What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 128 Stack Exchange communities.

I see the Accept rate as being counter productive to the site. I (incorrectly) answered a question about whether the accept rate should be shown on the Mobile version of the Site as that I was glad it wasn't because it's not very helpful. -- It was the wrong place to suggest this. However, it was this from the OP that really sort of annoyed me:

To me the accept rate immediately indicates whether i should answer the question or not even think about it. It would probably be a waste of time to answer a question of somebody who doesn't care about voting, accepting or even reading your answer.

To me, this is the exact opposite of what these sites are about. I don't think it should matter whether or why the user has a low accept rate. The negatives of the feature I see are:

  • (as above) Where users are deciding not to help other users out
  • Continue rubbish/noise comments about Accept-rate values
  • A general negative view or feel both for the site and the user

I get what it is trying to do, but couldn't this be done just for the user. Only visible on the user's profile page or displayed at the top like earning a badge etc That way it targets the user (the only one who can improve the situation) without the negative things I listed above...

thoughts?

share|improve this question

5 Answers 5

up vote 21 down vote accepted

I think people have a right to know whether or not to expect the rep and badges that come from accepted answers, as those are things that motivate people to answer questions on Stack Exchange.

I think the current method -- showing a percentage -- is somewhat counterproductive. There are posts here on meta about people getting bothered even with percentages above 65%.

My suggestion would be something along the lines of

  • Very High: 75%+
  • High: 50%+
  • Low: 25%+
  • Very Low: below 25%

which still lets people decide whether or not to answer if they care about reputation and badges, while giving positive meaning to any accept rate over 50%.

If you're not getting enough good answers to get your accept rate over 50%, you're either asking bad questions or questions that aren't a good fit for this site -- either off topic or too localized obscure.

share|improve this answer
2  
Going the Slashdot 'karma' route might not be a bad idea (which is what you're essentially proposing here), however many cases that I've seen aren't too localized. They are hard to answer, asked in low volume tags, or both. I usually can't comply with the delete requests due to the number of votes the question and subsequent answers received. –  Tim Post Aug 30 '11 at 4:48
    
That sounds like "too localized" to me -- if the tag is low volume, it's not of broad interest to SO, and if it's too hard to answer, others haven't commonly solved it, so it's unlikely a solution will help many people. If you have a lot of "too localized" questions, they probably belong in IRC / on a mailing list / emailed to the author of the software / etc. (Unless it's Facebook related in which case that place is here.) If you just ask a couple of questions before moving on, it shouldn't push your accept rate below 50% (partially thanks to the threshold before any accept rate is shown). –  agf Aug 30 '11 at 4:55
9  
You might be confusing obscure with too localized. Obscure questions will be of some use to someone in the future. Questions that are too localized will likely be of no use to anyone in the future. E.g. "Why is that blue car parked outside of my house?". If you find yourself in the unfortunate position of having to deal with Lotus Notes, there's nothing wrong with asking for help. –  Tim Post Aug 30 '11 at 5:09
    
I see the distinction, and I did really mean "obscure." However, if a question is too obscure to get a good answer on Stack Exchange, it's not particularly useful to ask it here, even if it's not "wrong." –  agf Aug 30 '11 at 5:17
    
I like this suggestion, but I would actually suggest changing Low to Medium and Very Low to Low, so that the categories are Very High, High, Medium, and Low. I wouldn't necessarily call an accept rate of 40% low. –  Chris Frederick Sep 21 '11 at 19:43
    
@Chris - There are exceptions, but most of the people I see with an accept rate of 40% are people who maybe accepted a bunch of answers once when someone prompted them, and then stopped again. An example of an exception: stackoverflow.com/users/756566/andrew-alexander, who I think just has a very high standard for accepting an answer. –  agf Sep 21 '11 at 23:34

I don't know:

On one hand accept rate bothers me a lot, it is a one-dimentional metric that does not really give me enough information about how upstanding a user is in the community. It is used frequently to bully users and leaves a horrible taste.

On the other hand, introducing accept rates heavily increased the amount of accepted answers in the system. Which in turn helps close loops and motivate the community.


When we introduced the accept rate stuff we were careful not to add too much of a value statement. On hover we do not say "horrible user, never accepts answers".

Going forward I would much prefer that we retired accept rate in favor of a more general metric that covers a variety of "citizenship" metrics that do not result in rep.

  • Does the user vote?
  • Does the user accept answers?
  • Does the user answer questions?
  • Does the user edit or suggest edits on questions?
  • Does the user flag stuff?

I don't know, perhaps we should go the other way and show a tagline of honor for the top N percent of users.


That said, I do not think it would be beneficial to simply retire accept rate if there is no replacement offered.

share|improve this answer
1  
I think rewarding 'good citizens' with something prominent in their user card would be a great step. Still, people really do care about the frequency that someone accepts answers and they're going to look for that .. making them work harder to get that metric just seems counter productive. –  Tim Post Aug 30 '11 at 6:01
3  
Where users are falling short in any of these citizenship metrics could the site itself not educate them via automated messages rather than leaving it to individual users to complain about it (often rudely) in the comments. –  Martin Smith Aug 30 '11 at 8:53
1  
@Martin hence the concept of possibly awarding awesome behaviour with a badge of honour, as opposed to a badge of shame –  waffles Aug 30 '11 at 9:16
1  
@waffles - That relies on users knowing (and caring about) badges. I was really thinking about those users with minimal participation that have never upvoted or accepted an answer. Perhaps because they are unaware of these features. –  Martin Smith Aug 30 '11 at 9:22
    
@waffles, I think those are some pretty sensible points...and it's a good reminder about the positives of the feature (it had to exist for some reason, which it seems has bee effective)... –  davidsleeps Aug 30 '11 at 9:26
    
@Martin Medal of Honor! –  davidsleeps Aug 30 '11 at 9:26

My only problem with accept rate is the pressure that it puts on people to take actions that aren't really appropriate. I've seen dozens of flags from users asking for perfectly good questions that have not received acceptable answers to be removed, just because the user wants to increase their accept rate but can't select an acceptable answer on several questions.

I also think some people are a bit too harsh with what is considered an acceptable rate. 60% or more from a user who joined just a month ago (to me) is perfectly acceptable. There is absolutely nothing wrong with leaving your proverbial line cast in the water a little longer in hopes that a new and better answer may emerge. Sure, the top answer might seem awesome to you and I, but perhaps the question author is looking for a more in depth explanation, example code that matches their problem better, or something else.

Still, hiding that value from everyone but the OP is not going to solve the problem you describe, it's just too easy to look at the user's profile and see for yourself. I remove all 'accept rate noise' on sight. The kind of user that we want asking lots of questions will hear crickets chirping as tumbleweed blows by and wonder if that orange number under their name has anything to do with it.

I can't support an idea that puts more work on users that want to answer questions. It's an important metric to many people, I'd rather that folks spend time on questions they feel good about rather than researching a metric. We encourage people to answer whatever interests them, but some people are reluctant to do that when certain rewards seem unlikely. There's nothing we can do about human nature.

Edit

If you think you've spotted a help vampire, flag one of their posts as 'other' and let us know with every detail you can provide. We will look into it. We might not always agree, but we will look into it. We want to keep leeches out, but we also have to be friendly to new users at least to a point (even with guard rails in place). Stack Overflow is quite unique to someone coming from a forum, and that is exactly the kind of user we hope will eventually become a constructive member of our community. At least until the point that it's obvious that they just don't, or can't 'get it'.

share|improve this answer
    
can't select an acceptable answer on several questions - that is the point Tim. We want to upvote helpfull info but not be forced to mark them if they aren't full answers or didn't resolve the problem. But as soon as someone upvote, seems that the script starts to see them as actuall answer and want that the user accept one of them. I have resorted to post placeholder answer, but I don't feel this is the solution. See my answer below on this same question, I would like to hear your opinion on it. –  SPArchaeologist Dec 24 '12 at 11:28
    
sorry - wrong page, my answer is not on this question. I was refering to this: link. –  SPArchaeologist Dec 24 '12 at 11:41

The negatives of the feature I see are:

  • (as above) Where users are deciding not to help other users out
  • Continue rubbish/noise comments about Accept-rate values
  • A general negative view or feel both for the site and the user

This is the core of your argument against showing accept rates. So consider this.

A low accept rate (defined as ~25% or lower) is indicative of one of the following:

  • A user who does not know how to use the site.
  • A user who does know how to use the site, but is either neglectful or malicious.
  • A user who has not gotten good answers, which may be indicative of:
    • Bad questions (bad questions can lead to bad answers)
    • Lack of user interest in these questions.

A user who does not know how to use the site is a user who needs to be educated. Informing them of how accepting questions works is neither "rubbish" nor "noise comments". It is vital information that the user in question may have missed. Informing them of this creates a positive view of the site for that user (assuming the information was appropriately phrased. IE: not condescending or coming off like rep trolling). And some people don't want to waste their time with users who couldn't be bothered to read the FAQ; that's their right.

A user who is using the site neglectfully or maliciously is not a user who needs to be here. So I have no problem with seeing their questions downvoted and/or filled with "rubbish/noise comments about Accept-rate values" and/or not being helped. Stack Exchange requires give and take. There are responsibilities placed on the person doing the asking just as there are on the answerers. Someone who isn't pulling their weight is a drag on the site.

A user who has not gotten good answers due to asking bad questions is someone who needs to start learning what good questions are. Bad questions contribute "rubbish/noise" questions to the site, which is far more damaging than someone posting a comment on a question about accept rate. The whole point of the site is questions, after all; comments are just a supplement. And if the user who asks bad questions has a negative view of the site, then hopefully that will help them improve or leave the site. Either way, the site gets better.

A user who is asking questions that nobody either knows how to answer or wants to answer is generally unlikely to fall afoul of a low accept rate. The reason being that unanswered questions don't count towards accept rate. In low-traffic tags/subjects, you are more likely than not to have a few experts drop in to comb through them for easy accepts/up-votes due to being experts in a specialized field. In low-traffic tags, you get a disproportionately higher rate of answerers who know what they're talking about.

That being said, there will be some idiots who think they know what they're talking about and don't. It happens. But does it really happen 3 out of every 4 answered questions? I'd want to see some statistics on that before I can be sure.

So no, I don't see a problem per-se with providing accept rate information to users.


This all being said, I do wish that people would only confront someone about their accept rate if it is actually low (no, 65% isn't low). And they've looked through the question history and have seen something disconcerting (ie: good answers not being accepted). Most people who have low accept rates don't have that many answered questions, so it would only take a minute or two to look through their history.

However, the fact that some users are jerks about accept rate does not mean that having the accept rate itself is a problem. It simply means that some users need to be educated about what bad accept rates are.

The absolute most I would accept would be something where only an accept rate beneath a certain threshold (say, 40%) is shown. That way, you deal with jerks talking about people who are in good standing accept rate-wise.

share|improve this answer
    
@davidsleeps: Because informing someone who doesn't even know what "accept" means that their accept rate is low is useless. They won't even notice the percentage indicator. That's why talking to them directly via comments works: because it's one person informing someone (likely with a link) of something they don't know. And without showing accept rates, you don't have a way of quickly spotting that someone needs to be informed or that someone may be a neglectful/malicious user. –  Nicol Bolas Jan 17 '12 at 22:29
    
@davidsleeps: To put it another way, how would your suggestion allow people to effectively deal with neglectful/malicious askers on the site? –  Nicol Bolas Jan 17 '12 at 22:30
1  
if a user asks a lot of questions (lets assume good, because bad questions are closed or edited) and doesn't accept any answesr, I fail to see where the problem is. Future users with the same problem that have arrived at the site via google (etc) are interested in answers...and voting provides the best way of indicating what is a useful answer... –  davidsleeps Jan 17 '12 at 22:38
    
@davidsleeps: That you do not see a problem does not mean that the problem does not exist. By your logic, we shouldn't even have accepted answers. If a user not accepting answers is fine, then what are accepted answers for at all? Acceptance has a purpose: it shows that the problem has been resolved to the satisfaction of the person asking a question by that particular answer. An asker who refuses to accept legitimately acceptable answers is a bad actor who is damaging the site by leaving answers unaccepted. –  Nicol Bolas Jan 17 '12 at 22:46
    
that's a good point, and I think having accepted answers is necessary, which does help me appreciate your point. maybe accept rate is only visible to users of a certain rep who should have the best idea in approaching those users (as well as known to yourself) so that it doesn't (imho) add noise to the vast majority of visitors to the site. –  davidsleeps Jan 17 '12 at 23:03
    
You are missing one very important possible raeason for low accept rate in your list: "Good, well thought and prepared - therefore difficult - questions." –  gorn Jul 22 '12 at 18:26

Even if you hide it, nothing will prevent someone using SEDE to calculate it1. In fact I guarantee that in the 10 minutes following this functionality's removal, someone will ask exactly that here on meta :)

1. I haven't checked, but there's a good chance it's there already!

share|improve this answer
4  
SEDE is quite a bit behind the normal cached accept rate value, I don't think someone would put much weight into that when simply viewing a user's history would suffice. –  Tim Post Aug 30 '11 at 5:23
3  
That requires a level of effort... –  davidsleeps Aug 30 '11 at 5:46

You must log in to answer this question.

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