# Revisiting our “never guess the game system” policy

We have a policy that says we never guess the game system someone's playing when we need to know it but it's not clearly stated in the question. Conditions for our site have changed since it was formed (sometime well before 2017) and it's come into question a few times recently by newer members. Now is a good time to take another look at this policy and discuss whether we want to continue with it, modify it, etc.

Current enforcement of this policy is uncompromising almost all the time: don't guess when it's inconclusive even if you feel 110% sure based on multiple signs; let the person clarify because it teaches them how to use our site.

## What had us decide on this policy originally?

From what we saw from the site's start in late 2010 up until ~2015 this policy made a lot of sense. If someone was unclear about their game they might be playing AD&D 2e or D&D 3.5e or Pathfinder or D&D 4e or D&D 5e (still brand new and not widely played). It was hard to know for sure (and we could rarely tell 3.5e or PF apart anyway) and fairly frequently something even weirder was going on like:

• They're playing a D&D 3.5e/Pathfinder hybrid, but they're asking us about a mechanic like Polymorph that works differently in both editions, in which case we'd have follow-up questions.
• They're quoting to us D&D 4e or AD&D material but their group plays D&D 3.5e. They just saw a book with “Dungeons & Dragons” on the cover in the game store's second hand shelf and they're super confused about why it doesn't mesh with their group's other rulebooks or character sheets.
• Their group plays a heavily homebrewed custom blend of AD&D 1e and Pathfinder or whatever. Chances are we'd leave this closed: “we can't help you here, you work it out.”

Our guesses were unreliable, and when we were wrong we were really wrong. If we'd reopened based on our wrong guess and were receiving answers we'd then have a huge mess to clean up and several annoyed users. If we were 110% sure we were still often wrong. Guessing was just a huge headache every time and never worth it.

## Nowadays this is all much less true

We've seen a convergence on D&D 5e in questions the site receives in the past couple of years. If someone's not clear about their game it's usually D&D 5e. If we have any basis to guess they're playing D&D 5e we're usually right. If they're not playing D&D 5e they're already aware of the importance of being clear about the edition they're playing, so they've specified it. These weird hybrid mix-and-match scenarios rarely come up.

Note that these aren't wild guesses being made: they're not “this person mentioned wizards, it's probably D&D 5e”, but instead educated guesses or inferences made based on the content of the question we do have. Taking the most recent example cited just earlier today, our users' guesses were based on an exact quote and page number from the PHB lining up with the D&D 5e PHB—and those inferences were correct.

The main pragmatic reason remaining to enforce this policy is because it requires people to learn how to use the site before they can get answers.

I've seen dissatisfaction with this rule from many members, especially those whose activity was mostly within the past couple of years, and I'm not surprised based on our recent track record.

# What do we want to do from here on out?

So here's the crux of this meta question. We haven't ever seriously re-assessed this policy within the new conditions the site faces, we've just re-affirmed it based on it having already been the policy up until then and working OK.

I'd like us to examine this policy as a community, consider our options and the pros and cons, and work out what we'd like to do from here on. That could be a modification or it could be continuing with the policy exactly as-is, modifying it to be OK with some level of inferences—whatever we choose is valid, but I'd like us to choose it because it's what looks like it should work well for us now in our current circumstances.

Given a fresh look: what do we want the policy here to be? How should we approach the idea of guessing the game system someone's using when it's unclear? If we're OK with some level of inference, what would that be?

• That linked policy’s from 2017, what exactly has changed? – mxyzplk - SE stop being evil Sep 23 '18 at 17:06
• @mxyzplk the linked Q&A from 2017 is an explanation of what was already the way we were doing things from before that. 2017 was just when it got written down that time. – doppelgreener Sep 23 '18 at 17:45
• And upheld by vote then too, but whatever, cut and paste is easy enough. – mxyzplk - SE stop being evil Sep 23 '18 at 18:22
• @mxyzplk I think dopplegreener makes a fairly good argument that the climate has changed enough that maybe the policy is no longer helpful. I mean, I totally disagree that the policy should be changed, but "should we change a policy in light of such and such new scenario" seems to me to be worthy and distinct from yet another duplicate complaining about the policy with nothing new to add. – Bloodcinder Sep 23 '18 at 20:09
• Previous discussions on this matter have also dismissed subtantiated inferences as no better than wild guesses and in fact effectively conflated the two, which used to be true a few years ago given all the wild stuff we dealt with, but is also less true now. – doppelgreener Sep 23 '18 at 21:18
• Bullet-point 2, by the way, is approximately how I played my first decade of D&D: a mixture of B/X, 1e, and 2e materials that would make a "purist"s eyes go cross-wise. (Except we weren't confused: we just bashed on it until it fit.) – nitsua60 Sep 23 '18 at 22:16
• That's a fair way to play, though I remember when that came up the person wasn't mixing and matching consciously and their confusion was based on the idea they expected the mechanics in these two books to mesh properly like all the other books and it wasn't. Usually because it was D&D 4e they'd bought for their not-4e game or vice versa. – doppelgreener Sep 23 '18 at 22:36

I'm a college instructor. I teach hundreds of students across multiple courses a semester in the same major. When I receive an email from a student asking what will be on an upcoming quiz or what they missed in the last lecture or what the assigned textbook is or whatever without stating the course they're asking about, I reply by asking for the course, even if there is enough context for me to hazard a guess or even if I could look them up in my roster to try to figure it out for certain.

Why? Because it's a waste of my time whether I guess wrong or hassle with my rosters, and if I guess wrong I might confuse them with an answer that doesn't apply to them. Because they need to possess enough theory of mind to understand that not everybody knows what they're thinking if they can't express themselves clearly. Because I don't want them to embarrass themselves in the future when they're equally presumptuous or ambiguous with another instructor who isn't as patient.

People coming to RPG.SE to ask questions need to put in the research effort (the determining factor for upvotes and downvotes according to the voting buttons) necessary to express themselves clearly or else be taught how to do so with minimal interruption in the site's process. It's not being pedantic to expect people to know what game they are asking about and tell us clearly if they want our help. Waiting for the querent to do their own work to clarify saves time and effort for all parties.

Guessing results in bantering and rollbacks in extreme cases and a failure to employ a teachable moment in all cases. Guessing does nothing to help train new members of high quality. For those one-off querents who are unlikely to return anyway and maybe don't know much about the ecumenical gaming community they are a part of, it also robs them of a chance to broaden their gaming horizons and to learn a greater appreciation for the community.

The policy should not be changed.

• Agreed on all points. If the system isn't stated, then we should wait for the asker to clarify (and put the question on hold in the meantime if need be). The alternative is risks adding to the confusion when a new user is trying to ask a question here, potentially leading to debates over how much information is enough to guess the system and possibly edit wars in particularly confusing cases. It's much simpler to wait and let the asker clarify it themselves. – V2Blast Sep 24 '18 at 4:42
• So I am going to address a point made in another answer (and I know this is only part of your point): How is is more appropriate to teach a new user to use the site by closing their question than to edit it in and comment on how the site works (and asking to verify if it is correct). This is for the case of a question that 110% sure to be X system. I'm just confused by the notion that the only way to teach proper tagging is to close the question. Can you explain a bit more maybe? – Rubiksmoose Sep 24 '18 at 16:29
• @Rubiksmoose Even if we think it’s 110% certain, we’re guessing at that too. It doesn’t happen often, but it still happens that askers are quoting the wrong game and that’s part of their problem. – SevenSidedDie Sep 24 '18 at 17:00
• @SevenSidedDie No, I get that and I am not disputing the validity of it. I am asking specifically about the argument about educating the user. There seems to be the assumption that closing and making the poster do it themselves is better at teaching than fixing it for them and asking if it correct and telling them how tags work in a comment. I'm trying to ascertain why people think one is superior to the other specifically. See also this comment. – Rubiksmoose Sep 24 '18 at 17:02
• I've been actively involved in (reversing) cases where a system guess was and would have been completely wrong so that is not my issue. I get that. – Rubiksmoose Sep 24 '18 at 17:05
• @Rubiksmoose Oh, I understand. In that case, it has equal teaching benefit, but worse failure modes. Imagine a new user who posts in the morning and checks back after work. Someone edited it to say [dnd-5e] and asked if that was right, no hold placed. It’s open, so people answer. Now the user comes back and responds. “Correct”: same as with asking & holding. “Wrong”: now there are a bunch of answers for the wrong problem. (It helps to always remember that holds stop answers. If having answers posted during a period of time would be bad, we hold.) – SevenSidedDie Sep 24 '18 at 17:22
• What SSD said. It's not that "edit first, then warn" isn't a teachable moment. It's that it's 1) not efficient (rather than everybody putting in minimal effort by immediately closing without argument, somebody has to go through the hassle of conclusively arguing for which system is right and then doing the tag and then also warning the user about the policy and about how they overlooked the policy this time while not confusing them about the policy) and 2) encourages an attitude of "if I do something wrong then they'll fix it for me" rather than "I need to make sure I follow the rules." – Bloodcinder Sep 24 '18 at 17:29
• Or, to revisit my experience as a teacher: when I clarify their problem for them and then tell them that I clarified it, they never improve in the future; when I refuse to help until they clarify their problem, they improve and I rarely have to do it again for them. I save time by following the policy, 100% without doubt in my experience. And I assert that that experience extends to this domain too. – Bloodcinder Sep 24 '18 at 17:30
• It's the same reason that paying your kid's speeding ticket for them isn't very effective. Teachable moment? Sure, but not the most teachable way. – Bloodcinder Sep 24 '18 at 17:31
• As @KorvinStarmast commented on another answer, "...getting people used to the system by doing (rather than someone else doing it for them) is a useful approach for the medium to longer term." I would say it's specifically a better teaching method. Teachable moments aren't binary. There are better and worse ways to teach. This is a better way to teach. – Bloodcinder Sep 24 '18 at 17:34
• @Bloodcinder thanks! well said. That does make your thoughts clearer on the matter and I find it very compelling. – Rubiksmoose Sep 24 '18 at 18:06
• If a student were to come to you and present a very specific instance, "I'm having trouble performing X procedure", and you know that you have only performed X procedure once during one specific course's most recent lab, are you going to tell them that you refuse to help them until they tell you the course name and lab number? What are you teaching them in that "teachable moment", the material that will help them succeed, or that you aren't interested in helping them? How "efficient" is that if they just stop emailing you because they don't think you care? – SeraphsWrath Sep 27 '18 at 17:04
• I don't tell them that I refuse to help. I ask for the necessary information. I've never had a student fail to respond to that inquiry. Once I know all the necessary information, I help them with the problem, and generally they are less inhibited about asking for help in the future because they know what to expect: if they're clear, they get a quick answer. – Bloodcinder Sep 27 '18 at 17:48

We should not guess the system, for all the reasons others have stated. However, we should make it clearer to new askers that mentioning the exact system is going to be a requirement for having their question answered.

Some kind of reminder or notification or prompt should appear on the "Ask A Question" page.

• One specific proposal I am curious to see what happened/is happening with is the one made here which is a step in the right direction I think. – Rubiksmoose Sep 24 '18 at 14:17
• +1 to this idea. We tell new users to take the tour, yet the tour doesn't mention the current policy about explicit system specification. This should be changed, or else we're going to keep running into this problem with newer users. – MikeQ Sep 24 '18 at 14:18
• @MikeQ Yeah this answer to that same meta addresses that and I honestly would like to see that idea get more traction since it seems like a pretty straight-forward change to me. – Rubiksmoose Sep 24 '18 at 14:25
• It's worth noting that Stack Exchange network may be developing a question wizard that RPG.SE might be able to use to walk querents through these sorts of steps. There's no timetable on it, but it's worth keeping an eye on. – Bloodcinder Sep 24 '18 at 15:49

It's not appropriate to edit based on a guess. Make the OP clarify the question themselves.

Why? Because we're also trying to train new users on how to use the site. And worst case, you bait people into bad answers when you guess wrong.

If the OP never bothers to come back to clarify their question - then what good are answers going to do them anyway? Patience can be hard, but it is rewarding.

If someone else guesses at the tag, revert or flag it and explain why.

• A million times this. A guess is still a guess, even if you're 95% certain, and the training of new users issue cannot be overestimated in its importance – Wibbs Sep 23 '18 at 17:44
• Exactly. You can be "100% certain" of something and still be wrong. – Jason_c_o Sep 24 '18 at 16:48
• Training of new users is a desirable outcome of the process. BESW mentioned that in chat during a recent discussion of this question-meta in chat. I don't think that it's emphasized enough that getting people used to the system by doing (rather than someone else doing it for them) is a useful approach for the medium to longer term. +1 – KorvinStarmast Sep 24 '18 at 16:52

I agree with the other answers saying that we shouldn't change our policy on this but have a slightly different slant on why:

Our current policy is very clear, easy for new users to understand (even if they disagree with it) and relatively simple (if not always painless) to enforce.

Any alternative to this policy, however well intentioned (and perhaps both sensible and justifiable in the abstract), will be much less clear, more contentious and harder to enforce.

Presuming that we don't declare guessing at systems to be acceptable in every circumstance (something that no one is advocating) any change to the current policy would mean attempting here on meta to draw a line somewhere as to when guessing was and was not acceptable.

The drawing of this line, however clearly stated it was here in meta, would in practice almost certainly lead to (potentially protracted and unhelpful) disagreements on the main site and require heavier moderator intervention.

If we say that 'we'll only guess a system when we're 100% certain that we're right' there will still be rare occasions when the most experienced users turn out to have been wrong, despite all of the cues. The much bigger issue however, is that, if we legitimise guessing in any form, less experienced users will also be empowered to guess when they are '100% certain' and they may not be as well equipped to make such an assessment.

We're already in a position where new users occasionally guess systems inappropriately - but currently the response 'Sorry, we never guess systems here', while potentially annoying, is pretty clear. Saying instead 'Sorry, we do sometimes guess systems here but you should not have done so in this instance' is going to be much trickier to enforce in a way that people find to be consistent, fair and reasonable (and don't take personally).

A categorical ban on guessing systems is obviously impersonal. Allowing some guesses and not others is always going to feel more personal to some users (whether or not it actually is).

• This is the first one I have been actually able to agree with. A good point well made. – SeriousBri Oct 4 '18 at 10:56

We should allow some leeway when a question is extremely obviously related to a specific system, especially (if not only) when new users are involved.

I agree that the majority of the time it's not appropriate to guess at the system, and we should avoid doing it and request clarification from the querent. However, it is sometimes the case that a question is so obviously based on a specific system that we can be overwhelmingly sure what the system is, but we refuse to connect the dots. It has been the case that a question can literally have quotes from and page references to a game manual which is easily enough for any expert (which we are meant to be) to identify the system it refers to, but we sit back as if we don't know what it is.

It is important to train new users in how to use the site. However, in such cases adding the tag and leaving a comment to that effect is still a way of teaching the user how we're meant to do things, and indeed even teaches them a second important lesson about how Stack Exchange works - that other people can and will edit your questions or answers to improve them.

It is possible in some circumstances that even these overwhelmingly likely inferences will turn out to be wrong, because the querent is actually playing some homebrew hybrid system, or bought the wrong book. But are these cases - especially now, as dopplegreener describes, that the D&D landscape has stabilised on 5e - really so likely that we must be so risk-averse? At present, this seems like an overreaction to avoid an extremely small risk.

Most of all, I feel that it is simply a more welcoming experience for new users when we are able to properly parse what they've written and make the inference to clarify their question for them rather than closing it until they fix it to this standard. I know putting questions on hold is not a bad thing, but that's always going to be unwelcoming to someone who's brand new to stacks, so where we can avoid doing so by assuming a very minor risk, I think that's a worthwhile risk to take. Where appropriate this also means these kinds of questions probably get answered much more quickly, which I expect helps with new user engagement and retention - if they come back a day later to find their question is closed, that's a quite different kettle of fish to coming back and finding some people have helpfully tried to answer the question.

begrudgingly written with a grim sense of obligation knowing that I'm one of the individuals that prompted this meta in the first place but also that my opinion is unlikely to get much traction.

• The "teach them by adding the tag and showing them" approach (and making sure that was correct by checking with them) is in line with most of our other teaching approaches too: show them how to format a question by formatting it, etc, not sitting back and waiting for them to figure everything out themselves. – doppelgreener Sep 24 '18 at 14:45
• +1 I really like the point about teaching the users by showing. I think we often maybe forget that there is more than one way to teach. I have been a proponent of the old policy, but I think you might actually have me reconsidering here. – Rubiksmoose Sep 24 '18 at 14:54
• I have done the occasional "teach by showing how to provide a reference/supporting point" in editing answers. I usually add a comment to let folks know what I did. Not sure if I am enabling or helping: probably depends on the user. – KorvinStarmast Sep 24 '18 at 16:57
• I downvoted in disagreement because as a teacher I believe strongly that teaching can be done well and it can be done poorly; it's not enough to consider it a teachable moment; it has to be taught well. And in my experience as a teacher fixing the problem after the fact is never as teachable as fixing the problem immediately or before it occurs. I explain that more thoroughly in the comments on my own answer. (I don't want to argue in comments on your answer. I'm just stating why I downvoted.) – Bloodcinder Sep 24 '18 at 17:36
• @Bloodcinder frankly I agree that technical improvements to the way the site works that allow us to emphasise the importance of tagging to the user while they are first composing their question is the best solution to this issue. However, that's not the site we currently have; we don't have a way to prevent the problem before it has happened as of yet, we can only react when the user does or doesn't do the right thing. This is what I think the policy should be for the site as it currently is; I'd happily be less tolerant if we had those kinds of changes made. – Carcer Sep 24 '18 at 18:20
• A zero tolerance policy on almost anything nearly always leads to those enforcing that policy, at some point, looking foolish. With that in mind, I think it's fine that the site values content over people, but I also think that people can value people over content to such a degree that there's room for a more welcoming new-contributor experience than Question held; learn to tag! (By the way, I think it's hilarious that you're a new contributor with, like, 19k rep.) – Hey I Can Chan Sep 24 '18 at 20:30
• @Chan That "new contributor" indicator is separately determined for mainsite and meta: see also greener's meta.se request on the matter for some more details. – nitsua60 Sep 25 '18 at 0:20
• @HeyICanChan A zero tolerance policy on almost anything nearly always leads to those enforcing that policy, at some point, looking foolish Yep. Fought two school boards over that crap. You nailed that dive. – KorvinStarmast Sep 28 '18 at 2:22

## No change as we're about to do this again

As stated in the original post, there were several reasons beyond just learning the site's policies and manner of operations. Among those were that upon site inception, there were plenty of opportunities for system confusion among the most popular games.

However, Pathfinder is progressing on their newest version and the opportunities for that confusion is about to happen again. Although that system doesn't seem to have ramped up yet, at some point it will and the chances of having confusing questions that could be original PF, new PF, or 5e will be abundant. Earlier today, someone was asking about a 'turrasque' (sic). Presumably they meant tarrasque, but it would be risky for me to assume that since there could be some game out there with a turrasque that wants to toe the line on copyright infringement.

And over time, the chance for confusion will be reduced as we sort out terminology and the like.

I expect over the long-term (which is the purpose of SE), these system confusion chances will ebb and flow and people searching for questions will benefit much more by consistent system clarity in all questions.

Let it ride.

• Yeah now that you mention it, we've got Starfinder out and Pathfinder 2e's playtest and people may be about to start mixing and matching especially since 2e is a bit polarising so far from what I've seen. We might be about to re-enter our earlier situation like you said. (I wonder if this is a ttrpg market cycle phenomenon.) – doppelgreener Oct 1 '18 at 17:21

# The policy doesn't need to change; the site's infrastructure itself does

tl:dr Instead of "soft-requiring" System/Edition tags via closing questions that don't have them, the site should change to "hard-requiring" them to be able to post questions in the first place. (This is different from requiring at least one tag.) The fact that it's required doesn't change, but the headache the first system causes is reduced.

### Why Closing is detrimental to the site

Closing Questions outright carries a punitive implication that simply doesn't give us a good rep. Despite whether new users should view it this way or whether they should just "toughen up", that is how it comes across. Nothing we do will change that because closing a question is inherently a snub.

Another thing is that we aren't "Experts" in the same way that a StackOverflow user is an Expert. In the vast majority of cases, we are not critical to the performance of someone's job, and a misinterpretation of the rules won't get lots of people fired. We're playing games.

When we close new questions commonly, it reinforces the stereotype of us as a Nerd-Clique, which is something we should be trying to avoid for the sake of the site's health.

### Some users have brought forth concerns of Burnout, or of preventing mistakes, not repairing them

I agree with these concerns, though I don't think that they should trump or be trumped-by issues of our image. I think that the current way our site works makes it difficult to appeal both to new users and to experts.

On the side of New Users, it is frustrating to get your question closed, especially if that new user is a new RPG player who might be unfamiliar with the breadth of how many RPGs actually exist, or how much context is important to their question.

On the side of expert answerers, it is both frustrating when we constantly have to ask for more info (both because of an incredibly strict rule and because of a lack of information) before being able to answer a question, especially if it is a question we want to answer because we can definitely determine what a user is discussing (as in, having pages from that RPG-edition's core book), and it does cause burnout.

Both of the concerns here are valid. And, something to note: Neither is less-valid than the other, because both pools are equally important to site health:

• If the experts leave, then the knowledge pool of this site falls apart. Some users will still create quality answers, but many of the top privilege-holders will go away.
• If the new users die, we'll just stop seeing as many questions. Experts will leave because the site will seem dead and we will see issues of stagnation and eventual death.

### Instead of "Optimizing" for one of these two critical groups, we should change the site

One thing we could do would be to make System/Edition tags (or the System Agnostic tag) required in a question. This would have several benefits:

• Avoiding having to "Optimize" for anyone by requiring that users performing tasks as-basic as adding tags for system and edition. This saves a lot of headache for everyone involved because it just takes a step that should be mandatory to submitting a question in the first place and makes it mandatory.
• It introduces new users to tags, and shows them how the tag system works. Many forums don't have tags, or they aren't that important, which is one reason why new users may not be familiar with them.
• It helps manage "Shopping Questions", namely, the kind which ask us to recommend RPGs. By not having a System/Edition tag (which would be required to post a question), it is more difficult to post these kinds of questions.

Note that this isn't merely "requiring tags," but it requires "System-Edition" tags. The color of these tags could be changed, say, from beige to green, to indicate their required status, and to differentiate the "Dungeons_and_Dragons" tag (which is not a System/Edition tag and is commonly misused) from the "dnd_[X]e" tag (where X is the edition number), which is a common source of tag confusion.

• Not all questions need a system tag - there are plenty on the site that don't – Wibbs Sep 27 '18 at 18:19
• @Wibbs Most of our questions do, unless they are System Agnostic, which would be treated as it's own "system" for the purposes of having a "system" tag. – SeraphsWrath Sep 27 '18 at 18:20
• Upvoted for describing a feasible mechanic to prevent the problem. It's not much different than how meta requires a tag. I'm fairly certain it's something the system can already do if setup to do that. But... 1) curating the system tags seems nightmarish, 2) system-agnostic is really non-obvious to new users, and 3) system-agnostic would need to be retroactively applied to all questions without a system tag. – Bloodcinder Sep 27 '18 at 19:10
• @Bloodcinder I think that last one is the most difficult and tedious, to be honest. After all, 2) can be fixed by editing the tag description, and 1.) is difficult, but once it's done it's over with. – SeraphsWrath Sep 27 '18 at 19:14
• What about when a querent asks a question about a game that's never been tagged before? (Asking both to clarify the intent of your answer and to ascertain whether the site supports it.) – Bloodcinder Sep 27 '18 at 19:16
• @Bloodcinder That one does seem to be the kicker, though new tags can be created and edited by any user with sufficient privileges. – SeraphsWrath Sep 27 '18 at 19:49
• My relevant feature request: Make it possible for certain tags (like game tags) to always appear first – V2Blast Sep 28 '18 at 3:00
• This isn't feasible for a combination of technical reasons and SE policy: SE will only write new features for the entire network, not custom features for only RPG.se, and they're never going to force all SEs have a required tags policy. It also misses that SE will never undermine easy closing (in fact, they have regularly added features to make it easier, not harder), because it's what separates us from Quora and Yahoo Answers and the quality nightmare that model results in. For better and worse, “elitist nerd-site” is the reputation SE intends us to have. – SevenSidedDie Sep 28 '18 at 4:39
• (It's also worth knowing that we're not hurting for new users at all. Our new user analytics are going steadily up year-over-year.) – SevenSidedDie Sep 28 '18 at 4:51
• @SevenSidedDie good to know. Out of curiosity, what mechanism makes required tags work on meta? Is it hardcoded just for meta's use case? – Bloodcinder Sep 28 '18 at 13:06
• @Bloodcinder Yeah, whatever it is appears to be hardcoded for meta, and the required tags are hardcoded and the same across every meta site. – SevenSidedDie Sep 28 '18 at 14:17
• @Bloodcinder Yeah, likely non-trivial to develop and test (plus touching other code like the privs system for who can do it). And required tags would probably break many desired usage scenarios, so it would need SE to be really, really strongly supportive of the idea. V2Blast’s feature request to make a sort-first list is more policy-neutral and has similar coding needs, but doesn’t do what this answer advocates (because it’s intended not to be policy-neutral). – SevenSidedDie Sep 28 '18 at 15:05
• @SevenSidedDie The "policy-neutral" observation is not one that was obvious to me but it's a helpful explanation. Thank you. – Bloodcinder Sep 28 '18 at 15:12
• @SeraphsWrath That’s only Meta sites. It’s not customisable though (what Bloodcinder & I we’re discussing). And not every Stack has an equivalent of our system tags to require anyway. So it would require a bunch of new features to select tags and turn it on/off per site (= significant developer time & pay), since SE would never force it onto every Stack. (That’s leaving aside that it’s a solution to a problem—questions being closed quickly for missing basic details—that SE doesn’t consider a problem.) – SevenSidedDie Sep 30 '18 at 4:18
• Relevant meta.se answer proposing this exact thing: Can our tag-prompt nudge toward including system? – Rubiksmoose Sep 30 '18 at 20:20

We shouldn’t guess.

One of the things mentioned against putting these questions on hold is that these are new users. I want to challenge the assumption that being a new user is reason for us to be reluctant to hold a question.

If a new user is put off by their first question being put on hold, they’re going to have a bad time at RPG SE.

Avoiding holds on questions by new users is not “being friendly”. It’s setting them up for failure. We should never flinch from holding a new user’s question.

If a new user can’t deal with a temporary hold, they can’t deal with the site at all. They will be a poor fit for the site, a source of conflict, and a drain on the energy and patience of the experts who give the site value to new users who do fit the site. It’s a failure all around.

Handling a new user with kid gloves to try to hide a basic part of site functionality from them in case it makes them unhappy only tries to keep users who wouldn’t actually want to be here if they knew. Users who can take a hold gracefully don’t need to have holding hidden from them, so hold or no hold is the same for the users we will keep.

Avoiding putting new users’ questions on hold damages the site and drains its resources, for no benefit.

We shouldn’t guess the system, even when we’re sure (we can be wrong about that too), and should put the question on hold. Not only does it teach new users how to hold the site, but it gives them an early chance to decide whether this site is for them or not. They deserve to know that sooner rather than later.

Holds exist as part of the site’s self-learning system for new users. It’s semi-automated to help people learn to use the site using their own energy, saving the energy of our experts for the important work of writing and improving our answers. Wanting to replace holds with our experts teaching people how to use the site — when they might not even become a regular — gets the site system exactly backward.

• I don't think it's fair to use a new user's initial reaction to having their post locked to claim that they would never be SE material. It is not difficult to imagine someone bouncing off the site immediately and never getting the chance to learn why it's okay, even if a bit more lurking would have enlightened them. – Carcer Sep 24 '18 at 17:19
• @Carcer It’s not about fairness, it’s about the fragility of our expert base. Experts are here to write answers. Writing answers doesn’t burn out experts, but dealing with new users who need help using the site does. The site from Day 1 was built totally around not having our experts burn out. There’s zero energy put into attracting or keeping question askers because, on the internet, the askers vastly outnumber the people who produce good answers. We draw them with existing quality… made by experts. Why should we risk burning out our experts just to (maybe) keep an asker who’s expendable? – SevenSidedDie Sep 24 '18 at 17:29
• @Carcer See Optimizing For Pearls, Not Sand for the official SE explanation of that position, and why it’s essential to a Q&A site’s survival. – SevenSidedDie Sep 24 '18 at 17:31
• That's an amazing policy, specifically re: preventing burnout. I somehow knew that but had never seen it articulated. Thank you for the link, SSD. – Bloodcinder Sep 24 '18 at 17:40
• @SevenSidedDie interesting read. So, completely flipping the focus, would then the argument that I, as an expert answerer of questions, get frustrated when I know what a question is asking but I am not allowed to actually answer it, because of an excessively strict policy about how explicitly a question must be composed, be a more compelling reason to want to change that policy? – Carcer Sep 24 '18 at 19:24
• @Carcer No. Curing that small problem would require weakening the system that’s in place to save us from multiple larger problems (discussions of how certain everyone is, metas about how certain is enough, comments confusing the asker, having to teach new editors to judge when and when not to guess, the rare but super messy times being wrong requires deleting our experts’ posts [which really drives experts away], & so on). It would do more damage than good. – SevenSidedDie Sep 24 '18 at 21:38
• We do have the "New User Stuff" queue for exactly that reason! We don't need to use kid gloves, we need to be even MORE careful with their behavior to show them how the stack works. – Trish Sep 24 '18 at 23:22
• I'll second this with an anecdote: my first question was held, received lots of comments, and scared the bejeezus out of me. (I'd already lurked a lot and read a lot of meta.) So I stepped back for a week, read and re-read lots of metas and comments, and gradually came back, a little more cautious. The process sure taught me effectively! – nitsua60 Sep 25 '18 at 0:17
• @SevenSidedDie for some reason even though I've read Optimizing for Pearls before, I've never understood it or internalized it nearly as well as when I read how you described it in your comment above. So, thank you! – Rubiksmoose Sep 25 '18 at 4:10
• The "setting them up for failure point" is one I agree with. +1 (I say this considering my long standing disappointment with how SE in general, and this SE, handle the new user experience). We want new users to adapt to our group norms. How we facilitate that adaptation (IMO) needs to be multi-faceted, not one size fits all. But on this particular point, it isn't that much to ask that new users bear with us as they get used to these norms. – KorvinStarmast Sep 28 '18 at 2:26

## Given the practical constraints of this universe, we shouldn't guess.

I don't actually think it's inherently wrong to guess - much as I don't think it's inherently morally wrong to volunteer to work for free, for example (as all of us are doing here, really.) But when drafting site policy, we have to include a utilitarian perspective and consider all the consequences. As others have suggested, it would be extremely difficult to create a reasonable guideline for "how sure" one has to be before guessing is acceptable - leading to arguments and confusion, all of which would be more trouble than it's worth. As with working for free, this is a this-is-why-we-can't-have-nice-things scenario - sure, it would be nice to say "that's obviously a mechanic that only exists in $system, so I'll answer for$system" sometimes, and I've certainly been tempted more than once, but in practice the value of being able to do so pales in comparison to the costs in time and complexity of trying to set up and enforce a policy that allows it while retaining members and members' goodwill and patience. We all want to jump in right away, but honestly, most RPG-related questions can wait 24 hours to go through the process with no harm done, so there's no real reason to change the policy.

## Bonus: Being strict actually helps the querent.

Obviously, when someone asks a question, we want to answer their immediate question, and that's definitely the core site goal. But if you look at it a little more broadly, we want to help users solve their problems and improve in general, which is one reason we allow frame challenges. Part of that process, then, is teaching fundamental question-asking etiquette. If you have an issue with your phone or car, the first thing any technician will ask is "What model is it?", and rightly so.

The same thing applies to game systems - some people forgot to mention it, but others may not actually know, and in either case, it's important for anyone trying to gain more knowledge about their \$object to first and foremost know what it's called and be able to readily convey that information to experts. It's not the main point of RPG.SE, but if one thing someone takes away from interacting with us is "not only are there other editions, but there are whole RPGs that aren't DND 5e, and they're common enough that you shouldn't assume without more info," I consider that a good day's work.

• And yes, even though I agree with Bloodcinder and d7, I still filed not one but two concurrences, because they use different reasoning and may garner different responses. I'm pretty sure that's what you're supposed to do, but correct me if I'm wrong. – SirTechSpec Sep 26 '18 at 21:56
• yep. my own experiences and adaptation process is in harmony with your points. – KorvinStarmast Sep 28 '18 at 2:28

### Don't guess, but: overzealous closing when the system is obvious is harmful to the user experience.

I disagree with the prevailing opinion that "never guess the game system" means "never add the missing tag where the game system is unambiguously clear from the text." If it's clear, it's not guessing.

In two particular cases, absurd arguments were invoked to justify removing an obvious tag:

• In Are 0-Level Spell Scars Free?, the question was tagged only (a Pathfinder-specific class), and described the Spell Scars ability working precisely as in Pathfinder. My tag was reverted just in case the user meant M.A.G.U.S, a Hungarian-only RPG from the 90s. The user later clarified that they indeed meant Pathfinder.
• In Can you use darts as improvised “light melee weapons” to trigger Two-Weapon Fighting?, the user quoted rules from pages 147 and 195 of the Player's Handbook, which matched the same rules on the same pages of the D&D 5th edition Player's Handbook word-for-word. My tag was reverted in case the user was playing some unknown hybrid of D&D 5e and D&D 3.5. The user clater clarified that they indeed meant D&D 5e.

The current system of handling questions like these causes several issues that run counter to the site's purpose:

• Without a system tag, the question is usually put on hold:
• This impedes giving the user a fast reply, which Jeff Atwood once described as "the entire goal of a question and answer site in the first place".
• Users who can provide an answer answer, and who may even be halfway done writing a correct and excellent answer, must wait (sometimes days) for the question to be edited and a reopen vote to be completed.
• The new user may accrue downvotes for failing to follow a site policy that wasn't fairly mentioned either on the Ask a Question page or the Tour:
• This provides a hostile environment to new contributors, who we are nowadays reminded to be nice to, and who have complained in the past that site policy is unfriendly to newcomers.
• Downvotes tend to stick even after the question is corrected. A good question is thus rated poorly, which impedes the site's purpose as a repository of questions and answers rated by quality.

Understandably, we can't guess at a system where there isn't enough information in the question. For example, suppose someone asks:

I have a tiefling fighter with scale mail armor. How can I improve my Armor Class?

This could be any D&D edition after AD&D 2e, or another RPG like Pathfinder which uses the SRD, so we can't guess the system.

Cases of accidentally mix-and-matching content from different editions still sometimes happens due to online sources; e.g. someone is playing D&D 3.5 but the first Google hit is a 5th edition's D&D Beyond, or playing 5e but the first hit is the D20 SRD. This happened earlier this year, but the user did tag in this case, and the erroneous nature of the mixed system was obvious.

In such a question, I believe that a frame challenge of "You appear to be mixing two incompatible systems" is a reasonable answer.

If someone is intentionally mix-and-matching systems, they're likely to be very aware of this, and are likely to mention the systems in the question. It's difficult to answer such a question at all without knowing which rules they use from which systems.

I'm not aware of any question (in the last year, at least) where the asker intentionally played a mixed system and didn't mention it until asked. Most mixing is done between largely compatible systems like 3.0/3.5 or 3.5/Pathfinder, and in those cases it's usually not possible to unambiguously deduce the system. You don't normally see e.g. a Pathfinder/D&D 5e mix because the two are so different that they wouldn't work correctly together without effectively homebrewing your own system.

Usually, a mixed game player is aware of the rarity of their combination as well as the existence of multiple games or multiple editions of their game, and the need to state it up front. The most common error is a new player who don't know that what they're playing is e.g. D&D 5th edition, or forgets that there are other RPGs than D&D.

The only issue I see is where the answerer or tag editor is overconfident in their ability to recognise RPGs. Someone who has only played 5e may see "dragonborn" and assume it's talking about 5th edition, unaware that this race could refer to 4th edition.

In any case, and I'm aware that this is a late answer which opposes the popular opinion, but if an editor can be sure which system the question is talking about, it's reasonable for them to add the tag, as readily as they would correct a typo where it's clear what word the author meant.

• How are we to judge when "sure" is truly sure? I sympathize with your point. Effectively, you're saying that the tag should be edited if it's certain beyond a shadow of a doubt. But do you have a recommendation for how to define/codify that? ...Much less, how to enforce that? – Bloodcinder Oct 3 '18 at 2:39
• "I'm not aware of any question (in the last year, at least) where the asker intentionally played a mixed system and didn't mention it until asked." FYI there has been at least one that used a mix of all editions and did not clarify until asked. – Rubiksmoose Oct 3 '18 at 3:34
• This is citing our “new contributor” notices incorrectly. It’s very important for our experienced users to understand our actual conduct policy instead of assuming what it means. I know it’s new, but I keep seeing “closing is unfriendly” and it cited to argue against closing. That is wrong. The Code of Conduct literally says to be clear about problems, be nice communicating them, and that askers are responsible for providing complete questions. No mention of us of not closing. We are not going to start avoiding closes based on a mis-cite. – SevenSidedDie Oct 3 '18 at 14:34
• @Bloodcinder the same way that we judge “right” is truly right? – fectin - free Monica Oct 28 '19 at 13:41

I think that any policy that forces us to default our guesses to a specific system will have the unfortunate effect of this place seeming even less welcome to questions that are not D&D 5e - many of which are already having trouble finding answers.

Keep the policy as-is