This is not "not an answer" by the strictest sense of the flag. It does attempt to answer the question, although it does so in a joking fashion. "Not an answer" flags should not be used to report incorrect or poor answers, that's what votes are for.

However, jokes are a different case and as far as I'm aware we've never had a community consensus on what to do about this. StackOverflow has a storied history with "funny" answers, this one in particular. How should we handle them?

In this case I accepted the flag but took no action other than downvoting the actual answer (sorry, Chimeara, I'm a grump).

In general, should we consider joking answers valid answers, or should we delete them in the future? Or what?

For reference, some related topics on MSE:

• I would say yes. What if someone's answer says to use Java as a solution. It might be removed as a Joke. – Lysol Jan 14 '16 at 19:42
• @AidanMueller to be really meta (on Meta, no less) you should post that as an answer. ;) – DMGregory Jan 16 '16 at 2:52

There's not one type of "joke answer". There are answers that are (attempts at) humour, and there are answers that address the question. Often the two are disjoint but when they overlap, when answers address the question, they shouldn't be flagged as not an answer, regardless of whether they are jokes.

The answer you mentioned falls squarely in the second category. It boils down to "you can prevent cheating by taking this action to retaliate against cheaters". It may be funny, it may be highly flawed (read the comments), but it answers the question. The famous parse HTML using regex answer also falls into this category; it begins with good reasons why regex and HTML are incompatible.

One example of a type of answer that doesn't answer the question is all the variants of "how many <type of person>s does it take to screw in a lightbulb" joke. All of them rely on creatively interpreting the premise of the question, so that the answer provides no practical advice on how members of <type of person> can effectively screw in lightbulbs, and instead provide humorous commentary on <type of person> itself. These are not answers.

So if the answer addresses the question, regardless of whether it's incorrect or a poor attempt at humour, they shouldn't be flagged as not an answer. They should be downvoted.

• Except by the fact that they are always (sometimes highly) upvoted . Besides, I fail to see how the linked answer even remotely tries to, in the most stretched, abstract, rhetorical sense, be minimally useful to that OP or to anyone with the same doubt. It couldn't be clearer that it's only purpose is to be funny trough "addressing the question", not to try to propose any real help. If a post aims on being funny-only, it's not an answer. It's so clear that it's the case that after the movie becomes widely watched, the pseudo-suggestion wouldn't work even hypothetically. – MAnd Jan 17 '16 at 8:08
• If they're highly-voted, that's fine. The reason what congusbongus posted here is a good policy to adopt, I think, is that it moves the subjective decision of what is "enough humor" from one person and distributes the decision to the community through up and down votes, which is a core part of the SE philosophy. All of the criticism you are leveling at the specific answer cited (and we should really be talking more generally here instead of nitpicking particular answers) makes the answer worthy of a down vote. Not a flag, nor a unilateral delete. – user1430 Jan 19 '16 at 1:46
• @JoshPetrie I only mentioned the cited answer above, and because congusbongus did and I was addressing his argument. That said, probably everyone would agree with you that the philosophy of SE is distributing decisions and that is awesome. It should be always the 1st choice. When it's working, it's great and no problem exists. The real question here is if community indeed downvotes joke-only answers or upvotes them for being funny, as others have reminded. I think that's what we should look after both in past cases and in future cases to see whether this newly defined policy works. Let's watch – MAnd Jan 21 '16 at 2:28

## Why jokesshould be accepted?

Being part of the SE community made me realize that we all to the very last one of us share at least one thing in common, that thing called Happiness. Can you really imagine yourself in life without sense of humor? Humor leads to happiness which plays big role of our life, it also prevents mental disorders such as depression, stress and anxiety - which obviously most of us developers suffer on everyday basis. Developers with mental disorders like those mentioned earlier can lead to suicide, murder and most of all, invent the most devilish crime in history with over billion dollars of pain and damage in the last forty years. Ha.

Why we are here? The StackExchange community is really the place to be serious about things. People get stuck, confused and frustrated heading here to shot out their problem hoping someone overseas have the precious solution. People actually come here and expect to get the most accurate answer relevant to the question. Without playing games.

Why it should be implemented? Life is too short. Let's not take away the joy out of our life, if you wish to make that corny joke into your post, DO IT! That smile that pops on your bearded head could comes up on another, but let's not make circus out of it.

TL;DR Allow the humor but not the joking. As long it's not abused, of course.

My thought on this issue, after thinking about it most of the day, is that we shouldn't do anything with them: they're regular answers that are subjected to the regular voting gauntlet and flag handling rules.

This means that:

• we won't delete or censor them (which solves the whole issue of subjectivity regarding humor, and also removes an additional burden on the moderators)
• we will reject "not an answer" flags on them unless they apply for other reasons (that is, the joke in the OP is an answer, a unrelated knock-knock joke is not an answer)
• we expect that down votes will be used to indicate that the answer is not very good or helpful if indeed that is the case (which solves of the issue of when humor crosses the line of utility)

I would use caution in allowing joke answers.

• What if the question OP finds a joke funny enough to accept that joke question? (I'm thinking about if a duplicate question eventually pops up, would we mark the newer duplicate and point to the accepted joke answer?)

Since it's not really possible to have a definite answer to these questions, it will make moderation harder, and may cause undefined behaviour from passing-by users; I'm afraid every question would get jokes as answers, that these answers get up-voted 'cause they're funny' and obfuscate the more serious content, ultimately leading to a joke Q&A site.

I don't mind jokes in the comments, though.

• This one joke answer is the only one I can recall seeing recently. I don't think we have a "drive by joke" answer problem. I'm more interested in how strict and fussy we want to be about this. We're already pretty exceedingly strict about things. – user1430 Jan 14 '16 at 17:35
• Also "what makes a joke" is still relevant; if we're going to say we don't accept jokes, that means somebody has to decide if a thing is a joke and delete it, or edit out just the joke. – user1430 Jan 14 '16 at 17:38
• As to your third bullet point, on MSE, the consensus seems to be that "if your answer was helpful, it was helpful." So that's something to think about as well I guess. – user1430 Jan 14 '16 at 17:49
• @JoshPetrie About that last comment, the accepted answer is the opinion of the OP (and 18 users), but it does not seems to be the community's opinion: the highest voted answer has 35 votes and the answerer felt that joke answers are not welcome. – Vaillancourt Jan 15 '16 at 2:13
• @JoshPetrie Take a look at this hilarious answer: gamedev.stackexchange.com/a/112293/72423 Even if it was indeed great fun to read, it does not address the OP question at all. – MAnd Jan 16 '16 at 1:28
• @MAnd Yes, I've seen it, but I'm not sure what your point is as far as bringing it up? – user1430 Jan 16 '16 at 1:38
• @JoshPetrie I brought it up in reply to you saying "This one joke answer is the only one I can recall seeing recently". – MAnd Jan 17 '16 at 7:52

My vote (if there was such a thing) is for not allowing joke answers. They muddy the waters, and make finding the real answers more work. Stack Exchange is a Q&A site, not a forum.

• They should name a subway sandwich after you... because you are full of baloney! – jgallant Jan 14 '16 at 19:38

I think that good humour should always be allowed. Some puns and jokes kill no one, it can even make your day. But, joke only answers, that helps in nothing the question, should meet the banhammer. The first example you gave even got locked because "it is not considered a good, on-topic question for this site". I think it's good, helps avoid newbies earning rep only trough jokes and bloating the site with unuseful answers.

• "that's actually a few years ago, I don't think it's valid anymore." It's valid until there's a newer discussion saying otherwise, which I could not find. – user1430 Jan 14 '16 at 20:04

I think we should start by differentiating the following. An answer that tries to address a question but has an incidental joke within it is very different from an answer that is only a joke - for the only sake of being funny.

The first case should be allowed because of the obvious reason that the primary function of an answer is trying to answer, no matter if it uses an academic, an informal or a funny tone. However, the second case should not be allowed simply because, well, if it's not even trying to answer, then it's not an answer (an example of a fairly hilarious answer like that is the following: How can I create borders in an open-world game that don't feel artificial?).

## It's subjective. But just as always

Of course, in practice the boundary that separates answers with jokes from joke-answers can become less clear. However, it is just as subjective as the similar decisions that are enacted when questions are flagged as "too broad" or when someone thinks a given answer looks more like a comment and then flags it as "it's not an answer". When it happens, a decision has to be made on whether in fact the answer was just a comment (i.e. related to the original question, but not trying to answer it). I don't see why that is so different from the case where jokes were what motivated the flagging.

Sure, one person deciding that is always a hard call. There is where the voting system for deleting answers would theoretically work the best, since multiple people summing up their interpretations about a given answer being or not an answer (due to whatever the reason) is always the ideal due to the highly subjective nature of the decision. However, the problem is two-folded.

## In practice, moderators are the only barrier against joke-answers

As @JoshPetrie has brought to my attention in the comments (I admitedly forgot that), only answers scored -1 or less can be voted for deletion. Considering that funny answers tend to be often upvoted, that alone puts them more often than not, outside the helm of possibly being voted for deletion.

If that barrier was not enough, we have to also recall that only users with at least 20K can vote for deletion. It means that in the rare event of a joke answer to have negative scores for long enough to be voted for deletion, even then, considering that gamedev.stackexchange currently has only 10 members with such rep (two of them are the moderators) and not all are necessarily active in reviewing tasks, that additional barrier to delete joke answers makes it virtually impossible that any joke answers ever get deleted.

This would open up a risky possibility. Therefore, it seems that in the vast majority of the cases, the moderators themselves are the only option for not letting a joke answer endure. In such a case, I sustain that yes, the moderators should have to make such decision when joke answers get flagged.

## TL;DR

And to sum up, as I argued above, the minimum criteria for it should be the same used in all other cases of "not an answer" flagging. I.e. deciding whether the answer really tries to usefully answer the original question, despite of the joke it carries.

• As far as I can tell, you are effectively arguing the position of "do nothing, let votes decide," right? Because if you're arguing for the position of "disallow" based on some criteria, I'd point out that I'd keep that open-world answer around because I do think it answers the question just fine, even if it is over-the-top. Which is why I think more-and-more than putting any kind of policy in place that requires moderators to think about deleting a question isn't worth the effort. – user1430 Jan 16 '16 at 1:59
• @JoshPetrie Well, that's just a regular case of "is it even an answer" decision - just motivated by a rarer event. As in any other case, if moderators happen to understand that an answer clearly is not an answer, (for whatever reason) it's plain normal and fine that they take action upon it. If they feel the contrary or are not certain, give the answer the benefit of the doubt and let votes decide. – MAnd Jan 16 '16 at 2:10
• @JoshPetrie For the records, as we are discussing precisely how to behave in each case, I would vote to close/delete the linked open-world answer in a second. There is no effort at all there to provide an useful, fine answer. It's more like making a joke to show the OP that his question could have multiple answers - what, if that was the case, should have been a comment to the question and/or a flagging. – MAnd Jan 16 '16 at 2:13
• FWIW, you can't vote to "close" an answer, only to delete (and only when it's at -1 or lower). – user1430 Jan 16 '16 at 3:17
• @JoshPetrie Yap, I knew it's a delete-only, I just got caught in the habit of using the word "close". But indeed I hadn't realized that deleting answers could only be voted for when they have a negative score and, as it seems, only by users with at least 20k rep. In that case, considering that funny answers usually get upvoted (so they are frequently going to be non-negative) and considering that the gamedev.stackexchange has only 10 members with such a rep (2 of which are the moderators), I think it changes things. I will edit my answer here accordingly – MAnd Jan 16 '16 at 3:33

I post this as a second answer because it's not in the same track as my other one (and the other one had already a vote).

After taking a look around, and thinking a bit more about it:

Moderating joke answers by removing them when they occur seems to be a hard and ultimately not really important thing to do.

From the links you posted and some more research, the consensus seems to be that joke answers are bad (mkaayyy?), but the common reaction to it is to not care much: the (down)votes will talk. The community seems to be aware enough that joke answers are not that good for the site. However humour in real answers or joke in comments are welcome.

To be honest, I don't think it's an issue (yet). As you said in your OP, you did not deleted because it was a legitimate answer to the question, a legitimate solution to the issue at hand. It might not have been a good one, but it did not deserve a flag. The results: it's way down in the answers list.

It still worries me that these joke answers could be accepted because in these case, there is nothing much to do about it, but then we'll have to rely on the readers' common sense to realize this and keep on reading to the second answer, which should get more upvotes. (This is the same situation than when an accepted answer suddenly becomes irrelevant because of a fix in an API or something.)

## In conclusion

From a moderation point of view, we should consider joke answers as valid: there is no need (yet) to handle joke answers in any particular way.