1

I have seen this question before, but it seems to be talking about the daily cap.

The idea came to me when I saw this profile. This is one guy who has only 1 answer (and 1 score 4 question) and 5k reputation. The reason being the fact that he has answered a question viewed by many people.

I believe this example shows there is something wrong with the scoring system. After all, by this single one-paragraph answer you cannot tell the person is an expert or has rights to edit questions and answers or approve tag wiki edits, or that he is one of the top 7% of all stackoverflow!

One way that seems to me that can address the problem is to limit the amount of reputation one can get from any of his answers and perhaps also questions. This is not a daily cap, but a forever cap. I'd say 200 for the answers and 100 for questions should be good, which correspond to 20 votes each. I believe an answer that gets 20 up-votes is a good answer and any number above that is just a matter of how often people have that question, not how great the answer is. Nevertheless, there are badges for answers with really high scores that still reward the answer, but don't give him unnecessary rights.

Note: I'm not trying to reduce other people's reputation. I'm trying to address the problems that arise since everything in the system depends on reputation.

30
  • He can already edit stuff.
    – user184498
    Commented May 9, 2012 at 23:44
  • 1
    @dunsmoreb, which he shouldn't. Answering 1 question shouldn't give you such a high privilege.
    – Shahbaz
    Commented May 9, 2012 at 23:45
  • 3
    I completely agree. A single answer doesn't really "qualify" you as knowing what you're doing.
    – animuson StaffMod
    Commented May 9, 2012 at 23:45
  • I agree totally dude...
    – user184498
    Commented May 9, 2012 at 23:46
  • I like, but maybe more than 20. 25? 50? We'd have to do some Data Exploring to get a good feel for what's an average regular good score and what is due to high view-count.
    – Double AA
    Commented May 9, 2012 at 23:47
  • 8
    I am, however, against anything that breaks something for everyone because of a tiny number of people. Commented May 9, 2012 at 23:47
  • what if that answer had been a bad one? Commented May 9, 2012 at 23:48
  • @AndrewBarber, what do you mean? How are bad answers affect this?
    – Shahbaz
    Commented May 9, 2012 at 23:50
  • 1
    If that answer had been a bad one, you might not be here posting this, as he would have lots of negative votes, or have deleted the answer. Popular questions cut both ways. Commented May 9, 2012 at 23:52
  • 1
    My point is, this was a fluke that could have gone either way (positive or zero), and the quality still counts for something. Commented May 9, 2012 at 23:55
  • 9
    I hate to say it, but this question seems like sour grapes. Ok, that user was incredibly lucky on their one answer. Sometimes you win the lottery on the first ticket you buy. But still, it was a good answer that has helped lots of people. Most of the people this change would effect have bought lots of tickets with only a few upvotes. Should we take away their biggest winnings because someone might get lucky without paying their dues? Commented May 9, 2012 at 23:57
  • 2
    @JonEricson I don't think these ridiculous numbers are a problem when isolated. The problem is that they mess up with other stats. This answer, which Google/Wikipedia could have given, sends its answerer in 1st position on the top users lists of all the tags of that question.
    – Bruno
    Commented May 10, 2012 at 0:01
  • 1
    Why should it be a problem if somebody answers a problem, which apparently many people have, and as a result gets a lot of credit for it? I'd prefer single focussed helpfull answers anytime as opposed to hunting all over the place for bits and pieces to get my problem solved. And the answer probably wouldn't have been upvoted as much if it weren't helpfull, right? Basically this is a similar problem as people gaining rep just by asking (good) questions without answering anything.
    – Devolus
    Commented Jan 8, 2014 at 13:47
  • 1
    @JoeW, If a user answers several simple questions, he becomes acquainted with SO enough to be trusted with some basic tasks and therefore given some rights. If a user answers a single questions and never looks at anything else, he shouldn't be as trusted as someone who has 200 very explanatory answers on [linux-kernel] tag.
    – Shahbaz
    Commented Jan 8, 2014 at 13:59
  • 3
    I'll just throw this user profile in as argument. Decide yourself for what side. Commented Jan 8, 2014 at 15:02

3 Answers 3

19

I agree with your sentiments about gaining editing privileges after only answering one question, but I disagree completely with your proposed way of fixing it.

First of all, is this really a problem? Are there users with edit privileges gained from one or two questions/answers running around editing stuff and doing a bad job of it? Are we looking to fix a problem that doesn't exist?

If someone lays down a great question or answer that continues to accrue votes, then good on them - they should not be penalized for that. We want great questions and answers, and moving to a form of communism whereby we stop awarding rep to a user because they already have "enough" feels totally wrong to me. Additionally, we shouldn't be in a position of asking "How much rep is too much?".

The edit privilege was set at a certain rep point to give people incentive and reward for attaining reputation. If this is a problem then possibly the best way to handle this is to require:

  • a certain number of approved edit suggestions

    OR

  • a certain minimum number of answers and/or questions, possibly even a certain ratio of questions/answers

5
  • 1
    There is a general issue, though, about some of these scores. There are general inequalities between tags and specificity of questions. Sometimes, the more expertise is required to write a question, the less likely it is to be highly upvoted. I think these lottery questions (often based on low-hanging fruits) will ultimately drive more expert users away.
    – Bruno
    Commented May 10, 2012 at 0:08
  • 1
    @Bruno: I'd argue (as someone who knows sed better than c#) that you are never going to solve the problem of the vagarities of the universe rewarding the lucky over the deserving once in a while. In the big picture, throwing out the advantages of luck will hurt the persistent more than the lucky, believe it or not. Commented May 10, 2012 at 0:23
  • I don't really care how the problem is solved (as long as it is). My problem is not with the number that represents your reputation, but with other stuff that come with it. The privileges are an example. The same user I mentioned is considered one of the top 9% users of this quarter even though he has done nothing. See what I mean? The side effects are many.
    – Shahbaz
    Commented May 10, 2012 at 10:18
  • 1
    By the way, the argument in paragraph that says: "If someone ...", can be applied to daily reputation cap too. If it's not ok to stop awarding rep to a user because they already have enough, then why is there a daily rep cap? If we shouldn't be in a position of asking "How much rep is too much?", then why was 200 rep for daily cap was chosen? The answer seems to be contradicting some of the choices already maid in the system.
    – Shahbaz
    Commented Jan 8, 2014 at 13:39
  • You lost me at "communism." Reputation is not a finite resource where if I hoard too many points, there's not enough left for everybody else. I think it's reasonable to rate limit reputation both on a time scale and on an answer scale; it makes exceedingly simple questions less appealing because, while they still generate a lot of views and votes in the mad rush for precious reps, the rewards are capped.
    – badp
    Commented Jan 8, 2014 at 17:08
5

Yes, I realize this is an old question, but since it's in the recently active list. . .

I keep seeing mention of a "problem", but no practical examples. As I understand it, the concern is the possibility of these lottery-winning mono-question users running rampant through the edit queues and calling down plagues of locusts o'er the land. This is a fine concern. We all want to keep the site clean and avoid vandalism. But I think the issues are being exaggerated.

First, where are the actual, real problems? Which mono-question users are actually running rampant? I'm all for fixing problems if there are actual problems to be fixed, but you provide no examples of practical impacts. Given that you're posting on MSO, it's fair to assume that you know how development priorities work. Potential problems go to the back of the queue, to be worked on when immediate problems are finished, or when they become immediate problems.

Second, if it's not about actual damage to the site, but just about a number next to someone's name, why do you care? A sense of outraged justice? A tingling sense that all is not right with the world? Yes, reputation can be considered a vague indicator of someone's theoretical ability with some form of technology or other. But what about situations like Remyabel? He's clearly a top C++ user if you take the time to go look at his profile. Yet from a casual glance at his reputation number, you wouldn't guess. (Note for posterity: Remy has 612 total rep as of this post's writing, due to 6,200 worth of bounties given away.)

Third, your statement about "everything depends on reputation" is flawed. For an example, one need only look at all the MSO grumbling about robo-reviewers. Those robo-reviewers, by and large, are people who received their reputation from more than one answer. But those multiple answers didn't prevent them being precisely the sort of problem you are concerned about.

Basically (and this isn't intended to be offensive, merely a statement of fact) it looks to me as though you're searching for a problem where there really isn't one, simply because you want a near-meaningless number to be "fair". If this particular user isn't doing anything to capitalize on his lottery reputation, the system is working as intended. (Or at least for a given value of intended.) And he's not. Look at his activity tab. In no way is he being abusive.

I would guess that Stack Exchange is not willing to invest valuable developer time in fixing an extreme edge case that is currently doing nobody any actual harm.

3
  • 1
    Just a couple of clarifications: As I mentioned, I don't care about the number, just the side effects. Your example shows another problem with SO. If Remyabel is a top C++ user, but can't cast close/reopen votes while a user with a single answer can? Does that look alright to you? Also, I don't know if those users go rampant with vandalism, probably not, but don't you agree that the robo-reviewers shouldn't have had review rights? Could it be that they gained their previleges from such reputation points? (Again, I don't know, it's for SO developers to actually check)
    – Shahbaz
    Commented Jan 8, 2014 at 15:00
  • Okay, I can agree with you on the specific point that Remy should have edit/close privileges. That's perfectly true. But again, if it's a problem, why hasn't Remy come to Meta to ask about it? Again, I'm not trying to say that this isn't slightly unequal toward people like Remy, but what can you say to the developers to convince them that this should have a higher priority if it ever makes it into their request queue? Commented Jan 8, 2014 at 15:10
  • When I had posted the question I hadn't thought about that. But then again I imagined adding an if (post.up_votes < 20) before increasing the reputation wasn't that much of a trouble either.
    – Shahbaz
    Commented Jan 8, 2014 at 15:21
3

My take on this: The clue is in the word "reputation". Your rep score is based (mainly) on the number of people who found your answers useful. And to me, that sums up what the word 'reputation' means. If an answer is useful to hundreds of people, then that is worthy of having a good reputation. That's kinda the whole point.

I think your issue of him having edit rights is unlikely to be a real issue. The user in question hasn't even visited SO for the best part of a year, much less done any editing, so what rights he may have are somewhat irrelevant. Even if a user in this position is editing things, it's still only a problem if he's actively abusing it. And that would probably be picked up and stamped on fairly quickly anyway.

Also, you take issue with him being in the top 7% of Stack Overflow, but frankly, being in the "top x% of Stack Overflow" is a completely meaninless stat.

There are vast quantities of one-time visitors with less than 20 rep who registered just to ask a single question and will never come back to the site. They massively skew that percentage figure -- anyone with any rep at all will get farly high on that percentage. So really I wouldn't pay any attention to that figure if I were you. Change the stat to "top x% of active users", and you'll have a completely different perspective.

1
  • Actually, "top 7% of Stack Overflow" means "top 7% of users with at least 200 rep." Otherwise it would be top 0.4% of Stack Overflow.
    – wythagoras
    Commented Feb 28, 2016 at 16:41

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