3

[Thanks to msh210 for telling me how to make these calculations.1]

I checked the average comments per post of the top 8 Stack Exchange Sites, to compare to Mi Yodeya:

Stack Overflow

  • 89,126,736 comments
  • 51,068,664 posts
  • 1.75 comments per post

Mathematics

  • 5,843,848 comments
  • 2,834,045 posts
  • 2.06 comments per post

Super User

  • 1,995,887 comments
  • 1,334,768 posts
  • 1.5 comments per post

Ask Ubuntu

  • 1,714,914 comments
  • 1,050,246 posts
  • 1.63 comments per post

Server Fault

  • 1,189,351 comments
  • 918,547 posts
  • 1.29 comments per post

English Language & Usage

  • 1,089,054 comments
  • 452,229 posts
  • 2.41 comments per post

Arqade

  • 475,061 comments
  • 334,319 posts
  • 1.42 comments per post

Ask Different

  • 421,383 comments
  • 328,856 posts
  • 1.28 comments per post

Mi Yodeya

  • 305,154 comments
  • 93,414 posts
  • 3.27 comments per post

As you can see, Mi Yodeya out-comments the other sites by a tremendous amount. The combined average of those 8 sites is 1.67 compared to Mi Yodeya's 3.27. I.e. Mi Yodeya has just under twice the average amount of average comments.

In case the results of the above survey are not representative because those sites are much bigger than Mi Yodeya, I checked a couple of similar sites as well

Photography

  • 184,484 comments
  • 99,626 posts
  • 1.85 comments per post

Philosophy

  • 141,149 comments
  • 53,390 posts
  • 2.64 comments per posts

Islam

  • 83,083 comments

  • 48,882 posts

  • 1.7 comments per post

Korean Language

  • 4,004 comments

  • 3,587 posts

  • 1.12 comments per post

Here too the numbers are well below Mi Yodeya, with a combined average of 1.83 comments per post. The total average of both groups is 1.71 comments per post.

  1. Is this a good thing, a bad thing, or neither?
  2. Is there a particular reason, or reasons, why Mi Yodeya generates many more comments per post? E.g. the nature of the content is such that it usually requires more clarification, or it easily generates tangential discussions.
  3. Should anything be done about this?

I personally like comments, and I think there are many comments that are very useful. (Note that while I don't have real data for this, my impression from checking the other sites was that Mi Yodeya also has a lot more active comments. I.e. it's not just that Mi Yodeya generates thousands of useless comments that are then deleted; it also generates useful comments that are kept. Unless that's just because of differences in moderation?)

What does everyone else think?

1. Note that the numbers may be off by a fraction of a percent because it is possible that I looked at the second-to-most-recent post instead of the most recent post, or it is possible that a few more posts or comments have gone up since I collected the data. None of that should affect the results in anything remotely close to a significant way.

||||||
  • Comments that were deleted may have been very useful at the time. Useful doesn't have to correlate with active. – Double AA♦ Jun 27 '18 at 21:45
  • @DoubleAA Good point. But my main point with that line was to show that in addition to having more average comments overall, Mi Yodeya also seems to have more average extant comments. – Alex Jun 27 '18 at 21:47
  • See also chat.stackexchange.com/transcript/468?m=45368460#45368460 – msh210♦ Jun 27 '18 at 21:49
  • Note that these are counts of extant comments. SEDE doesn't have information about deleted comments. Comments are deleted because they're obsolete (e.g. the post was edited to address them), because they were moved to chat, or because they were tangential or non-constructive -- so, both good and not-so-good reasons there. – Monica Cellio Jun 27 '18 at 21:51
  • @MonicaCellio Doesn't every new comment have a new number assigned to it? So even if there are deleted comments the number for the most recent comment includes all comments ever created? – Alex Jun 27 '18 at 21:53
  • @msh210 That's good to know. But it's possible that the skips are uniformly proportionate. I.e. the rate of skips per comment would be drastically different across sites. – Alex Jun 27 '18 at 21:55
  • @Alex I believe there can be gaps in the ID space; that is, not all IDs are actually used. I know that's true for posts; I don't know about comments but suspect it's the same. It has something to do with allocations of blocks of IDs in the database (waves hands vaguely here; that's all I remember). Also, sorry I didn't follow your link; I now see you weren't using SEDE for this after all. – Monica Cellio Jun 27 '18 at 21:56
  • 1
    See responses to #8 here. – Isaac Moses♦ Jun 27 '18 at 22:10
  • @IsaacMoses That post is actually what prompted this. That's where I saw the claim that Mi Yodeya has more comments than other sites. – Alex Jun 27 '18 at 22:12
  • 2
    Note that IPS.SE had 4.45 comments/post(!) and decreased to 3.79 c/p recently. Workplace.SE also has a problem with too many comments – Andrew T. Jun 28 '18 at 12:58
  • @AndrewT. Since posting this I calculated a few other sites, including Interpersonal Skills. Across all time they have 4.04 comments per post. A couple of other big ones were Workplace - 3.12 Skeptics - 4.3 and Science Fiction & Fantasy - 2.72. I suppose I'll edit the post to add those in, and continue to do so if I calculate more sites – Alex Jun 28 '18 at 13:06
  • See also this recent post. – msh210♦ Jul 2 '18 at 22:27
  • An old META answer recommended the practice of using comments for half-baked or unsourced answers. This has become standard practice on MY. – LN6595 Nov 4 '18 at 22:42
  • @LN6595 Why here more than than other sites? Is Mi Yodeya more conducive to half-baked or unsourced answers? – Alex Nov 4 '18 at 23:31
  • @Alex I was thinking the opposite. Mi Yodeya is less accepting of "because I think so" answers. MY users quickly learn that if they don't have something solid to say, they're safer in a comment. Other sites are full of answers without sources or reasoning; typical answers state "this is the facts" as a personal assertion. Such answers get quickly attacked here on MY. – LN6595 Nov 5 '18 at 2:20
4

Is this a good thing, a bad thing, or neither?

In my opinion, it's a bad thing.

First and most obviously, there's the signal-to-noise ratio. Comments have many purposes. (I mean de facto, not de jure). People use them to answer a question, to suggest a line of inquiry that can lead to an answer, to question the validity of the post, to suggest a broader/narrower question than the one asked, to link to a related post, to explain why the post should be closed/deleted, to explain why it shouldn't be closed/deleted, to make a joke about something in the post, to ask for a source for something in the post, to say they're glad the post was posted, to say the commenter had the same idea, to dispute some fact in or hypothesis of the post, to suggest a wording change to the post, to ask for clarification of a point in the post, and much more, plus to reply to any of the above, to reply to such a reply, and so on. Let's put aside for the moment the fact that most of those should not be comments and suppose it were true that they should be. The comments section would then be used for so many different things that it would be impossible to pick out the ones that were ever relevant and important, let alone the ones that still are, and the lasting value of all the comments is thus basically nil. (Trusting the up-arrows to know which comments are of lasting relevance or importance is a mistake.) And that's the actual situation on some of our posts (usually questions in my experience).

More importantly, comments on the question are often accepted by the asker as answers (anecdotally, I've seen many times that an asker referred to a comment as an "answer" or otherwise wrote as if his or her needs were filled). This is dangerous, even more so than the danger always inherent in communicating and teaching Torah in a forum like ours, because comments don't (really) have downvotes/edits, so those ways of determining/improving quality are basically gone, and because comments are even less likely to have citations to sources or subsequent comments seeking such citations than answers are.

Should anything be done about this?

Yes!

(Boy, am I glad you didn't ask what should. That I can't answer. But of course everyone should continue to flag comments for deletion as needed.)

||||||
3

I'm not sure if it is a good thing or not but I can think of 3 reasons this is so:

  • Jews love to argue. You probably heard about the journalist who asked a rabbi "Why do Jews always answer a question by another question?" - answered the rabbi "Why not?" !
  • more seriously, since the bar for answers is high, very often you see people slipping in answers without sources as comments to help out without needing the additional research
  • also, I find that the gap between the most knowledgeable members of the community and new users is very large. So some questions need a lot of refinement and this takes place in comments

Recently, I've taken to flag "not useful anymore" comments a bit more aggressively. I'm always a bit worried to increase the mod workload but feel it makes everyone life simpler if they don't have to browse through lots of unneeded comments.

||||||
  • Does Mi Yodeya discourage non-sourced answers more than other sites? Alternatively, is it harder to find a source to answer a Mi Yodeya question than to answer a question on other sites? Is there a greater knowledge gap between new and old users on Mi Yodeya than other sites? – Alex Jun 28 '18 at 5:08
  • 1
    At one time, I spent some time on Christianity and Islam (the most comparable sites). Let's just say I was not impressed by the quality bar, the depth of questions/answers and even the level of English. For other sites I don't know. Your analysis is interesting but I suspect that if you asked "Do we have too many comments/is it an issue?" the answer from the community would be "No we don't/it's not an issue" – mbloch Jun 28 '18 at 5:10
  • 2
    "The bar for answers is high" so people post answers in the comments is a bad thing. People need to get over the pain of receiving constructive criticism and post their best answers as answers. Relatedly, people who comment on answers need to stop saying "we require sources here" in favor of "this answer would be more compelling if you'd edit in your source for this information, as best you can." – Isaac Moses♦ Jun 28 '18 at 13:28
  • 6
    I think sources without answers (to help others find the answer) are more useful as comments than answers without sources – b a Jun 28 '18 at 16:31

You must log in to answer this question.

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