Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 158 Stack Exchange communities.

What is meta?
Here's how it works:
  1. Any Stack Exchange user can ask a question
  2. The community provides support, votes on ideas, and reports bugs
  3. Your voice helps shape the way Stack Exchange operates

There has been debate about whether a popular user (let's call him "Jon") might receive more votes for a similar, or even slightly worse answer than one from an unpopular/unknown user (let's call him "Charles"). For example, if Charles posts a beautiful answer and Jon posts a good answer to the same question at the same time, Jon's popularity on the site may mean that his answer gets more votes and eventually is accepted as the best answer, while Charles's answer may receive fewer votes and gets lost in the mix.

Although this is a nice benefit for Jon (and arguably a well-deserved benefit for being so super-dee-duper in the past), it is actually bad for the site as an answer resource, because the best answer is not highlighted for future users viewing the question.

Thus, I propose that all answers and questions remain anonymous until you cast your vote on them. This would be a relatively simple change (hopefully), but it would eliminate the potential of skewed votes based on user-bias entirely (unless of course you place a vote, see the user, and change your vote accordingly). Also, this would be an additional encouragement for users to vote, because voting would then have the additional benefit of revealing the poster.

Note I: I still think you should be able to browse all questions/answers from Charles's and Jon's profile, but when you get to their question/answer, it will display as normal (anonymous until you cast a vote).

Note II: Some say that this is a non-issue because users vote based on the quality of the answer rather than who posted it but that is not always the case. Note that I'm not saying the answer I linked to is not the current best one to the question, just that it received an exorbitant amount of votes purely because of who posted it, not because of its content. Also, if someone had posted a better answer to the general questions asked in the post, there is virtually no chance that it would have even come close to being the accepted answer.


Well, I really got pounded on this question. For the record, I am not one who believes that high-rep users are not high-rep for a reason. I know they are where they are for good reason. I just know that in my experience I am much more wary of a low-rep user's answer than a tried-and-true high-rep user's answer. Thus, at least in my experience user rep definitely affects my vote on questions and it is sometimes hard to separate the content from this user-bias. Thus, this is just an attempt at making the system more fair so that answer content is always king.

share|improve this question
The Carmack answer doesn't illustrate your point at all. It is a great answer, plus it is undeniably the most authoritative answer possible to that question. Show examples where a high-rep user received a disproportionate number of upvotes for an answer that was clearly inferior if you want this to be taken seriously. – Bill the Lizard Aug 14 '12 at 18:02
IMO the linked answer is the exception rather than the rule. And I'm not sure it's a big deal anyway. You might also be surprised at what ends up being the accepted answer (which can change over time, too)--it's not always the best answer, it's not always the answer with the highest votes, it's not even always correct. – Dave Newton Aug 14 '12 at 18:02
+1, I agree and I'm very likely to be moderated and/or downvoted for this, but IMHO it is actually true that OP says. – H2CO3 Aug 14 '12 at 18:09
@H2CO3 "I'm very likely to be moderated and/or downvoted for this" ... You do realize that comments can't be downvoted right? And expressing agreement is not something a moderator would get involved in at all, simply because there is no need to. – Bart Aug 14 '12 at 18:15
@BilltheLizard: I agree that John's answer was the most authoritative answer to what John Carmack meant. However, there are 3 questions asked in the post, and none of them are "What did John Carmack mean?". Thus, I think it fully possible for someone other than John to give a better answer to the questions, especially the last two. – Briguy37 Aug 14 '12 at 18:18
@Bart I do realize, but I have posts (answers and a question) that can be downvoted. And I have experienced that people downvoted my post for irrelevant reasons (incl. unsympathy). Also, this is a post with very low score - supporting it might cause the impression that I'm intentionally being rude or destructive against the community thus eligible for being moderated away. – H2CO3 Aug 14 '12 at 18:18
Where are those better answers? That link still doesn't support the idea that people get more upvotes based on their name alone. It's the best answer on the page, so it deserves to be at the top. Where are the links to questions where answers by high-rep users are ranked above clearly superior answers posted by low-rep users? – Bill the Lizard Aug 14 '12 at 18:23
@BilltheLizard: How is 919 votes and counting not proof that WHO answers the question can play a big role in the number of votes the answer gets? Any other Joe and that answer would have been lucky to get 20 votes. That is all I wanted to display by posting that link, and truly I wish I hadn't for all the distraction it has caused. – Briguy37 Aug 14 '12 at 18:33
It's not proof because there's not a better answer that should have scored higher than that one. It just doesn't illustrate the point you're trying to make. – Bill the Lizard Aug 14 '12 at 18:37
There was a link to that question from Slashdot or Reddit or someplace high-profile; that's what caused it to really take off. – Ernest Friedman-Hill Aug 14 '12 at 18:45

I think there might be a case to do some A/B testing in order to decide whether this phenomenon really exists, or put the canard to rest entirely.

I don't think John Carmack being voted highly on a question that was essentially "What did John Carmack mean by this?" is really a good supporting example, however.


Hypothesis: If there is a 'Jon Skeet' effect, it will manifest itself quantitively in the data. People asking questions are just looking for a good answer, and it is the community at large biased toward certain posters; therefore this problem is most likely to manifest itself as Jon Skeet having a higher rate of non-accepted answers scored higher than the accepted answer than the average user.


  1. Experimental Baseline
  2. Skeet-Only Results

Unfortunately, I wasn't able to get a baseline rate for all users without hitting the SEDE's timeout for that particular query, so I ended up with a very small sample.

The rate for Jon Skeet overall was 11.6% or 2470/21230 total answers. The rate overall for the admittedly bad sample of 10000 questions was 315/10000 or 3%.

Conclusion: This result seems to suggest that saying "There is no Jon Skeet effect" straight out would be naive. Further research is required in order to get a more accurate baseline, and eliminate potential confounding factors (such as high quality answers contributed long after the OP forgot about them).

share|improve this answer
I know what you're thinking, it's 'cuz Carmack should know best, right? Hogwash, I say. The internet will decide what Carmack thinks. – Dave Newton Aug 14 '12 at 18:03
Excellent work sir. – Ernest Friedman-Hill Aug 14 '12 at 19:40

There's absolutely no point in doing this, as you can undo and even reverse votes up to five minutes after casting them.

share|improve this answer
I have much more faith in the community than this. I believe there may be a few users who vote to see who made the post and then immediately unvote, but I think this would be a very small minority of users. Smaller still would be the percentage of users who change their votes once they see who the poster was. – Briguy37 Aug 14 '12 at 18:25
I'm not saying anyone would change his mind after seeing who posted, but I believe the vast majority would vote → unvote → read. I would. – Dennis Aug 14 '12 at 18:31
Yes, I probably would sometimes too. However, I definitely wouldn't do it for all answers on every page I visit, and I'd only do it for questions/answers I was already interested/curious about, and if I was interested/curious enough, it probably deserved the up/downvote anyway, and so I'd probably keep the vote in there for most cases. But you are right, there will definitely be the case where users are just curious and don't want to vote, and if they want to do that it is really up to them. – Briguy37 Aug 14 '12 at 18:43

The John Carmack post you linked to is obviously a special and rare case -- and of course, the question is about one of John's own tweets, so that's the best reason of all for his answer to be the best.

But I think in the general case, for every user who votes up because they recognize a poster, there's another petulant user who votes down -- or at least won't vote for those popular posters, because they feel like sticking up for the underdog.

share|improve this answer

This is based on the misunderstanding that high rep users got all their votes for being high rep users. Wonder how they got started?

Could it be that they are high rep users for getting many votes on their huge number of good answers? Or that even an average number of upvotes adds up to a lot of rep over 20k answers?

I can live with the fact that people having written 10-20 times more answers actually have 10-20 times the rep I have. ;-)

Hiding the poster can also give less votes to new users. I often give an extra upvote for a nice answer from a user with, say, 89 points.

share|improve this answer
You also need to factor in experience posting answers on SO, knowing how to properly format answers, and the standards the community holds, etc. – Servy Aug 14 '12 at 19:10

I can see where you're coming from. From my own testing with a sock puppet, I can confirm that contributions from high-rep users are given much less scrutiny than those from newbies.

However, in the same experiments, I couldn't find a serious skew towards high-rep answerers. What counts is mostly still the answer's quality.

The Carmack example is, as others have already pointed out, not really a good one because holy crap, it's John Carmack! :)

In conclusion: it's an interesting idea and I wouldn't be opposed to it! However, the problem it would serve to fix is (IMO) not a grave one. It would be a big change, and looking at the community feedback, it would have a hard time being implemented.

share|improve this answer

This is a fine idea (and it is too bad you got hammered for it).
I feel the need to post this to encourage a little perspective change to appreciate it.

  • All of the other answers here do a very good job of questioning the reasoning behind hiding user names.
  • None of them question the reasoning behind showing user names.

There is no good reason the burden of proof should lie with those who think user names should be hidden. Yes, since the status quo is showing them and changing to hiding them before voting would take more work, it makes sense that people want to see a benefit before taking on the cost of changing, but there is no logical reason to default to showing them, and you did well to point that out.

Actually, you did more than that. You painted a plausible scenario of harm coming from showing user names (and Pekka even provided evidence to support it), while I have not read a single argument for showing user names other than the exact opposite of yours (the "for" argument being that showing user names allows biasing votes toward newer users).

Rather than just assuming the biases cancel each other out, and rather, to build on Pekka's answer, than being content with "What counts is mostly still the answer's quality.", why not make a move toward answer quality being the only thing that counts?

(The fact that votes can be reversed does not kill this idea, by the way. Not everyone is that fickle (or has that much extra time on their hands).)

share|improve this answer

TBH, I think having a lot of rep goes both ways:

  • If I'm looking for an answer to something, I'm more likely to trust an expert with high rep than a new user with no rep.

  • Conversely, If I'm judging two answers, I'm much more likely to upvote the lower rep user to help him/her out.

Indeed popularity does have some effect. But it's definitely not the entire factor.

share|improve this answer

You must log in to answer this question.

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