There have been mentions of “rules-as-intended” (RAI) a few times on this meta, but always tangentially to other discussions, so I wanted to take this opportunity to provide a place to discuss it directly.

Specifically, there are two different things people generally mean when they use the term rules-as-intended:

  1. Literal: Actually what the designers of the rule were thinking when they wrote it. Why do spontaneous casters lose out on a spell level compared to prepared casters? is a recent, well-written question seeking this kind of information.

  2. Not-RAW: In this case, RAI is merely used as opposed to rules-as-written (RAW). That is, based on experience, fluff, verisimilitude, balance, fun, or any other considerations beyond the rules text itself. “What’s the best way to run this?” rather than “What does this rule mean?” Note that this meaning does not match the actual meaning of the phrase “rules-as-intended.” Rulings/Usage on Abrupt Jaunt is a recent question that (originally) used RAI in this fashion.

So, I wanted to ask the community: how should these two sorts of questions be handled?


1 Answer 1


Literal Case

This reference to intent is the more clear version: the querent literally wants to know what the authors intended when a certain rule was written. Oddly enough, in my experience, the term RAI is often not used here, possibly because of how common the other use is.


A question requesting this sort of information would typically be tagged , and/or possibly . These sorts of questions are, I think, definitely on-topic and appropriate, and as the designer no doubt did have something in particular in mind when he or she wrote the rules, there does, in theory, exist a correct answer.

Good Answers

The best answers asserting this sort of information (either in answer to a question specifically requesting it, or in answer to a question interested in any of a variety of perspectives) would cite specific quotations from the rules-designers. The problem with this is that the information necessary to answer the question well is often unavailable. Game designers speak about their products, and the rules in them, quite a bit, but there are a lot more rules that are never addressed specifically.

Alternatively, in theory, one might attempt to do a literature analysis, to try to determine authorial intent. This is difficult to do well, particularly in a concise enough fashion for SE, and RPG.SE isn’t necessarily where I’d expertise in such matters, but it can be done. It still requires citations: citations of meta-information such as the history of the product or the state of the community that played the game at the time the rule was written, citations of parallel cases elsewhere in the rules, etc. The overwhelming majority of attempts at this that I have seen have been fairly weak, but a few exceptions have been good enough that I feel compelled to acknowledge the possibility.

Bad Answers

Bad answers use a request for RAI, or an invocation of RAI, as an opportunity to post their opinions without Backing It Up! as required by Good Subject, Bad Subjective. This overlaps quite a bit with the Bad Answers section under Not-RAW, because the answers tend to look very similar: intent is asserted, but the answer does nothing to try to establish that assertion as fact.


This case is both more and less problematic. On the one hand, questions requesting this kind of information are easier to answer well (after all, it’s the default for this site!), and the problems it does have can generally be fixed without ruining a good question or answer. On the other hand, the term does not match the meaning, which produces ambiguity.


Usually posted by new users, who aren’t aware that “not-RAW” is our default. Occasionally used by established members when quoting rules-text to indicate that, despite the quote, they aren’t merely looking for analysis of the rules text (though this would, under my understanding of community expectations, not actually be necessary). The question may be good or bad regardless of the use of the RAI term; the term itself is almost meaningless in this case considering our defaults.

Good Answers

Much like questions invoking RAI, good answers invoking RAI as not-RAW would be good answers even if they never mentioned RAI. These are answers that Back It Up!, not by asserting RAI, but through other means (play experience, textual analysis, rules lawyering, whatever perspective they are using, assuming that perspective is appropriate to the question asked). They just happen to also assert RAI.

The only problem here is one of communication: by using “rules-as-intended” to actually mean “not-RAW,” at least in my opinion, these answers confuse matters. The term rules-as-intended implies more than just not-RAW, however often it may be used that way. Some people do say “rules-as-interpreted,” which would also be abbreviated RAI, but “rules-as-intended” is (at least in my experience) the much more common expansion of the abbreviation RAI.

Bad Answers

These are answers that are fail to Back It Up!, and try hide that fact by asserting intent. For many questions, intent does not automatically make for a good answer, but even if it does for a particular question, a good answer has to establish that intent if they are going to use it as the basis for Backing It Up!

To be fair, I don’t see this kind of problem too often on RPG.SE. It’s more of a problem I associate with discussion forums, but it’s a serious (and obnoxious) enough problem there that I wanted to pre-emptively discuss it. On discussion forums, these kinds of assertions of RAI often signal the end of meaningful debate, as one side is asserting that their interpretation is what is intended, and refusing to consider any other perspective, while the other side has nothing they can actually debate because no case was actually made for intent.

While I haven’t seen it happen with RPG.SE, I can imagine this compounded by the fact that, as an SE, we ideally want a “right” answer. Raw opinion doesn’t really have a place here. Good Subjective requires that we Back It Up! So there may be a temptation to try to use an invocation of “RAI” as a fake way to Back It Up!


A question requesting intent needs answers that cite sources. This isn’t necessarily a special rule, as all answers must Back It Up!, but I think it is important to emphasize it in this case. Thus I suggest:

  • Questions requesting intent should specifically require sourced quotations. Editing the question and/or commenting to remind answerers is appropriate. If the question seems to actually want speculation, the question should be closed as Primarily Opinion Based.

Meanwhile, questions that use RAI to mean not-RAW might be confused with the above, and therefore:

  • Edit questions using RAI to mean not-RAW, such that the term RAI is no longer used. In the basic case, just expanding it to “rules-as-intended” could work, but preference should be for something less jargon-y, or nothing at all. A comment explaining to the querent that not-RAW is the default for this site is appropriate.

On the flip side, answers asserting intent (whether in response to a question that specifically asks for it or not) must Back It Up!

  • Answers that assert intent, but do not Back It Up!, require specific fixes:

    • If the question is specifically requesting intent, the answer should be flagged as Not An Answer.

    • If the question is not, and the answer otherwise Backs Up! its position well, editing the assertion out may be appropriate.

      • This is particularly true where “RAI” has been used to mean “not-RAW” – since that is the default for us, it doesn’t really need to be a part of the answer. If nothing else, expanding it to “rules-as-interpreted” eliminates any ambiguity, though the preference should be for something less jargon-y.

      • On the other hand, an answer that is clearly speculating that a given interpretation was intended can probably be left as is. E.g., “I also think this was the intent, though I cannot back that up,” doesn’t need to be removed so long as the rest of the answer is good. Being clear about what is and isn’t objective/backed up can go a long way, so long as the meat of the answer is Backed Up!

    • If the answer is merely asserting RAI as a way to avoid Backing It Up!, it should be downvoted. In particularly egregious cases, flagging as Very Low Quality may be appropriate.

In general, we should default to taking “RAI” to mean rules-as-intended, and then take that literally, and hold people to Back It Up! We should behave this way because we want to protect those questions actually requesting intent by making it clear that if you are going to assert intent, you must Back It Up! This isn’t really different from usual, but because of the way these terms have historically been abused in RPG-discussion communities, I think a somewhat more vigilant/pro-active approach is appropriate.

  • 2
    \$\begingroup\$ This has not come up nearly enough to take all kinds of actions to address it. \$\endgroup\$
    – mxyzplk
    Commented Jun 18, 2014 at 3:56
  • \$\begingroup\$ @mxyzplk I'm not suggesting "taking action" so much as a "policy of action" as it comes up. \$\endgroup\$
    – KRyan
    Commented Jun 18, 2014 at 3:57
  • \$\begingroup\$ @mxyzplk I suppose I did recommend the banner; you're right that was overboard. \$\endgroup\$
    – KRyan
    Commented Jun 18, 2014 at 4:03
  • 1
    \$\begingroup\$ "If the question is specifically requesting intent, the answer should be flagged as Not An Answer" - This seems like an improper use of flags. IMHO, an answer that asserts intent but fails to back it up is a very low quality answer (indeed, an incorrect answer), but it isn't Not an Answer. \$\endgroup\$
    – Tim C
    Commented Apr 16, 2016 at 4:14

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .