tea.mathoverflow.net - Discussion Feed (On "let's build a really big list!" questions)

Jon Bannon comments on "On "let's build a really big list!" questions" (9393)

2010-10-04T11:33:53-07:00

I'm a recent big-list offender, but nevertheless think that there needs to be a better place for big lists. The trouble is, MO is a great place to have many mathematicians weigh in on topics (imagine how easy it would be to replicate Hadamard's interview of mathematicians on MO) and so sometimes the temptation to post a big list question becomes too much to resist.

Perhaps there should be some posted standards for what constitutes an appropriate big list.

Or, as was suggested, perhaps big lists can be banished to meta.

Anyhow, sorry for posting a big list! I hope there will be a better place to handle such "survey questions" in the future.

Yemon Choi comments on "On "let's build a really big list!" questions" (1124)

2009-12-16T20:27:52-08:00

@Reid: thanks!

rwbarton comments on "On "let's build a really big list!" questions" (1123)

2009-12-16T19:46:04-08:00

@Yemon: I added 'big-list' to my ignored tags and selected the "Hide Ignored Tags" checkbox in my preferences, and I'm much happier now! I see any given "What are some..." question at most once, and add the tag. You can do the same, since you have > 500 rep.

Yemon Choi comments on "On "let's build a really big list!" questions" (1112)

2009-12-16T13:10:17-08:00

Chipping in belatedly: I too am not a big fan of these kinds of questions; but I don't know I'd go as far as to take measures against them. It is slightly annoying to see them bobbing back to the surface but ca c'est la vie perhaps.

Harry Gindi comments on "On "let's build a really big list!" questions" (1089)

2009-12-15T18:08:01-08:00

Community wiki is not a tag, it's a setting.

davidk01 comments on "On "let's build a really big list!" questions" (1087)

2009-12-15T17:48:58-08:00

I thought the wiki tag took care of these kinds of questions.

Edit: Ok, so the community wiki is not a tag. So maybe the wiki setting should be more obvious from just skimming the question list so people who want to avoid "wiki"-like questions can easily do so.

Harrison Brown comments on "On "let's build a really big list!" questions" (1069)

2009-12-14T20:17:16-08:00

Ilya Grigoriev says: "Also, here's a random idea: maybe we should close such questions as 'too general', and create several questions in their place? For example, a question that asks for 'A list of blah ___' can be split into questions 'A list of blah in algebra', 'A list of blah in geometry', 'A list of blah in analysis', perhaps also basic mathematics/ applied mathematics. This will make reading through the answers and finding the answers you are interested in more manageable. Also, the old question can be made into a link to the new ones."

I like the idea in theory, but since most people's complaint about big-list questions is that there are so damn many of 'em already, making five such questions instead of one is probably a Bad Idea.

I'm starting to think that it might be best to have a separate area for that sort of thing, or figure out what it is SO does that stops questions from being bumped. I mean, SO has some ginormous list questions as well, but SO naturally has a higher turnover rate than MO probably ever will, so...

rwbarton comments on "On "let's build a really big list!" questions" (1068)

2009-12-14T20:01:42-08:00

I went through the 200 most recently modified questions and tagged a bunch more "big-list". I happened to notice that none of them have any of the arxiv tags. Unfortunately, there are also a few substantive questions without arxiv tags, often because the topic seems to fall into a gap in the classification, but viewing only questions with arxiv tags seems to be a pretty good approximation of what I would consider "real" questions.

Oh, and @Anton: I do sort of like your question, and have some pet peeves of my own, but terminology is about as subjective and argumentative as math gets...

Harry Gindi comments on "On "let's build a really big list!" questions" (1067)

2009-12-14T19:35:00-08:00

Hey now, that's going a bit far, isn't it?

Peter McNamara comments on "On "let's build a really big list!" questions" (1066)

2009-12-14T19:15:52-08:00

The problem with the internet is that you continually get a batch of new users who come across the same "big list of crap question" and want to throw in their two cents (just like on other forums, arguing vi v emacs, or creation v evolution). After a while, the value that these new posts give to the discussion (remember MO is *not* meant to be a discussion board) is negligible at best. So I propose that we close off allowing any more answers to a "big list of crap" question once the number of answers reaches a certain predetermined number (that should be no more than 50).

@Anton. Apart from the fact that your "big list of crap" question is a "big list of crap" question, I would vote against posting it on mathoverflow because I see it as being subjective and argumentative. The examples you give, such as renaming group scheme as scheme group, are really quite pedantic.

Harry Gindi comments on "On "let's build a really big list!" questions" (1065)

2009-12-14T12:47:45-08:00

Voting down in community wiki does not require reputation, fortunately.

Akhil Mathew comments on "On "let's build a really big list!" questions" (1064)

2009-12-14T12:38:30-08:00

Stackoverflow apparently does not bump a question up to the beginning every time a new answer is posted. I don't know if this is a good idea in general, but it would keep, say, "Famous mathematical quotes," from staying permanently on the front page.

Ilya Grigoriev comments on "On "let's build a really big list!" questions" (1063)

2009-12-14T12:37:51-08:00

Another random idea: make voting down in community wiki posts require less reputation (15?) and have no penalty.

(Sorry for the spam guys, I promise I'll stop now).

Ilya Grigoriev comments on "On "let's build a really big list!" questions" (1062)

2009-12-14T12:34:52-08:00

Also, here's a random idea: maybe we should close such questions as "too general", and create several questions in their place? For example, a question that asks for "A list of blah ___" can be split into questions "A list of blah in algebra", "A list of blah in geometry", "A list of blah in analysis", perhaps also basic mathematics/ applied mathematics. This will make reading through the answers and finding the answers you are interested in more manageable. Also, the old question can be made into a link to the new ones.

Ilya Grigoriev comments on "On "let's build a really big list!" questions" (1061)

2009-12-14T12:20:32-08:00

I'm strongly bieased towards allowing such questions. I think, barring homework questions, if there are people willing to answer the question, and it could be of interest to a non-trivial number of mathematicians, it should be allowed.

In particular, some such questions (e.g. "Cocktail party math", or better - "Single paper everyone should read") seem to me like some of the best content of this website. They are precisely the thing that attrackted me to the website, and made me think "Wow, this place is fresh and cool".

Of course, there is the fact that the "let's make a really big list" questions are difficult to write well, and are more likely than the average question to be crappy. However, I think the rating system should take care of this -- they really don't take up that much space on the page, and if there is no interest, they'll just slowly slide into oblivion. The only real danger they pose is that it's too easy a way to attract reputation; so I think that people who don't mark them as "community wiki" should be frowned upon.

Moreover, I think there are guidelines that could make such questions better. A "big-list" tag is a great idea. Putting "a list of rules" into the question statement also seems to be a good idea. By all means, people should speak up when such questions are not well-worded. The great thing about "community wikis" is that people can edit the question to make it better (especially if there aren't any good answers).

Two more thoughts: I think the flood of such questions will decrease when people will get them out of their system, and the new ones become duplicates.

Secondly, there is one problem with these questions that I don't know how to fix: once there is a million answers with score zero, it becomes very difficult for a new (possibly great) answer to the question to be found by moderators. It'd be amazing to have a way to somehow make old score-zero answers fall into oblivion...

Harry Gindi comments on "On "let's build a really big list!" questions" (1060)

2009-12-14T11:49:46-08:00

I feel like you should hold on to it until this situation with the _abundance_ of "big list" questions is dealt with regardless of whether or not it's good, just because people (Reid, for example) seem to be getting somewhat fed up with all of these questions.

I agree with your second statement that it doesn't seem exactly cut out for MO. These kinds of questions seem like they belong in meta or some kind of discussion forum. Meanwhile, what is the policy on meta? Is it really only for posting things _about_ MO, or is it just for everything (not homework help) that doesn't belong on MO but has to do with the community in general? That is to say, is meta "strictly suit and tie"?

Anton Geraschenko comments on "On "let's build a really big list!" questions" (1059)

2009-12-14T11:32:36-08:00

@Harry: yes, that's a very witty thing you said about "scheme group", but I was hoping to also get opinions about whether the question is any good for MO, or if it would just turn into another annoying question with lots of answers that aren't very interesting to look at.

I think the question is interesting to talk about with people (though I usually talk about specific examples rather than trying to list a whole bunch of them), but I think it may be too much of a "forum/discussion question" for MO.

Harry Gindi comments on "On "let's build a really big list!" questions" (1058)

2009-12-14T11:21:49-08:00

I was trying to think of something funny to say about "Scheme Group", but I have neither the will nor the presence of mind to do so at the moment (finals). Just imagine that I said something witty about how "Scheme Group" sounds kinda silly.

Edit: What if mathematicians are sued by the league of super-villains for infringing copyright on their monthly workshop, on, of course, scheming.

I'm not sure if that was good enough, but I at least tried.

Anton Geraschenko comments on "On "let's build a really big list!" questions" (1057)

2009-12-14T11:08:14-08:00

There's a [big-list] question I've been thinking of asking, but I'm not sure if I should. Since there are people thinking about [big-list] question here, I'm going to use this opportunity to ask you to screen the question and tell me if I should ask it. Would you post an answer to this question? Would you be interested in seeing answers to it?

Incorrectly named things (with corrections)

Sometimes (semi)-standard mathematical terminology is wrong in some way, and it drives me nuts. I'd like this question to be a place to vent about it, but also a place to suggest solutions to the problem. Moreover, since I believe terminology is important, this should be a place to see how people reason about choosing terminology.

Here are the rules for what your answer must include:

Clearly state what the offending bit of terminology is.

Explain clearly why the terminology is wrong/misleading.

Suggest an alternative you'd like mathematicians to adopt. If you can't think of an alternative, but think others might be able to, at least make an attempt. Give some indication of what properties the correct terminology should have.

Example

In representation theory, a representation is said to be "irreducible" if it has no proper subrepresentations. The problem is that non-representation-theorists get confused and think that "irreducible" is synonomous with "indecomposible" (meaning that the representation cannot be expressed as a direct sum of proper subrepresentations). In most other areas, an object with no proper subobject is called simple, so I try to use the terminilogy "simple representation". More generally, since representations are always modules over some algebra, the language of modules should be used for representations.

Here are some other examples that I would turn into complete answers:

"The proper base change theorem" should be "the base change theorem for proper morphisms" since the base change morphism isn't the one that's proper (c.f. the correctly named "flat base change theorem").
"group scheme" should be "scheme group" or "algebraic group", the idea being that you should specify the category before specifying the construction.
when the same letter in different fonts is used, the scriptier font should have it's long and short vowels interchanged, so "Let X\to\mathcal{X}" be a cover" should be read "Let ex to eeks be a cover."

Anton Geraschenko comments on "On "let's build a really big list!" questions" (1056)

2009-12-14T10:58:09-08:00

@rwbarton: you may be interested in this hack. Make a bookmark in your browsers bookmark toolbar with the following "URL": javascript:void($(".question-summary%20+%20:not(.tagged-interesting)").hide()) Now visit the home page or the questions page and click the bookmark (or just past that line into your location bar and hit enter). All the questions that don't have an interesting tag should disappear.

I like the idea of a [big-list] tag. I'm going to start retagging such questions when I see them at the top of the home page.

Harry Gindi comments on "On "let's build a really big list!" questions" (1054)

2009-12-14T10:50:38-08:00

Sounds wonderful if it can be done.

rwbarton comments on "On "let's build a really big list!" questions" (1053)

2009-12-14T09:47:03-08:00

Actually, combining some of these ideas, how about an option to hide questions I've voted down from the Questions list?

rwbarton comments on "On "let's build a really big list!" questions" (1052)

2009-12-14T09:44:42-08:00

Right, so this reminds me: is there really no way to follow questions marked with any of my ~20 interesting tags, without manually building a URL which contains all of them?

Ilya Nikokoshev comments on "On "let's build a really big list!" questions" (1050)

2009-12-14T09:38:37-08:00

It's instrumental to think about the reasons Stack Overflow doesn't have this problem: they have too many questions to follow them all! This forces people to follow a particular tag, which, by definition, means they're as close to happiness as possible.

It's the same reason I don't care much about the quality of average item on EBay or an average Wikipedia article: since I'm nearly always coming to either of these sites with a specific search, and continue from that place, again, to specific pages of interest to me. That works fine!

Will this "noisy martketplace" model work for MO? Maybe not. Anyway, for those who are interested: here are questions tagged AG, AT, MG, RT, or NT.

Harry Gindi comments on "On "let's build a really big list!" questions" (1049)

2009-12-14T09:18:19-08:00

Maybe the, "Big list of crap" questions could be moved to meta?

rwbarton comments on "On "let's build a really big list!" questions" (1048)

2009-12-14T09:15:08-08:00

I get annoyed even by questions like "Generalizations of Planar Graphs", so I'm not sure what can be done to mollify me. The one feature I would like is to be able to see from the Questions list which questions I've up- or down-voted, so I don't have to check whether I've remembered to do so every time someone adds yet another answer to such a question, pushing it onto the Active page.

Harrison Brown comments on "On "let's build a really big list!" questions" (1047)

2009-12-14T08:46:35-08:00

Ben Webster mentioned the growing trend of CW questions whose purpose is basically to generate a big list of some sort. A couple of these kinds of questions have amassed upwards of 70 answers.

MO's had these kinds of questions from the beginning (mathematical jokes, good writing) but there's been something of a perceived upswing in their density recently. I don't personally have a problem with this (they take up, like, three spots on the front page maximum at any given time, and besides, they can be fun!) but Ben postulated, and I think it's probably true, that some people are getting annoyed with the number of these questions being asked.

So: what can we do to mollify people who don't want to see a bunch of "ginormous list" questions all the time, while allowing people who want to contribute to such questions the opportunity to do so? One idea that just occurred to me is to create a special tag for those kinds of questions, and people could ignore it if they were so inclined. Thoughts?