Is the History Stack activity experiencing a major downward trend?

Question

When looking at the year in moderation post, I did a little digging, and checked the question counts over the last 10 years using the search is:question created:year. We get the following numbers:

Year	Questions Asked	Closed:true
2022	755	86
2021	962	132
2020	1342	216
2019	1402	161
2018	1720	144
2017	1812	230
2016	1420	238
2015	1250	225
2014	1103	193
2013	1127	127
2012	830	102
2011	364	35

(Note these numbers do reflect both open and closed questions from each year, but do not (can not) include questions which have been deleted.)

Update: Added a new column which represents a portion of the questions asked, but are closed in the system (closed:true). This still doesn't represent questions which have been deleted.

A graph of the above data looks like this:

It appears there was a steady increase in questions asked up until 2017 and we have been losing activity ever since. It also seems our closure rate has remained fairly consistent.

---

Grabbing a look at the sites analytics as suggested by @LаngLаngС seems on the surface to confirm this:

Note the highest post count corresponds with the same year mentioned in the table above.

One question we need to consider: is is this trend just ours, or is it overall to the SE community in general?

If it is just our issue, what do we need to do to fix it?

(My son, half in jest, mentioned that it could be a diminishing returns type issue-we have so many answers posted already that new questions are mostly duplicates or are unnecessary).

Answer 1

Yes, it is experiencing a downward trend, from all angles available to me:

• personal impression of ad-hoc/heuristic mental 'stats'
• 'official statistics' as presented by site-analytics

Apart from the things mentioned, perhaps needed for more insight:

• How many unregistered users posting decide to register?
• How many users start earning reps from non-HNQs, 2-or-more posts?
• How did numbers like these develop over the years?
• How many users beomce active for a while, but then seem to leave/become inactive again, just by pure visits to the site?

What would certainly be interesting to know:

Are driving perhaps factors:

• 'mere' question-closures (plus subsequent 'staying closed'),
• 'perceived as nasty' comments or downvotes,
• sub-par answers, unanswered questions, (what else)?

One factor probably being at work here is a variation of your son's suspicion: the higher the rep of a user, the lower the tendency of that very user to ask questions.
So that might be a harvest effect while SE itself and overall approaches a slow-down or even ceiling in new user recruitment? Certainly this downward trend occurs to me on some other sites on which I have analytics privileges.

But this and things like SEDE queries need more insight from staff. Or some clever queries by more SEDE experienced users. If anyone with the skills and motivation for more insight comes along and reads this: please share your results!

One aspect apparently lacking from the statistics in this question is 'how many questions were deleted?'. That would be quite a distortion of post mortem analysis. If we received much more posts but also got even more so 'much better' at deleting questions, then the observed trend might not be an indicator of 'reduced activity' at all but a rather different situation: simply 'us' sieving out more posts.

But It seems that this is covered by current site-analytics. But one aspect for this, albeit not broken down between questions and answers, is to be gleaned from "20xx a year in moderation posts", which sadly only started appearing in 2018:

Posts Deleted in	by Moderators	by Community	Total
2018	582	855	1437
2019	409	1143	1552
2020	483	1289	1772
2021	238	857	1095

So, while we see a steady increase in post deletions prior to 2021 (with a perhaps seen as welcome trend of community taking a bigger share from 'moderator-duties'), the drastic slump in 2021 seems to rule out ''better' at deletion' as a big explainer for this trend.

At this point I can only repeat that 'hidden' statistics, available to staff, need to be analysed as well.

One step into this direction might be to suggest or even demand a more fine-grained statistic section from "20xx a year in moderation"? Like: also adding a section on general statistics and site activity?

Ready-made SEDE-queries that seem relevant on this would be

• Total Questions and Answers

For comparison, other sites and their development look like this:

— MedicalSciences (still 'Health' on SEDE) seems to contradict most theories about 'pandemic effects, early 2020'…

— Politics

— Skeptics

— History of Science and Mathematics, noteworthy in its difference

Stable sites seem those like WorldBuilding, still in upward trend are for example Islam and Christianty…

• Bounties and Questions by Month

Those queries can be compared with other SE sites.

For Stackoverflow the bounties graph looks like this:

For 2022, we see in terms of participation in terms of Meta voting for the entire year so far (–September) only five users casting 80 votes altogether so far!