Bringing back identification questions

Question

As you might know, the topic of identification questions has been a rather controversial one over the site's entire history. About 6.5 years ago, the site decided to ban these questions as off-topic. However, as time went by users have grown concerned that this might have been a wrong decision. Therefore, we recently had and have a discussion about what we want to do with these questions and specifically if and in what capacity we might possibly want to welcome them back on the site again.

Over the last month people shared various viewpoints in that discussion and the most popular ones argue for bringing back identification questions in some form. However, a part of the userbase is concerned that this won't actually happen, due to a lack of trust in the moderation team as well as somewhat of a felt dependency on clear moderator statements about these issues, as also recently addressed in a follow-up question. So first of all, in an effort to clear some of these doubts, let the moderator team acknowledge that we have seen this discussion and that the community would like these questions back in some capacity. So let's get this process rolling then!

Another concern by some users was that the past decision for banning identification questions hadn't involved enough community participation. So we would like this process to be as transparent and community-driven as possible. It is clear the userbase has changed its mind to a significant degree, so we want to try and involve everyone to participate in making these questions work again.

---

We know that some users have been waiting for a change in this regard for a while and would prefer to see swift work made of it by the push of a button. However, we also have to see and acknowledge that a significant part of the discussion has been about regulating these questions in some form or another, in order to reign in the quality problems that once caused them to be banned. And in the same way as we can't ignore the general consensus of having these questions back, neither can we ignore this part of the discussion. So rather than outright unbanning and reopening all the questions immediately, we'd prefer the community would work at least some of these issues out, so we have a base to work with and can find some agreement on how to proceed with these questions once they're back. Let's use the current state to prepare for a smooth arrival of identification questions once we open the gates for them again.

We are in the privileged position that we already have years of experience with these questions and possible methods for improving them. This is not to say everything we tried back then worked (if it had, we wouldn't be in this situation) or that we didn't make mistakes. But we don't have to start at square one and see what questions we get. Many of us who've been there still know and can share this experience together with some constructive solutions. In addition to that, we also have some new tools at our hands, which arose in the meantime.

• First of all, we adapted the identification close-reason to better reflect this transient state of the policy and to encourage working this stuff out. We know it's a bit weird having the close-reason there while not being fully clear on which questions are actually okay. But in fact you are encouraged to also use specific examples of them to show what should or shouldn't be appropriate in a future policy and also close- and/or reopen-vote them accordingly. Understand this as kind of a reorientation phase similar to the early beta site period, where we are finding our exact policy with the help of specific examples. However, please try to ultimately have the meta discussion find its way back here, instead of engaging in lengthy comment threads under each and every individual question.

• In addition to that, also consider specific guidelines you think might make sense for these questions. Do we want to require specific hard facts about the films in question with regards to time frame, language, actors, scenes? Anything else you can think of? How many of these facts are enough? How precise is precise enough?

• Then also think about the technical tools at our hands. Do we still need an explicit close-reason or do we rather want to handle specific questions on a case-by-case basis with the existing close-reasons? In the former case, what should it look like? Back in the day we also had a tag warning (well, we still have, only it's an error right now). It seems a good idea to bring it back, but what should go into it? Then we also have the newer feature of a customizable question asking wizard, as also addressed in this discussion. Do we want to address identification questions specifically there? And what should go into it?

---

Yes, this process requires some work, but it doesn't have to take months, like some of you are afraid of. We don't have to have all the answers immediately and we don't need a completely fleshed-out plan for each and every technical site feature, but we would like the community to share their ideas on this and involve them in the process of getting these questions back as much as possible, in an effort to have the transparency and community agency some still see lacking here. If noone shows up at all because noone actually cares, then that would be unfortunate. If you say "Requirements? Nah, don't need them at all!" and suddenly get unanimous applause, then that's what it is. And if all we get are some rough ideas of what could possibly be done, then the site has to work from this. But at the very least we want you to think about it and to talk about it while the site is still preparing for the questions' arrival. We know it's a bit confusing with all these related discussions going on right now and some of this could have gone into the more general discussion. However, now that we know the questions are coming back, let's focus specifically on how they're coming back and which ones we want to actually bring back.

We'd like to close this with an appeal to be nice. This has always been a controversial issue and it will likely stay one for a while. On the one hand, this means being nice and welcoming to the new users who come here with their question (who might also not be aware that we're currently in somewhat of a transition phase). But this also means being nice to experienced users who might not necessarily share your specific views on this issue in all aspects. Working this out as a community is the goal here and with as diverse an issue, this is quite a journey. So let's stop walking and start running.

Answer 1

First, I think a good way to attract and then evaluate quality ID questions would be emulate a site that has this process in place.

I would start by borrowing from our sister site's, Science Fiction and Fantasy, Tips for asking Story Identification questions. Then a tailored approach can be taken since this site is different in that 1) it accepts all genres but 2) it only accepts movies and television. This was also largely copied by Literature in their discussion on IDQs as well.

Below, I have quoted out the parts that came from the SFF post and then my thoughts underneath.

• No detail is too insignificant. You might think it's ever so obvious that your characters are white or teenagers or whatever, but you still need to tell us. If you don't tell us, we don't know it.

Okay as is. If a question came to us, it obviously means the user is having trouble identifying, but that doesn't mean we are clairvoyant. In addition, prompting for details may help the user solve/find it on their own.

• Try to offer unique details. That means you need to think about character names, details of any incidents that occurred in the plot, any dialogue that you can remember, any words invented by the author, etc. Anything that can help distinguish your story is key to finding it. Thus, don't refrain from revealing "spoiler" parts; somebody else might remember the same thing you did. If you want, you can partially hide them by preceding the spoiler with >!. For example >! but he dies at the end.

Again, okay as is. Many movies have similar themes and character designs, anything unique narrows down that field.

• Be available to answer questions about it, and answer them in a timely fashion. Story Ident questions will often lead to comments asking for more details. Those users who enter into a dialogue about their ident are almost certain to get more interest than those who post and run away, never to be seen again.

To me this should apply to any question asked. If you want to receive a quality answer, be ready to help obtain it.

• What is it? It may seem obvious to you that your question is about a book/film/TV show, but harking back to "you need to tell us", if you don't say what it is, you've hurt your chances of having it identified. If we have to spend ten minutes puzzling that out, that's ten minutes less time we'll spend actually helping you.

This one probably does not apply as much since we are strictly dealing with specific media types (Movies & TV, duh). However, it might help to say where it was seen. Was it in a theatre, a made-for-tv movie, a feature film that was later aired on television, etc. It would also help to identify what the production type was, be it animated, live-action, computer graphics, etc.

• When did you see it? It's absolutely no help to know that you saw it when you were younger. Be specific. Tell us when you actually saw it and whether you think it was new or old when you saw it.

This one is relevant as well. With all the remakes, reboots, and running series out there knowing when you first saw it or what period it seemed to come from can narrow down options.

• Tell us what made it memorable. I've noticed that very generic ident questions can often become answerable when people share why they remember it so vividly. This can also lead to further details emerging about the way that the story was portrayed, the audience it was intended for or where you saw it.

Same as offering unique details. Anything that helps it stand out is welcome.

• Tell us what it isn't. If you've already spent years posting the same question around the internet, make sure you tell us. There's nothing more frustrating then thinking you've found a mention of the story only to find a post from 2007 on Yahoo Questions that's identical to the one that was just posted. If you've ruled anything out already, let us know so that we don't waste time following the same lines of enquiry.

This is a big one. Help eliminate some upfront grunt work and unnecessary back-and-forth of "is it X? -- no... -- is it Y? -- no..."

• What did it look like? Some questions have been answered through the askers describing the appearance of the cover of books or through rough sketches of the work they are describing, such as this one.

While book covers would not relevant, still identifying film styles or techniques would be useful; or as mentioned before if it is live-action, animated, etc.

• You've never finished asking. After posting, come back and review what you've put. Try to add at least one extra detail, no matter how measly. The more you edit, the more likely you are to remember something else. Don't forget that each quality edit will mean more attention for your question when it gets bumped back to the top of the front page.

Again, original poster engagement is key. We don't want drive-by questions. We want to help users and then to keep them involved in the site.

Oh, and if someone posts the right answer, remember to be polite and accept the answer, even if that means having to spend a few minutes finding your login details!

A nice reminder that the person answering question wants a little validation too! It would also help with closing duplicates.

---

Next, for the tools and techniques:

Do we still need an explicit close-reason or do we rather want to handle specific questions on a case-by-case basis with the existing close-reasons?

I think an explicit close reason is not neccessary, but it could be helpful since the space to keep it would still be available. It could also point to the relevant meta discussions to help new users understand the feedback they are receiving.

"Low Quality ID Question" - This question does not meet our site standards for asking identification questions. Please add in further details that may help our experts search what you are looking for. Take a look at our Tips for Asking an Identification Question [link to meta].

Back in the day we also had a tag warning (well, we still have, only it's an error right now). It seems a good idea to bring it back, but what should go into it?

Yes, we already know IDQ's will be popular. Anything we can do upfront to point new users to resources to help them provide quality posts is helpful. This will also provide a first line of defence in identifying low-quality posts or drive-by users who may not come back; if they didn't take a small amount of time to read the guidelines before posting, why will they do that going forward.

"Our site has minimum standards for asking Identification Questions. If you have not done so already please take a look at our Tips for Asking a Good Identification Question [link to meta] and provide as much detail possible.

Then we also have the newer feature of a customizable question asking wizard, as also addressed in this discussion. Do we want to address identification questions specifically there? And what should go into it?

Again, yes, anything to help user provide quality posts is going to be beneficial. Although I am not as familiar with this tool, so I am refraining from further details.

---

Lastly, one other thing I will note, is that we have to embrace the "non-googlers". We have to assume a good faith attempt (until proven otherwise) and that something is preventing the user from locating the work on their own. We can't assume a user has access to reliable search engines or sites like IMDB.

Answer 2

We can streamline our handling of identification questions by adopting a flexible strategy similar to Science Fiction & Fantasy SE, using an existing close reason, and creating a "Tips for asking identification questions" guidelines meta post.

Do we still need an explicit close-reason or do we rather want to handle specific questions on a case-by-case basis with the existing close-reasons?

Like in SF&F SE, instead of creating a new, specific close reason for these questions, we should consider using our existing "needs details or clarity" reason and handle them on a case-by-case basis, based on the details that OP provided. This approach would be particularly useful for questions that lack sufficient information to narrow down to a single or very small number of potential matches, as Valorum mentioned in his answer.

Using this existing close reason has several advantages. First, it avoids the potential misuse we've seen with other specific close reasons in the past, such as the "trivia" close reason we removed. There's a risk that an ID-specific close reason could be used as a super-downvote by users who generally dislike identification questions. The "needs details or clarity" reason requires close voters to specify why the question lacks sufficient details, encouraging more thoughtful close voting.

Back in the day we also had a tag warning (well, we still have, only it's an error right now). It seems a good idea to bring it back, but what should go into it? Then we also have the newer feature of a customizable question asking wizard, as also addressed in this discussion. Do we want to address identification questions specifically there? And what should go into it?

A tag warning for ID questions and updating the "Ask Question" modal similarly could significantly improve the quality of ID posts. By adapting guidelines from our community and borrowing ideas from Science Fiction & Fantasy SE, we can create a concise yet informative warning. I've drafted a potential tag warning that we could use as a starting point (we can create a separate meta post to discuss and fine-tune the tag warning later):

To increase your chances of getting a helpful answer, please include as much detail as possible:

• When did you watch this movie or TV show?
• What genre does it belong to?
• What plot points, tone, setting, characters, actors, or other unique details do you recall?
• Which movies or shows have you already checked and ruled out?

For more tips on how to improve your question, see our guide on asking identification questions.

This tag warning serves as a gentle reminder rather than a strict requirement. Like in Science Fiction & Fantasy SE, it's not intended to be grounds for closing questions, but rather to encourage users to provide more comprehensive information. We could use similar content in the "Ask Question" modal to guide users as they compose their questions.

It's important to note that implementing these changes may take some time. The process typically involves Community Managers (CMs) and could stretch over several months. Additionally, our identification tags are currently blacklisted, which also requires CM intervention to resolve.

Given these potential delays, I suggest our moderators proactively reach out to the CMs about our policy change. A good starting point would be requesting the CMs to unblacklist the [identify-this-*] tags.

This early communication could help expedite the unblacklisting of relevant tags, the implementation of tag warnings, and the "Ask Question" modal update.

---

To conclude, and most importantly, the discussions regarding the possible addition of a close vote reason, requiring specific details, tag warnings, and other aspects can be ongoing. We don't have to wait for these discussions to conclude before removing the ID close vote reason and allowing ID questions. We can continue to refine our approach as we receive new identification questions, using real examples to inform our decisions. Reopening and undeleting ID questions can be addressed in separate meta posts after the close vote reason is removed. This approach allows us to adapt our policies based on practical experience rather than theoretical discussions alone.

Answer 3

I propose that we base our moderation of ID questions on the principle that a...

A suitable question must contain, either singly or in combination, a level of sufficient detail that the field of possible answers can be narrowed down to a single or incredibly small number of properties^∗.

If a question contains this much info, it should stay open. If it doesn't, it should be immediately closed and work should be undertaken to rehabilitate it to the point that it can be reopened. This has the benefit of being simple, clean and very easy for all users to understand.

Conversely, demanding a level of detail beyond this is simply going to result in good, answerable questions being closed, defeating the purpose of re-allowing them in the first place.

---

^∗_{For example, a combination of factors could be some common facts that are very unlikely to match anything other than a single target:}

_{"This early 2000s film was about an environmental paralegal}

_{Or it could be one highly identifiable fact.}

_{This film had a main character called Earring Brokovitch".}

_{In either case, a relatively committed searcher could narrow the field down to a single film.}