I came across this question a few weeks ago. The question How to properly document S4 class slots using Roxygen2? has 199 upvotes (and 1 downvote). While that's not a surprising total for popular questions, and I'm not saying it doesn't deserve +198 score, it seems odd because

  1. It has a mere 6978 views. Similarly-upvoted questions from the same time period have tens of thousands to hundreds of thousands of views (with one exception that I see). Oops, I got the year wrong
  2. This looks like a fairly specific topic (a documentation tool for the specialized statistical language R). Most top questions from that period are about very-widely-used languages/software like .NET, Java, the IPhone, Ruby, Git, etc..
  3. Nobody bothered to answer the question. Moderators pointed out that there were several spam answers that got deleted.
  4. Arjan pointed out that the upvotes were suspiciously evenly distributed over time. I'd expect that questions normally get the bulk of their upvotes shortly after creation, and then lose currency and only get upvotes on occasion. (am I right about this?)
  5. Some other questions on similar topics asked around the same time have +16, +15, and +3 upvotes. In particular, How to properly document S4 methods using roxygen2 looks very similar in scope (just with methods in place of class slots) and topic (writing R documentation), but has far fewer votes.

Are these observations cause to be suspicious about this? or has it just legitimately accumulated the votes?

  • 5
    Also 6 deleted answers. This is apparently a magnet for... some sort of attention.
    – Shog9
    Jun 6, 2012 at 5:34
  • 1
    Rather umm - interesting deleted answers too...
    – Paul
    Jun 6, 2012 at 5:39
  • @PaulPRO: what do you mean? Jun 6, 2012 at 5:40
  • 1
    @Mechanicalsnail - None of the deleted answers are actually answers. Two are questions, two look like the content was edited so it's hidden from 10k users, and one is just nonsense.
    – jmort253
    Jun 6, 2012 at 5:44
  • 3
    Unless you have 10k+ rep on SO, @Mechanicalsnail, You can't see the six (profane, useless, and spammy) deleted answers to that question. Jun 6, 2012 at 5:44
  • 2
    The edited ones were both spammy ads for a website that sounds sunglasses.
    – Paul
    Jun 6, 2012 at 5:47
  • The SEDE query is wrong—the question you're asking about was posted Sept. 2011, not Sept. 2009.
    – waiwai933
    Jun 6, 2012 at 5:51
  • @Paul website that sounds sunglasses? That's.. interesting! ;) Jun 6, 2012 at 5:54
  • @ShaDowWizArd Certainly sounds that way. Jun 6, 2012 at 6:01
  • @ShaDowWizArd Haha wow I'm tired. <-- And I had to reread this because I put "wow I'm tied". Time for bed haha
    – Paul
    Jun 6, 2012 at 6:06
  • @Paul been there, but taught myself to always read twice whatever I post to avoid such things. :-P Jun 6, 2012 at 6:09
  • 6
    The votes are (too?) nicely distributed over time, if I take a quick look: at least one vote each day since February? So, no Reddit or whatever, apparently basic search, if legit.
    – Arjan
    Jun 6, 2012 at 6:11
  • 1
    @Arjan exactly my observation as well. My first thought was something here is fishy but I have learned my lesson yesterday :) Jun 6, 2012 at 6:17
  • @Arjan You are being paranoid ;)
    – yannis
    Jun 6, 2012 at 6:18
  • 1
    The reason that the question has not gotten any substantive answers is indicated in the first comment: the "real" answer is one that has not been implemented yet. This comment was from (one of) the author(s) of the roxygen2 package under discussion, so can be considered definitive (until such time as the feature is implemented). Jun 6, 2012 at 18:24

1 Answer 1


It's a well formulated, interesting question, and beyond Stack Overflow's nowadays noobesque norm. The fact that has gone unanswered for a while could imply that it's a hard problem to solve, although I'm not familiar with the subject matter to be certain. If it is indeed a hard and common problem in the R world, it's quite reasonable that it got some vote love, isn't this what upvotes should be used for? If you check the question's timeline, the votes are spread quite evenly, there isn't any spike that would hint the question was shared on Reddit or anywhere else.

What made you think there was something more here?

  • 5
    Also its author keep editing it every couple of weeks for the past couple of months with rather minor edits, he even admits it - "no added substance", "minor fix", "esthetic improvements" etc - so this help keep the question on top of the Active view and gain more attention. Jun 6, 2012 at 6:15
  • @ShaDowWizArd Good catch. There was also a bounty at one time. I don't mind the minor edits, even if they are only to bump the question, it's still unanswered and since the bounty failed, it's a way to get some cheap attention. That said, the question is very close to wikification...
    – yannis
    Jun 6, 2012 at 6:17
  • @YannisRizos: I don't mean it's a bad question or that it doesn't deserve the upvotes, just that it struck me as suspicious for the reasons listed (I made some changes per the comments). Jun 6, 2012 at 6:33
  • Also, if anything the evenness seems more suspicious; being shared on Reddit would explain the upvotes. Jun 6, 2012 at 6:34
  • @Mechanicalsnail Why do you care? Moderators have an arsenal of tools to catch vote fraud, given that the question survived I'd say it's safe to assume that there's nothing fishy going on there. You're overthinking this, some questions are just good. And the minor edits explain why it kept getting upvotes after a while...
    – yannis
    Jun 6, 2012 at 6:35
  • 3
    @YannisRizos, a tool is only as good as the algorithms which power it. It's very hard to codify "intuition" as an algorithm, so saying "it's safe to assume that there's nothing fishy going on there" is quite a bold statement to make =) Yes, some questions are just good, but some others are just not caught by tools, but instead by people seeing patterns that the tools don't.
    – Rob
    Jun 6, 2012 at 9:52
  • @Rob Given that the question doesn't look fishy at all and has been seen by 6K people, and probably a horde of SE mods after being featured on MSO, I'd say that the chances for vote fraud that hasn't been caught by the tools are extremely small. If however it's vote fraud and OP managed to fool everyone and the tools, well, they probably deserve a badge or something.
    – yannis
    Jun 6, 2012 at 9:59

You must log in to answer this question.

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