2
$\begingroup$

I recently received the popular question badge for this question. I was surprised to learn that it had been viewed 1,000 times, given that that would mean that about one in fifty people to view the question upvoted it. I went and looked at other people with this badge, and found this is not at all uncommon, and that in fact having 20 upvotes seems rare among questions for which this badge was awarded. Many of these questions that just recently were awarded the badge have been inactive for over a year and have very little interaction. For example, we see here a question that is 2 years old, inactive for 1 year, and was just recently (the past few hours) awarded the badge. They all seem to have almost exactly 1,000 views.

My questions:

  1. Do badges for number of views track unique views in some meaningful capacity?
  2. Are the links I've provided evidence of widespread manipulation to obtain badges, and if so are any steps being taken to counteract this?
  3. If the answer to the above questions are "no" and "yes" respectively, what measures might we take to prevent this abuse?
$\endgroup$
  • $\begingroup$ I suppose that you are aware that voting (and number of views) for your question might be rather atypical compared to most questions since it was for a long time (and at the moment still is) in the network-wide hot network questions list, right? $\endgroup$ – Martin Sleziak Jan 23 '17 at 17:01
  • $\begingroup$ I found meta.stackexchange.com/questions/36728/… that new views from the same IP address within 15 minutes, but it makes little sense to me that this question might have garnered (on average) 100 views per hour since its creation. $\endgroup$ – Stella Biderman Jan 23 '17 at 17:04
  • $\begingroup$ @MartinSleziak I actually didn't know that! That makes sense for my question then. I still worry about posts like math.stackexchange.com/questions/108547/… which was just awarded the badge 8 minutes ago despite being last active 4 years ago and there not being anything on the question that is upvoted more than 6 times. Unless we have a massive viewership that doesn't vote on our posts, this seems difficult to explain to me. $\endgroup$ – Stella Biderman Jan 23 '17 at 17:07
  • 3
    $\begingroup$ I'd consider it exactly to be expected that most users who view question do not vote on it. Both from experience with the site and considering the fact that large part of the traffic are users who do not have an account here and so cannot vote at all. Similarly, if the number of views grows slowly, the number of views for questions with recently obtained badge is expected to be about 1k. $\endgroup$ – Martin Sleziak Jan 23 '17 at 17:21
  • $\begingroup$ Of course, if you want to investigate this in detail, you can try to find all sort of stats, for example what is average number of upvotes on questions or what is average ratio votes/views. For the latter you can use this SEDE query which was taken from this question: What is the average question score per view? $\endgroup$ – Martin Sleziak Jan 23 '17 at 17:21
  • $\begingroup$ @MartinSleziak thanks for the resource. Re: "wow they're all close to 1,000 votes, "that was definitely an oversight of mine. Of course it's the case. I'm still mildly surprised that so many people visit this site and don't have accounts, but I believe you. $\endgroup$ – Stella Biderman Jan 23 '17 at 18:15
5
$\begingroup$

You found the correct resource regarding the technical aspect, which is:

How are the number of views in a question calculated?

I would say the tracking is not very rigid but rather inflates the count. It would be easy to manipulate. Somebody could just log out and visit their own question trice each day and would get the badge after a year. (Actually, I think they would not even need to log out.)

Yet that "[t]hey all seem to have almost exactly 1,000 views" is not at all evidence for any abuse. As Martin already remarked it's just an artifact of you checking questions that recently got a badge that is awarded exactly at 1000. So of course they are now very close to 1000.

I do not think abuse to get that type of badge is a major concern. I really do not think it is all that wide spread, and even if it were I'd only care that much as it is rather inconsequential, compared to vote fraud or editing for badges.

$\endgroup$

You must log in to answer this question.

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