Please charge rep for questions after threshold

On one hand people are gaming the system, but the flip side is that people are asking questions that the community sees a value in, that is why they voted the questions up. I think a stream of great well thought out questions should be encouraged.

There are a couple of things that I think should be done to help avoid abuse.

1. No rep should be gained by asking a duplicate question, in fact any question that is asked and then closed should give you no rep (which should be fixed retroactively). I would go as far as to say the the answers should also get no rep, because they dilute from the value of the original question.

2. The community should be given the ability to revoke ownership of a question, in case it is abuse, abandoned or the asker has obviously lost interest. When ownership is revoked all the rep gained should be revoked as well.

The solution to this problem seems simple to me: Don't award rep for asking questions at all. There's just no need to.

People that genuinely want a question answered have plenty of incentive already to ask it. People who just spend all their time asking questions (so many that they can't possibly be actually involved in software development) are just wasting the time of people who answer questions. They should not be rewarded for this. Look at my record on StackOverflow and then look at "Thanks". 399 questions and hardly any upvoted answers and yet they have four times my rep!

• U62
• Thanks

Just remove rep for asking questions (and do a recalc for existing accounts). Still have voting so people know which ones are interesting ofc.

Just to follow this up - this contributed to me dropping stackoverflow.com as an interest. I was just looking over my posts on the subject just now.

I'd like to spout my tech credentials but I think it would be patronising, but 19 years and a majority in serious C++ dev, and working at the very top pay bracket of the field - I think I am qualified to have an opinion. I still work as a top end senior developer/manager.

I respect Jeff Atwood's blog immensely, his blog is great.

And so, I can't understand how this place is made a haven, positively incentivised for disruption, for the people who have no answers and only want to pollute the discourse.

The most silly thing I see in general is that a genuinely hard question, something that really needs the collective good to contribute to, never gets any votes and frequently submerges underneath questions which regard IDisposable or the operation of virtual destructors in C++ - pathetic really.

But a post (or an answer to) "what are the best lessons you've learnt from comments you've read in code you've worked on?" gets gazillions of votes for the questioner or reader. Some answerer rise immediately in stackoverflow status for a comment regarding Klingons and Commander Data which takes the fancy of a bunch of scifi fans (for instance.)

In that sense, it's just a reddit for nerds (I do like reddit) rather than a site which encourages the contribution of the best to avail themselves. But the points system ends up being an offense to good sense if you have something on the sharp end to contribute.

I like my reddit distinguishable from my stackoverflow - otherwise I may have to start "IamA, AmA" and see how many stackoverflow votes I get for absolutely no technical contribution whatsoever.

I think my MBA degreed department manager could get more stackoverflow votes on here for asking the utterly inane, and wallowing in purile discussion.

I had to argue extensively to some folks about how a (smart) pointer to heap memory was significantly different to a vector (particularly that default copy semantics meant buffer copying was inevitable.)

It was a straightforward argument really, what is the difference between a stack allocated array and a dynamically allocated array. But they went back and forth for 20 responses and still gave the points to someone that wanted to allocate BITMAPS!!!! of massive screenshot dumps!!!! in vector !!!

How would I ever have the points to be credible when you don't earn points for applying yourself to the hard things?

However, as I said months ago, the points scoring rules are there to encourage gross participation - but this will be at the expense of the gifted and experienced in general, that can actually add rare value.

But maybe nobody's looking for rare value on here - and maybe it's meant to be beginner answers to beginner questions, asked ad-finitum.

I find it hard to think of good questions, and when I do, they don't get many votes, so I'm fairly amused to see that there are people that can come up with a few hundred questions.

More power to them - just because they're trivial or silly to "you" doesn't mean that they have no value to others.

One thing I haven't really seen mentioned is how this waters down the value of SO rep outside of SO. I've seen Joel in person touting the new careers site by noting that you can find ace programmers with thousands of SO rep, so they must be great, right? Um, it depends. It could mean they've asked 2 or 3 questions about "favorite programmer joke" or "funniest code comment" that everyone loves to upvote (or, more likely, 200 questions on how to set breakpoints). Yes, we're programmers, and we loves us some geek humor, but that certainly doesn't qualify the person asking as an expert developer. I've asked 4 questions (100% accept rate) and answered 90+ with lots of accepts and upvotes, yet I continually run into people with much higher reps who have contributed no meaningful answers. That brings down the value of reps in general. If I was an employer searching the careers site, I'd want to know that rep points = capability/experience without having to research answer history for each person.

I should note that I do believe (counter to some arguments in this answer thread) in the free-market's ability to set the value of the questions themselves -- that's fine. I just think question value and member rep should be two completely, unrelated things.

Here's one suggestion for solving this: Keep track of reputation points received from questions and points received from answers. Reputation from questions should only be given if the asker already has achieved that amount of reputation from answers. In other words, limit each user's question reputation total so that it is always less than or equal to their answer reputation total. That seems like a fair solution, that should not penalize any active, productive members of the community.

Good questions will happen naturally with good people running into real problems for which they need real help. Awarding rep for questions does nothing but encourage rep-whoring question saturation making it hard to find the truly good questions.

It has now become perfectly routine to have to sift through dozens hundreds thousands of questions on SO to find one worth putting time into answering.

Personally, I think you should get 10 questions up front, then 1 additional for every 5 - 10 answers you provide which are up-voted.

Actually, what I really think is that questions shouldn't be awarded rep at all.

I know this is an ancient question, but I don't see this point in any of the other answers:

The solution to rep from crap questions is to delete the questions. At the time this question was asked, I'm pretty sure people already lost rep from deleted posts; a change was made later to keep the rep if the post was old enough and upvoted enough.

Deleting a question takes its answers with it and sometimes we refrain from deleting a poor question because of good answers, but in the case you're describing, where questions are utter garbage, do you really want people answering those in the first place? On The Workplace we've had an occasional broken-windows cleanup, where we go through recent closed questions to fix them, delete them, or decide explicitly to keep them, and one thing we've seen a lot is people answering bad questions when they should have been voting to close instead. In a situation like that, I'm more willing to let a decent answer, and its attendant rep, go away by deleting the question. (We've done our reviews in batches of 10, chosen from closed questions less than 60 days old. You could, of course, adjust those parameters to suit your site.)

(I know that when this question was asked it was specifically about SO, which due to its size has special problems. But it's here on Meta.SE now, so we need to consider everybody else too. On some sites it's perfectly fine to ask lots of questions so long as they're good, so requiring a certain ratio of answers wouldn't work.)

I partially agree with U62's answer. I would award less reputation for upvotes in questions (maybe 5 instead of 10?), and I'd do a massive rep recalc after that

Please note that today's change is one of the ways we're addressing this problem.

https://blog.stackoverflow.com/2009/08/new-question-asker-features/

More to come.

EDIT: as promised...

• Should the weight of downvotes be increased?
• https://blog.stackoverflow.com/2010/03/important-reputation-rule-changes/

Marking this completed, since this is a substantial reduction (half!) in rep from questions.

I have to agree - you don't see people spamming bad answers and getting lots of reputation for it, but you do see people spamming bad questions. That tells you there's something wrong with the system.

You've already made the answers gain non-linear rep to stop great answers from providing too much rep; why not do the same to questions to stop bad questions? Only instead of starting out high and getting lower, start out low and get higher

eg. First three votes only get +2 rep, next three get +3, next three +5, next three +10, then drop back down so that overall, a 15+ vote question will still gain just as much rep as it does now, but a 3-vote question will gain much less.
Also, judge -1 on questions much harsher than on answers, since they are much rarer (compared to -1 for answers). You may want to consider making those non-linear as well (starting out low and getting higher).

This is tough. The value of StackOverflow, to everyone involved, is intricately tied to the number of valuable questions as well as the answers given to those questions.

It might be interesting to see privileges doled out based on question/answer ratios or even question/answer score ratios. As programmers, I think a lot of us would love to see some beautiful, holistic number or icon that resolved all of these intricate nuances into a single score.

However, to gain the potential benefit of making one's Rep more reflective of their expertise, the Rep score could become more complex & opaque to new community members. The simplicity of the existing scoring methods have a great appeal.

I think a user should be capped at something like 500 rep if he/she does not have at least something like 50% of their rep from answering questions.

It is too easy to copy questions from other web sites so the benefit of doing so should be limited. However good questions do have value so users should got rep for asking questions – just pot some limit on it.

These users do not deserve the system's trust and instead cause nothing but problems.

If by asking questions and posting answers and getting reputation doesn't construe the system giving you "trust", I'll throw in the towel. The community voted up the questions, not down. So the system being the community obviously has given the trust.

If a user is spamming the site, they should be warned and tossed into the penalty box if their behavior does not change. There is value in having questions posted here, so why limit this?

Please stop these people from becoming trusted members of the community. It will not be good for anyone.

Overall, I say NO. This site needs people asking questions and posting answers to stay alive. If we start taking this elitist attitude about users who are getting reputation from the community, there will eventually be no community.

I don't like the idea of a threshold for questions, because SO is a Question/Answer site.
Couldn't you address this problem by reducing the rep for questions significantly? Say 3 points for an upvote instead of 10?

I understand what you are saying about meaningless questions gaming the system, however to discourage or possibly penalize (charge for) question askers would severely undermine the supply and demand of questions to answer on SO.

It really comes down to the ratio of good vs gaming users, and in reality - what is the worth of reputation after a certain level?

On the other side of the coin, I tend to ask more than answer these days, unless I see an unanswered question I can help with.

This is partly as I am usually too busy to write up answers to my satisfaction, but also as I would prefer not to flood questions with correct, but not particularly good, answers just so that may 'game' 1 or two votes from the system.

Why award rep (positive OR negative) for questions? QUESTIONS don't benefit the community unless they are ANSWERED. Therefore, arguably, the value isn't added by the question, but by the answer.

Now when SO was just starting out, there was value in encouraging people to ask questions. Now, however, there is less of a need. I imagine stackoverflow as less of a large database, and more of a peer-help site. In the peer-help model, there's no need to encourage asking questions: people want answers! In the database model, it makes sense, though.

I think SO has grown more toward the peer-help system. As a result, rewarding asking questions but not providing answers seems wrong. Even if a reward for good questions is there, the contribution of someone who asks 400 questions with 2 votes each, with each vote having an equal chance of being an upvote or downvote should not reward 3.8k rep!

Eventually, everyone who wants it and is persistent enough will have mod capability -- even at 10 points a day, it would take you just under three years to become a mod. The solution is to have some way for rep. to exit the system or to raise the requirements for moderation capability on a regular basis (year 1: 10,000; year 2: 15,000, etc ...) In theory, if you can gain rep. by questions, then you should be able to be the Jon Skeet of questions. Just a bad-a$$ question asker. After all, somebody is up-voting those questions.

I've been on SO for 8 months and got 1300 rep (or 6 Skeetdays worth) in that time. I don't especially care for the powers endowed upon me as my rep (slowly) rises; I'm here for Q&A! If someone wants to game SO then good on 'em. Like others say it's up to the community as a whole to realise the value of their contribution with votes.

Surely this will only be a real issue if someone uses their "ill-gotten" moderation powers for evil, at which point the other moderators will kick their arse...

There are a couple of issues here. Whether they should get or not rep and whether they should or not get moderation abilities.

I don't care about rep points, but moderation capabilities are a tool and a powerful one (I just learned that with enough rep points you can close a question with no other votes).

Now there is the correlation between the questions/answers and the power they yield. There is a general assumption that due to the fact that questions are not good (by some definition of good) the user will misbehave or at least not be competent as an administrator. Is that so?

I don't believe it. I have known bad programmers (well, better average programmers) that have made great managers and at the same time great programmers that could not deal with a team. To me both competences are not directly related. Coming back to the reputation problem, I cannot say in advance that a user with low or no technical competence whatsoever will be a bad moderator.

Having said that, the worst can come true: we can end up with horrible administrators (not only the users we are talking here about, but also good technical people that arbitrarily decides to close questions, or change the contents... (it is curious how the threshold for editing is rather low and still people respects others contents). There should be tools to deal with it: administrators, moderators being able to block the malicious user powers...

But that is a general problem with administrators misbehaving, not specific to the set of users this question deals with.

Oh, BTW, each so often I also feel aggravated each so often when I see that a dumb question/answer brings more rep points than some answers that took me a time and effort to work out. And I don't really even care about rep!

In theory, the goal of the SE sites are not to give answers to the questions of the OP. The goal is to create the possible most useful Q/A catalog by volunteer cooperation.

The difference is that making questions is not a free service, but a valuable contribution.

In exchange, the sites are not expected to accept the questions, to answer them, and so on. There are hard, probably the hardest requirements of the Internet, how a question can be asked.

Although the practice is that the motivation of most created questions is to get an answer for the OP, the site rules admittedly don't consider it. Questions are judged - in theory - by their usefulness to reach the goal of the site, the goals of the OP (to get an answer) is a no-issue.

On this reason, if someone creates a lot of questions, what we consider useful (upvotes, left open, gets answers, views and so on), then these are valuable contributions and the OPs should get their well-deserved reward for them.

If the questions are low-quality, then they will be downvoted, closed, deleted, they cause rep loss, they become undeletable taint on the profile page of the OP, and in extreme cases, they can also lead to automatic Q-bans, or manual suspensions for low-quality contributions.

I see no reason to punish someone for asking too many questions. If the questions are good, they are so valuable contributors than the people answering them. If not, there are already the mechnisms to handle them.