7

Does Stack Exchange collect, or can calculate, any metrics which would give an indication of which site an individual user is likely to both ask and answer questions?

I ask because in the brief time I've been perusing the various sites, it seems with a few notable exceptions (e.g. Puzzling) that there isn't much overlap between the askers and the answerers.

| improve this question | |
  • 2
    All the data is availble for you if you can construct your own query: data.stackexchange.com – ben is uǝq backwards Feb 10 '19 at 21:46
  • You might see this on beta or newer sites, since they usually have less organic traffic and fewer total questions. – TheWanderer Feb 10 '19 at 21:48
15

I've constructed a SEDE query which compares the top X askers (by # of questions asked) to the top X answerers (by # of answers posted). Running it with X = 100 declares Mi Yodeya as the winner with a 55% overlap, followed by three sites with 52%.

enter image description here

The query excludes child meta sites; those generally have a much larger overlap than main sites, as can be seen in a previous revision of the query.

The outcome will certainly depend on the value of X; if you increase it to 1000, Meta Stack Exchange leads the board with 59%.

| improve this answer | |
  • 1
    That's intriguing especially as X tends to large numbers, if anything it tells me I need to formulate the question more carefully. Thank you for taking the time to construct the query. – james Feb 11 '19 at 1:18
  • Yeah, if X is bigger than the actual # of users, you're guaranteed to hit 100%. This method was the best I could think of (and model in SEDE) on Sunday evening; it's just one possible way to measure it. – Glorfindel Feb 11 '19 at 6:48
  • It would probably be best to make X site specific (eg. 2% of total unique askers on that site) – Double AA Feb 11 '19 at 13:01

You must log in to answer this question.

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