12

There have been a number of discussions on tags where it seems like the *-identification questions simply throw a wrench into the works. Genre tags, media tags, oddball topics/genre hybrids like and . There's any number of tags that don't necessarily stand up to scrutiny as helpful tags, unless you combine them with *-identification tags (or stuff like or ), such as , , and more.

I feel like there's a regular debate about whether or not a tag is worthy enough to exist based on how/when it's used outside of *-identification questions. Most of the tags that pop up, including the 3 I mentioned above, are mostly *-identification buddies, but have no usage guidance indicating that's the case. We also don't have any strong guidance on when we should apply such tags if the OP doesn't.

Here's some facts:

These questions are a massive portion of our entire stack, for better or worse. Simply tagging them with + genre or + media doesn't scale. Searching by + still returns over 600 questions to sift through and + is still over 1,500, in no way helpfully guiding anyone to a question due to sheer scale. They used to be sufficient, but it's no longer the case, and more guiding tags are often met with pushback or even scorn.

Should we have separate guidelines specifically for making or allowing tags meant specifically to pair with identification questions, outside of the few that have already been grandfathered in? As opposed to a stream of discussions on whether or not we should nuke tags whose primary purpose is to facilitate , as seen below.

See:

And more, where we have have discussions on tags that are heavily or primarily used by ID questions, and many of them fire fighting tags that are otherwise bad, partly because they're being used inconsistently and for story-id questions, and we have no tag guidance on making supplemental genre or topic tags.

The purpose of the discussion is to identify whether or not the community feels that this is a problem and we should put effort into addressing it, or it's fine and we should just continue to let things develop as they are.

Coming up with guidelines to focus on types of tags to make, how to word usage guidance, and how we should enforce usage could take considerably amount of energy that's simply not worth it if the community doesn't want to do anything.

Additionally, others may have noticed issues with the system that I haven't, or have counter-arguments to what I've called out as issues.

  • If this would mean using tags like science-fiction-genre on story-identification questions as a matter of policy (thus bringing back the feared and blacklisted science-fiction), I’d be wary of it. – Adamant May 17 '17 at 20:20
  • 1
    @Adamant It'd mean, I would hope, coming up with actual guidelines for creating tags meant to assist story-id, instead of our current process of: "Oh, we noticed this apparently crappy tag.... oh but it's used on a bunch of story-identification question, what do?" that happens regularly. – user31178 May 17 '17 at 20:23
  • In this instance, the additional tag is helpful in guiding the questioner to an answer. – Valorum May 17 '17 at 20:29
  • 4
    Also, erm. What precisely is the question here, other than should we think about doing something about something? – Valorum May 17 '17 at 21:05
  • @KutuluMike - Except that that would make it immeasurably harder to answer the damn things. I get that not everyone likes Story-ID questions but that's no reason to hamstring those that do. – Valorum May 17 '17 at 21:14
  • @kutulumike Where do I argue they're not useful for story-id. I'm saying they're useful, but the ones we have, simply high-level media tags and such, aren't usefel enough because of sheer scale, so I'm hoping for less pushback against additional types of tags meant for story id – user31178 May 17 '17 at 21:20
  • @KutuluMike - Many questioners use media tags to identify what they're actually asking about. – Valorum May 17 '17 at 21:21
  • 3
    @Valorum 1) tags are not a substitute for asking good questions, and 2) if media tags are only meaningful when paired with story-id... then perhaps they shouldn't be separate tags. – KutuluMike May 17 '17 at 21:23
  • @kutulumike I did write this hastily, so I've updated as such. But I disagree about tags not being useful and also your lack of investment, given your involvement here already ;) – user31178 May 17 '17 at 21:24
  • @KutuluMike - I don't entirely disagree, but what you seem to be suggesting would require systematically retagging nearly ten thousand questions to make them into Story-ID-Movie, Story-ID-Novel, etc etc – Valorum May 17 '17 at 21:25
  • @CreationEdge - I'm honestly not sure what the purpose of this meta is. I feel like I've wandered into the mid-point of a train of thought. – Valorum May 17 '17 at 22:06
  • @Valorum "I would like to stop fighting against the wave and instead start making guidelines to assist in making more and better tags for story-identifaction et al.? What should our guidelines be?" Not sure how that purpose is unclear. Current guidelines for tags don't separate out ID as a whole other beast. But they are a separate beast, and they need separate considerations. – user31178 May 17 '17 at 22:08
  • @CreationEdge - What wave? – Valorum May 17 '17 at 22:09
  • 1
    @KutuluMike I have included links to several examples of tags where some users wanted to or have nuked them because being used for story-id wasn't justification enough. On top of that, many of the tags that we seem to mainly just use for story-id offer no usage guidance that it's the case. – user31178 May 18 '17 at 0:24
  • 6
    I often find myself searching for an old story-ID question. It comes about this way. As I randomly read old stories, from time to time I come across one that seems to remind me of a story-ID question I've seen somewhere, not necessarily on this site. Then I have to try and find the question, and it isn't easy. Intelligent use of keywords could make it easier. As it is, I seldom use tags for searching, because tagging is too hit-and-miss. – user14111 May 18 '17 at 7:37
9

First, I'd like to apologize for having my question itself be unclear. I noticed a problem, and I'm having trouble figuring out the best way to call out that problem as a "question" for meta, while my thoughts are also filled with possible solutions and related problems. Anyway, let's start.


Intro

Here's our list of identification tags:

, , , , ,

The comprise 22% of all of our questions. / adds another 1%.

These types of questions serve several purposes, which I'll address individually.

Helping the asker find what they're looking for

This is the most basic and obvious use of the questions! They are helpful to the person asking, because they get the information they want and leave.

The primary source of criticism from these questions comes from comments similar to Jeff Atwood's on his infamous blog post about the subject. But guess what? Jeff Atwood hated the idea of SciFi.SE, and left the company partly because his vision of what the Stacks should be didn't match up with what they were becoming and what they are now.

I don't mean intend this as an ad hominem, but I believe the criticisms outlined there aren't valid, simply because we as a community have long ago decided otherwise.

On the premise that these questions are useful at the very least to the asker, then we can increase their ability to receive an answer by increasing the visibility of the question to those who might have an answer.

Helping people with expertise find the question

We've done this so far using media tags such as or . And, we clearly have experts in some of these tags. Whether they're experts because they've seemingly read every short story known to man, or because they're adept at searching their resources based on small clues, having these media tags helps to dramatically reduce the scope of the search through memory or data.

You can browse unanswered questions in a genre, such as or target audience such as , to help narrow things down.

Yet, even by combining those tags, sometimes there's simply too many to sift through. Plain search terms can also be a mess, something easily seen when looking for questions, where searching slytherin on its own raised far too many false flags to be useful (this is a direct counter-point to the argument that character tags aren't useful).

This means we have a need for some more specific tags, to help categorize questions. We already do this to some extent, which I'll address later, but it's been mostly slipshod up to this point, with a clear goal of aiding identification questions in mind.

Helping people who read the question before find it again once they have an answer

Sometimes you read an identification question, don't have an answer at the time, but later re-read the work or come upon it in a search and realize it's the answer to that question. The trouble is, finding that question again later can be a problem!

If you remember it was about a female character do you narrow it down by searching "female" or "girl" or "chick" or "lady"? Or if the main focus was about a UFO, do you search "ufo", "space ship", "spaceship" "aliens", "first contact" or what? Do we resort to Google to try and find stuff on our own stack that we know exists but simply isn't cataloged well enough? Maybe. But I think we can do better.

Helping people come across interesting stuff that they might want to look into

This is perhaps the most controversial purpose of the usefulness of identification questions. It's certainly not part of the advertised or intended purpose of our Q&A. Yet, intent or not, it happens. Perhaps I'm alone in this, but I have read issues of comics or short stories simply because they showed up on the stack. Similarly, I've gone through and read passages from Tolkien's writings after browsing through tags such as .

This has provided me two uses, then.

  1. I've increased my knowledge, and thus realm of expertise, should any similar questions come up in the future
    • Duplicates are not uncommon, there's 618 of them for alone
    • I could also rule out the book if I know it's not a match
  2. It's given me something fun or interesting to read (which also describes 75% of my interaction with the stack).

So... what?

So... there's a need to make it easier to find topics within identification questions, just as we've done with our other large tags. If you look at our tag map you'll see tags like , , and more. Each of these work-specific tags has been somewhat controversial in the past, yet are widely used today. It's a need that seems to be easier to pinpoint or put to paper for specific works, but it exists for identification (and history) questions, too.

New(?) guidelines

Perhaps not new. It seems to me that we have some loose idea of guidelines for identification-supplementing tags, we just don't put them into practice consistently, or appear to make the tags with that specific intent. Rather, a tag gets made by whomever and it stays being poorly-defined or used for a long time. A recent specific example is the ongoing discussion of , a tag that has been around for over six years, but was only brought up this week. Or which was completely burninated, removing it even from questions where it was correctly applied and used the most (and the problem appeared to be the usage guidance, not the tag).

What I'd argue is that these two tags are both examples of genres which can significantly help narrow down the scope of the questions needed to look at to find a question or its answer, and aide users in the same manner as and .

The key is to be explicit in that purpose by:

  • Making tags clearly refer to the genre/sub-genre
  • Provide clear tag wiki guidance
    • medieval from: Use this tag for works written or based in Middle Ages (5th to the 15th century).

    • medieval-fantasy to: For identification or history questions about a work in the medieval fantasy sub-genre

    • female-characters -> female-protagonist: For identification or history questions about works notable for having a female protagonist

    • Update other genres, such as to have new verbiage indicating "For identification or history questions..."
    • Always keep in mind that tag wiki excerpts are not meant to be definitions, but instead provide usage guidance
  • Proactively removing misapplied tags to non-identification questions (here, would get an exception)
    • This would have saved , make have a clearer scope, fix and by contrast show us that none of the current questions are valid (since it's already slated to be axed anyway, and under these guidelines, we'd be expecting the tag to be used for works told from the villains point of view)

I would also propose that, over the normal course of things, if an identification questions gets answered, we should go back and edit in those genre/sub-genre tags, which adds to the usefulness of the whole endeavor.

However, this is not a call to start tagging things with lettuce. But there are a wealth of sub genres available to SFF, and categories that cross genres such as female protagonists, anti-heroes, alternate histories (a note, currently is about 1/2 ID questions that'd fit these guidelines, and 1/2 questions where it's used poorly enough that we'd normally considering axing the whole thing), and more.

Closing

I don't see that any of my suggestions fall outside the line of what we have long accepted as useful or expected, but I think the tags are certainly neglected and we have a predisposition towards nuking tags instead of improving them. Sometimes we consider that they add use to identification tags, but some of the discussions I've seen or participated in have been about whether or not existing solely for the purpose of adding it to is enough. Rather than trying to see if the tag can stand on its own we should should consciously address it from the viewpoint of:

"Are a high proportion of these questions identification questions? If so, should we make this an identification-only tag? What steps do we have to take to improve the tag for ID-only?"

Input and feedback are appreciated. I'm also working on compiling a list of some of the tags that I have found that are the most obvious examples of needing to be categorized as "identification and history" tags, which may or may not help my argument. (As you've seen here, there are some tags that on their own are bad or simply poorly used, but when paired with make a lot of sense. There's quite a large number.).

  • Incredibly enough, the entire premise is based mostly on what the asker find easier, rather on what the answerers interested on these topics want. – Braiam May 31 '17 at 13:38
2

I've been a little confused about the point of getting more tags on this. There are only two possible advantages that I know of currently. Initially, I was thinking of how it would help to find questions that you remember exist, but aren't sure what the name of the book was (especially useful if they are duplicates of a more recent post). Then, in the comments, CreationEdge pointed out another possible benefit I wasn't focusing on: guiding people who want to write answers to questions they might be able to answer

Search for finding answered questions you know exist

I don't think it's practical to expect someone to be able to find a particular story-identification question only through tag search no matter what types of tags are used, so I don't really see the need for coming up with some kind of better system.

Searching by story-identification + novel still returns over 600 questions to sift through and + short-story is still over 1,500, in no way helpfully guiding anyone to a question due to sheer scale.

The obvious way to narrow down these search results is to use search terms as well as tags. People searching through these questions can type key words that fit the story they are thinking of. I don't see how tags would make searches easier.

Search for finding questions you might be able to answer

In terms of searching for questions to answer, filtering by hasaccepted:no is definitely also useful. It looks like there are around 5000 story identification questions without accepted answers currently, 824 "short story" identification requests without accepted answers, and 363 "novel" identification requests without accepted answers.

However, I feel like it's never going to work too well to use search to find story identification questions that you might be able to answer. It's not very scalable—you're getting one person's eyeballs on the question (your own) and no matter how much of an expert you are, it's likely that you won't be familiar with the story. To me, it seems like these questions usually get answered either through being on the front page of the site and getting lots of eyeballs on them—some of which might belong to people who are familiar with the work—or through the steady stream of visitors who arrive here from Google search.

  • It's not really about finding a question that's the same as yours, but a selection of questions that may be in an answerer's realm of expertise or interest. Simple search terms are pretty terrible at acting as categories. – user31178 May 18 '17 at 4:58
  • @CreationEdge: I see. From that perspective, it seems like it would make sense to filter out questions that have accepted answers. – sumelic May 18 '17 at 5:02
  • That doesn't accomplish much if you still can't find one that might possibly be in your area of interest, considering there's still 500 more unaccepted ID questions than all Star Wars questions – user31178 May 18 '17 at 5:07
  • @CreationEdge: Searching for story identification questions in your area of interest doesn't seem like a very effective strategy to me--I've edited my post to explain why. – sumelic May 18 '17 at 5:12
  • Anecdotally, it's worked well enough with short-story and comics, so why not additional tags? It's never going to be a perfect system, especially since we'll never get tag hierarchies, but we shouldn't avoid doing something because it won't be a 100% solution when the system is designed to not give us 100% solution. If it's helpful to the people that would use it... Then it's useful. – user31178 May 18 '17 at 5:24
  • @CreationEdge: I'd be interested in hearing more (like a post or something) about the way it's helped with short-story and comics. Right now I kind of have a hard time understanding how these tags have been beneficial for getting story-identification questions answered – sumelic May 18 '17 at 5:28
  • Well, you've given me stuff to focus on as I work on an answer. And I'm pretty sure tagging story id with comics or short story naturally draws our resident experts to those tags. Sometimes I even just browse combos to see if I know anything (or to find interesting stories I've never read). Not sure there's any data on it, except maybe our gold badge users. – user31178 May 18 '17 at 5:31

You must log in to answer this question.