32

It seems that all good questions that could be asked have been already asked. Also it seems all good answers that could appear have already appeared. Existing questions/answers base probably covers most of the programming topics. And, therefore, it is more difficult to ask an interesting question that bothers everybody than, for example, couple of years ago when SO was comparatively empty. Do only I get this feeling or not? Are there any metrics for proving/disproving that it is more difficult to get more reputation on SO as time passes? I want to get rid of the feeling that all questions have been asked and answers have been already answered.

UPD: Maybe Jon Skeet would be great candidate for the experiment of getting statistical data about SO reputation dynamics? We can track dynamics of his reputation over the last (two) year(s). Maybe it could tell us much. Is there any right way to get such data easily?

UPD2: Here is some info about Jon Skeet dynamics: http://stats.blogoverflow.com/2011/08/does-jon-skeet-have-mental-powers-that-make-us-upvote-his-answers-the-effect-of-reputation-on-upvotes/

11
  • 2
    It depends on the site, but I agree that some particular questions (regardless of the site we're taking into account) get asked fairly soon. Maybe it's still possible to ask challenging questions, but certainly it gets harder.
    – Alenanno
    Feb 14, 2012 at 11:51
  • 5
    Wouldnt' this question be answered better by statistical data, instead of speculation? Sure, you'd have to define "difficult", but that's not a bad thing. Maybe some stats on reputation increase per user or something like that?
    – Nanne
    Feb 14, 2012 at 12:42
  • 1
    Also depends on the tag you prefer to answer in as well. I'd recommend filtering on your tag with answers:0 as well and select the newest tab and bookmarking it; this way, you will always see the newest unanswered questions. It's just a way of life at SE, that people who answer first (with a real, quality answer) tend to score highest.
    – casperOne
    Feb 14, 2012 at 13:47
  • @Nanne: that's exactly what I asked for: Are there any metrics for proving/disproving that it is more difficult to get more reputation on SO as time passes?
    – altern
    Feb 14, 2012 at 15:41
  • @Nanne: I even think that Jon Skeet would be great candidate for such experiment. We can track dynamics of his reputation over the last (two) year(s). Maybe it could tell us much :)
    – altern
    Feb 14, 2012 at 15:42
  • @altern That is a fantastic idea. One assumes the quality of his answers have probably been pretty consistent, and he has a long history of consistent activity. Feb 14, 2012 at 17:15
  • @AndrewBarber and altern - Very related discussion here Feb 14, 2012 at 17:26
  • I think it's not much harder to get rep, but much, much harder to get gold badges for questions/answers. Feb 14, 2012 at 19:08
  • 1
    Can you define what you mean by more difficult? And more difficult for whom? New users? Experienced users? Feb 14, 2012 at 19:26
  • 1
    @AndrewBarber - Wouldn't his statistics be hard to judge from, given the frequency with which he hits the reputation cap? It would seem that he would have hit a ceiling at some point. I posted my own numbers for the last three years below, and I've never hit the reputation cap over that period. I've seen a steady increase in reputation gained per year, despite spending less time answering in the last couple of years. Feb 14, 2012 at 20:57
  • @BradLarson Good point! The reputation cap would be a big issue. I expect he would still hit the cap easily, each day. It would have to go by his votes, not his reputation Feb 14, 2012 at 21:04

5 Answers 5

6

I don't know how its been for everyone but if you examine this query Jon Skeet he has been receiving more votes per week with roughly the same number of posts.

This of course doesn't really mean much since some of this is money from jam and it only correlates posts and votes. Not effort and votes

enter image description here

1
  • Thanks for the "money from jam" link, it's fun (and a bit humbling)...
    – sarnold
    Feb 14, 2012 at 23:41
19

There are still some diamonds appearing from time to time.

I agree that over time, there is general decrease in votes - not the total votes, but "votes per question".

As the tools and programming languages themselves keep evolving, people will always have new stuff to ask so in my opinion "all good questions that could be asked have been already asked" is not the real reason for the decrease in votes.

I think the real issue is the overwhelming amount of questions, and major part of them are junk. It's just too hard to wade through all the junk to find the good questions.

When Stack Overflow was younger, there were much fewer questions asked so the "hard core members" could read them all and quickly upvote the good ones. Now going through all new questions is a full time job. Nobody can do it anymore.

Jeff has done everything in his disposal to fight this and improve the quality of the questions, but in a place so big it's mission impossible.

Bottom line - yes, it's harder to gain reputation from questions but your main concern of "all questions have been asked and answers have been already answered" can be dismissed in my opinion and is not the reason for this difficulty.

7
  • 1
    Your example "diamond" question probably has a higher propensity for upvotes since it was written by a mod. People seem to be more generous with their upvotes when the question comes from high-rep users. Of course, maybe that's because mods and high-rep users know how to write and ask good questions. :P
    – Rob Hruska
    Feb 14, 2012 at 12:54
  • 1
    Also, nothing like joining SO and hitting the rep cap almost immediately. Actually, I guess he's still unregistered, but still...
    – Rob Hruska
    Feb 14, 2012 at 12:57
  • Good point about high rep members having more "trust" but still, anybody can come with a good question and any good question can be overlooked or not get the attention it deserves. As for Steve Souders, I believe Sam invited him, don't think it's just a coincidence. Feb 14, 2012 at 13:03
  • Ah, that would explain. Seemed interesting that he'd just randomly come and answer so shortly after it was posted.
    – Rob Hruska
    Feb 14, 2012 at 13:09
  • @Rob yes indeed, especially when he's "heavy gun" in his area of expertise, not just a simple person. :) Feb 14, 2012 at 13:16
  • @RobHruska I have lots of questions, even a 0 vote one I asked in the last week. The question linked is clearly an outlier, I do not think the fact I have a diamond played much here.
    – waffles
    Feb 14, 2012 at 20:55
  • @waffles oh my, didn't realize until now that Sam Saffron is you! Not that it matters, lol.. :) Feb 15, 2012 at 7:43
14

It's indeed more difficult to earn reputation than a couple of years ago.

No, this is not caused because all questions are already been asked. They are still asked. It's just caused because there are way much more users who you have to "compete" with.

I myself was previously more active in [java] tag, for example. But these days I can barely enter a complete answer to my satisfaction, or I see the orange notification bar and I see that someone else (often more than one!) has already posted (and gradually edited) an answer along the same lines and even already received a bunch of votes on that. So I just press Ctrl+W at that point.

Well, try finding a "niche" tag where you're really good at and continue there. You can just post answers on questions which have not received answers for hours, so that you can take the precious time without hassling to be the FGITW. For me, that's among others [jsf]. I don't earn much reputation per post on a daily basis, most answers have 0 or 1 votes and rarely reach 2, 3 or more votes per day. It are the accepted answers which counts and reputation earned by votes on one of those >10K old answers.

6
  • Thanks for the suggestion about the niche. I try to get a badge on [svn] tag. But it is still so damn difficult even though there are almost no competitors. It seems that noone interested in the specific topics. And... all good questions that are interesting for most part of the community have been asked
    – altern
    Feb 14, 2012 at 15:33
  • 6
    If you post a "world class" answer which makes the answer interesting, understandble and useful for more people than only the OP, then you will slowly but surely continuously get votes for it, even though it's only one vote per week.
    – user138231
    Feb 14, 2012 at 15:37
  • even though answer might be great, nobody digs for all these stupid questions, therefore nobody sees both the question and the answer.
    – altern
    Feb 14, 2012 at 15:45
  • This is a very, very good perspective. Most of my answers are in [asp.net], but I started out just answering questions tagged with [AjaxControlToolkit], because no one was answering them. +1 from me. Feb 14, 2012 at 15:50
  • 3
    @altern Maybe everyone is using git :D... runs away fast! Feb 14, 2012 at 16:16
  • this is pretty much why I live on the dapper tag and occasionally will answer an abandoned question in my interesting stream
    – waffles
    Feb 14, 2012 at 20:57
8

I think this will depend on the areas that you frequently ask and answer questions in. For more established, slower-moving topics like languages ([java], [c++], etc.) it might be harder to find good questions that haven't been asked yet, or compete against faster answerers. However, I believe there are plenty of opportunities to earn reputation in niche or more rapidly evolving areas.

In mobile development, the frameworks and operating system features keep changing on an almost monthly basis. Every time something new is added, there are plenty of new questions that can be asked about fundamental aspects of this technology, or common problems that people face. Adding definitive answers can rack up a lot of votes for you.

For example, within the iOS arena, Apple rolled out a new language feature called Automatic Reference Counting (ARC) in mid-June of 2011. Within a few days, people were asking all sorts of questions in the tag, and any good answers were receiving plenty of votes. In ~8 months, I've received 219 upvotes on 21 answers just in that narrow tag, with the best answer sitting at 73 upvotes by itself. This happens every time Apple updates something. Despite this, many of these questions can sit for a day or more without a good answer because of the lack of people with experience in the new thing.

Beyond rapidly moving areas, there are plenty of niche or more technical questions that are sitting there unanswered. While these may not get the views of simpler or more common questions, I've found that they can have very long tails for voting activity. I have answers that I provided three years ago still earning votes as people stumble across the same obscure issue I posted a solution for.

Looking back at my own personal history over the last ~3 years (since I joined in September 2008):

Year  |  Reputation Earned That Year
------------------------------------
2008  |  226
2009  |  10687
2010  |  16622
2011  |  22460

Despite being more selective with what I answered starting in early 2010, I still see an acceleration in reputation earned. If 2012 keeps up the pace from the first month and a half, I'll earn 28000 points by the end of the year. This is despite never having hit the reputation cap a single time due to votes.

Much of this due to a combination of the two factors I describe above. The answers I've provided in more technical areas accumulate votes regularly over the years, and the few questions I've answered in hot new areas have led to some large vote counts. As the popularity of Stack Overflow grows, I imagine the number of votes from these areas will grow as well.

In the end, though, you should just focus on providing quality content, both in questions and answers. While the imaginary points act as a fun motivator, what's really important here is the act of helping out both current and future readers.

2

I agree with you that it is harder to earn reputation now than before (I haven't been using SO for that long, but long enough to see something)

I don't think its as much that the good questions have already been asked. It's not like there aren't any advances in the technological field. New stuff will come out and there will be more questions to ask about such and such.

I think the problem with SO and reputation as of now is its users... Theres so many of them. Its not the users fault, but theres so many questions being asked on SO that one hardly has time to spot out a good question, and because there are so many questions, most questions receive ALOT less attention than they should.

Questions hardly have any time to get peoples attention before they are on page 5. I think something should be done about this... Like maybe a holding area for questions, and allow questions to be asked and maintain their spots on the website for a while before, after a period of time, new questions are added on the page. So this would stop the page from being full of way too many questions...

1
  • 2
    So, uh, having a rate-limiter that limits the site to 50 questions per hour or something and hand out question tokens to users on a lottery basis? Auction basis? Reputation basis?
    – sarnold
    Feb 14, 2012 at 23:37

You must log in to answer this question.

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