69

Observations, suggestions and some questions after 4 days on Stack Overflow.

I've been hearing about Stack Overflow from more and more people, and it turns up more and more often in my Google searches. This weekend I decided to finally give it a try, see how it compared to e.g. the original experts exchange where I participated a lot (although it's more than a decade since I last answered anything there), and perhaps to get a chance to find knowledgeable people for a little competition.

Sign-up's fine, site concept looks straight forward, interface is very friendly and understandable. After reading up a little I start answering questions, and getting some accepts, some upvotes, some downvotes. I quickly pick up on the need to be very exact in answers, and the tendency among users to quickly first-post a one-liner followed by an edit with better information. A few hours pass, I get the initial badges, and then the first real surprise when I hit the daily limit. This is kind of a discouragement, so after answering a few more posts I leave for the day.

Next few days progress like this: I watch questions, try to answer exact and quickly, start doing upvotes on others, discover that I'm not getting more reputation and must have hit daily cap, leave the site.

Then, I discover what will make me leave Stack Overflow as soon as the initial fun's over. I cannot compete. No matter how many questions I answer, no matter how good the answers may be, I will never get on the top list. Doesn't matter if I stay here around the day for years, it's simply not possible to catch up. And I, as I believe many others also do, answer questions equally for recognition and as a desire to help.

So, the questions.

  • Is the one-liner followed by edits a desired behavior? Accepted behavior? If not, are there any mechanisms in place to discourage this?

  • Is it intentional that newer members should not be able to compete for the top spots? If so, is there a rationale behind this decision?

  • Why is the reputation cap as low as 200? Answer 3–4 questions properly, and then not much more point for me to stick around?

  • Why is the reputation cap system not favoring active over long-term users? Why not e.g. "1000 for stuff you posted today, 500 for stuff you posted this week, 200 for anything older"?

And to sum it up:

  • What incentives do I, as a new user with technical knowledge in several fields, have for sticking around and answering questions?

I've seen most of these discussed in various forms here, but still post this in order to present my impressions as a brand new user, and in the hope of a discussion of the summary question, "Why stay?". My pardons if this is an inappropriate use of meta.

16
  • 17
    The "long-term reputation cap" idea I find interesting.
    – Pekka
    Feb 17, 2011 at 10:33
  • 16
    However, "I can't make it to the top" is incorrect - it is possible to make "it to the top" if one's heart is in it; The most serious threat to Jon Skeet's #1 position to date, Nick Craver, started a year after Jon and was catching up massively until he got hired by SO. It just takes a lot of time and effort. Remember that the top users have to keep contributing every day to stay there
    – Pekka
    Feb 17, 2011 at 10:47
  • 49
    +1 for trying to compete with Jon Skeet after 4 days on SO Feb 17, 2011 at 10:48
  • 2
    @systempuntoout: IIRC, on old experts exchange, points expired after a year or so. When I joined there, I spent 6-9 months getting to the top of a few categories, and then had to work to stay there. That was OK, because it wasn't just in theory I could reach the top. I don't intend to compete with the top rank here, I'm merely observing that even if I had the spare time to try competing it'd be quite pointless.
    – Erik
    Feb 17, 2011 at 10:55
  • 10
    @Erik actually my +1 was sincere; when I started on SO a year ago I remember that I was shocked for the difficulties of giving correct answers. If you are already hitting the daily caps in the first 4 days, you surely have talent Imho.. so, an honest +1 for the fearless idea to get to the top and for the nice review. Feb 17, 2011 at 10:58
  • 2
    A comment on the culture here: arguments on the lines of "I liked how [hyphenated-site] implemented this feature" will have to be backed up pretty strongly. There is an extent to which SO defined itself by the ways in which it differs from that other place. Feb 17, 2011 at 12:19
  • @dmckee: IMO, using expiry, the rating reflects how good you are at answering questions + how active you are now. With the current reputation system, the rating reflects how good you are at answering questions + how long you have been around. I prefer the first, as it allows the field to change and gives a more current reflection of who's contributing. That said, I'll take note of your comment on the culture :)
    – Erik
    Feb 17, 2011 at 12:30
  • 4
    @Erik Reputation is not meant to be an activity indicator, but an indicator of expertise. See meta.stackexchange.com/questions/21458/…
    – Pollyanna
    Feb 17, 2011 at 19:29
  • 9
    "Answer 3–4 questions properly, and then not much more point for me to stick around" - if that's all it takes to hit the rep cap, I'm insanely jelous. The other sites in the trilogy 3-4 questions answered properly usually nets a whole 1 or 2 upvotes on a good day. Feb 17, 2011 at 20:42
  • 4
    @Farseeker: It massively depends on the tag. C# you can hit the cap in 3-4 questions easily. Perl takes a little more doing. Hadoop, well, the top user for all-time only has 55 upvotes in the Hadoop tag.
    – CanSpice
    Feb 17, 2011 at 21:32
  • 4
    @Farseeker - Even a fairly popular tag like [iphone] takes a lot more to even come close to the cap. I'm the #1 answerer there, and I've never hit the reputation cap. Also, I've never seen the "post something quick first and refine later" tactics that have been complained about here and elsewhere, so I think it's entirely dependent on the popularity of the technology here. Feb 17, 2011 at 21:43
  • Great synopsis of a new users experience. Some interesting thoughts/ideas in there. But not giving +1 due to the seeming over-emphasis on participating just to earn rep rather than participating for fun and greater good of the community.
    – Dhaust
    Feb 17, 2011 at 23:05
  • One nit-pick: You CAN get reputation after you hit the rep-cap, bounty awards are completely exempt from reputation cap, and also accepted answers are exempt
    – Earlz
    Feb 17, 2011 at 23:36
  • 1
    @BradLarson: Yep, hitting the rep cap in any of the mac/iphone related tags is pretty hard. I recently hit hit 40k rep and I've hit the rep cap 23 times in nearly 2 years (only 11 of those times are from only upvotes). Feb 18, 2011 at 1:20
  • 2
    Erik, this is an excellent question! I think that given the feedback and upvotes, you can see that your 'pardons' are totally unnecessary :) It's always interesting and useful to get thoughtful feedback from new users. So thank you.
    – Benjol
    Feb 18, 2011 at 6:52

9 Answers 9

28

If you are interested in comparing yourself to others there are always the Reputation Leagues.

rep league link

top new users

These show you your position on a weekly, monthly, quarterly and yearly basis so while you can't "compete" (if that's an appropriate term here) on absolute terms with users who've been around for a couple of years you can see if you're doing better than the this week or this month.

6
  • 7
    I think a lot of the feeling of being unable to compete would go away if year totals were the default reputation display, and you had to dig for all-time totals.
    – Erik
    Feb 17, 2011 at 11:19
  • 4
    And @Erik is considered the top new user on that list.
    – Borror0
    Feb 17, 2011 at 19:39
  • this is also linked directly from the /users page as well Feb 18, 2011 at 3:40
  • 1
    I would just like to point out that the screen shots were added by Jeff and I am therefore not responsible for the lack of freehand circles contained therein.
    – ChrisF Mod
    Feb 18, 2011 at 12:50
  • -1000, no freehand circles (hey, I'm just voting on the post as it is, it's not my fault you are the OP on record...) ;-D
    – Pollyanna
    Feb 18, 2011 at 13:09
  • What do you do when you are not listed as a user? Jun 20, 2014 at 14:51
27

What incentives do I, as a new user with technical knowledge in several fields, have for sticking around and answering questions?

I think you are approaching it from the wrong point of view in regards to this question specifically.

Most people answer questions because they want to help other people, enjoy answering questions in their field and learn through that process.

The rep generation is just a small benefit. It appeals to my competitive side like you say it does for yours. But it's not that important at the end of the day.

I came on originally wanting to generate huge amounts of rep, but now I enjoy learning more. I think SO taught me that behaviour. If it was a rep machine, I'm not quite sure if I would have learnt so much, I might have just harvested easy to answer questions to get to the top.

4
  • 1
    This doesn't really make sense - in one thread you complain about how rep is everything and in parallel thread you say rep doesn't really mean much after all :) I agree on the key point though so +1. Feb 17, 2011 at 10:34
  • @mzab The other thread is about how rep influences upvotes. And because the highest upvoted answer appears first, the rep influence factor might be pushing the higher quality content down the page. So it's more about quality of content than the actual rep system.
    – Tom
    Feb 17, 2011 at 10:36
  • 1
    I enjoy answering questions in my field, I do this a lot on specialized forums (e.g. IRC). There's not much competition there, and the questions/discussions tend to be focused within my primary interests. In short, for me this is a more attractive place to just enjoy my field, and that might lead me and others like me away from the very good concept stackoverflow has. I'm looking for an incentive for people like me to spend time here, in addition to my existing specialized forums. Reputation and competition would be this incentive.
    – Erik
    Feb 17, 2011 at 10:39
  • 6
    Gullen I think Erik is right though, I think it is speed which has more effect than rep of the poster, fast answer often generates 3 or so rep within a minute of being posted, once you have 3, people often don't even bother reading (and upvoting) other answers. Feb 17, 2011 at 10:39
14

Is the one-liner followed by edits a desired behavior? Accepted behavior? If not, are there any mechanisms in place to discourage this?

Fastest Gun in the West Problem --> Good description of the problem
Is the fastest gun in the west solved? --> Good overview of the data and statistics behind the problem

Is it intentional that newer members should not be able to compete for the top spots? If so, is there a rationale behind this decision?

Can the top contributors ever be challenged? --> You can join the top users, but to eclipse them you will need to put significant effort into the game. However, it's certainly possible.

Why is the reputation cap as low as 200? Answer 3–4 questions properly, and then not much more point for me to stick around?

Will reputation spiral out of all control? --> Without the cap reputation becomes linked to time on site, rather than expertise.
The rep cap forces self-leveling of user participation --> Stack Overflow prefers steady users, rather than binge users.

Why is the reputation cap system not favoring active over long-term users? Why not e.g. "1000 for stuff you posted today, 500 for stuff you posted this week, 200 for anything older"?

Reputation doesn't expire. What you are looking for is an activity indicator, and that can be had at the Reputation Leagues.

What incentives do I, as a new user with technical knowledge in several fields, have for sticking around and answering questions?

I could pontificate, but the reality is that the reasons one sticks around are their own. However, it turns out that the system is very addictive for many people.

4
  • Interesting data on your second link, but said data says "if you post the 5th answer you only have a 2% chance of getting your answer accepted." The implication is that posting later decreases the chance of being accepted; that may be so, but you can't tell, since it could be that questions with 5 or more answers mostly have identical answers, and that harder questions only attract on average three answers.
    – user142852
    Feb 17, 2011 at 21:41
  • My point being, you can see a trend, yes, but you can't directly attribute position to likelihood without understanding all the other underlying variables. Sorry to use the big voice, but correlation does not imply causation. Ergo. Vis-a-vis. Concordantly.
    – user142852
    Feb 17, 2011 at 21:45
  • @Nine Yes, there's certainly a lot more factors to deal with than a simple first glance might suggest.
    – Pollyanna
    Feb 17, 2011 at 21:46
  • In fact if I had to bet, I'd guess run of the mill questions attract lots of answers mostly identical, to which the OP just says the first one whilst harder questions attract one fast but hasty answer, then a few better answers. No statistics to back that up, of course, but if it were the case you'd be looking at classic confounding: en.wikipedia.org/wiki/Confounding
    – user142852
    Feb 17, 2011 at 21:48
11

Is the one-liner followed by edits a desired behavior? Accepted behavior? If not, are there any mechanisms in place to discourage this?

Depends on the scenario. Posting something quick to be in there first then editing is one such intention, but bear in mind sometimes people think: wait, I should have expanded on that, what I gave isn't enough, or I feel I should add more. People also correct mistakes they think they made and add more at the same time. Editing post answer isn't necessarily a bad thing; I've pointed out mistakes and downvoted, then removed said downvote when the question was fixed.

As for the fastest gun in the west problem, well, posts are optionally displayed in various sortings, one being random, so being there first is no guarantee.

Is it intentional that newer members should not be able to compete for the top spots? If so, is there a rationale behind this decision?

I don't understand what you mean by this. I can give an example of one user I first saw post an excellent answer in december with about 200 rep who now has 12k. There is nothing stopping you earning rep except the time you can commit to be here and the quality of what you post.

So live with it. Rep / answering questions here doesn't make you a good programmer. Writing code does. When you come here, remember that. There are experts on here with less than 10k rep.

Why is the reputation cap as low as 200? Answer 3–4 questions properly, and then not much more point for me to stick around?

Something's gotta stop Skeet!

Joke aside, this prevents the above to some extent. If all you have to do is sit on here answering questions all day, it's much easier to grow your rep to massive proportions. So this helps hold you back. There are badges for hitting this on 150 days total, if you're so inclined.

Why is the reputation cap system not favoring active over long-term users? Why not e.g. "1000 for stuff you posted today, 500 for stuff you posted this week, 200 for anything older"?

That's an interesting idea. However, my reputation very rarely changes over old posts. Sometimes I get the odd +10, or the odd -2, but not very often. And it's a vote that counts today, not for when the post was cast.

What incentives do I, as a new user with technical knowledge in several fields, have for sticking around and answering questions?

Absolutely none. It's a choice. You either choose to come, with all the good things and bad things about SO included, or you decide it isn't for you.

That said, some of us have reasons for sticking around. You've got one of mine in your question:

I quickly pick up on the need to be very exact in answers

This has two consequences: usual junk that comes with internet forums just isn't present here. By junk I'm talking about having your question answered by 20 people all with an opinion on the technology you're using, how much "Micro$oft" suck, how great Ruby is and how rails has truly given them the ability to, y'know, just focus on features and other such noise.

(Ruby used because that's a tweet I saw somewhere. The point is fanboyism of languages in forums is at best mildly irritating, at worst it stops your post being answered. I have no opinion on ruby/rails/Microsoft, I just don't like users who think I should be using {x} instead of {y}.)

The second consequence is that, well, be prepared to learn. I reckon you can learn more on here about programming than in any degree programme in my country of residence and then some. I've never seen any other internet forum where people so regularly quote the standard for their language, or know about compiler differences, or... in short, the expertise is at a very high level here.

That's why I stay. I'm in the company of people who are much better at coding than I am, and I aspire to learn from them.

5
  • 3
    +1 for "be prepared to learn". That's the fundamental answer to the main thrust of the OPs question. Feb 17, 2011 at 19:24
  • Also, I'd add that a one-liner followed by edits is definitely intended behavior. One of the things that makes Stack Overflow so great is how well it does at promoting both speed and quality, while sacrificing on neither. You get good answers in minutes or hours (sometimes even just seconds) instead of hours or days, and this behavior is a big part of what makes that possible. Feb 17, 2011 at 19:27
  • One more possible advantage one-liner + edits. If you write a longish answer, you might find you have just duplicated someone else's efforts. The one liner might reduce the number of redundant answers you get.
    – Aryabhatta
    Feb 17, 2011 at 20:43
  • @Joel Interesting. I do think it is absolutely the right idea. If you couldn't edit to fix those typos or that nonsense you asserted because you're still on SO at 2am, or to further explain that point you realise in hindsight isn't very obvious, people would be less inclined to post answers and the pace of things would slow up massively.
    – user142852
    Feb 17, 2011 at 20:54
  • 1
    @Moron True. Also, to say there's no incentive not to better a question once answered is slightly inaccurate. If you're not adding noise and are genuinely improving, there's always: stackoverflow.com/badges/62/populist
    – user142852
    Feb 17, 2011 at 20:58
6

Part 2, way too long for comments :)

Several of my questions have been answered, and I'm understanding more of the reasoning behind certain mechanisms now.

@Joel Coehoorn: Your explanation of why the first-post with edits are encouraged makes perfect sense, thanks.

On the 3-4 answers per day: Given crazy upvotes on dead simple questions like What's wrong with this C++ program? it's not very hard to hit the cap, at least for C and/or C++ tags. (BTW: I had already hit the cap when I answered that one, too bad :) )

I've now figured out why I have this impression of Stack Overflow as a very competitive site, as opposed to what many have stated is the intent. First of all, there is a reward system. That's the impression a new user gets of reputation, it's a reward. So, this leads me to believe the users should strive for rewards. Next, there's two major types of questions, the simple ones that are easy to answer, and the hard ones that require thought and research (I'm ignoring the ones that cannot be answered but still are). A little experimentation quickly tells me that you get a lot more reward by answering several simple than you get by answering one or two hard. So, since I should strive for the offered rewards, I start answering simple questions. And now the competition becomes visible. I need to answer faster and better than the next guy in order to get this reward. I adopt first-post + edit, I watch and refresh for new unanswered questions, I'm generally in a "react fast" mode, I'm competing.

After reading answers posted here I have a much clearer understanding of what Stack Overflow intends to be, and of what reputation is intended to be. But for me, a random new user, it doesn't work like that. So, in my original question I provide opinions on why the reputation system doesn't do what I believe it is intended to do. Now I'm shifting more towards thinking that the reputation system doesn't do what it's been claimed it really is intended to do either, it's stuck somewhere in the middle.

Looking around at other meta discussions, I see that some considers it somewhat problematic that there's no way to differentiate in between simple and hard questions. Such a differentiation, although I have no idea how it could be implemented apart from question vote count, would have given me a different impression of Stack Overflow from day one. I'd have focused on questions that stretched my knowledge without overextending, spent more time on them, and tried to come up with higher quality answers. I'd find my level, and with a good "scoring" system I'd earn rewards indicative of my level, whatever it is.

On the various variants of "You either choose to come, with all the good things and bad things about SO included, or you decide it isn't for you". Yes, that is quite obvious, and of course I'll stay or leave based on my own thoughts about SO. But, I would believe that one of SO's goals is to attract as many users as it can?

Making a parallel to the industry I work in, MMO game development: We have our share of veterans, who've played a lot and put work into their characters. They expect and deserve to get something back, and that is reflected in developed characters that can easily beat the newbies. And we have our share of new players, who we want to stay and play. These need to see that it's possible to catch up with the veterans at some point, or they will ragequit :). We also have a share of casual players, who will stay because they enjoy it, regardless of whether they can catch up or not. Making this work is a balancing act, which is made simpler because the veterans understand and accept that they cannot and should not stay ahead for ever, as that would end up killing the game for lack of new players.

But then, if user competition isn't a driving force behind SO's success, this parallel doesn't apply much.

On the suggestion of differentiated caps, I still think this would work. It would reduce the time needed for an active new member to "catch up" with an inactive long-time member, and change nothing else. It's values such as these we would balance in a game to keep both our veterans and newbies happy. What I'm unsure of is whether this should even be a goal for SO, if the competitive user group is large enough for this to be worth the effort.

Finally, if it is acceptable, I'll leave this open for a while and revisit when I have more than 4 days of experience. My intent with the original questions, and the ones I pose in this post, is to give a perspective from a new user, perhaps exaggerate a little to make points, but not really to critizize or ask for changes as much as trying to understand and perhaps provide useful feedback. I expect some of my opinions to shift as I learn more about SO, and so far I enjoy this place quite a lot so you're not rid of me just yet :)

3
  • 3
    Your observations are very insightful, and I appreciate the time you've taken to post them. I think your point, "We have our share of veterans, who've played a lot and put work into their characters. They expect and deserve to get something back... we have our share of new players, who we want to stay and play... We also have a share of casual players, who will stay because they enjoy it" is very important, and we do need to work on "balancing the game" so that people with all these play styles (and others) feel they are getting something out of it.
    – Pollyanna
    Feb 18, 2011 at 1:47
  • Votes wouldn't help you differentiate between hard and easy. The most upvoted questions are generally the most popular (and even most silly). The hard ones are (possibly) the ones which are unanswered and still getting a trickle of upvotes even months later.
    – Benjol
    Feb 18, 2011 at 7:04
  • Erik, I think you're also hitting on some of the problems specific to StackOverflow. The StackExchange model has grown beyond just SO; the other sites (ServerFault for example) don't have nearly as many users. I don't see the first-post edit situation there or the easy vs hard questions nearly as much. (Easy questions yes, but the audience is very different) So my point is, you're experiencing some of StackOverflow's "growing pains", things that much longer terms users didn't have to deal with when we first found the site. Feb 18, 2011 at 12:57
4

I think you are right on the "one-liner" aspect, speed really has very significant impact on the rep awarded. I don't see a way to fix it though (nor do I think it is actually wrong). I have seen inverse effect several times just in last days - I posted a quick answer, then detailed it a bit, got a few rep and then (after hours or even later) someone came with uberdetailed answer which then reaped the upvotes (including mine) and got accepted - so quality still matters.

What bothers me more is that the easiest questions reap the most rep - I got 7 upvotes for this one-liner trivial answer on rather dumb question (it got closed after all), but the really challenging answers (which often require some digging through my source code archives and even some research) often get exactly one upvote (from the asker, I assume).

This makes harvesting easy questions so much more effective than answering the really tricky and interesting questions (which are what make me come back to SO, because offer the greatest opportunity to learn).

1
4

The main thrust of the question seems to be one of motivation, and while earning lots of shiny points is intended to be a motivator I think for most big contributors it's really a small part of it. The main or real motivation here is two-fold: learning, and helping others. You can also use Stack Overflow to help build up your personal brand. I know several stack overflow users have been able to turn good Stack Overflow profiles into better jobs or other opportunities.

I also want to address this point:

Is the one-liner followed by edits a desired behavior? Accepted behavior? If not, are there any mechanisms in place to discourage this?

This is, to my mind, definitely a behavior worth encouraging. One of the things that makes Stack Overflow so great is how well it does at promoting both speed and quality, while sacrificing on neither. You get good answers in minutes or hours (sometimes even just seconds) instead of hours or days, and this behavior is a big part of what makes that possible.

Finally, you wonder why the reputation seems to favor long term vs active users. That's just silly. Those at the "top" are there because of dedicated work over the period of years, and the "competition" is fierce enough that it wouldn't take long for most to fall rapidly if the participation declined. This kind of dedication is definitely worth rewarding, and changing this to favor active users would greatly devalue their work.

2
  • 2
    "the "competition" is fierce enough that it wouldn't take long for most to fall rapidly if the participation declined." I stopped using Stackoverflow for nearly a year, and dropped from the first page to the third page. Now I'm on the 7th page and dropping with no new contributions, despite getting 1-2k per month on old answers. There are a lot of new sharks out there. So yeah, stop participating and expect a precipitous fall.
    – Pollyanna
    Feb 17, 2011 at 19:41
  • Heh, I've been using SO/MSO for about 1.5 years now, and I've seen many new people pass me by. I don't contribute a whole lot though.
    – Earlz
    Feb 17, 2011 at 23:46
4

This behavior you've observed, which tends to reward fast answers over good answers, is specific to the high-activity tags on Stack Overflow. I haven't seen it on low-activity tags or on other sites.

Do not fall under the delusion that high reputation means better answers. High reputation is primarily a sign of high activity on popular topics. The more difficult a question is, the fewer views it gets, and so the less reputation you can expect to get for answering it.

There are users who don't care about reputation. These people take whatever time it takes to write a good answer, or eschew answering if they don't think they can give a better answer than the existing ones. They go looking for questions where their answer will help, not for questions where their answer will garner reputation. They like reading other people's answers when they can learn something from them, and happily vote for competing answers when they're good. They will research their answers, even it “wastes” time, because they like to be challenged. I find this way of life a lot more rewarding. (Especially as you can't convert reputation into dollars or ponies.)

(Oh, and if you want to be the top user somewhere, my advice is to find a topic that has a Stack Exchange site of its own that matches your expertise well enough that you can rise to the top.)

3

The reputation cap is only for upvotes. You can still receive reputation for accepted answers and bounty - so if your answer is accepted in question with 500 bounty, you will receive 515 within single day + max 200 for upvotes.

There is also a lot of place for competition. Each tag has its own top users league - you can compete in the last 30 days or in all time. If you are expert in several technologies you can easily get your "competition".

5
  • Are reputation cups issued to Meta users or something? I didn't get one!
    – Stephen
    Feb 17, 2011 at 19:27
  • @Stephen: Pretty sure Ladislav meant Reputation cap, but since SO doesn't like suggested edits of less than 6 changed characters, I'm leaving it be.
    – Powerlord
    Feb 17, 2011 at 19:51
  • I forgot to wave my facetious flag before making that comment. ;)
    – Stephen
    Feb 17, 2011 at 20:54
  • 3
    @Stephen Reputation cups prevent accidental reputation injuries when worn properly.
    – Pollyanna
    Feb 17, 2011 at 21:40
  • 1
    @Stephen, you can only flag as "noise, offensive or spam", maybe you want to open a feature-request? :)
    – Benjol
    Feb 18, 2011 at 6:57

You must log in to answer this question.

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