34
$\begingroup$

Our site is growing quite rapidly: currently each day there are

  • about 150 new questions asked
  • about 250 new answers posted
  • about 150 edits made to posts.

Each of the above actions changes the front page. One screen-full (not scrolling down) of questions on the front page has around 10 - 15 questions. And during the day in, say, East Coast Time in the United States, the lifetime of a question on the first screen-full (where it gets the highest exposure) is about 30 minutes, with most questions gone from the front page within 3 hours of its last edit.

Sometimes questions go by so fast that the relevant experts who can answer the questions don't get to see them.

Question: Is it time we asked the StackExchange overlords to turn on for us also the question limits of 6/day and 50/month?

Note:

  • This shouldn't affect most users, as most people ask far fewer than the proposed limits of questions. It does help, however, prevent the front page being crowded by one or two users and could potentially help with spam (though we don't have much problem in terms of that at the moment).
  • For comparison, the question limit control was put into place on StackOverflow, SeverFault, and SuperUser (the "big three" in the Stack family) when they reached steadily over 100 new questions per day. Currently we generate about the same number of new questions per day as ServerFault and SuperUser, while having a much smaller traffic (less than 10% of their visits per day) and smaller demographic (less than 40% of their numbers of users).
$\endgroup$
23
  • 2
    $\begingroup$ This question is motivated by Martin's from about a month ago with GEdgar's comment there. Also related is meta.math.stackexchange.com/q/2464/1543 $\endgroup$ Jul 25, 2012 at 13:13
  • 2
    $\begingroup$ @Jay: it has been observed on several occasions. See Martin's question that I linked to, and t.b.'s comment here. I don't, however, have any statistics to answer what you asked specifically (some clever soul maybe able to figure out a query on the Data Explorer). $\endgroup$ Jul 25, 2012 at 13:31
  • 3
    $\begingroup$ @Jay: one should perhaps think of it as a limit so that one user by himself will not take up more than ~4% of the daily new Qs. I don't see why it is a problem for two users to each ask fewer than 7. $\endgroup$ Jul 25, 2012 at 13:34
  • 1
    $\begingroup$ @Jay: Ah, I see. What was left unsaid in my question is the issue of perceived fairness. Given that questions are blazing by so quickly, some users maybe less happy with certain other user taking up lots of screen real estate. This was the argument put forth during the past couple of weeks about why high frequency edits annoy people. $\endgroup$ Jul 25, 2012 at 13:46
  • 12
    $\begingroup$ YES!${}{}{}{}{}$ $\endgroup$
    – Asaf Karagila Mod
    Jul 25, 2012 at 14:37
  • 1
    $\begingroup$ As a parenthetical to Martin's comment: remember that the Data in the Data Explorer is asynchronous, so you won't be able to query, say, the data from this past week. $\endgroup$ Jul 25, 2012 at 15:12
  • 4
    $\begingroup$ The question influx rate is already quite overwhelming, at least for me. I'm all for a quota being put up. $\endgroup$ Jul 25, 2012 at 15:47
  • 1
    $\begingroup$ Would a user who is posting tons of questions not just create a new account to ask more questions if they reached a limit? Any one who has enough reputation or badges or whatever (whether on this site or in real life if they are not anonymous) to care about their account probably isn't going to be asking more than 6 questions a day, or 50 in a month. So, the people affected would be new users who have no attachment to their accounts. $\endgroup$
    – GeoffDS
    Jul 25, 2012 at 17:51
  • 2
    $\begingroup$ @AsafKaragila That's a good point, thanks. But, how are they identified? Is it by IP address? I mean, I'm on a university campus and I have access to hundreds of computers in various labs. I could, with no technical knowledge, have a different IP address every time I logon even if I logged on every day for a year straight. $\endgroup$
    – GeoffDS
    Jul 25, 2012 at 17:54
  • 2
    $\begingroup$ @Graphth: Behavior patterns are a lot harder to fake and change. $\endgroup$
    – Asaf Karagila Mod
    Jul 25, 2012 at 18:12
  • 5
    $\begingroup$ @Graphth: One could have all the best reasons in the world, but this site is still a shared resource. If one tries to hog too much of the timeshare, one gets punished. You can see it with many people that are asking a lot of questions. Often the answers are accepted in mere minutes, even minor hints. I support this limitation because I think it will push more people to chew on the questions longer before posting them. $\endgroup$
    – Asaf Karagila Mod
    Jul 25, 2012 at 18:16
  • 3
    $\begingroup$ @cardinal: I think that setting the one exception aside, most people are very reasonable in their edits. I hope that once the current issue has past, it will not return. Especially with Bill's sandbox idea. $\endgroup$
    – Asaf Karagila Mod
    Jul 25, 2012 at 20:34
  • 6
    $\begingroup$ Question for moderators: Let's say the limit is enacted; would a moderator have the power to waive the limit for someone "on petition"? (I'm thinking along the lines of the automatic CW-ing of a much-edited question, in which moderators can un-CW it). Say someone reaches the 50-per-month-limit due to some singular circumstance; could they contact the moderators (or start a meta post) and explain the reason and request that the counter be reset for them, and would that be possible? If moderators have the power to grant the occasional exception, that would address many of the issues raised below $\endgroup$ Jul 26, 2012 at 21:04
  • 2
    $\begingroup$ @Arturo: I am pretty sure the answer is no. But unlike the CW thing, note that the question limit is not a permanent change: if you hit the 50 / month limit, you'd be able to ask questions again after at most $21 = 30 - \lceil 50 / 6 \rceil$ days. $\endgroup$ Jul 26, 2012 at 21:25
  • 2
    $\begingroup$ @Arturo: If the issue is really urgent, a user who has used up hir questions could get someone else to ask he question for them. $\endgroup$ Jul 27, 2012 at 19:11

4 Answers 4

28
$\begingroup$

We have enabled the 6/day, 50/month limits.

Ideally we'd like to keep those numbers to avoid too many special cases throughout the network, but if they do turn out to be problematic, we can revisit this discussion.

$\endgroup$
3
  • 7
    $\begingroup$ Thank you, Adam. $\endgroup$ Jul 28, 2012 at 3:10
  • 6
    $\begingroup$ The most important question now is: will there be a badge for hitting these limits, and if yes, what should it be called? $\endgroup$
    – user31373
    Jul 29, 2012 at 21:19
  • 1
    $\begingroup$ Marking this as accepted so that it will top the answers list. $\endgroup$ Sep 27, 2012 at 10:59
28
$\begingroup$

Yes, I think this is both reasonable on general grounds and will help to ameliorate some "site flooding" issues we are currently experiencing.

Are the exact numbers of 6 per day and 50 per month negotiable, or is it an on/off situation? To me 6 questions per day sounds quite reasonable, but 50 per month is small in comparison. It seems to allow users to paint themselves into a corner: asking, say, 5 questions a day for the first 10 days of the month is regarded as acceptable, and then they are locked out for the next 20 days? That seems slightly weird to me. But I think it will be an improvement on the total lack of question limits we currently have!

$\endgroup$
13
  • 4
    $\begingroup$ You have expressed concern that: It seems to allow users to paint themselves into a corner. If we introduce this limit, I have no doubt that users asking many question will be noticed and warned by more experienced users. (Similarly as a new user who does not accept answers or does not use homework tag usually receives comments about his behavior with pointers to faq or meta.) $\endgroup$ Jul 25, 2012 at 14:12
  • 1
    $\begingroup$ @Martin: that's a good point. (And my concern is very mild, actually.) After I wrote my answer I realized that for many "counters" on this site, you get reminders as you approach the limit: e.g. votes per day. Does anyone know if something similar (already) occurs for answers per day or per month? $\endgroup$ Jul 25, 2012 at 14:17
  • 3
    $\begingroup$ It doesn't appear that such a warning is available, nor does it appear that SE will implement it. See meta.stackexchange.com/q/96610/155238 $\endgroup$ Jul 25, 2012 at 15:14
  • 1
    $\begingroup$ The problem that mildly concerns you comes from the 6, not the 50, I guess... I think that 6 questions per day is quite a big number, in fact! Can anyone with SQL-fu come up with a query showing how often that 6 is reached in practice? $\endgroup$ Jul 25, 2012 at 19:02
  • 2
    $\begingroup$ My guess is it would be hard to ask 6 questions in one day which could not have been better asked in a smaller number of questions... $\endgroup$ Jul 25, 2012 at 19:03
  • 3
    $\begingroup$ Data: (@Mariano) There have been 52 instances this year (through 27 Jun 2012) where a user asked at least 6 questions in a day. The maximum was 13 on 13 Jun 2012 (user name redacted). There have been 6 instances of a user posting at least 50 questions per month in 2012, with a maximum of 70. The same person that posted 13 on a single day accounts for 2 of these 6 instances (April and June, respectively). $\endgroup$
    – cardinal
    Jul 25, 2012 at 19:52
  • $\begingroup$ 6 questions is still within the limit. If I did not made mistake somewhere, there where only 17 days when the "most curious" user posted more than 6 questions. (Namely 13.6., 1.5., 8.5., 10.5., 6.5., 3.6., 4.6., 6.6., 17.6., 12.1., 4.3., 5.3., 18.3., 25.4., 26.4., 25.5., 26.6.) $\endgroup$ Jul 25, 2012 at 20:53
  • 4
    $\begingroup$ I agree with @Mariano that $6$ questions a day is a lot, considering the current limitations of the front page exposure already mentioned. I guess that a limit of $3$ questions per day is a much more reasonable number. $\endgroup$ Jul 25, 2012 at 22:13
  • $\begingroup$ @MartinSleziak: Yes, if we make it strictly greater than 6, then the number of instances drops to 20. There were two each on May 10th and June 4th. There was another from an unregistered member on January 11th, which may account for the final discrepancy in the counts. :) $\endgroup$
    – cardinal
    Jul 25, 2012 at 23:25
  • 1
    $\begingroup$ @cardinal Could you please give a link to the database query so that others can browse the results. $\endgroup$ Jul 26, 2012 at 3:31
  • $\begingroup$ @BillDubuque Here are the data I have used: data2.xls (Excel sheet); data2.txt (tab delimited text file). They only contain top user for each day and are based on csv.zip obtained from single-day query I mentioned above. If posting this is against no names policy at meta, you (or some other moderator) can delete this post. (It seems like cardinal's post was edited by a mod for this reason...?) $\endgroup$ Jul 26, 2012 at 5:35
  • 3
    $\begingroup$ I think that three is a bit pressing. Four questions per day sounds better. $\endgroup$
    – Asaf Karagila Mod
    Jul 26, 2012 at 8:13
  • $\begingroup$ @Martin: In my comment, I imposed a self-redaction in accordance with my understanding of the no-names policy. I will post a link to the query in a bit. :-) $\endgroup$
    – cardinal
    Jul 26, 2012 at 19:06
13
$\begingroup$

Yes, let's ask for this to be done soon, please.

Asking for the same limits of 6/day & 50/month seems easiest and good enough.

$\endgroup$
12
$\begingroup$

I support the limits with the current numbers (6/day, 50/month), because those limits support a mission of providing help for self-study.

At first, I too was worried that 50 might be too small and might affect my ability to self-study (that is, study with everyone's generous help). However, looking at my own peak stats, I do not believe I would have been seriously affected by these limits in the past 15 years, including about 2 years here.

The rate of questions asked should actually be self-limiting: a good question takes time to compose and requires some self-reflection: what is it really that I don't understand? One should strive to find a “minimal working example” of one's own ignorance. I like to think Pólya said, “If there is a big problem you cannot solve, I bet there is a smaller one too.”(Actually I believe this was a snarky paraphrase, Pólya's writing is usually a lot more positive.)

Secondly, getting answers to questions is also self-limiting: a good question should be at one's "zone of intervention". It should be hard enough that it cannot reasonably be done on one's own, but close enough to one's experience that the solution will be a growth experience, not just a TV show. In other words, good questions take time for people to answer, and time for the original asker to digest the answer.

Taken together, these should limit questions to comfortably within the proposed limits (at least in my own experience, there is still quite a bit of room). Thus good questions should naturally never reach the proposed limits.

The technological/forced law limits should only apply when these self-limiting features are ignored. Perhaps the question asker is not really examining their questions so that we get poorly thought out questions or photocopies of unread homework assignments. Perhaps the question asker is not taking time to digest the answers, either because the original question was too easy and should have just been done one's self, or because the original question was too hard and none of the answers provide a learning opportunity.

$\endgroup$
1
  • 13
    $\begingroup$ I've often found that in the process of trying to write a good question (making it definite, precise, specific...), I end up answering my own question anyway. The cogitation for writing a good question takes time, and it is unfortunate that a number of people can't or won't spend this time. Just dumping your question on math.SE is sooo much easier than thinking... $\endgroup$ Jul 27, 2012 at 0:16

You must log in to answer this question.

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