What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 122 Stack Exchange communities.

I was wondering whether SO proved any of the claims used when it was built? Claims and mechanisms such as

  • How the community will do things right and the law of large numbers will select the good solutions (~)?
  • The property of no-registration required causing OPs to never find back to their questions and no one accepts it.
  • Do the cost associations of given and received downvotes achieve the desired effect?
  • Doesn't upvotes draw attention of those who simply want the badge for +300 votes and vote for answers without even thinking about it?
  • Were the SO mechanisms created based on any scientific (human, phsychological, behavioral, etc.) research?
  • etc. you name it.

(PS: Maybe a dupe, maybe too early to ask.)

share|improve this question
    
why the bounty? –  Kyle Cronin Jul 17 '09 at 7:11
    
+1 Reciprocated! –  Marko Carter Aug 5 '09 at 11:25
add comment

5 Answers

up vote 1 down vote accepted
+400

Do the cost associations of given and received downvotes achieve the desired effect?

If I am understanding your question correctly, I think there is some evidence for this from this StackOverflow blog post:

I reformulated my queries, and I found that there were many, many more suspicious upvote patterns than downvote patterns. More than five times as many!

I'll repost a comment I left on that same blog post:

I would like to point out the fascinating fact: an upvote changes a user’s rep by five times as much as a downvote (+10 vs. -2), and upvote-fraud was about five times as prevalent as down-vote fraud. It may be a coincidence, but the economist in me says that it is the invisible hand of the market at work.

share|improve this answer
    
Thanks for the link, very good read. Quote: "if we decide the upvotes are questionable", emphasis on all words - Oracles. Quote: "we’ll silently remove any vote fraud we detect" - yet another slap in the face for the person who loses 'fraud' upvotes from his answers: your answer is not that good... See my answer here: meta.stackoverflow.com/questions/6460/… –  kd304 Jul 17 '09 at 17:51
    
There is another kind of malicious downvote: you post a far superior answer than the others and they downvote you for your confidence. Happened to me several times. –  kd304 Jul 17 '09 at 18:16
    
Thank you for showing me the game theoretic aspects of SO. –  kd304 Jul 21 '09 at 6:11
add comment

How the community will do things right and the law of large numbers will select the good solutions

I think perhaps one of the problems with SO is that the asker of the question chooses which answer to accept. Now, this can work very well, if the answer quite obviously does or does not answer the question. For example, if a question is posted asking why a particular function doesn't work and somebody replies with the explanation and a fix which makes the function work, the user can obviously judge that to be a good answer; afterall, their function now works and they understand what they did wrong.

However, for those questions which are more subjective, such as the user asking for advice or an explanation, they're probably amongst the worst people to be judging what is a good answer. I've seen so many accepted answers that contain incorrect information and bad advice, where there are perfectly good, better answers that were not accepted.

Obviously, the voting system really helps here. If the asker of a question has accepted an answer while a different one is getting twice as many up-votes, that may be an indication to them that they haven't selected the best answer, but unfortunately not all questions are actually that popular. Some only have 2 or 3 answers with a few tens of views and 2 or 3 up-votes distributed around the answers. It's not quite so obvious to the asker which is the best answer between three that all have just one up-vote each.

Also, in the case where some of the answers do only have 1 or 2 up-votes, there aren't enough votes there to balance out. All it takes is for a couple of equally misinformed people to up-vote a slightly incorrect answer over a better one in an unpopular question to make it appear as though it's the best one.

Not that I have any suggestions on how to solve this, unfortunately.

share|improve this answer
    
Thanks. I just want to add, the words you quoted I was thinking about the up/down votes, not the actual questions and its accepted answer. –  kd304 Jul 12 '09 at 9:40
2  
The solution to the problem you mentioned is already in the system: You can express your disagreement with comments, downvotes and in your answer. If the original OP is still around, he is at liberty to change his mind and switch accepted answers (except bounty?). I fear to point out that SO's greatest strength and greatest weakness is the same thing. –  kd304 Jul 12 '09 at 9:45
add comment
  • How the community will do things right and the law of large numbers will select the good solutions (~)?

I think this is true; but the "right" answer is not always the accepted one. This often appears via large upvotes; think "populist", or those occasions where the accepted answer has -ve votes and there is an answer with +15 upvotes; I think that expresses the "good" solution, even if the OP disagrees

  • The property of no-registration required causing OPs to never find back to their questions and no one accepts it.

In general, it doesn't matter too much if we have a question without an accepted answer; the answers are there, with a (subjective) voting indication of quality

  • Do the cost associations of given and received downvotes achieve the desired effect?

IMO, downvotes don't quite carry enough... -5 downvotes and +1 upvote (for sympathy) yields zero points overall; or it could even yield +10 points for a new user! I'd prefer to see, say, -4 points? -5 points?

  • Doesn't upvotes draw attention of those who simply want the badge for +300 votes and vote for answers without even thinking about it?

The badge isn't that exciting; I don't think I've seen this, but it isn't necessary to agree with every vote cast...


I also think it is interesting to see how much UV and MSO can help shape the community.

share|improve this answer
add comment

I was wondering whether SO proved any of the claims used when it was built?

Why are you asking us when you should be asking god Google? What does Google say about Stack Overflow?

It's the most -- and arguably the only -- widely accepted metric of success on the internet that I know of.

share|improve this answer
    
I already talked to god about this. God promised to show me some proofs so I can decide myself. –  kd304 Jul 20 '09 at 10:32
2  
tell Sergey and Larry I said hi! –  Jeff Atwood Jul 20 '09 at 10:36
1  
Anyone else think this question is either a troll or a joke question? Maybe the bounty is up there so it won't be closed? –  George Stocker Jul 20 '09 at 12:41
    
@Gortok: No, I put the bounty on 17 july, 5 days after my initial question, and the bounty is from my own reputation. If you want, you can have an answer too. 4 days left. –  kd304 Jul 20 '09 at 13:23
    
Thank you. Btw, they said Hi: flickr.com/photos/cs___/166226907 –  kd304 Jul 21 '09 at 6:13
add comment

I asked somehow similar question - Is SO able to create stable, live ecosystem?
I think answering either one gives much of an insight to the second...

share|improve this answer
    
There is no impossibility but only inability. –  kd304 Jul 17 '09 at 13:46
add comment

You must log in to answer this question.

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