Looking at my rep tab it feels like I am getting a lot of upvotes for older answers recently. In the last month I have only given about a dozen answers, which earned barely above 30 upvotes, yet my rep has increased by more than 1000. I don't feel like there was a comparable rep reap on old answers a few months ago.

Yet I was discounting this as superstition, but then I remembered that I had seen others mentioning the same.

Is this still just the four of us having superstitions or has there been any change I missed that tends to promote popular old answers more than they used to be promoted?

  • 14
    The beautiful part about SE is that people usually have the same problems over time, so they keep finding old answers and upvoting them for helpfulness. I know that I routinely upvote answers that are anywhere from a month to two years old, and don't see any problem with that. Congrats on an extra 1k for free. – jcolebrand Jun 12 '11 at 20:04
  • 2
    @jcolebrand: But that isn't something which started a few months ago. I am asking about an increase in this. – sbi Jun 12 '11 at 20:08
  • 1
    @sbi When I end up on a question on stackoverflow from google and it solves my problem I upvote the question and the helpful answer. I believe others do the same. – steenhulthin Jun 12 '11 at 20:21
  • 1
    @Steen: <sigh/> Really, what's so hard to understand about it? (Could someone please improve my question to make this clearer? I don't know how.) I bet hundreds of thousands of users have done so 4 months ago. Why are they doing it even more now? – sbi Jun 12 '11 at 20:34
  • @sbi Are those upvoted answers related to similar topics perchance? (both for you and the others?) – trutheality Jun 12 '11 at 20:48
  • 2
    @sbi All I can guess is that the network has gotten more popular, or that things are better shown on google now, or something similar. – jcolebrand Jun 12 '11 at 20:56
  • 1
    I've been seeing some upvoting of my older answers too. I'm not unhappy about it. My immediate take on it was "I don't think it amounts to 70 upvotes a month". On second thoughts, though, over a 30 day month, that needs just over 2 upvotes per day; maybe it does. It is not self-evidently gratuitous upvoting, I'm glad to say. – Jonathan Leffler Jun 12 '11 at 21:05
  • 2
    Maybe you can give us informations about the topics, which were upvoted? Maybe increasing interest in such topics? Or could it just be by chance? How many upvotes did you have last month? How many the month before? It might be normal fluctuation, but that's only a guess. – user unknown Jun 12 '11 at 21:22
  • @user: I'm mainly to be found in the c++ tag. Through the last three days, I got 6 upvotes for stackoverflow.com/q/2652223#2652223, stackoverflow.com/q/4669936#4669936, stackoverflow.com/q/2139254#2139254, stackoverflow.com/q/3626541#3626541, and stackoverflow.com/q/4172961#4172961. I haven't checked the upvotes I got a few months ago against those I got the last month (which is why I asked whether it's just superstition). How do I find out how many votes I got, say, January 2011? – sbi Jun 12 '11 at 23:58
  • @trutheality: The upvotes necessarily are closely related, because I'm the opposite of a generalist and the overwhelming majority of my answers are in the c++-tag. But I have seen what might be a proportional number of upvotes for old answers in other tags (like svn), too. – sbi Jun 13 '11 at 0:09
  • 1
    @jcolebrand: If there'd be a sudden jump in the popularity of SO that can be seen in other numbers (like page hits, questions, answers, etc.), I'd gladly put this to rest. That is exactly what I was asking for. – sbi Jun 13 '11 at 0:10
  • Your Userpage, Reputation, Graph: (meta.stackoverflow.com/users/133368/…) includes all reputations, for questions too, and downvoting penalties are included too. You can view rep-by-time, and browse the pages, or modify the scope of the graph and use mouseover, to find out more details. – user unknown Jun 13 '11 at 0:16
  • @user: I can see the graph just fine, but I miss a sum of rep for the marked timespan. – sbi Jun 13 '11 at 0:31

Maybe related to the Google page ranking changes which makes SO appear higher than it used to compared to the copycat sites. Apparently the changes were rolled out to all English language google users about 2 months ago, and they seem to work for me, so if I search for any issue now I'm much more likely to get the SO site as the top result, which means that I might give an upvote if it's useful and/or interesting.

  • Thanks. That would make sense, and it does seem to fit my feeling for when this happened. It's the first answer that addresses my actual question (why the increase?) rather than trying to lecture me on where upvotes come from generally. – sbi Jun 13 '11 at 21:04

I noticed some of my older answers got a bunch of attention when the Vox Populi badge was introduced on May 9th 2011, but it's leveled off since then.

I find that most passive rep comes from questions or answers where it's a frequently-brought-up issue. This is particularly true if a duplicate question is asked where you've posted in the original. The more paths you have back to your posts, the more opportunity there is for people to vote on those posts.

  • I'm a member for almost two years. I know about these "frequently brought up issues". I was asking because I felt they are suddenly more frequently brought up. That badge thingy, however, was exactly the kind of information I was after. – sbi Jun 13 '11 at 0:11

Unless you see some relationship among the topics of the answers, I would guess that there really isn't an answer to "Why more upvotes over the past month than over the month before that?"

It's coincidence, serendipity, chance, emergence, whatever name you'd like to give it. It may decrease next month. It could simply be due to increasing SO membership and usage. Maybe your number of answers and the number of members have hit some critical value such that your answers get viewed many times per day.

Either that or you have a fan/freak, who is carefully upvoting you a few times a day to avoid the vote fraud filters.

  • (A) It's not a fan, because I got several upvotes for the same questions in short time. (B) The number of answers of mine have most definitely not exceeded some threshold (a concept I doubt anyway), because I haven't answered much at all since about last December. (C) I did file this under coincidence until I felt like it's been this for two or three months now. Even then I meant to disregard it, but then I vaguely remembered seeing others reporting this, too. – sbi Jun 13 '11 at 0:05
  • It's certainly an interesting phenomenon. It wasn't quite clear from your question that you thought it was becoming a trend -- sounded like you were just talking about this month. I wonder if it's worth the time to try to analyze it using the SEDE? I'm not even sure how you would go about it... – jscs Jun 13 '11 at 0:08

By my rough calculations, in the last thirty days you earned 941 reputation, of which 360 was from current upvotes, -2 from current downvotes, and 45 from current accepted answers. This gives you 538 background repuation gain or roughly 18 background points per day.

Less than two background votes per day doesn't strike me as surprising given that you have 26 questions (one of which is starred by 59 users) and 1372 answers, some of which are probably referenced by current answers. Heck, I have a fraction of that and I probably get a background vote per day.

In light of these numbers, I think your tenure and the size of Stack Overflow are all that is needed to account for this as just normal voting patterns.

  • Really, is that so hard? I was not asking where my upvotes come from. I'm not exactly a newbie and have had time to form my own opinions on that. I was asking where the increase in those upvotes comes from. You haven't added anything to this question. – sbi Jun 13 '11 at 21:03
  • 1
    @sbi: I did add something: exactly how much background upvoting are you getting? Is it surprising? I believe your perceived increase in background upvotes coincides precisely with your decrease in activity, i.e. the past few months. You have only an impression that it has increased. – Rick Sladkey Jun 13 '11 at 21:39
  • 1
    @sbi: For comparison: In the month of march (by my rough calculations), sbi received 45 upvotes on old answers, 4 on old questions, 12 on new questions and 97 on new answers. And 4 accepts. The "old" part of this is about the same level as now, no rise here. – Paŭlo Ebermann Jun 13 '11 at 22:25
  • @Rick: No, these two don't correlate. I have the feeling this increased about two to four months ago, whereas I stopped contributing so much about six months ago. – sbi Jun 15 '11 at 17:59
  • @Paŭlo: I'm not sure March wasn't already showing an increase. – sbi Jun 15 '11 at 18:00
  • @sbi: I am not a SQL guy but I bet a query could be written against data.stackexchange that could provide the data for an accurate graph. That would be interesting viewing! – Rick Sladkey Jun 15 '11 at 18:02
  • We would first have to define the exact difference between votes for new and old answers. In the ones I was looking at, I found some dubious cases. – Paŭlo Ebermann Jun 15 '11 at 18:15
  • @Rick: That would indeed be. Unfortunately, to say that I am "not an SQL guy" would be a so much of an euphemism, it would almost be lying. I'm the opposite of an SQL guy, and I hate DB stuff. :( (That and my aversion against web programming pretty much shuts me out of 90% of the job market.) – sbi Jun 15 '11 at 18:17
  • @sbi, @Rick: I tried it a bit, have a look at this query. Does one of you want to create a graph from this? (Do you also want a monthly aggregation, or a sliding window?) – Paŭlo Ebermann Jun 15 '11 at 20:57
  • @Rick, @sbi: Here is the monthly aggregation (for your user ID and 10 days cutoff). It looks like in february you got most votes (and reputation) for old questions and answers, actually. – Paŭlo Ebermann Jun 15 '11 at 21:10
  • @Paŭlo: That's cool! Favorited. :-) – Rick Sladkey Jun 15 '11 at 23:38
  • @Paŭlo: That's indeed cool. February, however, is likely the month when someone reddit'ed one of my C++ FAQ entries. Any real analysis should leave that one out, because it's certainly not representative for the trend I think I'm seeing. However, your query at least makes clear that, if there is such a trend, it seems to be less dramatic than I felt it is. – sbi Jun 21 '11 at 17:07

You must log in to answer this question.

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