18
$\begingroup$

In the last week and a half, I've had three reputation decreases of 10 points each due to "User was removed." I understand why this happened; in three cases, users deleted their accounts (or were banned) and the reputation gain I saw from their upvotes on my answers was rolled back. I am fine with this policy.

The recent frequency of these events does strike me as unusual, though. I've been around MSE at varying levels of participation for a while. I'm not tremendously active right now; I will usually answer 0-2 questions per day when I have time. I've seen this kind of reputation decrease before, but never three times in rapid succession.

My question is: Are more users deleting their accounts / getting banned right now than usual? I'm aware of a typical influx of questions at the beginning of academic years, for instance; perhaps this is known to coincide with that? Or maybe I just experienced a low-probability event.

Edit: There has now been a fourth instance of this in about 10 days! Very interesting.

$\endgroup$
  • 7
    $\begingroup$ A great share of the self-deletions are probably related to a generalized insatisfaction with Stack Exchange. (See the post featured on meta, for example.) $\endgroup$ – Aloizio Macedo Oct 25 at 1:43
  • 2
    $\begingroup$ @AloizioMacedo Perhaps, but I'd be surprised if that were true for the users I referenced in my question; I think (but am not quite sure) the three users who I referenced were all quite new with <100 rep, and I'd be pretty shocked if they were invested enough in the SE ecosystem to be aware of any of that. I realize that stuff is posted prominently on the right sidebar, but I have to think it'd just be noise for new users. But, maybe. $\endgroup$ – Aaron Montgomery Oct 25 at 1:45
  • 4
    $\begingroup$ Not sure, but I have personally been considering deleting my my account due largely to the remarks by SE staff to the media (and the total lack of taking responsibility) and not really want to implicitly support/participate in that. It is possible that some people have removed their accounts, but "couldn't", and came back before removing again. $\endgroup$ – Paul Plummer Oct 25 at 3:32
  • 15
    $\begingroup$ I suppose we need data, rather than anecdotes, but my guess is that Aaron is seeing a random statistical fluctuation. I haven't lost any points that way since 2 October. $\endgroup$ – Gerry Myerson Oct 25 at 5:02
  • $\begingroup$ @GerryMyerson: I had two "user was removed" in the last week. After a while of having none. $\endgroup$ – Martin Argerami Oct 26 at 15:39
  • 1
    $\begingroup$ My streak is over; I lost 10 points to "User was removed" on 2 November. $\endgroup$ – Gerry Myerson Nov 3 at 10:09
  • $\begingroup$ @GerryMyerson I think we have enough anecdotes to qualify as anecdata! (Kidding, of course.) Martin Sleziak's answer was excellent and, I believe, the best answer that could be given with existing tools. However, I think it's right to wonder whether it captures the true answer. What's really interesting is that while I was well aware of the CoC controversy when I posted this question, it never even crossed my mind that some users would connect the two. (I'm still not convinced of the connection, but now I wonder more than I did.) $\endgroup$ – Aaron Montgomery Nov 3 at 18:23
11
$\begingroup$

As far as I know, we cannot get some data on dates of account deletions from the Stack Exchange Data Explorer (SEDE). Still, for deleted users we can find the date when they posted their last post. (Assuming they made at least one post. The same is true for comments. And also for some other actions - but I took posts and comments, since they are the most typical actions of users on the site.)

If we take the last post (last comment) as at least rough estimate when the user deleted their account, we can obtain at least some data.

In the following queries I have checked for posts where the OwnerUserId is NULL. (Similarly for comments.) Such posts also include post which were migrated from other site and the original poster did not create an account here. But probably this does not make significant difference - and we would need a more complex query to eliminate those cases, so I decided to go for what I can get with a simple query.

Another thing to keep in mind is that deleted posts are not included here. And also a user who never made any post/any comment won't show up in those stats.

Possibly it could make sense to look only at users which made more posts than some given threshold - users who posted one or two questions and never returned are less interesting for stats like hits.

You can check the numbers and graphs when we look at:

As I mentioned above, this does not actually gives date of deletion of an account. (Probably it would be better described as the date when the user stopped their activity.) Still, if we take this as an approximation of account deletions, I do not see a significant increase.


It is not very difficult to modify the queries above to look at the non-deleted users by the time when they made their last post/comment. (If users haven't posted anything for a longer period of time, that might mean that they have left the site. So this data might give us an idea whether number of such users changes.)

You can look at such queries:

$\endgroup$
  • $\begingroup$ These stats ignore one important factor. Users do not "delete their accounts" on SE. They request that the "managers" (using the term as a collective noun for staff/admins/moderators/whoever) delete their accounts for them. The "managers" can process that request queue at any speed they choose. If there were 1,000 outstanding requests to leave, they could still process one request per day if they felt so inclined. And "checking in every day to see if the account has been deleted yet" probably counts as "activity"! $\endgroup$ – alephzero Nov 1 at 14:06
  • $\begingroup$ @alephzero Looking at the FAQ on account deletions, this does not seem to be true. Only some accounts need human involvement. The FAQ says, that accounts are deleted "by a scheduled task ... provided you haven't cast a lot of votes. If your account has cast a lot of votes, it will not be deleted approximately after the 24-hour timer expires, as above. Deletion will be held up while a Stack Exchange employee checks your account to determine if your votes should be preserved..." $\endgroup$ – Martin Sleziak Nov 1 at 15:41
7
$\begingroup$

Unfortunately, I think the stack exchange sites are being designed in the way that, this is the direction it will go (not necessarily a bad thing though). It is designed to be a Q&A site where eventually most commonly asked question can be easily searched. After years of effort of all the contributions from people, it's getting closed to the goal and that is good. However the flip side of this is it becomes harder and harder to contribute new things, making participation much harder and less fun for both new and old users. I can summarized a few points here,

  • Over time, it becomes harder to ask a not too difficult question at the same time that is not duplicate. As a result, the average difficulty of "legit" questions is getting higher over time (fewer people are capable of answering).
  • As a result, active participants find it harder and more time consuming to answer a new question (and usually to only get a little reputation) on average, therefore it becomes less motivated for them to further participate.
  • The flood of harder questions also mean that fewer people have the ability and knowledge to be an active participants. They are almost exclusively math professors/senior Ph.D. students at least, these group of people (including myself) on average tend to be harsher to "dumb mistakes" or "easy question", and therefore...(next point)
  • New users often don't like the site after asking an easy question and getting down vote/closed/harsh comments/etc. .

When people don't like the site, the most natural thing to do is to leave and close the accounts, and therefore caused the thing you experienced. I was once a several thousands reputation user on this site that decided to delete all my stack exchange accounts (and only using this secondary account because I saw your question and want to answer!) because I also find it harder and harder to contribute new things.

In a different stack exchange sites where an answer could be more subjective, there are other culture problems that might make participants unhappy (for more detail, just google it and you can find a lot of complain on various stack exchange sites for many years from many people), and eventually decide to leave.

$\endgroup$
  • 10
    $\begingroup$ "When people don't like the site, the most natural thing to do is to leave and close the accounts" - since actually closing an account is a hassle, IMO the most likely thing people do is just walk away. Recent events may have prompted some people to leave and delete their accounts as a form of protest, of course. $\endgroup$ – alephzero Oct 26 at 10:56
  • 5
    $\begingroup$ The idea of a limiting state of MSE is interesting to me. I wonder how many undetected duplicates there are on the site? I suspect many. $\endgroup$ – Aaron Montgomery Oct 26 at 13:58

You must log in to answer this question.

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