Is it feasible to keep "identification-request" questions going as they are now?

Question

Let's start fresh

Let us take a moment and start a fresh conversation about the value of identification-request questions to our community. Let's start with a fresh, clean slate with identification-request questions. While they are popular... do they really contribute to the site and community?

Before we make a decision, let's take a look at some data points from the SE Data Explorer...

Let's take a look at the site activity and vote graph.

What is it that we see here? Clearly we have a lot of unanswered questions (zombies).

Poor zombies... they're being neglected and could use some slaying.

What's causing these zombies?

Let's take a look at the total unanswered question history by tags.

What do we see? Why our old friend, identification-request!

The amount of question does seem troubling, but what about the quality of those questions?

Let's take a look at the data for question and answer score by tags. (Sort them the amount of questions, descending.)

Even compared to the most popular tag, naruto, the average score for identification-request is disappointing.

From this data, we see that our community tends to have a polarization towards specific series tags, much like programming languages on Stack Overflow. This along with our various meta tags like culture and anime-production are valuable to our community for the content they provide.

But what about identification-request? Do these questions contribute anything valuable and worthwhile to our community? Do they attract so-called "help vampires" that plague much of the Stack Exchange network of sites (especially Stack Overflow)?

I personally believe that the current state of identification questions diminishes the value of the site to the average user. If things continue at this pace, our site might be flooded with mostly unanswered or low-quality identification-request questions.

identification-request questions are typically easy to ask, topically neutral and initiated by new users. But is this a good thing or a bad thing?

Robert Cartaino brings up a convincing point about these types of questions in Movies and TV SE meta.

In a not-too-distant future, most of those questions will be asked by hit-and-run users who will never return to this site. And you'll get bored having to tease out a decent question and provide answers to a post that will not add one lick of value or interest to this site.

If you go through unanswered identification-request questions, you'll notice a good portion of them to be don't seem to have much response from the OP as the age of the question increases.

Have identification-request questions helped us attract new returning users? Are there many good identification request questions that got you interested in this site or a particular series?

I'm interested in hearing your thoughts on the topic of the value of identification-request to our community. Give us good examples, bad examples, or those in between. This is an important topic of discussion and should warrant more attention. If possible spread the word on current and new identification-request questions, to attract more attention to this discussion.

Now what do you think?

I have given you my opinions, so please take some time to consider the topic and give your thoughts. You're free to agree, disagree, or be undecided. You're an important part of this community, regardless of how much or how little reputation you have. Your opinion matters because it lets us know how you feel and what kind of experience you would like.

I'm not necessarily saying all identification-request questions are bad, nor that we should get rid of all of them. I'm suggesting that we should perhaps be more stringent on the quality of these questions. The human memory can be very unreliable (witness testimonies are a good example of this), especially when a fair amount of time has passed. Your memory might be affected by current trends, media, or even your day to day, if you're going by just memory alone, unless you never forget anything, you'll probably be missing bits and pieces.

For an example, if I'm trying to remember a certain ecchi/dirty joke anime series about an office lady and her talking hamster, Ebichu. I might confuse the characteristics of Pikachu and Ebichu, and because they share similar characteristics, being mousy and cute animal creature. I might mistake Ebichu as being yellow or have stripes, because I watch the anime regularly. If I give such an erroneous description to someone trying to help me the mismatch actually makes it hard for that person to find the series since they're going to be looking for a yellow hamster.

This discussion will be open awhile to garner more attention feedback, so take your time. If you feel that you have a better solution, feel free to post your suggestion as an answer.

As a side note, the discussion primarily focuses on identification-request as a standalone tag. Whether or not we should be tagging questions about identifying a specific reference in a specified tagged series (e.g, who is being parodied in X series in Z screenshot) should be left for another discussion.

tl;dr: identification-request questions are not of good quality (by nature), piling up, they don't seem to actively contribute to user retention. A complete ban on such types of questions seem counter-intuitive as there have been some good questions, but the bad ones seem to be spoiling the rest of the bunch. If you're indifferent about these types of questions, it's probably because you don't feel that they actively contribute anything, unlike the series tags or meta tags like anime-production.

Answer 1

There are currently 440 questions asked with this tag, with 57 unanswered. Naruto, as you have mentioned, has 407 questions, but only 8 are unanswered.

As you mentioned, we are polarised towards series specific tags and Naruto also has a large fan base in A&M so I feel this is a slightly unfair choice.

So taking a different question tag, like anime-production (132/11) or Tropes (92/5) which are also non-series specific - we can see percentagewise that we are not far off the ~11% unanswered that identification requests has.

On Area51, it says that a healthy beta is 90% answered rate, so ideally that's what I'd like to aim for with these, and its not too far off it

There are a multitude of issues with these questions

• Generally not as interesting to answerers. No identification requests normally exceed any significant upvote count. I don't think this will change regardless of rules enforced

• Identification requests have a much larger scope. Naruto consists of several media types, but identification requests could be targeting an individual scene in one out of every anime ever. This means this questions can be very hard to answer, and I think this is a large contributing factor to the unanswered stat

• It probably does attract zombie users, but as long as they're here I don't think there's much to do to mitigate that

I think the rules defined earlier in the year are on the right track and don't need immediate rework, maybe we need to be more strict with closing questions that don't fit the requirements though

I personally don't like the questions, but they serve a useful purpose and I still think we should keep them, as long as it does not overtake the site.

TL;DR: I think the rules we have in place are adequate, we have closures for under defined questions, perhaps maybe we should be more stringent if it is getting out of hand

Answer 2

Until these kinds of questions (and answers to them) are strictly prohibited on SE, I will answer them if I know the answer and if I feel like typing it and finding some references, even though I know that the poster might not ever come back to the site.

I like helping people when it's as easy as googling a link to a credible source about something I know and just typing a short description, a 3-minute job. Gratitude in response to a helpful answer is a reward enough for me. Of course, I might not feel like doing this at times, but it's only natural.

Answer 3

I actually have a few ideas on how to alleviate this problem, but I think to see most of them implemented would require effort from the part of Stack Exchange. Before I propose these ideas, I will cut to the chase and offer my opinion: it is feasible to keep identification request questions going as they are now, at least until this site leaves Beta. Granted, we need to be more stringent and add more QC guidelines and be more clear of what they are.

Option 1: Have a separate forum for identification requests

This could be a chat room or another Stack Exchange site reserved for identification requests. If such an SE site is formed, we could even merge Anime & Manga identification requests with those from Arqade and Movies & TVs, and just migrate all questions having only the identification-request tag to that site.

The current threshold to participate in the site's chat rooms is 20 reputations, which is not asking too much from the new users. We can even make this easier for them by rewarding reputations to users for completing the Tour. Now about the Tour: I really dislike it in its current unhelpful form. The site Tour should be more unique and tailored to Anime & Manga, and provide more incentives (e.g. privileges) for new users to complete them. More about this in the next section.

Option 2: Require new users to complete the Tour before they are able to ask identification request questions

When a user asks their first question, after they click on the Ask Question link, they should be brought to a page asking What is your question about? and a list of the 3–5 most popular tag names on the left and tag wikis on the right. The first in that list should be identification-request. If they check that, they will be brought to the updated Tour page, in which they will learn about some useful tools for identification request questions such as Google Reverse Image Search for images and Shazam for music. Then give them some tests similar in nature to the review tests (audits) that could test them some identification questions that could be answered using Google Reverse Image Search to check that they know their tools. If they fail too many times, they will be banned.

Option 2.5: Further require some reputations to ask identification request questions (does not apply to trusted users)

This is if we adopt Option 2 and want to be more strict about the qualities of identification request questions. We can also add one more restriction that they can not ask a second identification request question unless they accept or leave a comment on their first one, given that it has an answer with more than 2 upvotes. This is going a bit extreme, so it should be reserved for the time when the number of low quality identification request questions skyrocketed.

Option 3: Downvote and ignore more the bad questions

Unlocked, unanswered questions that have negative scores will be automatically deleted by the system after 30 days, and questions with -4 score or lower won't show up on the front page. Plus, downvoting a question doesn't cost one reputations, unlike downvoting an answer which does. Voting is easier and less time consuming than flagging and also saves the reviewers and moderators some time from unnecessary hassle.

Answer 4

In my opinion the main reason so many ID requests are unanswered or with no answers is due to the fact that users who ask these types of questions are 'no shows' and often don't return back to the site to accept the answer and simply because this sites majority only like/have knowledge in areas such as Naruto, Fairy Tail and other popular tags and anime series or image only identification requests because of Google's, yes, annoying Reverse Image Search.

To me reverse image search is likely to get a new first user a -1 or a -2 on an image id request due to how easy it is to find what anime it is from just a 'right click on picture' in Google Chrome or a 'upload picture by URL' on iexplorer, Google. Personally I tend to upvote. There's nothing worse than just getting a -1 straight away. Also image identification, whether they are easy to find or not hardly get upvoted anymore. The most you might get is if it is a moe picture and that will be +5 max. I was surprised to see how many ID requests (image only) and other identification requests got +11 and even more and the answers got ridiculous + 15 - +20 in some cases last year and even nearing the end of the year last year. (might be exaggerating a bit, but I'm sure I have seen some).

I do know that there was Google Image Search then and these answers mostly all used Google Image Search too.

Generally I thought Stackexchange encouraged more questions. I have to say that being more stricter on identification requests is a bit silly. It sounds like you want to force and strain visitors/question askers into writing 60 paragraphs of description. Sometimes even 3-4 lines is enough for some person who knows it to answer it if you include the Genre, a specific scene and some other details and it kinda reduces people from just looking away to the next question and losing interest due to how long it is.

Finally the last thing I would say is that if we want more answers to the unanswered id requests then we need a wider and bigger Anime & Manga community with people who watch and read different types rather than Naruto and other animes/mangas that fall under the category Shounen. That way, if someone asks a id request for a yaoi manga for example, we have some other users who can answer them on the site and if we have more people, with a variety who watch different things then there is more of a chance of these getting answers and possibly good ones too. To be honest, I didn't really notice this. To me it doesn't seem to be spoiling anything, its just that there aren't any answers to them....yet....otherwise I think I might just stay undecided for now.