What we'd like to do about those gosh darn "identification requests" questions

Question

This discussion is now closed.
Here's phase 1 of what we've decided to do about identification-requests (deprecating).
Here's phase 2 (blacklisting).
Here's phase 3 (delete and lock).

TL;DR for those who want to find alternative resources for ID request: What identification requests are acceptable here? Otherwise, where else can I possibly get helped with it?

---

The topic of id-request is a very subjective one. I believe that it's difficult to provide an objective data for analysis. Afterall correlation does not imply causation.

Id-requests are something we struggled with dealing with since they first came up on our site. Objectively on paper, they can make decent questions. But in practice, this is far from true. Some common complaints about these types of question are that:

• They are too localized and oftentimes vaguely worded, and typically unhelpful to other users. Someone trying to remember the same thing as you might recall certain details more vividly than others, this oftentimes results in different descriptions of the same thing.

  There's no easy way to resolve this. Answering identification question is not unlike giving directions: if you don't give someone good directions they'll have a hard time finding what they are looking for.

  Take a look at this map, made by a directionally-challenged character from a popular manga:

  

  

  Would you find what you're looking for or where you need to go? The place you're looking for based off of these directions? Probably not. However more often than not we get users the ask question that feel like these maps, despite our attempts to help them include better details with our guidelines.

• Users make little effort, but expect the world in return. We've seen a lot of these types of id request questions. Most of them are in the form of image-only id requests, which got out of hand really quick, mostly because the ones asking the question almost never tried doing a reverse image search themselves and expect us to do it and tell them. We're not here to be someone's personal search engine/concierge service, we're here to ask and answer questions about the world and culture of anime and manga. Filling our site with these questions can be as bad as watching filler episodes of a long-running anime. It gets us nowhere.

  Since then we've disallowed media-only (only a link to an image, audio, video) identification questions, but we've made some provisions to allow users to ask questions about identifying specific merchandise, cosplay, and the rare but occasional itasha, using actual pictures that you've seen or taken yourself.

  As of this post, the help center topic for what's on-topic and our tour page have been updated to note that users should follow these guidelines when asking identification request questions, including our updated stance on media-only identification requests.

• Rarely do users that ask id-requests stick around or bother to come back. Statistically speaking the retention numbers on id-request askers are low. Not only that, but a fair number of them are asked by unregistered users, whose accounts are cookie-based and volatile, meaning that once they lose their session cookie, they have no way of logging back to this account (without registering that account). This results in many id request questions being essentially abandoned.

While we care about making more of an effort to downvote and close these questions, the fact that these post continue to pop up and with the same issues, is a bit of a cause for concern.

I, for one, do not like to see the current trend of where identification questions are going and would like to start deprecating them. We've made many efforts and guides to help users write better identification questions, but as time goes by, reminding them of such gets pretty tedious. Our previous stance was to downvote and close these questions based on our guidelines. That's all fine and done, but it requires users to actively pay attention to these questions as they come up via the review queue or on the main page. Instead of devoting our time to dealing with these questions, I'd like to see our community's efforts direct to community development between users. We've done gift exchanges, viewing sessions, and even a multi-discipline project like game development: I'd like to see us nurture this spirit more. Therefore, I believe it's in the community's best interest to deprecate and eventually remove identification questions in their entirety from our site.

I do understand that there is opposition to removing the questions completely, which is why I'd like to give them one last chance to not only be heard but also defend their position on this matter and what they believe would help improve the situation of dealing with the influx of low quality and abandoned id-requests. Maybe we've overlooked something, maybe we could have done something better, or maybe we went about things in the wrong manner.

You may discuss or scrutinize the plans as much as you'd like here on meta or in our main chat room. Regardless of how much or how little reputation you have in this community, your proposal will be heard equally, as long as there is consensus on it within the community. However, note that this is likely to be the final time to voice your opinions. I would like us to finally reach a decisive conclusion on this matter once and for all.

You have until May 8th to make your case (for, against, other) and leave any comments.

However, while such question are off-topic on our site, they will still be welcome in our main chat for users with with at least 20 rep (globally on StackExchange).

Answer 1

I am against keeping identification question on the site and I believe that there are quite a few users within the community that feel that same way on this matter. I would like to propose the following plan for deprecating and getting rid of them:

1. Start closing all NEW identification-request questions with a modification to the existing off-topic close reason to remind users that all identification questions are off-topic. All users are encouraged to vote to close them as they pop up.

2. After 30 15 days, blacklist the identification-request and associated tag synonyms, close ALL remaining identification-request questions, and apply a historical lock to them. This will most likely be done via a feature-request to the CMs.

3. After 60 30 days, delete all identification-request questions with a score of 3 or less less than 4, without at least one answer with a score of 3 or more. The ones that remain will serve as a reminder to other (new) users that while we have allowed these question in the past, these question are no longer on-topic on our site. These questions will remain and be untagged and locked permanently. Future questions should not only be locked, but also deleted.

   Doing so will is meant to preserve some of the better identification request questions for purposes of keeping historical artefacts. The three-step process is meant to be less abrupt and ease everyone into the removal of the identification-request tag and questions tagged with it.

Removing the tag will most definitely have a rather significant impact on our question and answer count. This is why, in return, I would like to us to start organizing more community events to talk about, recommend, and watch anime in a monthly basis (or weekly, if the initiative gains enough traction).

Some examples would be short weekly collaborative watch sessions for on-going seasonal anime (two sessions, each one for people in different time zones), and a monthly binge session of an entire complete series or set of movies following a theme (there will be 2 time slots to accommodate different time zones as well).

There is also the possibility of hosting themed community events involving asking question on certain tags: we've done similar things in the past, both when with seasonal tags and themed tags.

I'll give you guys until the end of the next month (May 8th) to mull over this, before implementing anything.

This is a modified version of @Makoto's data query that shows the question that will be kept.

However, while such question are off-topic on our site, they are still welcome in our main chat for users with with at least 20 rep (globally on StackExchange).

Answer 2

Id requests suck, I've been convinced for quite a while that we should get rid of them, and I agree with ʞɹɐzǝɹ's plan to do so.

While there are many reasons they suck, and many reasons to get rid of them, I want to focus on one in particular--it takes way too much time and energy to police them.

On a typical day, I visit the site four or five times. Each of those times, there's usually at least one new id request to look at (I view the site from the "All questions" tab, ordered from newest to oldest). It's usually written as one long sentence all in a single paragraph in shoddy English, with awful punctuation and incomprehensible word use. Sometimes this is because the asker is a non-native speaker of English, but sometimes the asker is a native speaker of English who is just too lazy or ignorant to put any effort into decent writing. I have no way to tell, so I have to go in and make paragraphs and use commas and fix typos to make it minimally presentable. I've also started cutting out all the useless fluff that people add to id requests—"Thanks in advance", "When I was young I saw this show I really liked", "Please excuse my poor English", "I'm going crazy trying to find this", "It was really cute", "It had great art", "I met my spouse because of this manga", etc.

There's also been a rash of users recently who don't understand how tags work for some reason, so they just toss a bunch of random tags on their question instead of tagging it identification-request. naruto, bleach, pokemon, dragon-ball are all common but I just voted to close a question which had been wrongly tagged satisfaction-guaranteed. So those have to be changed. (Since so many of our questions are id requests, it boggles my mind that they didn't see another question tagged identification-request and figure out how they were supposed to do it. But id questioners do keep finding new, inventive ways to suck.)

Then the question is usually horribly vague, so I leave a comment asking the OP to add more information, like when they saw it, where they saw it, what exact year it was, etc. Typically I also downvote, and sometimes vote to close. Since I like to retract downvotes and close votes if it's justified, I come back later to check on it and see if the OP has done anything to improve the question. Very rarely is that the case. Usually the OP has posted some kind of whiny comment asking why everyone is complaining when that's really all they remember. Sometimes I try to explain to them why their question is bad, but it's usually a waste of time.

This process becomes draining, tedious, and frustrating after a while. I imagine it's also frustrating for the question asker. Also, the time I spend doing this is time I'm on the site, not asking or answering questions--in other words, not actually building up the site, but rather just trying to stop it collapsing under the weight of bad questions. Not making a better site, just trying to prevent it becoming a worse site.

I know it's not just me. I can see other users doing the same thing—users who once contributed great questions and answers, some of the best on the site, now spend all their time on the site sifting through garbage on the off chance of finding, not a diamond, but maybe an AA battery that isn't quite dead yet, or a jar that could make a good container for nuts and bolts and screws. I can see these users making the same edits, leaving the same comments asking for more information, voting to close, arguing with whiny questioners who expect service two hours ago, dammit, and an apology from the manager for the waiter's rudeness.

It's just not worth all this trouble. Whatever value id requests add to the site—and, as we've discussed before, many of us believe they don't add much—that value is not worth all the time it takes to police them according to the strict, baroque guidelines we've set for them. Instead, we could have users out there writing better questions, or answering better questions. If we ban them, they'll keep showing up, and we'll still have to vote to close them, but just automatically casting a vote to close isn't nearly as draining, tedious, and frustrating as that entire process of trying to goad an unwilling OP into writing a decent question.

Answer 3

Originally I would be for keeping identification-request on-topic on the grounds that it can be a good starting point for new users.

HOWEVER!

I have now changed my stance. I am in agreement with making them off-topic and the removal of low scoring questions.

My original view was that making an id-request question wasn't hard, and that new users would be able to post good quality questions. However this has not been the case, and the community has had to

• Show how to do reverse images searches
• Establish minimum guidelines
• Remind users about these:
  • When using the tag
  • In the tag
  • In the Tour
  • In the the help center

in an attempt to try and improve the quality, but it appears that new users — and sometimes normal users who never used the tag before — seem to ignore all of these.

However, we can't put all the blame on questions. Stack Exchange is a Question and Answer Site. Making a lot of high quality questions is one thing, but we also want high quality answers too.

Like questions, I thought that new users could easily answer an id-request and get some rep — since 1 upvote = 10 rep, and accept = 15 rep, compared to questions which were 1 upvote = 5 rep. And writing a good answer to an id-request isn't hard: you name a title you believe that matches, give a description of the title then list how the title matches and possibly what parts may have been misremembered.

However, too many times, when I see an answer to an id-request by a new user, they are just one-liners, sometimes with a link, sometimes an illegal site and sometimes with padding crap like:

it's Sakamoto desu ga!!!!!!!!!......................need 150 characters¹

and even when the community posts a comment notifying them of this, it usually falls on deaf ears and another user has to use their time to edit the answer to include even a description of the mentioned title.

While other users editing answer isn't new, it's normally formatting, re-organizing text to flow better and removing banter. But for id-requests, users are more than often editing answers to add in new information (such as the case with this answer). Aside from the fact that the editing user is giving up time for another user who clearly doesn't want to put any effort into their own post, a third-party adding in more info into an answer is traditionally an edit rejection on the ground of conflicting with the author's intent (with the exception to quoting content from links).

---

Another thing is that id-requests are also the only tag which one can't ask and answer at the same time. Sometimes when a user discovers something interesting they post a good question and answer it, sharing their knowledge with the community.

However with id-requests you can't do that. As per what Krazer has said in the question and in JNat's comment, how one person might recall a title could be very different from how anyone else might do it. Self-answering an id-request doesn't really add any valuable knowledge when compared to this self answered question.

Even when the user doesn't abandon the question, when they ended up finding the titles they wanted identified they end up self-answering with a one-liner, which is one of the problems.

---

^{1: can't remember if it was 150 characters or not, but you get the idea}

Answer 4

I'm on the fence about this but because no one has posted a contrary opinion yet, I'll offer this up. I'm uncertain about this because as I see it, there's one good reason why identification questions should be a banned, and few admittedly not as good reasons why they should be allowed. Given the balance between the two, I'm of the opinion we should err on the side of caution and so continue to allow identification requests.

One good reason to get rid of them

Identification requests have one fundamental problem, and it's a pretty damning one. It unlikely that when answered they'll be all that helpful to anyone other than person asking the question. That's not what these Stack Exchange sites are supposed to be about. If questions only helped the asker then it wouldn't be worth the expense of running a site like this. The value in these sites comes from the fact that one question can end up helping a lot of people.

Now, I'm not saying questions should judged based on how many people they might help. There is a good reason why the old "too localized" close reason was removed, there just isn't a good way to set objective criteria for judging whether individual posts could help "enough" people. This is different though, this is an entire class of questions that don't seem to add much value to the site. That's a reasonable justification for not allowing them.

A few reasons to keep them

The one saving grace here is that while identification requests are unlikely to be helpful to someone else trying to identify something, they still are of some value to people browsing the site. I read them to see if they mention anything interesting, some show that I never heard of and may like. It's not big thing but they do seem to gather views at rate similar to other kinds of posts, so they must have some value beyond helping the person asking the question.

What I think is the most compelling reason to keep them is that we seem to do a relatively good job of answering them. There doesn't seem to be another place that does a better job. It's not a question of the site lacking the expertise. There's also nothing that makes these questions fundamentally unanswerable or outside the scope of the site. This isn't a case where these questions don't fit the Q/A format. They aren't ones we can say are better suited to a more traditional discussion forum.

While this doesn't matter much as it used to, getting rid of identification requests would take a big chuck out of the activity on this site. The good thing is that since this site is no longer in beta, there's no fear of closure as a result (plus they changed graduation criteria so it's no longer activity based anyways). Still, it's not obvious this would help improve the overall health of the site. It's easy to say that people asking identification requests almost never post again and aren't contributing to the site, so banning them won't hurt, but that is actually hard to measure. While it's true not many accounts post again, it's hard to say how many people are posting using new accounts, having forgotten or lost the old one.

The lowered activity can have snowball affect as well. People who see a less active site are not as likely to become active. If you're someone like me, who isn't interested in shounen fighting anime, then this site effectively becomes a lot less active. I've probably read more identification questions here than not. Active users may end up becoming less active.

So I think there's actually some risk to banning these questions. There's lots of hard to predict ways their absence could negatively affect activity, like less word of mouth, lower search rankings, and the site not showing up in the "Hot Network Questions" list as much (though this site's id requests don't seem to show up there as much as other sites).

Finally, another reason to keep them is the fact that things have actually gotten better. The site now has better criteria for closing bad identification questions. The growth in these questions has plateaued, and no longer threaten to take over the site. So, just when they've gotten under control they should now be eliminated completely?

Answer 5

Just my 2 Cents.

1 - This was a ridiculous small time frame. For a matter of this magnitude you should have allowed at least six months discussion

2 - ID requests are basically all I'm interested in answering on this site, since the rest of the questions seem of even lower quality, like questions about some boring stuff from Naruto, Bleach, One Piece or whatever kids watch these days.