What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 131 Stack Exchange communities.

Although there is more to Stack Exchange than just points, that doesn't counter the fact that the point system is what makes it thrive, and is valued by the majority. It would be hard to deny this, considering all the discussions on the reputation system here on meta. ('reputation' is the 3rd most used tag, after the obligatory ones)

I find that this ideological selflessness, shouldn't stop the search for a more 'honest' point system.

Isn't the value of reputation proportional to the quality of the reputation system?

Ideally, higher reputation should be a linear mapping to effort and knowledge put into the community. The better this mapping, the higher value reputation has, the more you will value it.

The more you value your reputation, the higher the motivation to participate in the community.

When noticing certain flaws/unfairness in the reputation system, your motivation to participate will decrease. Again, why would it otherwise be such a hot topic on meta.

More motivation, higher quality of Stack Exchange. And isn't that what we are aiming for in the end?

To summarize:

  • The better the reputation system,
  • the better your reputation represents how much you aided in the community
  • the higher the motivation to invest in the community
  • the higher the quality of SE?

That's why I believe it's important to have constructive discussions on how to improve the system, instead of shrugging them off as unimportant. On a topic on whether answering difficult questions should earn more points (something which I believe relates to a fairer system), paxdiablo correctly commented that the biggest challenge is to find a good measure for complexity.

Relating to the point mentioned here, wouldn't it be useful to find this measurement (and others)? Perhaps SE's data explorer can be a very useful tool in this quest.

The goal would be to find correlations between measurable (current or to be implemented) data on questions/answers, and 'fair reputation'.


The real 'underlying' question isn't the last mentioned example, but rather the title of this question. "Shouldn't reputation be valued more?"

I feel the title of Erik's post "Why should I keep answering on Stack Overflow?" is spot on, as an opinion of a new user (and most likely many others). Perhaps a majority of opinions aren't represented here on meta, as the people to which this subject might appeal aren't here, or simply can't upvote as is the case for me.

share|improve this question
4  
You make good points, but what is your suggestion where to go? The system is highly imperfect, no doubt. But all in all, there seems to be some kind of a Churchillian consensus about it - it's the worst form of reputation system, except for all other forms that have been tried from time to time... What exactly do you suggest doing in the quest for a better system? –  Pëkka Feb 17 '11 at 23:41
3  
We've had troublesome users in the past who insisted on using "XP" to describe the rep system, so I'd suggest not doing that lest you get up the hackles of various old hands. If you are serious about trying to improve things there is no point in bringing up bad memories. –  dmckee Feb 17 '11 at 23:46
1  
@Pekka: finding correlations between measureable (current or to be implemented) data on questions/answers, and 'fair reputation'. –  Steven Jeuris Feb 17 '11 at 23:53

3 Answers 3

up vote 5 down vote accepted

That's why I believe it's important to have constructive discussions on how to improve the system, instead of shrugging them off as unimportant.

We don't simply shrug off such discussions. We generally agree that the current system isn't perfect, and then sit back and ask, "So what do you suggest, and how will that improve the system?"

The critical juncture that most suggestions fail at is that while it may make the reputation system more "honest" or more "true to the professed intent" the suggestions would actually not increase user participation.

The reputation system is one of many carrots built into the system to encourage people to play the game and disgorge their wealth of knowledge to the system.

Any change to it must show that it positively affects user participation - and not in a theoretical way, such as, "If the system were more honest, people would feel better about what it means to them, and they would participate more." This is an unverifiable, subjective theory.


CAUTION: Snarkiness ahead!

Relating to the point mentioned here, wouldn't it be useful to find this measurement (and others)? Perhaps SE's data explorer can be a very useful tool in this quest.

It would be! And perhaps SE's data explorer can be coaxed to provide some useful data for this!

The goal would be to find correlations between measureable (current or to be implemented) data on questions/answers, and 'fair reputation'.

I'm excited by the possibilities. I hope that you start on this endeavor immediately, and return with your results ASAP.

I hope you didn't come here to ask us to do it for you, plssendmetehcodez questions are frowned upon in these parts.

share|improve this answer
    
First off, thanks for the first truly constructive feedback! Initially, I wanted to try to find some working 'formula' to measure complexity, and post this as an answer to the mentioned post. Upon reading other reputation discussions however, I realized this might be a wasted effort, hence this question. The real question is perhaps, "Shouldn't reputation be valued more?". This is shrugged off, just look at the +1 upvote of this question vs it's 100 views. Or the only upvoted answers/comments here. ... not motivational. ;p I don't even get the rep to upvote you. –  Steven Jeuris Feb 18 '11 at 11:25
2  
@Steven if it's motivation you're looking for, Meta is not really big on that :) That's not to say that good suggestions aren't welcome, but from experience, they need to propose very specific steps. –  Pëkka Feb 18 '11 at 12:14
    
Thanks Pollyanna -- this was what I was trying to say; except, you communicated it much more clearly. –  The Unhandled Exception Feb 18 '11 at 12:43
1  
@Steven The lack of upvotes for this and other proposals is that there's nothing to agree with. There's one upvote, and no downvotes. One person agrees with you. No one disagrees with you. Everyone else either thought it was too long to read, or felt that since it made no concrete proposal, then there wasn't any reason to upvote it. You won't get a strong response either way without presenting data to support a conclusion, and/or suggesting a specific change. Discussions that say, "We should be doing more about this" are only good for finding out whether people agree that there is a problem. –  Adam Davis Feb 18 '11 at 13:15
1  
@Steven At this point, you've essentially got your response: People are ambivalent about this. There may be a problem, there may not be. If you want to progress you'll need to propose some concepts or ideas beyond this. At that point people will show that they agree with or disagree with your ideas. If you can back them up with data you'll be better off. But there's been a lot of discussion about reputation and what it means over the last 2.5 years, so you'll probably want to go through all the questions and answers regarding it to make sure you aren't rehashing old ideas. –  Adam Davis Feb 18 '11 at 13:18

Counterargument:

The real value of SE comes from people who have some other motivation for being helpful than just watching a counter increment. People whose egos are fed by the role of Oracle. People who are building their general online reputation. People who learn by teaching. People who think that perhaps they'll get better answers from the quasi-deities if they have made a conspicuous effort to be helpful amongst the hoi polloi. People who love to push the 'delete' key and are willing to tirelessly answer questions to get a finger onto the biggest, reddest, button at SE.

The rep system as it is provides a signal, as they say in the evolutionary fitness business. Like a peacock's tail, it's an imperfect signal. Nonetheless, it works well enough to sort people into them that put effort into the site and them that don't. There are people with thousands of rep who, by some arguments, 'don't deserve it', but there are perishing few with 10s of thousands.

(See 1Sam10:8: 'Saul has slain his thousands, and David his tens of thousands.')

This, I submit, is all the system needs to do.

share|improve this answer
    
+1 the bloody hands of Skeet –  Shog9 Feb 18 '11 at 2:16
    
Oracle feeds SO users? wow! :) –  belisarius Feb 18 '11 at 3:03
2  
Case in point: Jon Skeet didn't get to the top just because he wanted rep. He got to the top because he's insanely good at what he does. And because he has a pathological pre-existing propensity for answering questions. Well before StackOverflow was even a twinkle in the eyes of J&J, he was out there on other forums, answering, answering, answering. –  Benjol Feb 18 '11 at 7:22
    
I'm not seeing the counter argument, everything you 'counter' refers to what I already stated in the first sentence. "Although there is more to Stack Exchange than just points," Actually it's a prime example of what I classified as "it's important to have constructive discussions on how to improve the system, instead of shrugging them off as unimportant." I agree the system works well enough, but 1,149 questions tagged with 'reputation' (the 6th highest rated tag) lead me to believe this topic is a valid one. –  Steven Jeuris Feb 18 '11 at 10:55

[...] the point system is what makes [StackExchange] thrive

I think there's much more to it than the reputation system. I won't argue that reputation is at the core of the StackExchange model, and it does a lot to help the success of the sites and the quality of the answers given, but I do not think it's the only thing which helps StackExchange sites thrive...

Isn't the value of reputation, proportional to the quality of the reputation system?

What is the value of reputation anyway? You can spend it on bounties, so if you have a question you want to publicize, then reputation points have value. You use reputation points to earn more privileges on the sites, so if you want to earn more abilities on the site then reputation has a value. But, the value of reputation comes from the site itself. That is, you can't spend rep points anywhere else. You can only use them to become more entrenched in the site as is, so you have to like the site to begin with to value rep. It's a bit recursive...

True, if the reputation system were totally flawed or of terribly poor quality then reputation would have less value; so I see your point. But also look at Meta. Reputation could be thought of as essentially meaningless here because upvotes mean "I agree" and not necessarily "Good answer". Sometimes upvotes on meta mean "you included a freehand circle"! Yet people still value their MSO reputation...

More motivation, higher quality of Stack Exchange. And isn't that what we are aiming for in the end?

To summarize:

  • The better the reputation system,
  • the better your reputation represents how much you aided in the community
  • the higher the motivation to invest in the community
  • the higher the quality of SE?

I think there's a significant limit to what quality reputation alone can bring to Stack Exchange. I highly value quality in the SE network. I have reached a point where I don't care so much about my reputation on most sites. I'd love 10K so I could have access to the 10K tools, but the work it would take to get there is daunting. I answer questions because I like to help. I edit questions a lot. There's no rep for doing so (as a 2k user). I do it because I like quality. I chat a lot. I do it because I love the community. There's no rep involved for that (and I strongly feel there should not be.)

To summarize: Rep drew me in. The Stack Exchange system as a whole kept me for life. I am all for improving the reputation system but I don't see it being the end-all-be-all improvement engine for Stack Exchange.

share|improve this answer

You must log in to answer this question.

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