In my proposal to repeal the DGtS policy, I wrote:
we can and should craft some guidance to nudge users in the right direction when approaching untagged or under-specified questions.
This is intended to be a community effort, though I have done my best to put out a high quality first draft for us to work with. Once we’ve settled on something we’re all mostly happy with, I’ll remove this notice. That said, we should consider this post to be a living and evolving document. And if you guys think I’m just totally on the wrong track, I’m open to being told “Thomas this is garbage rewrite this.”
The “Don’t Guess the System” policy was repealed under the platform “Let experts make expert decisions”, and what we want to do here is put to paper some of the ways an expert can identify the system being used in an underspecified question, and give some guidance for steps to take when approaching such a question.
Always ask the querent to confirm anyway.
This was our standard practice by way of policy before, but it should continue to be our standard practice, even if we are sure beyond any reasonable doubt of the system being described in the question. Even if the question meets the highest tier of evidence described below, we should still drop a line in the comment section:
I’m assuming this is [system-edition], since you mentioned [this and that], could you confirm that?
This helps to remind the user that they still need to be specific about the game, and when they confirm, there is no longer room for anyone to question our expert judgment. And in rare circumstances, maybe we were wrong.
Tags: to add or not to add, that is the question.
In Pyrotechnical’s proposal, we are given some basic guidance that I want to expand upon:
Do not edit tags into the question unless there are clear indicators on the querent's intent via comments, language in the question (including things like a link to D&D Beyond), stating the system in the title, etc.
It is these “clear indicators on the querent’s intent” that need a bit of exposition. To this end, there are two tiers of “clear” evidence. The first tier will be those things that are generally just as reliable as an explicit statement of the system, and the second will be things that are somewhat less reliable, but still good enough that a system expert should be able to make a confident judgment.
Tier 1: Sourcebooks, quotes from sourcebooks, and system specific tools.
This tier is occupied by things that we had already made policy exceptions for because the community agreed that these things should be just as reliable as an explicit statement. These exceptions were solidified in this Q&A:
In her response to that question, Doppelgreener wrote:
As experts operating with considered judgement, we should be considering that when someone unambiguously indicates their system by indicating system-specific documentation while providing nothing to contradict such an indication, they have indicated their system sufficiently to us.
In my response, I (Thomas Markov) wrote:
Citing materials is a way for someone who doesn’t know there are different editions of the game to tell us what edition they’re using.
Referencing a sourcebook by name is one way this can happen. For example, if a question is prefaced with “In Xanathar’s Guide to Everything...”, we know this is referring to D&D 5e material. However, we must be careful, as sometimes a book title may not uniquely identify a system or edition: Dungeons & Dragons has twelve volumes with “Player’s Handbook” in the title, twelve iterations of “Monster Manual”, and nine Dungeon Master’s Guides.
Mentioning a sourcebook is even more reliable when accompanied with quotes (the longer the better) and page numbers. “Playing D&D, how do critical hits in the PHB work?” doesn’t help me figure out which of the 12 D&D PHBs is being referred to; but if we instead have “Playing D&D, How do the critical hit rules on pg. 196 of the PHB work?”, now we can narrow it down to the 5th Edition PHB, which is the only PHB featuring critical hit rules on page 196. Alternatively, suppose I see a question like this:
How do these rules work?
When you score a critical hit, you get to roll extra dice for the attack's damage against the target. Roll all of the attack's damage dice twice and add them together. Then add any relevant modifiers as normal. To speed up play, you can roll all the damage dice at once.
No system is mentioned, no sourcebook is mentioned, but a system expert would be able to identify the quoted passage as the critical hit rules from page 196 of the 5e PHB. We are looking for these mentions of and citations to source materials that unambiguously identify the game system and edition being played.
This first tier of evidence also has system specific tools. This exception was first established in the wild with this Q&A:
In the case discussed there, D&D Beyond is a D&D 5e-exclusive tool set, so questions that mention D&D Beyond can be safely assumed to be about D&D 5e. And any other tool set that is associated with a single game system should be treated similarly.
Tier 2: Some combination of terminology, concepts, and mechanics that identify the system.
This is where the guidance gets far less concrete, and where we have to rely on our experts to make decisions. What we are looking for here is some combination of terminology, concepts, and mechanics that, when considered together, uniquely identify one system. Unfortunately, there is not really much concrete guidance here. As always, the more details we have, the more sure we can be. Ideally, we would have multiple system specific game terms, such as the name of a feature, spell, or mechanic, but sometimes a system expert will be able to correctly identify the system with a single piece of identifying information.
Of course, keep an eye out for terms and things that don’t fit, which might suggest that the OP is confused or playing some odd combination of systems. This could make otherwise clear information unclear.
To reiterate the first section, always leave a comment briefly explaining the decision you made. As mentioned earlier, we still want to have OP confirm the system, and we need to make our reasoning available to other users. If I determine that a question is about dnd-5e, you shouldn't have to ping me to ask why; you should already see a comment from me saying something like:
I've added the [dnd-5e] tag since you mentioned bonus actions, Tasha's otherworldy disguise, and Wild Magic Sorcerers. Could you confirm this is correct?
[Note from Thomas: This section is obviously quite brief, more brief than I think I would like, but that may just be the nature of the beast. The whole point here is to rely on the expertise of our users, and you cannot really model diverse system expertise in a few brief guidelines. If you have some good ideas for fleshing out this section, please offer them, but I think this might be about the best we can do. And if it is, I think that’s okay, I’m going to trust our experts until I’m given a really good reason not to.]
Without these things, we’d prefer not to add a system tag.
Without any of these elements that uniquely identify a single system in the question, we should not add a system tag. Questions like this generally fall into two categories: (1) the question is so underspecified no one really knows what the question is even about, and (2) it could be one of a small number of game systems. This first type of question isn't much of a concern for the guidance here, it will already be closed in short order for "needs details or clarity", and is usually the type of question that would be closed for that reason even if it did specify a system.
But sometimes we get a question that is written in a way that a system expert can determine that it is consistent with two or more systems. To handle this case, Pyrotechnical continues:
State your assumptions: When guessing the system, it is important to indicate what your assumption is within your answer, so a header stating, "I'm assuming you're referencing 5th edition," would be adequate before proceeding with your answer.
When a question could feasibly concern any of two or more systems, instead of tagging the question, we should instead state the system in the answer. Ideally, this would be a brief explanation that the answer is only applicable to the one particular system. We would like to avoid having a user misunderstand this and think an answer for one system applies to a different system. Just be clear about what game your answer applies to. In addition to this, we can also include a system tag using markdown to go along with this brief explanation: fate-core is had by using the syntax
[tag:fate-core]. So an answer which makes an assumption about the system in question might look like:
pathfinder-2e This answer applies to Pathfinder 2e only.
Main Answer Header
Rest of answer.
Answer underspecified questions at your own risk.
There are two elements of risk to consider when answering an underspecified question. The first deals with making an assumption about the system that is later determined to be incorrect by the question's author. On this point, Pyrotechnical provides further guidance:
I do not anticipate that this will occur very often, but in the event that the assumed system is incorrect, the answer must be deleted regardless of score. This is necessary to prevent confusion for others who may find the question at a later date.
An answer is considered wrong if the querent has edited in the system and it doesn't match the assumed system.
Deletion can occur via any of the following methods:
- Deletion by the respondent is preferred, but not always viable.
- If the score on the answer is negative, deletion can be handled by the community.
- If the score on the answer is positive, deletion must be handled by the moderators. Members of the community can assist by raising a the, "In need of moderator intervention," flag option and citing that the assumed system was not correct.
Overall, this proposal creates a slight increase in the moderator's workload, but I don't get the impression this will be more than a handful of additional tasks for them every year.
If you take a shot at figuring out the system and post an answer, and your system assumption turns out to be incorrect, your answer will be deleted. I know I included a disclaimer that this post would not include any policies, but maybe it should say it does not create any new policies. This is just the natural application of the general Stack Exchange policy that answers which do not answer the question are deleted. If a question turns out to be about D&D 3.5, and your answer assumes it is about D&D 5e, your answer is not an answer to the question and should be deleted.
The second risk element here deals with question closure. We are not going to police closure of underspecified questions. Recall the two categories mentioned previously - questions that could be about anything, and questions that could be about one of some number of systems. The difference between these two things is not always going to be particularly clear. One user may think they have a questions narrowed down to two systems, and one user may think the question should be closed for "needs details or clarity". This second user is free to vote for closure on the question, and the community will work it out as we do with all questions that are put up for closure review. If you choose to answer such a question, you are assuming the risk that the question may end up closed before you finish writing your answer. This is the cost of doing business with underspecified questions.
Avoid edit wars and arguments in comments.
This is guidance that applies all of the time, but it is worth inserting a reminder here. Edit wars are discussed in this Q&A, but the main take away from the answer there is:
If a short comment exchange works, disagreements are resolved, things that needed to be pointed out were, clean up your comments (encourage the other to do too) and move on.† If it doesn't, take it to meta. It doesn't usually matter which of you starts that. We're ok with a meta discussion about whether a specific question should have a specific edit, especially once it's clear some users disagree on it. Remember to focus on the edit, not the users.
†: No, there's no hard length for acceptable comment discussion. Single back and forth should usually be sufficient, and if it isn't and it feels like banging two rocks together, move on.
There is nothing wrong with trying to resolve disagreements in comments, but if things don't seem to be heading toward resolution pretty quickly, simply disengage and start a meta discussion. This is okay, we should be talking things out as a community to work together toward resolutions.