8
$\begingroup$

Since CV is about statistics and machine learning, and this forum has been gaining popularity, can the organizers please share some statistics about CV. I'm sure a lot of folks would be interested in this.

Specifically,

  • number of registered users,
  • number of active participants,
  • ave number of posts per day,
  • no of Qs and As, by month, from past n yrs,
  • the most prolific user,
  • growth rate (based on any metric)

Others please feel free to add to this list.

$\endgroup$
  • 4
    $\begingroup$ I am voting to migrate this to Meta. Are you aware of the information at stats.stackexchange.com/users? Define "active", "prolific", etc. $\endgroup$ – Nick Cox Jul 2 '16 at 17:21
  • $\begingroup$ Ah, not aware of that at all. Will look into it. Thks guys. $\endgroup$ – horaceT Jul 2 '16 at 17:24
  • $\begingroup$ hey Nick, how to view posts on meta? $\endgroup$ – horaceT Jul 2 '16 at 17:29
  • $\begingroup$ @horaceT If you click through to look at this comment, you're reading meta $\endgroup$ – Glen_b -Reinstate Monica Jul 2 '16 at 22:51
  • $\begingroup$ @Glen_b Got it, thank you. But I still like to see if someone has already come up with these statistics. Save me some work...... $\endgroup$ – horaceT Jul 2 '16 at 23:03
  • $\begingroup$ @horaceT, please consider accepting Glen_b's answer if you think that it answered your question (it sure seems that it did). $\endgroup$ – amoeba says Reinstate Monica Jul 6 '16 at 15:00
  • $\begingroup$ @amoeba Pls tell me how to "accept" the answer. Being a newbie, I'm not sure if that's the same as upvoting? $\endgroup$ – horaceT Jul 7 '16 at 20:42
  • $\begingroup$ No, it's not the same. There is a tick sign to the left of Glen_b's answer. Click on it "accept" (click again to de-accept). $\endgroup$ – amoeba says Reinstate Monica Jul 7 '16 at 20:46
  • $\begingroup$ @amoeba Ah, got it. Is accepting just a book keeping thing, or does it serve any other purpose? $\endgroup$ – horaceT Jul 7 '16 at 21:13
  • $\begingroup$ On the main forum it brings 15 reputation points to the author of the accepted answer person and also 2 reputation points to the author of the question (for the act of accepting). On the Meta forum (where we are now) there is no reputation gained or lost, so it's more of a book-keeping. $\endgroup$ – amoeba says Reinstate Monica Jul 7 '16 at 21:15
15
$\begingroup$

Some of the sources of information:

  1. The "Users" tab, which stares at you every time you use StackExchange:

    enter image description here

    That should be on screen every time your read or post on the main site.

    If you investigate it, it has a wealth of information -

    enter image description here

    reputation shows how much net reputation (upvotes + acceptances + bounties received - downvotes - bounties awarded) a user has received over the last week or month or quarter or year or in all time (depending on which item you click on in the next row of links under the circled link)

    new users gives reputation stats for new users

    voters gives stats on the number of votes awarded (for the last week/month/ etc)

    editors gives stats on the number of edits made (for the last week/month/ etc)

    moderators tells you who the current diamond-moderators are

  2. Reputation leagues:

    While still on the "Users" page, scroll down to near the bottom of the page:

    enter image description here

    This has similar (though slightly less accurate) reputation information as from the Users page -- but you can find it going back in time. You want to know who had the highest reputation-gain in December 2015, or the most reputation at the end of May 2013? That will tell you.

  3. The Stack Exchange Data Explorer (SEDE). The one specific to CV is here

    This is an interface to a queryable database of all the publicly available information about posts, posters and so on. If you know SQL (T-SQL) you can write your own queries. If you have no clue, almost any query you would care about has probably already been written. If you can write a little SQL you may be able to fork an existing query to modify it.

    This database is updated each weekend (so unless you query it right after it updates it's not quite up to date -- if you look at it on Wednesday it won't have information about a post you made on Monday afternoon).

    Of the questions you ask, many already have queries at SEDE you can look at; try a search there. I describe some examples of search terms I used and queries that they locate below.

    Just over a week ago there was a SE-blog about SEDE here. Among other things it conveyed the important information that there's an introduction written by Monica Cellio here

    You can even download the whole database SEDE uses and look at it at your leisure.

    Numerous queries have been pointed to in questions or answers here. For example, this post points to some. This post points to some others.

  4. If you gain sufficient reputation you get some direct summary information you can look at, but a lot of it can be had through SEDE.


Let's look at a few specific examples of getting information about things you mention, to get started:

  • number of registered users,

    Go to SEDE/stats. Search "registered users". Click "popular". Top query link is "Total number of registered users over time"... sounds like a good start. Try it, once it has run* scroll down. A bit over 88000 registered users by late June.

    * you'll need to answer the CAPTCHA first if you haven't made an account there

  • number of active participants,

    I'd suggest looking at the Users tab for this -- for example:

    as I post, there are 9 pages (9 x 36) + 4 users (=328) with more than 200 reputation this year. Not all are active, though some just have continued to gain rep from old answers. That's an upper bound.

    What about SEDE?

    In SEDE, go back to the search bar, search for (say) active users, and again choose the popular tab (popular queries tend to be better written).

    On the first page of results is this query, for example:

    Number of active users with >= 1 post in last 30 days

    Which gives 289.

  • ave number of posts per day,

    Again, in SEDE, type "questions and answers" as a search. Among the list of popular searches is this query:

    Total Questions and Answers per Month for the last 12

    which gives monthly figures. For example, for the month from 27 April to 27 May there were 2576 Q's and 1746 A's, or per day about 86 Q's and 58 A's were posted. (Note that a substantial fraction of new questions posted will close.)

  • no of Qs and As, by month, from past n yrs,

    see above query. There are others that go back 2 or 3 years

  • the most prolific user,

    In what sense? Reputation? Posted questions? Posted answers? Comments? Edits? Votes? Bounties? Total actions?

  • growth rate (based on any metric)

    One measure can be computed from the Total Questions and Answers query above


Edit: There also appears to be a site that presents a variety of statistics for stackexchange sites (based on SEDE data by the look) -- CrossValidated's page is here.

$\endgroup$
  • $\begingroup$ +1 That's comprehensive! $\endgroup$ – Tim Jul 4 '16 at 8:13
  • 2
    $\begingroup$ Totally off topic here. i suspect CV and to a greater extent stackoverflow have taken "mind share" from other statistics/programming forums. One that I follow is the R list, which has a blend of programming and stat Qs and As. At its peak couple yrs ago, there were > 200 posts a day; I just look at it today, and it's about 20. $\endgroup$ – horaceT Jul 7 '16 at 23:30

You must log in to answer this question.

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