Is the Only-Tag Test for meta-tag-ness broken here?

Question

We've talked about meta tags a lot here. One thing that the last round highlighted is that we have a lot of tags that are impossible to use without a system tag. It's been said that tags like spells or rules-as-written maybe shouldn't exist because they can't be the only tag on a question. Perhaps those are debatable, but more system-specific tags like aspects and prestige-class are more clearly unable to stand alone.

Normally, the wisdom is that this makes them meta tags. But is that true here?

The thought I'm chewing on is that, perhaps, we can't use that test for meta-tag-ness unless we give a 1-tag discount for a system tag. Why we might have to do this, is because many questions require a system tag before you can reasonably talk about all sorts of things, and tagging for all those other things is incredibly useful. (q.v., spells's existence.)

It's worth noting that the original SE Blog post on meta tags formulated the test as a heuristic, not as the stronger hard rule that we sometimes treat it as (my emphasis):

If the tag can’t work as the only tag on a question, it’s probably a meta-tag.

Is the strong only-tag test for meta-ness broken on RPG.SE? Can it be un-broken by this hack or is it more complicated than that?

Answer 1

Correct, the system tag can be considered an exception to the Only Tag test with respect to other tags.

This is pretty similar to Stack Overflow, where the language is the equivalent of the system tag. Like all the SO questions tagged "inheritance," well, inheritance works differently per language so they pretty much all have <language>, inheritance as tags. A lot like our system tag, and you'd see <system>, spells as the equivalent tagging structure here.

Therefore (and to avoid tags like java-inheritance or dnd4e-feats), the system tag can safely be assumed to not count towards the one tag rule.

Answer 2

A meta tag is a tag that describes irrelevant aspects of a question that can't stand alone, according to the blog post.

The reason meta-tags are a problem is that they do not describe the content of the question. They describe some other aspect of the question, like the author’s skill level, or the author’s motivation for asking it, or generally what “kind” of question it is (poll, how-to, etc.).

Meta-tags are actually a subset of a larger problem that I usually call dependent tags. These are tags that don’t say anything by themselves – you can’t tell what the question is about unless they’re paired with some other tag (or several of them). These tags are a problem because people don’t realize this and will often use that as the question’s only tag.

As long as a tag describes the content/nature of a question it is not a meta tag. It may still be a dependent tag, though, which is what that test checks for.

Dependent tags are not actually a problem for this site, and seem to be nothing but helpful here. System tags are an excellent example of a dependent non-meta tag as they definitely cannot stand alone yet are very helpful (note that all questions that have a single system tag appear to implicitly be tagged rules-clarification). In general questions tagged with a single dependent tag do not seem to be a problem for us and the existence of most of our current dependent tags is beneficial so i think we should ignore the MSO advice on that one. Meta tags, of course, are discouraged and probably not helpful, but I am unsure we have any tags like that. Occasionally someone asking a bad question tries to suggest a just-for-fun tag, but we seem to be aware that that is a dangerous idea.

Answer 3

Those tags could — and, arguably, should — be renamed to, say, fate-aspects and dnd-prestige-class, which would allow them to stand on their own.

In fact, there's a practical reason for using such a naming scheme — it's not just policy-wanking for its own sake. The reason is that while, say, the term "aspect" has a clear and well defined meaning within Fate, we really just happen to be lucky that no other popular RPG has chosen to prominently use the same word. As soon as that happens (and I expect that, "aspect" being such a generic word, sooner or later it will), the tag will become a complete mess, with people using it for two different concepts from two different systems.

(SO, for example, is full of such useless muddled tags. Just glancing at the front page, I can see a dozen examples like field or select; does that mean the SQL SELECT keyword, or the HTML <select> element, or the POSIX select() syscall, or one of the two Perl built-in functions of that name (one of which is a wrapper for the POSIX function), or...?)

Fortunately, for now, RPG.SE is still small and cohesive enough, and our question volume is low enough, that we can probably just wait and deal with the problem if and when it arises. But renaming the tag now would keep the problem from arising in the first place, and might save us a considerable amount of manual retagging labor later.

---

Ps. I wasn't sure if I really needed to address this, but since KRyan mentioned it: Unless some future edition / variant of D&D decides to completely change what they mean by "feat" or "prestige class", we'll still only need a single dnd-feats or dnd-prestige-class tag, not separate tags for 3e, 3.5e, Pathfinder, etc. The point is that within the scope of D&D and its variants, "feat" and "prestige class" are clear and well defined terms, just like "aspect" is for Fate and all its variants.

Now, D&D is still kind of the 900 pound gorilla of RPGs, so it's pretty unlikely that some other RPG system would decide to use a specific term like "prestige class", except maybe as a deliberate nod to D&D, anyway. Thus, prestige-class may be a poor example here, as it's unlikely to become ambiguous. But a word like "feat" is just about generic enough that some other system might decide to use it for something else, confusion be damned. And I could definitely see that happening to "aspect" some day.