I'm wondering how much a reputation score would make people trust an answer or not? Does my current (167) make it more likely that you believe me or distrust me?
Reputation is how much the System Trusts you. That is all. There are some high rep people who regurgitate other people's answers. It is common for me to come across an answer that has been upvoted that does not answer a question or is wrong.
It's a problem caused by voting up 'cause you like the answer vs. voting up because it is the right answer.
The issue I have is when an answer is wrong or has one line that is wrong, you can't edit it. The only people who can edit it are people who have the wrong answers.
I'm less likely to base my respect solely on your reputation score, but more on if I've seen you post/answer questions before, as well as your rep. It also depends on whether I know you're right or not. I had 1 rep when I posted my first answer (obviously) but it was voted up and accepted within the same day. It's all about a) confidence, b) correctness, c) and finally reputation.
If I have absolutely no idea if your answer is correct or not, I'll judge yours based on the surrounding answers, whether I recognize you, and finally your rep.
Reputation is how much the System Trusts you. That is all.
This is true -- we need the system to trust you, so that you guys (and gals) can run it. In a perfect world there should be almost no difference between the "moderators" and highly trusted users.
If Geoff, Jarrod, and I have to moderate, the site is a failure, because we don't scale. The community does!
(oh, and I voted Brian's response up, of course)
In all honesty I think the reputation system is worthless, almost as worthless as the badge system.
I'm not looking for an answer from someone who is deemed an expert by this community! As a programmer I am looking for one thing when I find a question that interests me.
THE RIGHT ANSWER!
Some people are here to get involved with the community and to build up a high reputation, and when some features are locked down by reputation (voting down) it is important to answer a few questions from time to time to become a more established member. Other than that, the majority of visitors of this website will come here to find a correct answer.
Time will tell.
I'm not sure if rep, or badges might make more sense down the road. I think we've all belonged to message boards where people with tons of posts are often not the best resource at those boards.
You can already see that you can get a lot of rep for answering simple opinion questions in popular threads, and no rep for a good, detailed answer that is very useful and based on years of specialized knowledge.
All I have to do to find out if your answer is good is to try it. The great thing about programming is that "trust" really has nothing to do with it. I try your solution, and it either works or it doesn't. If two people answer with competing solutions, I compare the pros and cons.
We're not dealing with politics here. It's computers. They work, or they don't.
Pretty much repeating the opinion of others here, but I base my judgement on the quality of the questions and answers of posters. I've seen some bad answers from high rep people and good answers from low rep people.
One thing I'd say, however, is that somebody with a higher rep is probably far less likely to be completely clueless as they have proven they can ask and answer questions sufficiently competently to gain the respect of the community.
These are the types of questions that will lose you rep, so yeah, I suppose so.
EDIT: Well, maybe I was a bit too blunt, sorry. This forum is intended for programming questions, not questions related to the forum itself. As we are still in beta, maybe I am a bit wrong on this one (As the upvote shows).
At this point, in the beta, I have not found myself trusting people even with very high (3000+) reputation any more than someone who is new -- everyone in here is pretty knowledgeable and engaged, and we all know how the beta is being populated manually by request.
That may change, I guess, as the site is opened to the public and there are more fly-by users.
I agree SO is too new to establish much any trust and name recognition. I imagine, over time, the badges may be a better indication of trust.
I think rep will play a role eventually.
One of the other forums/boards I used to frequent was perlmonks, they seemed to have a pretty good balance with their rating system, sure there were lots of people at the top tier that were not the best on the site, but to make it to the top tier you needed to participate frequently.
Given a community moderated rep, I think that the people that participate frequently almost have to pick up at least some knowledge, either by reading other answers, or by having their answers corrected, or accepted, etc.
So, yes, a high rep will make a difference on how I view your posts. It doesn't mean I'll blindly accept what you're saying, but someone with a high rep presenting a different point of view will probably be taken more seriously because of the rep.
I have a ton of rep, but I've posted my fair share of dumb answers.
My take on it is that rep means "This person knows stuff," however, it doesn't mean they know the particular stuff you are asking about!
Reputation is not meant to measure expertise in a particular area. Reputation is an incentive for those with expertise in a particular area to share their expertise with others. To the extent that questioners and other users of the site are able to judge for themselves whether a particular answer is correct and helpful, the reputation of an expert who gives such answers will tend to increase over time.
So, given two people who have answered a similar number of questions of similar difficulty in similar topic areas, if one has, say, double the other's reputation, that will indicate his answers are correct and helpful more often than the other's. Otherwise, don't put much stock in reputation per se. If you are going to compare reputation, do it per month per question answered in the topic area.
The reputation of a member is indicative of two things: quality of answers and length of membership. If someone gives a detailed and well-written answer to a question, I'm not even going to look at their reputation.
Everyone starts with a small reputation, and it takes a while to build a higher reputation. Still, it's not a good indicator. You lose very few points for a down-vote compared to an up-vote; so long as someone only gets less than 8 down-votes per up-vote there is a net positive gain. So, someone with 100,000 reputation could have 10,000 up-votes and 0 down-votes or 100,000 up-votes and 450,000 down-votes. Still, those 100,000 up-votes are good answers and the person deserves credit for them. Both members' reputation represents an immense contribution to stackoverflow and it would be foolish to ignore either.
So my answer is no; the only time I look at reputations is when I see two identical answers; if the time difference is a matter of 2-3 seconds I favor the person with lower reputation.
I don't think the reputation system holds much water. It doesn't tell you how good their answers are, more that they are "keen" and have given popular answers. I think it just encourages people to rush in and post answers whether they know what they are talking about or not. The problem with this is you then have to sift through the "me too" posts to find useful information.
Also, I'm noticing some people plagiarising answers from other sites without attribution.
I think voting on individual answers is good, and over time this will hopefully lead to later, better answers rising above the quick answers the initial people have given. As for reputation, the number means next to nothing.