55

I struggle to pick up rep points even when answering questions and giving good demos of code via a pastebin - sometimes spending 15 mins+. Then, along comes a noob question, for instance What does $ mean in jquery and before you know it the simple answer has got 26 upvotes, never mind the 11 for asking.

Can we do anything about this? Maybe give the question a complexity rating which goes towards the possible amount of points that can be earned? Probably contentious I know but its just an idea.

7
  • 7
    You don't seem to be struggling. You have over 5K on SO. That's a really high score against those new to the site. How do you think they would feel?
    – random
    Jul 17, 2009 at 14:56
  • 4
    Along the same vein...being new to SO and not as versed in the finer dynamics & mechanics of C# and .Net, I am pretty well stuck at looking for the newb questions. Then, along comes a veteran with uber rep. We essentially say the same thing, but I get nothing for it.
    – IAbstract
    Feb 8, 2010 at 10:47
  • 1
    I recently had to spend over an hour, editing spelling and grammar on questions for 2¢ each, to reach the 100 points needed to put a bounty on own of my own questions, bringing me back to 1 point only.. then I'm frustrated that I can't upvote ANYTHING cuz I need 15 points(is anyone listenin? please change this loophole). Thus is life, we all can't be Skeet. But I realize now that I need to be careful how I spend my rep. I will save my rep for a rainy day. Sep 21, 2011 at 23:55
  • @gnat: It indeed isn't. Mar 26, 2015 at 15:12
  • @PatrickHofman did you check it's dupe target (Diminishing returns / non-linear reputation from “great” answers)? "My proposition to improve that situation would be to make the reputation gain from a single answer non-linear..."
    – gnat
    Mar 26, 2015 at 15:21
  • 1
    @gnat: My bad. I read your comment, clicked to the link, and assumed it was the same as the other proposed duplicates... Mar 26, 2015 at 15:23
  • OK, why is this marked as a duplicate of another question asked four years later?
    – Jason S
    Oct 8, 2015 at 13:15

11 Answers 11

62

rep points kinda remind me of that whole "Zen and the Art of Archery" business. I've gotten caught in that trap of reputation-score-addiction once or twice: poring over questions over and over until I answer one that gets upvoted. It's not worth it. Post questions or answers because you feel like it or because you think you can provide value to the community.

4
  • 5
    @Jason S: you speak the voice of reason! Jun 28, 2009 at 16:28
  • 18
    blasphemy, it's all about rep ;-) Jul 17, 2009 at 19:47
  • This serves as a good reminder!
    – snmcdonald
    Dec 20, 2010 at 2:13
  • 6
    Upvoting this feels strange :D Dec 30, 2011 at 13:09
28

I think that answers to simple questions get you more rep points because more people can understand both the question and answer & hence are more likely to say "yes this is a clear and informative answer".

Answers to more obscure questions won't get you as many (if any) rep points because less people are interested in either the question or answer and even less people (perhaps only you and the OP) are interested or even understand the answer.

2
18

The funny thing is; on SO some of what I consider my best answers are on very obscure topics - intricate details of System.ComponentModel, for example - that few people ever get involved with.

I know full well that if I'm lucky I'll get maybe 2 votes for it (simply because most people don't dabble in those areas, are unlikely to read the question, and even less likely to know if the answer is good or gibberish), but there is a certain pleasure in simply giving a detailed, robust answer that I can refer to later (and hopefully comes up in google when somebody wants to know how to implement the undocumented IFlibbleHoop interface*)

*=not a real interface, don't look for it...

3
12

The hardest, most obscure questions I answer earn the least points. I get one upvote from the asker, and that's it.

1
  • 7
    + 15 for the accepted answer if you're lucky Dec 29, 2009 at 22:08
2

A very controversial feature would be to award rep for up-voting corresponding to the up-voter rep points. The more rep the up-voter has, the more rep point is awarded.

For example, you always award 5% of the up voter rep (10 for a 200 rep voter, 100 for a 2k rep voter, but only 1 for a 10 rep voter)

The rational behind it is that the system trusts more high-rep-users votes.

Caveats

  1. I don't know how it would relate to rep-farming.
  2. It would also rapidly have stellar-rep users. May be just have an extra bonus if the voter has a much higher rep than the author.
4
  • 1
    I think your math is wrong. 0.5% of 200 is 1. Jul 17, 2009 at 14:53
  • Ouch... :) Corrected Jul 17, 2009 at 14:53
  • Nor is 0.5% of 10 = 1. Jul 17, 2009 at 14:58
  • Ian: Yeah... but 1 is the smallest unit, so I sorta "round up" :-) Jul 17, 2009 at 15:03
1

Perhaps the more generic answers are useful to more people and thus get upvoted more. So it's not really a problem, it's just the system working correctly.

I wouldn't worry about it; simply answer those questions you can add value to. The reward you get isn't really the upvotes; the reward you get is the knowledge that you've made somebody else's life better by helping him or her with a tough problem.

1

Esoteric questions are only of interest to people working with that item. For example, there are many questions about WPF Minutae on Stackoverflow which I don't look at because I am not using WPF. I have asked questions about SANs that got no answers because very few people care about or work directly with SANs.

That doesn't mean to say that WPF isn't important to people working with it. I just don't work with WPF so I don't read WPF questions. SANs can be very important to those working directly with them but many people aren't directly involved with them.

Populist questions are the most accessible so the most take interest in them. Therefore they get the most traffic. The rep system is just a heuristic - it's not perfect. Bear in mind that it was designed to encourage people to answer questions, not to be an objective metric of their 1337-ness.

1

Not all questions are created equal. Is there a reason noobs should not get good reputation for participating? After all, that is the spirit of the SO trilogy and it's the magic that got us all hooked to the point we're now chatting about it on a meta site.

Likewise, I see negative favoritism for noobs questions that involve responses ranging from downvotes to outright arrogance (e.g. RTFM, don't ask here, etc.). I would sooner reward someone who contributes to the society before I'd ding them based on lack of relevance based on my own experience level.

To that, +1 to the asker.

1

Huh.

That author only got 150 points from that answer, as of right now at least. She has under 400 total reputation on the site. Think about that: can't retag questions, can't edit anything (not even community-wiki posts)... Yes, it's a good amount of reputation points for a single answer, but hardly outstanding. And since a different answer was accepted, she hasn't seen many additional upvotes in the days following.

It's a beginner answer to a beginner question. No harm done.

Frankly, I'm much more concerned about the folks gaining reputation points via sympathy votes. But ultimately, it's a meaningless number - there's just no point in getting too worked up about it...

0

My answer is relating to the non "noob" questions...

You can consider doing answer refinement. Give a good short answer to the poster's question, then refine it better and better.

Answer refinement helps both your reputation and the original person that asked the question. The person that asked the question may not need a code example, the extra 10 minutes you spend doesn't need to go into your original post.

Answer refinement also helps other questions on the site, because people that would give a very similar answer to you won't waste their time, and they will move onto another question that they can help with.

I am not aware of any penalty for editing, and I do not believe it is frowned upon.

11
  • not sure I am with you Brian
    – redsquare
    Jun 28, 2009 at 12:28
  • By the way, even after I answer a question and it is clear that it is the highest voted up, I will still refine it better and better. Sometimes even the next day. Jun 28, 2009 at 12:33
  • 1
    Brian's saying that you should submit an answer right away, and then edit it and keep working on it. That way you'll be more likely to get a fast-mover advantage, and dissuade others from answering the question you've already answered.
    – Brent Ozar
    Jun 28, 2009 at 12:35
  • ***dissuade others from answering that would be answering along the same lines as my answer. Better they spend their time answering a different question if they were going to give the same answer as me :) Jun 28, 2009 at 12:38
  • and of course, by refining your answer, you keep the topic bumped, so more people will see it. A cheap trick, but I've used it a few times when I was juuuust a few votes short of a badge. ;)
    – jalf
    Jun 28, 2009 at 12:48
  • 1
    as far as I have understood, there is a penalty for editing and that is that after n edits (30?) the answer will become a community answer. imho edits by the original author should not count towards this threshold! Jun 28, 2009 at 13:04
  • @jalf: I would only edit if I had something of value to add though. Jun 28, 2009 at 13:06
  • @Oliver Giesen: I agree with that, but I guess maybe it has to do with what jalf brought up, some people may abuse it. Jun 28, 2009 at 13:06
  • Easy solution - set a limit to how often an edit causes bumping: If EditCount > X AND BumpsPerDay > Y THEN DontBump Jun 28, 2009 at 13:10
  • I can usually find something of value to add though. Of course I wouldn't edit a post just to add another space or something, but reading through an old post of mine, I can usually find something that's unclear, or some relevant detail I forgot to mention.
    – jalf
    Jun 28, 2009 at 13:22
  • 2
    @jalf: In that case then you deserve the extra upvote(s). I really like @Peter Boughton's idea of a bump threshold. And @Oliver Giesen's idea of edits from original author not counting towards wiki answer. Jun 28, 2009 at 13:32
0

I think the solution is simply to adapt. Some questions are pretty much read once, by a small handful of people, then it drops off the face of the Earth and is never visited again. In those cases, not much rep is at stake, and it generally goes to the first acceptable answer. It doesn't matter if you write a better answer, if by then the question is answered and no one will ever read the thread again.

Others are continually bumped, as new answers are added, people comment, and the question gets edited. Those tend to attract much more rep, and in those cases, it also tends to balance itself towards the good answers. It doesn't matter that someone else posted first, because people keep coming back to the thread, so once your superior answer goes up, everyone will see it, and you'll get plenty of rep.

Which means it's not really a huge deal IMO. Sometimes you get fewer upvotes than your answer warrants, but that's usually only in the cases where no answer in the thread gets more than 3-4.

In the questions where some answers get 20+, later answers have plenty of time to get caught up.

Of course, the type of question also influences this. Questions that are tricky to answer, but which interest a lot of people are great. (I had one answer jump to 74 upvotes within a few hours the other day. Not because it was a stunningly good answer, but because it did the job, and a lot of people were interested in the subject, and kept commenting and keeping the question (and my answer) visible.

And sometimes, I write detailed answers and get nothing for my trouble because by then, no one else is going to read the question again.

That's just life. :)

As for the JQuery question you mentioned, it's worth noticing that it is

  • asking about a popular technology (so more people will read it, answer it and up/downvote it, and
  • is easily answered. It's hard to give good or bad answers to it. It's not a question where people are inclined to read down past the first answer to see if there are any subtleties they'd miss otherwise. The simpler the answer, the more the "quick" answer is favored. Which means it's usually more rewarding (in more ways than one) to answer the more complex ones.
3
  • 3
    This brings into question what does rep mean. Getting 110 rep points for asking such a question.....doesnt fit right with me
    – redsquare
    Jun 28, 2009 at 13:03
  • funny you should mention it, I just asked this: meta.stackexchange.com/questions/139/… But yeah, at a guess, I'd have to say "because it was a question a lot of people were interested in, and it was clear and concise"... Something like that.
    – jalf
    Jun 28, 2009 at 13:21
  • 1
    maybe split a users rep points into questioneer and answerer?
    – redsquare
    Jun 28, 2009 at 13:31

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