Does this answer about towns protecting themselves against dragons really need to be deleted?

Question

A new user recently left this answer to How did towns mundanely protect themselves against Dragons or bring them to the ground?. The answer is currently deleted so is only visible to 10k users. However I would like to discuss what led to its deletion and whether it should actually remain deleted.

Timeline

The sequence of events till now is:

1. A new user posts their answer to the question.
2. A few clarifying comments are left pointing out issues or oversights in the post but none worthy of deletion.
3. Post is deleted by community review.
4. User reposts answer as they don't know how else to handle it. (Note: new user)
5. Repost gets flagged for mod attention.
6. I delete the repost and undelete the original while trying to seek clarification on why it was deleted in the first place.
7. I posted in chat seeking understanding of why the post was deleted.
8. Some further comments were left explaining the issues with the answer.
9. The answer was deleted again. 2/3 deletion voters being the same users that deleted it in the first place.

The Answer

The answer itself is too long to reproduce here entirely. It answers the question with mechanical analysis of a battle between an Adult Red Dragon and a fictional town. Most of the relevant issue is related to a table of numbers that appears unsupported. The section of the answer is shown below.

The makeup of a town is highly variable, and there's no standard. But I'll talk about several common types of people in the towns. I'll spare you the math but I ran the odds vs an adult red dragon to see their hit chances with/without hiding, and how many hits it would take for them to kill a dragon.

Keep in mind, the goal is NOT to kill the dragon - if the dragon is taken to bloody (50%) or even 75% hp, it's going to have to seriously think twice about attacking the town.

I've also multiplied each "person count" by 10, because a town will have quite a few people, and I wanted to make sure it was obvious that we aren't talking about 1 knight fighting a dragon here.

People x10	Weapon	Stealth*	Rounds to bloody**
Commoner	sling	28%	21
Guard	short bow	36%	8
Thug	heavy crossbow	28%	6
Archer	longbow	58%	1
Knight	heavy crossbow	28%	6
Veteran	heavy crossbow	36%	5

*Stealth is the best odds to hide, assuming they can get advantage on stealth and impose disadvantage on the dragon's passive perception (e.g. by being both far away and hiding behind an object). Also remember to think about when the DM would rule that they simply cannot be detected by the dragon - e.g. breaking line of sight plus the sounds of battle are too much for the dragon to use hearing, plus they are too far for blindsight.

** The number of rounds it takes 10 people to reduce the dragon to 50% hp, if they can hide a proportion of rounds based on the Stealth column. The people need to be in range for this total number of rounds.

Discussion Points

I do not believe this answer should be deleted. Yes, it needs to support how it generated the numbers in the table but that advise should be left as a comment not simply deleting the answer. The question asks how a town could defend itself against a dragon, this answer attempts to run the numbers to show that the battle isn't quite as one sided as it appears.

I fail to see how this qualifies for any of our deletion reasons. From the deleted answer help article:

Answers that do not fundamentally answer the question may be removed. This includes answers that are:

• commentary on the question or other answers
• asking another, different question
• “thanks!” or “me too!” responses
• exact duplicates of other answers
• barely more than a link to an external site
• not even a partial answer to the actual question

The answer clearly meets none of the condition above. Furthermore as a post by a new users we should be giving the benefit of the doubt and trying to workshop it into a better answer rather than deleting it and potentially driving a new user away.

Does the community believe this answer deserves to be deleted? If so on what grounds?

I would like to undelete the answer again and provide guidance for improvement; however, I understand that my opinion does not out-weigh that of the community. Therefore, I won't take any further actions until this meta has been resolved.

---

Update

The answer has since been un-deleted and significantly improved via edits. However this meta remains as a valid discussion of the events that lead to its deletion and whether that was the correct action to take.

Answer 1

Deletion is entirely inappropriate

Deleting answers has historically been, and should continue to be, reserved for truly egregious cases. When it says “Not an Answer,” that is very literal—it is for cases where the Answer in question is literally not an answer at all, to anything, but instead is another question, or some kind of commentary, or otherwise makes zero attempt to answer anything at all.

Misunderstanding the precise details of the question and answering based on that misunderstanding? That’s still an Answer. Maybe it’s not a good one, but the tool we have for dealing with answers that aren’t good is the downvote—not deletion. And for that matter, “answers that don’t respond with precisely the information requested” are an established and expected thing around here—frame challenges are valid and allowed, and often times are considered good answers. If you don’t think this particular case is, fine—but that’s a matter of your vote, not moderation.

Moderation tools are intended to be used only in exceptional cases. They are for glaring, obvious problems that need to go away. Otherwise the intended functioning of the site is that the community consensus—as determined through voting—determines something’s fate. Doesn’t answer the exact question as you see it? Downvote. Isn’t a valid frame challenge in your opinion? Downvote. Doesn’t convince you of its accuracy? Downvote. Isn’t supported the way you would like to see it supported? Downvote. Doesn’t hold to some policy? Unless that policy is Be Nice or rules against spam or piracy, downvote. These are not concerns for which the moderation tools of this site are suited, and the overuse (if not outright abuse) of those tools has been an ongoing problem that needs to stop.

There is a wider community of users on this site than can be found on Chat, or on Meta. A consensus there, or here, doesn’t really mean very much at all—not compared to a consensus in main site voting. Therefore, unless there is a very good reason to do otherwise, we should in all circumstances favor solutions that allow main site voting to handle problems. This is something trivially handled by downvotes, and therefore it should be. To turn it into a matter of moderation-tool override is a significant (un-agreed-upon) divergence from historical practice, and in any event certainly not a good idea.

Answer 2

A perfect example of why we shouldn't pin the highest voted answer to the top

Hold still folks, because I've got opinions and have been waiting all week to have enough time to unload them!

For this specific instance, we have a querent who is asking a question about what happens in the Forgotten Realms (FR) for a town that gets attacked by a dragon. The querent specifically focuses upon how these towns might go about bringing a dragon to ground.

From reviewing the comments here and elsewhere, I've further understood that the querent really wants an answer that's focused upon FR specific elements. To be clear, I didn't get that impression from reading the question, and tags alone are not sufficient in my opinion (sometimes folks just like to throw tags around).

The most highly voted answer (71-0 at the time of writing) does not answer the question with the requested information, rather it is a bit of frame challenge answer. But with a score of 71 (at the time of this response), it's definitely what the RPG.SE community deems the best.

The second highest voted answer (14-0 at the time of writing) provides some very specific 5e lore from a specific adventure module set in the FR, which details what happens when towns try to fight back against a dragon. This answer helps, but it doesn't actually answer the querent's question of how does a town ground a dragon.

The third highest voted answer (8-1 at the time of writing) provides some info from the FR wiki as well as 5e specific information. Additionally, the answer provides some details from the DMG to attempt to quantify the problem but even goes on to discuss a potential means by which to ground the dragon by using the low-level entangle spell. Curiously, of all the answers provided thus far, this is the first to actually directly answer the querent's request for a grounding method.

Finally, the fourth highest voted answer and the subject of this thread's contention (9-6 at the time of writing) provides a detailed breakdown of the overarching 5e mechanics associated with the issue at hand to put forth an answer which posits that the effort of grounding a dragon may be more trivial for the game's definition of a Town than previously thought.

---

What I find fascinating about this whole thing is that every answer provided, in some way or another, does NOT fully answer the question being asked. And for the record, I think this is fine, because I think that every answer is correct for a certain type of playstyle. Furthermore, I think that every respondent may've seen the question a little bit differently and that influenced the nature of their answer.

Sometimes a querent doesn't know enough to ask the exact right question and as a result, they are relying upon the opinion of experts to help guide them. If they already knew the information, they either wouldn't be asking or would maybe be doing a self-answer.

Therefore, the 'correct' answer to this question is going to be whatever helps the querent most. But the sum of all answers provides valuable insight to various different ways that people may read the question. Given we're not dealing with immutable things like the gravitational constant of the universe, this is a very valid spread of answers and none should be deleted.

Answer 3

Thoughts quoted from Someone_Evil on a different post

I am going to quote the entirety of the following answer, even if it doesn't all strictly apply here:

• https://rpg.meta.stackexchange.com/a/11159

It is site practice to delete unsupported answers

Admittedly it's never really documented/discussed on it's own, but mostly in the discussions around fun topics like game recommendations and designer intent requests. Specifically, when the questions type really needs supported answers, aren't getting them, and we need to talk about how to moderate them.

But it does also show up in other guidance: What are the citation expectations of answers on RPG Stack Exchange?

We don't want your opinion; we want your expertise. If you do not have experiences you can bring to the table in that particular case, do not answer the question. Answers not doing so may be downvoted and/or deleted.

(It talks about experiences as citations because it's from the section on subjective questions, however I see no reason to distinguish the two on this other than what type of support the answer needs to bring.)

But haven't we got a post notice for this?

Yes, there's a post notice for needing citations, but post notices are a poor tool for moderating content. They are fixed wordings and only available to a very small subset of users. Comments are better at pointing out problems because they are customizable (by default you write out each one) and availible to many users. The post notice does also not exclude deletion, but explicitly warns about it:

Want to improve this post? Provide detailed answers to this question, including citations and an explanation of why your answer is correct. Answers without enough detail may be edited or deleted.

Let me try to explain why that post notice get used at all, if those answers should get deleted. Firstly, deletion might not be the correct thing. If the answer is probably salvageable, users should get a chance to edit it, or comment for the poster to do so. Also, diamond mod deletions are instant, and harder to revert (can't be undeleted by non-diamonds). There's also an important learning and healthiness aspect to community doing moderation and evaluations, so we often try to not immediately delete (nor close, etc.) borderline cases.

So, we prefer to let the low quality review process do its thing. We might flag the post (if it wasn't already), add a comment with out concern (again, if it's not already covered), and add the post notice for increased visibility of it (if we remember to/think it's worthwhile). Expecting the post notice to be used beyond this is putting an aspect of site moderation too much onto the diamond mods, I hope you want us to put our efforts into more useful things.

So for the specific answer

Without support being added to it, I don't see it having any value. And it being deleted does not prevent the poster (nor 10ks I believe) from editing it. The user could then vote to undelete and/or flag for undeletion (if diamond deleted, as this one retroactively is, the flag is necessary). No such efforts have been made on this answer. I don't personally see the reason to undelete it, but I'd be happy for someone to change my view on the quality of an answer at a score of -6.

---

My own thoughts

I have a distinct memory of the day I saw the answer I quoted above, because I also wrote an answer, one that I have now undeleted as of today:

• https://rpg.meta.stackexchange.com/a/11157

It draws on SE guidance and I felt confident in the idea that bad answers should be downvoted and not deleted. Then I received comments such as:

The problem wasn't the answer being wrong, it was lack of support. And those can be deleted as per the notice.

And that made sense to me, especially given what the post notice actually says. Someone_Evil even wrote that entire answer about how answers lacking citation/support are deleted and so I rethought my stance. I looked into it and found that Skeptics does this same thing, as do other sites, and some sites even have tags that indicate answers must have actual citations/sources.

It is clear that people do not agree whether unsupported answers should actually be deleted or not, just look at the answers here, but currently, and for the past while, I have thought that they should be deleted. The answer, as it was originally posted, did not have support; it grabbed at numbers without backing up where any of those numbers were coming from and I voted to delete it.

---

That all said, if we want a new Meta on what we want to do with unsupported answers, I welcome it. To that end, have some links to questions that I could find that have discussion about unsupported answers:

• Two Answers - One Deleted - What's the difference?
• What are the citation expectations of answers on RPG Stack Exchange?
• When do we close a question for drawing opinion-based answers and how can we try to get ahead of such closures?
• Why was this answer to the question "Can a player choose to fail a spell's ability check contest?" deleted?
• Why was my answer to a question about the naming of oozes deemed "not an answer"?

Answer 4

Yes, it doesn't answer the question - not even as a frame challenge

The answer itself is a good attempt at a theoretical question regarding how a populace could defend itself in an actual battle - but that isn't the question asked.

N.B. I was the editor who worked with OP to update their question.

While the question originally was targeted more towards an actual in-game problem, after I submitted an answer it became clearer that a practical solution wasn't what OP was looking for.

After discussing in chat, and a bit in comments, we settled on that OP was most interested in knowing how towns responded to dragons in lore. This could provide them a basis for a better understanding of how the historical lore could apply to their current situation.

I may have been incorrect in my understanding of their goals, but that was what I had in my mind when I edited the question to focus on that aspect.

Unfortunately, the answer in question here doesn't address that need at all. A frame challenge in this instance isn't appropriate, because OP is looking for actual (fictional) events that exist in lore and theoretical mechanical situations.

The querent was also very specific in wanting information on ways to ground the dragon. Unfortunately, the theorycraft answer doesn't address that need at all.

Because the answer doesn't address any of the stated needs in the question, it really isn't appropriate to keep it up as an answer. It simply isn't an answer to the question asked because it doesn't address the issues in the question and this follows the general guidance on the use of the Not An Answer flag.

Answers that do not address the question asked, or are otherwise off topic of the issues presented in the question.

From further down the page, one of the Not an Answer options is:

answers that attempt to answer another question

And that's exactly what this is. It's not answering the question of how have towns responded to and/or grounded dragons in FR lore, it's answering how could a town respond from a purely attack perspective. It's not a bad start to an answer, it's just an answer to a different question.

Answer 5

To me, the answer fell into a pit that we usually close the question for: Idea generation.

It might have tried to sail the good subjective bad subjective cliff if it had made proper references, but it did not even back up the basic ideas it prompted, making it pretty much an answer without any substance that proposed something that didn't even match up on a basic analysis of the situation. Nothing on the quoted table was backed up by even pointing to official stat blocks, making it impossible to even check the claims for verity!

As a result, it read like a random "I just made this up on the fly" idea-answer, which are not only discouraged, but also have been deleted in the past.

---

After the more recent edit is does back up its numbers, but the math given isn't good. Even with a cursory knowledge of the system, the numbers don't add up: The dragon doesn't get disadvantage on the perception rolls because the town is already burning (which is well lit conditions) and hiding takes up the action, thus the "28% chance for a commoner to hide" as well as the number of calculated turns are wrong.

Answer 6

Undelete, add the post notice, and delete later if the post is not improved.

Want to improve this post? Add citations from reputable sources by editing the post. Posts with unsourced content may be edited or deleted.

You left out a reason for deletion that we use on the stack. Unsupported answers may be deleted. But you are right, we were hasty in deletion, and we should have used the post notice first.

The answer is wholly unsupported, so the post notice is warranted.

The question asks:

What methods are expected to be feasible in the combination of system (as of fifth edition) and setting (northern regions of Forgotten Realms), that towns employed to protect themselves against dragons?

The question asks explicitly for answers to be situated in two important contexts:

• 5th Edition
• The northern regions of the Forgotten Realms setting

Thus, there is a reasonable expectation that answers be supported through reference to 5th Edition material in the Forgotten Realms setting. This is what the question asks for unambiguously. Does the answer do this?

No.

The answer fails to provide mechanical support from 5th Edition rules for the claims made, and the answer makes no attempt to demonstrate that the claims made in the question are feasible in the Forgotten Realms setting.

The answer needs to do two things to be properly supported:

• Provide the details of the calculations made so they can be verified by the reader
• Demonstrate that the assumptions made in the answer are actually relevant to the Forgotten Realms setting

Because that’s what the question is asking for: methods that are “expected to be feasible in the combination of system (as of fifth edition) and setting (northern regions of Forgotten Realms)”. As the answer does not provide either of these things, it does not provide what the question asks for. This is what the post notice is for.

And if the answer is not improved (in some generous time frame), it should be deleted until it is.