# dealing with many poorly motivated questions from a single user

This character seems to be asking asking many high-level (meaning broad and vague) questions in rapid succession. As well, the replies from this person are essentially excerpts from (searchable?) texts or online references.

Joking aside - IMHO, these kinds of questions are no good for the site. How to handle repeat offenders?

-
If only there were some way of making him edit his questions :) meta.math.stackexchange.com/questions/540/… –  Tom Boardman Aug 10 '10 at 16:23
I think his questions are good - not sure what the problem is. –  anon Aug 10 '10 at 16:40
Questions like these appear to be motivated by a casual skim through books on one's shelf that have been gathering dust. They do not indicate that the asker has been doing any mathematics at all. As well, the answers seem to come from searches in an index and then spat out onto the page rather than having come from experience in working through the material. @muad: If Chandru1 is an AI, you will have pie on your face! –  Tom Stephens Aug 10 '10 at 16:49
@Muad- regardless of any fact about the OP, the questions themselves are broad and unmotivated- questions like these stop the site feeling like a community and turn it into a faceless grindfest of well known algebra results. As soon this site stops being about people answering people on questions that feel like they matter to both parties, it ceases to be the kind of community I want to be part of. –  Tom Boardman Aug 10 '10 at 16:50
You should probably change the title :/ it is funny but may incur some backlash. Unless of course you genuinely want to give him the Turing test. –  BBischof Aug 10 '10 at 17:21
@BBischof: Warning heeded, thank you. –  Tom Stephens Aug 10 '10 at 17:23
Tom Stephens, AI or not, I've still learned from his posts. –  anon Aug 10 '10 at 17:24
Tom Boardman, I am becoming increasingly unsure what sort of questions are the right sort for this site.. –  anon Aug 10 '10 at 17:26
The whole discussion makes no sense now that you have changed the title... –  anon Aug 10 '10 at 17:55
@muad: the linked-to questions above are not each individually terrible enough to warrant this thread, but the many back-to-back-to-back-to-back-... questions by the same user, in the same questionable style, is what I am trying to bring attention to. I contend that you would have 'learned' just as much math in the past two days by randomly flipping pages from a stack of books and reading the first theorem that appeared - and then closing that book, and repeating. I also contend that this user would have 'learned' the same amount of math by that method. This is all I will say. –  Tom Stephens Aug 10 '10 at 18:06
Tom Stephens, please see here for what I was referring to math.stackexchange.com/questions/1956/… –  anon Aug 10 '10 at 18:09
Thanks for raising this issue @Tom. I've added a link to the user's profile in the question. –  Larry Wang Aug 10 '10 at 19:16
@Tom I just worry that an AI will come along and thing that we don't want their kind here. Personally, I would be very interested in any questions an AI might have... –  BBischof Aug 11 '10 at 16:19
In fact some of his questions are quoted verbatim from problem books with complete solutions -- see my comments here math.stackexchange.com/questions/2471 –  Bill Dubuque Aug 14 '10 at 16:50

Here are some tips for improving your posts:

1) Make your title as descriptive as possible. In many cases one can actually phrase the title as the question, at least in such a way so as to be comprehensible to an expert reader. For instance, a recent question on the site is titled Why is the Hilbert cube homogeneous? This is an ideal title: if you know what the terms mean, you probably understand the question already. Among Chandru1's previous titles, Dimension of Vector Space is quite vague (although it is admittedly not so easy to come up with a good title for this question), whereas Cancellation of Direct Products is pretty good, but could be even better: what's the question?

Since you can, and should, put a lot of content into the title, what goes in the body of the question itself?

2) First, in many cases it is appropriate to restate the question in less succinct language, defining or linking to terms as you feel appropriate. (On MO, we like this but realize that it is not required and not even always helpful: if you are asking a sufficiently technical question, it may be best simply to write for an audience which has a certain technical background, which is part of what the tagging system is for.)

3) If the question is homework, you should absolutely say so up front. If the question is not homework, you should take steps to convince us of that. A good way of doing this is by providing context: e.g., I was thinking about Theorem A, and I wondered what happens when you change/weaken the hypothesis / I have done work on Topic B, and this raises the question / I was reading Document C [tell us which one, specifically! link to it, if possible] and I started to wonder...

Especially, if you did not come up with your question yourself but it is taken from some particular source, I would feel most comfortable if you cited the source. One feature of mathematics is that a good question can be a contribution just as valuable, or even more so, as an answer or a proof. You should not try to pass off others' questions as your own.

4) After you give background and state the question clearly, it's best if you say a little bit about what you've done to try to answer the question. If the question is asking for a reference to something, tell us where you've already looked. (I hope you have already tried wikipedia and google). If the question is asking whether something is true, what do you think and why? If the question is asking for a proof of something, what you have tried so far?

-
Everything above is great advice! –  BBischof Aug 11 '10 at 22:01
It sure is. We should get around to compiling our own how to ask good questions guide soon. –  Larry Wang Aug 11 '10 at 22:36
Why don't we just use the one from MathOverflow? –  97832123 Aug 12 '10 at 2:46
I have created a meta question specifically on this topic. –  Larry Wang Aug 12 '10 at 3:52

Everyone, I am sorry for what I have done. I hope not to continue in this fashion. Just tell me as to where should i improve upon. Should i stay away from here. Seeing people's views i really feel ashamed as to what i have done. I am sorry, but honestly speaking, i really don't know where i have gone wrong.

-
This question on meta, like others, could have been posted without hyper-focusing on specific individuals. –  T.. Aug 11 '10 at 21:34
Chandru1, I found your questions very interesting actually (although they're beyond my ability to answer) and really appreciate you answering my question about $\pi$. It would be a shame and a loss if you were driven away so please don't be! –  anon Aug 11 '10 at 21:41
Thank you for engaging in the conversation. Don't leave. Stick around and help this site become a great resource for excited and motivated undergraduates and graduate students. The regulars (which certainly includes you) will benefit by learning about others' mathematical strengths and the numerous opportunities to assist each other and passers-by. The less-frequent users will benefit from having such efficient access to our collective hard-earned knowledge. –  Tom Stephens Aug 11 '10 at 21:52
@Chandru: agreed, please don't take this as a sign to stop participating. My answer contains some specific advice on how to improve your posts. –  Pete L. Clark Aug 11 '10 at 21:57
I think this is an indication we need to be more careful about letting users know what we are thinking instead of assuming they are just ignoring policy. I know I feel sheepish. –  BBischof Aug 11 '10 at 22:02
@Chandru1: As you can see, nobody is asking you to leave. Stick around, continue to ask and answer questions, and try to follow the wonderful guidelines that Pete L. Clark gave. That you've taken enough interest to come here and read the feedback from others is a great start. @Pete: I have updated your mention of your answer so that it points to the right place. –  Larry Wang Aug 11 '10 at 23:40

For posterity, as the originator of this meta thread, I should clarify what is on my mind.

My personal policy on asking questions is to carefully pose a problem/cocept that I have been stuck on (preferably for days), and whose replies would increase my understanding of mathematics.

As far as answers go, it seems to me that the most helpful answers are provided by someone who has worked through the posed question before, or has taken time to truly consider the question and their response.

As you can see, I have not posed many questions here, nor have I answered many, so I am not a model member of the community.

My complaint with the style of questions that Chandru1 has been posting is that they appear as problems or theorems would in a text. Many appear as commands rather than questions (i.e. solve this, or prove that), and none of them give the appearance of having been thought out in advance. I will not take issue with the content of Chandru1's answers, as I am not an expert answerer and I think just about all relevant answers are useful - or should simply be subject to our built-in voting system.

All of that said, I personally think each of the pieces of mathematics Chandru1 has posted have been interesting in themselves.

I am confident that poorly-worded, poorly-titled, not-thought-out questions will appear on this site frequently, but from the same user over and over is polluting. Typically comments aimed at the question will fix most first-time offenders. In the case of Chandru1, I feel the community has been essentially promoting what I consider to be poor behavior.

One of our moderators has posted a thoughtful response concerning this broad issue, please take a look.

the questions themselves are broad and unmotivated - questions like these stop the site feeling like a community and turn it into a faceless grindfest of well known algebra results. As soon this site stops being about people answering people on questions that feel like they matter to both parties, it ceases to be the kind of community I want to be part of.
-

If one user in particular seems to be causing trouble, posting the issue on meta is a good way to draw attention to the issue. Flagging one or more of his questions with a brief explanation would also work if the user is exceptionally badly behaved (I don't think this particular case is bad enough).

This particular user seems to be trying to participate in the site as intended, just not putting in a whole lot of effort. If this is cause for you to downvote his posts and refuse to answer his questions, that's your own choice.

There was one incident where his behavior did cross the line of what was acceptable, but since he has been fine so far, (unless another moderator has been deleting his bad comments) I don't think any extreme measures need to be taken.

Also, see this answer and this one for examples of Chandru1 demonstrating his own attempt at a solution. Voting these up/leaving positive comments may help improve this user's behavior.

-
Thank you for highlighting the redeeming traits of our friend Chandru1 - I certainly failed to do so. I concur with your conclusions. –  Tom Stephens Aug 10 '10 at 21:32
Is it possible to discretely point Chandru1 in the direction of this meta thread? I don't see any change in behavior and our comments to Chandru1's posts have not altered this person's style in the slightest. –  Tom Stephens Aug 11 '10 at 12:06
@Tom: Until this thread contains concrete suggestions for improvement, I see little point. All it would achieve is to say "Hey. Some people don't like you." and downvotes+comments do the same thing. Of course, any comment on one of his questions linking here would probably be on-topic, I just don't see much good coming from it. –  Larry Wang Aug 11 '10 at 15:28
Yes, I see. –  Tom Stephens Aug 11 '10 at 16:53
I have seen this. I like this discussion, and to be honest i accept it. One thing which i don't like is "the one incident" whic h Kaestur Hakari is refering. I didn't like the way by which Charles Stewart. One generally improves by Experience and thats the same with me. I am sure that you shall see a better Chandru1 in the time to come. –  Chandru1 Aug 11 '10 at 21:20

The person you are referring to is posting technical questions of mathematical content. The questions are not broad. They are precise statements which are not easy to solve. This user is an asset to the site. Do not denigrate him.

-

Admittedly, I am saying this with the benefit of hindsight, but was this really the best way of dealing with the problem. While Chandru's participation hasn't been perfect, I don't think his behavior been that bad (only 3 downvoted questions). As for how to deal with it - downvoting seems to be a good way to get peoples attention, but make sure to let them know how they can improve as well.

-
Downvoting isn't always a good way of getting people's attention. For downvoting to be a useful signal, the person who gets the downvote has to (1) care, and (2) interpret it as the reasoned act of others, not just something which randomly happened. I imagine it's only those of us weaned on the internet who are receptive to such anonymous signals… A direct comment, addressed by name, is a more primitive human interaction and has a better chance of getting even old-school people's attention. –  ShreevatsaR Aug 11 '10 at 22:55
@ShreevatsaR: Downvoting without leaving a comment is certainly not helpful. But if the user actually understands how the system works, or can learn (it's pretty simple, after all), then adding a downvote to your comment may provide incentive for them to improve the questions according to your feedback, so that they can get back that lost rep. –  Larry Wang Aug 11 '10 at 23:43
@ShreevatsaR, @Kaestur: As I said, let them know how they can improve (which is distinct from letting them know what they did wrong) –  Casebash Aug 12 '10 at 1:19
I agree with both Casebash and Kaestur, BTW. It was already implicit in the answer, but I was just pointing out that voting by itself can be insufficient. –  ShreevatsaR Aug 12 '10 at 1:24

It may be worthwhile to try to come to some consensus as to how to deal with problems like this since this is one way that someone could try to game the rating system. For example, a user could have two accounts and could post older or obscure problems from journals under one account and then post the published answers from another account. This would be a quick way to gain high reputation. As things stand, I don't see any way to guard against that. Afaik this hasn't occurred yet here or on MO, but it probably will at some point.

-
Mods can see users' IPs. Stackexchange has other methods in place to detect voting fraud. If you suspect this is occuring, flag the posts saying so and we can investigate. –  Larry Wang Aug 11 '10 at 22:37
This is a complicated issue and if we want to discuss it seriously, then it would need its own thread –  Casebash Aug 11 '10 at 22:39
But IP's are worthless - one can use proxy's etc. I agree it is very complicated. Does the software currently look for any patterns of collusion? –  Bill Dubuque Aug 11 '10 at 22:49
It sure does. See this meta.so question for the latest success? story. –  Larry Wang Aug 11 '10 at 23:45
This is a really good point, and something I've also been thinking about. One should also keep in mind that it only takes one account and a resource with many questions to game the reputation system. In a perfect world we would all like the questions to be motivated by a genuine desire for understanding, but I guess it is a sad reality that the possibility exists for people to abuse the system. –  Glen Wheeler Mar 23 '11 at 11:10

In effect, you (and we) are observing that a one-dimensional rating system is inadequate.

Once multiple rating dimensions are available for the postings (intrinsic interest, difficulty, prior effort made by poster, etc), the filtering will be that much more efficient.

There is also the question of how to aggregate multiple ratings into user reputation. This can be left as-is or be made sensitive to additional dimensions of ratings from the postings; this is a separate design decision. But there is no obstruction to numerically revealing more of the math.SE user opinions on what makes particular questions good or bad. This is what the Badges attempt to do, but less sharply.

-
IMO this voting system would destroy the site. Is this truly on the horizon? –  BBischof Aug 10 '10 at 20:03
How is it that one-dimensional voting is good but two- or three-dimensional voting would destroy the site? Usually more information is better. –  T.. Aug 10 '10 at 21:03
I don't really see how this system would work. As we currently have it, higher reputation grants you more powers, because it is assumed to be a measure of how much the community trusts you. How would your proposed system deal with this? Harder questions = more trust? More interesting questions? These statistics would be very useful to keep track of, just like we keep track of a user's accept rate. But I don't think it makes sense as a replacement for reputation. In brief, this would make the system more complicated. Changes that add complexity require very compelling reasons. –  Larry Wang Aug 10 '10 at 21:20
User reputation can be left the same if desired, based on the "overall" (which could be the only) rating parameter. The point is to allow additional dimensions for numerical measurements of questions. For example, if users can click on something that would rate a posting as elementary, high school, university, or research level many users would have a preference for selecting postings according to that measure. This does not imply that how many of one's postings are research vs elementary-school level has to appear in the user page or the reputation. –  T.. Aug 10 '10 at 21:27
@T..: There has been talk of tagging questions by difficulty in the past. If this were carried out on a widespread scale, a user's profile page would display the number of middle-school, lower-div-undergrad, or whatever level questions they were involved in (it counts both answers and questions atm). I think that your idea is a good one, we just need to come up with a good implementation, and this discussion would be better served somewhere else (maybe even on meta.so). –  Larry Wang Aug 10 '10 at 21:44
Several problems with difficulty rating through tags. Tags are set by a single user; they are not continuous numerical ratings ("average difficulty rating: 6.5"); there is a small limit on the number of tags, so tags-as-ratings compete with tags-as-subject-classifiers. Giving users additional rating options for a question or answer beyond UP, DOWN, or NOTHING, can be aggregated in all sorts of useful ways. The alternatives to increased efficiency of filtering are (1) chaos as userbase grows, or (2) keep branching new sites for subfields (statistics, math-ed, etc) to stem the tide. –  T.. Aug 10 '10 at 22:04
Some sites work fine with 1-dimensional rating system. And for better or for worse, we also have to live with one. –  Grigory M Aug 11 '10 at 8:57
I don't use StackOverflow, so I can't judge whether your example "works fine" or not. We currently have a 1-d rating system, which is different from having to live with one. Certainly one can discuss desired changes to the site technology on meta. SE is alleged to have deep-pockets venture capital funding, and they have access to as much programming capability as exists anywhere. –  T.. Aug 11 '10 at 19:09
MO would also benefit from multidimensional evaluation of postings, but much less than math.SE, because of the more uniform background of the posters, and the also much narrower range of any dimension one might want to measure: difficulty level, novelty of question, effort made by poster, etc. (I didn't realize you had posted two links, by the way.) –  T.. Aug 11 '10 at 20:41
@T (sorry, my formatting was misleading) anyway, the point is not that 1-dimensional system is good, but that we'll have to live with it at least for some time; so let's think how we should live with it –  Grigory M Aug 11 '10 at 20:47
There's room to think about both. I am pointing out that examples like the one discussed in this thread are (part of the) evidence that a multi-dimensional system would have considerable value, at least for a broad site like math.SE. MO may be able to go without it indefinitely, I don't know. But equivalents of math.SE have been tried, and failed, when the mass of heterogenous material could not be filtered. –  T.. Aug 11 '10 at 21:30