9

Recently, the tags:

have been created.

What is the rationale for these creations?

Most recently, tags were discussed here, but later here. More generally for Stack Exchange there's a conversation here.

I personally don't see the value in any of these tags. There are so few questions on these topics, there's no point in creating a feed for these topics, nor do they help with searching as general keyword searches would pick them up anyway.

| |
15

There's more than that: I've also created , , , , , and .

The rationale is that our MTG tag hierarchy has been super undeveloped (borderline nonexistent) which makes it extremely difficult to find and reference past answers or find past duplicates. Very often I simply cannot find those questions at all even when I know they're there to be found. Our canonical questions on a great many topics are lost in a giant uncategorised pile of questions tagged only .

We have a curation tool for this scenario—tags—which has been underutilised and which I'm now applying to help us resolve this problem. There are many ways to slice up the metaphorical tag pie, so what I've been doing is working out distinct, significant, and unusual topics where we see questions very specifically about only that topic and where there is a lot of distinct confusion about the mechanic or topic itself.

The most concrete case where this is valuable is when this helps us find and make duplicates. You highlight as not having value, but in the process of adding that tag we found 3 never-before-closed duplicates in only 14 questions:

These duplicates were never identified in the beginning because it was extremely difficult to find questions about the indestructible mechanic. (You say keyword searches should pick them up, but I'll get to that shortly.) Now it is trivially easy to find questions about the indestructible mechanic, and when we see future questions about it, we can check for duplicates again then too. So they are already giving value.

Similarly when I added we found these to be duplicates:

I think it's noteworthy these questions are the only duplicates in those tags. We didn't just find new duplicates—rather, before consolidating the topics under tags, duplicates were simply not getting discovered and closed at all.


There's other ways these tags add value to the site:

  • We gather topics together to help bring them attention to the focus of experts who they need to reach. Because we have , , and , people who are experts on Rules Enforcement Levels can watch these tags to have them highlighted. (And indeed each of these tags has one watcher at time of writing this.)
  • We gather topics together for newbies to learn about. If you've just visited our site confused about the commander format, color identity, regeneration, indestructibility, planeswalkers, emblems, double-faced cards, or rules enforcement levels, you can dive into previous questions on that topic to learn other things you might have been unclear on.
  • We make it easier to find prior duplicates.
  • We make it easier to find and reference canonical questions and answers on a topic.
  • We make these topics more easily searchable and discoverable because they're actually not by keywords alone.

There are so few questions on these topics, there's no point in creating a feed for these topics, nor do they help with searching as general keyword searches would pick them up anyway.

Not actually true. When I run a search for [magic-the-gathering] indestructible is:question I find 55 questions. Despite that, there are only actually 14 questions directly about the indestructible mechanic, now tagged as such.

75% of those keyword results are noise!

This is because most questions showing up in search are proposing a scenario in which indestructibility just tangentially comes up, but the questions are not, themselves, about the topic of indestructibility.

The tag lets me cut through the noise to find the questions fundamentally about the mechanic.

I don't want to look like I'm cherry-picking though so I'll examine the other two tags you listed as well:

These noise levels are why we weren't finding duplicates. It makes the topic difficult to dive into because by keyword alone we're finding tons of things that are not actually about the topic but are just tangentially mentioning it.

These noise levels aren't true of all of these tags. There's only 10 MTG questions that mention emblems and four tagged . It's not hard to sift through those. Same goes for . But these are giving us value by gathering together a topic of focus and making it trivially easy for newbies and experts alike to find questions on that topic.

There's also usefulness when a tag isn't easily keyword searchable. has a lot of equivalent keywords: do you search transform or double-faced or dfc or dual-faced or back face (catches a lot of questions about morph) or flip (a common misnomer that also catches flip cards)? Have you missed any keywords? Having a tag categorising the topic resolves this set of problems.

| |
  • 8
    Just a small addendum/note, there are over 3.1k mtg questions on the site of around 9.7k total questions. It doesn't really make sense for ~1/3 of the sites questions to be under a single tag, especially when using sub-tags adds so much searchability without really costing anything – Malco Oct 30 '18 at 16:36
  • 1
    Brilliant answer. I certainly see the value now! The use case of, "using tags as an efficient search" had never occurred to me. – John Oct 31 '18 at 16:46
  • 9
    Maybe we could talk through general guidelines for mtg tags? It sounds like you've got a philosophy here (e.g. tags for keyword abilities of the form "mtg-{keyword}") and it'd be great if we could consistently follow it. If you're up for it, a separate Q&A summarizing all this would be really helpful. – Cascabel Nov 2 '18 at 23:25
  • @Cascabel I'd be up for doing that. I could have something out as time is available over the next few days. – doppelgreener Nov 3 '18 at 12:33
  • 3
    Specifically I think it'd work as a self-answered question - what are our conventions for mtg tags? and you can preface your answer with a note that it's not final (though based on actual existing tagging, so it's not just a proposal either). – Cascabel Nov 5 '18 at 18:57

You must log in to answer this question.

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