be more effective than automated nag emails. Users should get an e-mail notifying them when they have a new message. And of course this would also have added benefits outside the scope of this issue.
| |

You must log in to answer this question.

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