3
$\begingroup$

Whether its # of questions, # of answers, # of posts, # of upvotes, # of total votes, # of visits, or # of page views (and maybe other things too, but I removed them since they were affecting the scales on the graphs), there's been a downward trend in the last few months (these graphs happen to start from 1 Feb 2021, but the trend continued a bit in late January too):

enter image description here

In the one-year period from 1 Feb 2020 to 31 Jan 2021, there was an extremely big upward trend (especially in the last few categories):

enter image description here

Would anyone like to offer theories about what might have happened?

The upward trends circa January 2021 were observed before and may have been attributed largely to a user who ended up getting a network-wide ban for voting too much across the entire SE network (I have permission to say that), and I'm not surprised that the voting would drop after January 2021 for that reason, but I do not see why we're getting record-low numbers (over a fairly large interval of time) of posts and votes and views and visits lately.

$\endgroup$
12
  • $\begingroup$ people going on summer holidays perhaps? $\endgroup$
    – glS Mod
    Commented Aug 24, 2021 at 11:27
  • $\begingroup$ From February to August? $\endgroup$ Commented Aug 24, 2021 at 13:01
  • $\begingroup$ no I meant the last few months, July and August mostly. That's where I see a noticeable drop. Before that I don't see that marked of a decay. Maybe a little bit, but not too much outside of the usual parameters. It also looks more pronounced here because there was a peak just before the time period in the graphs on top. See e.g. i.sstatic.net/jZucg.png $\endgroup$
    – glS Mod
    Commented Aug 24, 2021 at 13:35
  • 2
    $\begingroup$ It's like we grew and then decayed. $\endgroup$ Commented Aug 24, 2021 at 15:17
  • 2
    $\begingroup$ To me, questions are getting more and more boring ... a significant fraction can be binned into "How does linear algebra work?" and "Random python problems". Not really the level I would like this site to have. $\endgroup$ Commented Oct 1, 2021 at 9:01
  • $\begingroup$ @MarkusHeinrich unfortunately, that's hard to avoid without setting stricter limits to the type and level of questions we want to accept. Otherwise, it's not surprising that the number of people asking questions about basic linear algebra or programming significantly outnumbers those asking about more specialistic topics. But it doesn't seem like most people on the site want to go that direction. Or at least, those that do haven't made their voice heard much on meta. $\endgroup$
    – glS Mod
    Commented Oct 1, 2021 at 9:27
  • 1
    $\begingroup$ @glS I totally agree with your conclusion. It is also ok if the majority of users want to keep it that way. But to be honest, I do not feel very motivated to answer such questions. So this certainly has an influence on my activity here. $\endgroup$ Commented Oct 1, 2021 at 11:27
  • $\begingroup$ @MarkusHeinrich the only thing I can say about that is that (1) it does make a difference to voice that opinion when it comes up on meta, and (2) tags are a great way to filter content and alleviate those problems. You can for example follow the site via email subscriptions to specific tags, and the ignore tags functionality will make some types of questions you don't particularly care about completely vanish from your point of view. They are not perfect tools, but can be very useful $\endgroup$
    – glS Mod
    Commented Oct 1, 2021 at 12:48
  • $\begingroup$ @glS thanks for the suggestions! I'll try to participate more on meta. Concerning (2), I already found those ;) $\endgroup$ Commented Oct 4, 2021 at 8:06
  • $\begingroup$ @glS Unfortunately, SE does not want tags to be a great way to filter questions by their level. (The HW tag on physics is a noteworthy and very helpful exception, I'd certainly favor having sth like this here.) And after all, there are quite interesting LinAlg-related questions, so filtering all of those is not a good method either. $\endgroup$ Commented Dec 29, 2021 at 15:49
  • $\begingroup$ @NorbertSchuch I agree to some degree. Though I wouldn't say that SE itself is that strict about tag management. We could probably get tags such as "research-level" if the community really wanted to. I'm not convinced these would work that well though. I personally think the best compromise is to blacklist tags you are not interested in, and subscribe to more "specialistic" tags, which indirectly also manages to somewhat filter out by "question level". Of course, it also means possibly missing out on interesting questions outside such filters, but that's hardly avoidable $\endgroup$
    – glS Mod
    Commented Dec 30, 2021 at 8:14
  • $\begingroup$ @glS On physics I'd say filtering homework-and-exercises works quite well, combined with subscriptions. Here on QC, subscriptions make less sense if I am interested in QC questions broadly. I filter the textbook-and-exercises tag, but I guess one issue here is that it is not used/edited in very consistently, at least, there's still a lot of textbook/exercise-level questions showing up. $\endgroup$ Commented Dec 30, 2021 at 19:12

1 Answer 1

4
$\begingroup$

To me this seems simply explainable as due to statistical fluctuations, rather than some overall trend we should be worried about.

Here's the total number of posts from conception to 29/08/2021, averaged weekly:

enter image description here

Here's the same as above, but averaged monthly (note that the last datapoint is for July; August data should be available in a few days):

enter image description here

Aside from slightly lower than average numbers in the first two weeks of August (which, again, I think are simply attributable to summer holidays) I don't see a particularly worrying trend. The numbers just slowly go up (or at least don't go down).

From the mod dashboard we also get statistics on the number of people visiting/acting, and the trend is pretty similar to the one on the total number of posts. I'm not sure I can post these here though, as the information is not available with the standard analytics tools (I think), and they explicitly say to not share the specifics of this data in public, for whatever reason. Going for example by the number of new visits, in July we got the largest number (7725) since the conception of the site.

$\endgroup$
1
  • 1
    $\begingroup$ (1) Voting and traffic have also gone down. (2) If you could have added August, you'd see a more pronounced decline in posts. (3) Last August did not see such pronounced declines as in this August (maybe fewer people were vacationing though since traveling wasn't as easy in the first COVID August as in the second). (4) New visits will go up over time as the site hosts more and more questions and therefore shows up more in search engine results, but overall visits and page views are at record-low levels when comparing over a fairly long time. $\endgroup$ Commented Sep 1, 2021 at 1:03

You must log in to answer this question.

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