7

It looks like several people have started creating tags for various mechanics in Magic (protection, bestow) and retagging existing questions with them. On Arqade, these types of tags have been removed because they mean different things in different games, and thus it doesn't make sense to have a "protection" tag for Magic that doesn't explicitly state that it is for magic (such as "magic-protection").

Is this something we're encouraging on this site, and if so, why are we choosing to disregard the lessons learned on Arqade?

| |
  • What are the lessons from Arqade? Could you provide an explanation or links for those of us who don't use that site? – doppelgreener Dec 3 '14 at 2:13
  • 1
    You can get to all sorts of discussion on the topic from meta.gaming.stackexchange.com/questions/3800/…. Of note is the answer from Jeff Atwood, one of the founders of Stack Exchange. One could argue that Magic now meets the question number threshold, which is why I think a specific tag like magic-protection might make sense. – bwarner Dec 3 '14 at 2:36
  • 1
    The full name of the game is Magic the Gathering so any subtags should use the syntax "mtg-[mechanic]". – Chrismas007 Dec 4 '14 at 20:41
  • Also, I think a good (accepted) answer to this post should be a suggestion on how to "sub" tag MTG. As a user mentions below, there are too many mechanics in MTG for each to have a subtag, however, there could be a good comprehensive answer which breaks MTG down into quality subtag components. – Chrismas007 Dec 4 '14 at 20:45
  • At this point, I don't know that I can accept an answer unless a moderator steps in and makes a decision. I just wanted to raise the discussion. As often happens in cases like this, it seems the community is divided. I'd at least say that we shouldn't create more tags for Magic at this point until we've agreed on a plan of how best to do so. – bwarner Dec 11 '14 at 14:19
  • @bwarner Sure, there's not consensus, but given the voting on the answers, I think it's pretty clear where the majority opinion lies - and it's not really the place of moderators to override voting on meta (without really strong reasons anyway). We can always revisit it if, say, someone comes up with a concrete proposal for which tags we'd want. – Cascabel Dec 13 '14 at 4:23
20

I agree with you, but Arqade is also my home site, so that's not surprising. I think we should not have a tag, for the reasons you mention.

I also think we should not have a tag, or any similar tag that refers to a specific rule of the game. This is for three reasons:

  1. Nobody is an expert in "protection." Playing Magic and knowing where to look for rules references effectively makes someone an expert in Magic in general, but there's no reason that someone would know just the rules for one specific part of the game without having significant general knowledge of how the game works.

  2. There is no reason to search for . It's not necessary to find a particular question about protection; that could just as easily be accomplished by searching [magic the gathering] is:question protection, plus other relevant search terms. And it wouldn't really be useful for getting an overview of how protection works. That could be more easily accomplished by reading What does protection from [quality] do?.

  3. As Jefromi points out, tags may not be added consistently, which can actually worsen search experience. If we have tags for some keyword abilities but not others, and for some ability words but not others, and that choice is made based on site activity, it makes this problem much worse. If a new user asks a question, they would have to attempt to tag it with every ability the question is actually about to properly tag the question. As question volume grows, this could even make it difficult for experienced editors, who would have to know every keyword that has a tag to recognize when a question is missing one. In addition, a question that asks about the interaction between five different keyword abilities (I'm sure it's possible) would be impossible to tag correctly if all those abilities have tags. All of these are parts of the same problem: in the context of search, a tag should allow someone to find all questions about that topic; with these sorts of tags, that may be difficult or even impossible to accomplish.

| |
  • Your answer doesn't describe even a single benefit. It only discusses the downsides. Is this because you did not consider the benefits, or because you truly think there are none? – Rainbolt Dec 2 '14 at 18:08
  • 4
    I don't see any benefits to having the tag. This answer is an honest representation of how I see that tag on this site. – murgatroid99 Dec 2 '14 at 18:10
  • 1
    Another downside: these kinds of tags will tend to be really unevenly applied. People posting questions won't generally think to tag beyond the name of the game, so we'd be setting ourselves up for a lot of retagging for little to no benefit. – Cascabel Dec 3 '14 at 1:24
5

Does it add value to the site?

Yes.

If I want to search for all questions that are tagged and , I currently do not have a way to do that. Therefore, creating a protection tag would add at least some value to the site.

Nearly half of the questions on this site are about Magic. Contrary to the other answer, some players are much more intimately familiar with some parts of the game than with others. For example, I am probably more familiar with the Infraction Procedure Guide than I am with the Layer System in Magic.

You could argue that additional tags decrease the value of the site by polluting it. I would counter this with two points:

  • You don't have to favorite the protection tag, so it won't pollute your sidebar.
  • Most magic the gathering questions only have one tag.

Can it stand alone?

No.

Across all games, protection generally means "defense against something". However, when paired with a supporting tag, the meaning changes drastically. For example, "protection" means something incredibly specific in Magic.

Stack Overflow has a tag. This tag is almost always paired with a language tag such as or . However, unlike "protection", "garbage collection" pretty much means the same thing in every language.

So what's the conclusion?

I think that although it adds value to the site, its inability to stand alone really kills it as a tag, and therefore it should not be added.

| |
  • 3
    A question about protection really should have the word "protection" somewhere in the body, which is what that search looks for. And my point was that you said "I currently do not have a way to do that." I disagreed with that statement. – murgatroid99 Dec 2 '14 at 18:59
  • @murgatroid99 Well, my answer no longer says that, so we should clean up our comments. – Rainbolt Dec 2 '14 at 19:34
  • 3
    Now it effectively says "Without the [protection] tag, I can not search for questions tagged [protection]" which is tautological and therefore meaningless. You could just as easily say "If I want to search for all questions that are tagged [magic-the-gathering] and [foo], I currently do not have a way to do that. Therefore, creating a foo tag would add at least some value to the site." – murgatroid99 Dec 2 '14 at 19:45
  • 1
    Unless you give a reason that someone would search for questions tagged [protection], the analogy is perfect. You say that some players are more familiar with some parts of the game than others. But most questions are answered while still on the front of the new questions page, and there is exactly one unanswered [magic-the-gathering] question right now, so finding unanswered questions to answer in a more specific area of expertise is extremely easy even without a more granular search. – murgatroid99 Dec 2 '14 at 20:20
  • 1
    Regarding your latest edit: the search I give in my question has 32 results. It's easy enough to look through all of them without having another tag to search with. You say this tag would add value, but I don't see how it's better than what we already have. – murgatroid99 Dec 2 '14 at 20:27
  • @murgtroid99 I didn't say it was better. In fact, my conclusion clearly states that I think we should not add this tag. Instead of recognizing that adding the tag has its pros and cons, you are just aggressively attacking every pro you see in my answer. I just wanted to write a more balanced answer. – Rainbolt Dec 2 '14 at 20:35
  • Only about half of the questions in that search are actually about protection, so I'll concede that some people may see that as a benefit. I didn't address it in my answer because I don't see it as a pro. – murgatroid99 Dec 2 '14 at 20:40
  • @murgtroid99 Your very first comment is invalidated by your most recent one. "my answer provides a perfectly good way to search for MTG questions about protection" followed by "Only about half of the questions in that search are actually about protection". Since when is about half considered perfectly good? I've rolled back my answer in light of this new discovery. – Rainbolt Dec 2 '14 at 20:48
0

Yes, I think we should have these. (Naturally, since I added a couple of them.)

Having categorisation tags for MTG would be helpful, whether of the format of [protection] et al, or [mtg-protection] et al (subtagging, as suggested and adopted on Arqade). People here seem concerned about picking up the former format, so we can go with the latter: Jeff Atwood suggests the latter would be legitimate under circumstances that are the case for MTG questions here (loads of questions, loads of potential tags).

I added those tags because we have virtually no categorisation among our MTG questions, and that makes them very burdensome to sort or search through, and that isn't a good situation. A similar tag system has worked exceptionally well on a site I'm a long-term user of. I would consider it a major loss in potential gain for the site if we weren't able to apply a similar system here, and it was enforced that MTG should have no mechanical tags at all, considering it also does no damage to the site for being present.

It adds value to the site

Like Rainbolt described. I can use tags like this to search for particular questions I'm after that are about Protection or other mechanics.

murgatroid99 suggests searching for MTG questions that mention protection is just as good, but it isn't: a lot of false positives turn up which are questions that are not about protection, but just happen to mention the word somehow. Heck, Are there any good intermediate deck building resources available in magic turns up as a result, because the person is talkng about protecting against various strategies. The tag itself adds distinct value not achievable without it.

This searchability is value for regular SE users, who can use it to dig through questions on a topic for reference/linkage/citation or to aid in identifying duplicates. It also has value for new players, who could do the same thing to check if their kind of question has been asked before. (Many will not do this, but for those users there is no net damage caused by the tag's presence.)

The non-standalone nature does not matter much

I say that from experience on RPG Stack Exchange. Take a look at our front page, and you'll see loads of tags that describe categories of mechanics that exist in particular games. Many of these cannot stand alone. stress, aspects, natural-weapons and so on are close to meaningless outside the context provided by another tag, for instance. However, this is fine: the tags still provide excellent categorisation value, and we find them very helpful, and the fact we might never answer a question with that tag alone doesn't really cause any trouble. (We have a very active and self-reflective meta community. We're keenly aware of problems with several of our tags. These aren't among them.)


This could set precedent for adding tags for various ability keywords. Since there's a lot of those (106!) and many other named mechanics (e.g. creatures, legendaries, and so on) this could open up us to an awful lot of tags.

However, that does not necessarily mean that all of these potential tags will ever be created. Like RPG.SE, we should still follow the policy that tags are only added when someone finds it useful to add them, and tags aren't generally useful until there's a lot of questions on a topic.

For many ability keywords, we simply do not have any questions at all (e.g. amplify, rampage, horsemanship, modular, fear) or will be unlikely to get very many (e.g. banding, phasing — the more of these I searched for, the more keywords I found we just don't have questions about). That leaves us with many ability keywords that will simply never have a tag anytime soon.

| |
  • 4
    I notice that the tags that have been recently added are [bestow], [heroic], and [protection]. Would you say that this should set a precedent for adding a tag for each of 26 ability words and each of the 106 keyword abilities? – murgatroid99 Dec 3 '14 at 0:50
  • 3
    I would also like to note that tags that mean different things in different contexts are not considered to be acceptable on most StackExchange sites. StackOverflow (where this is most easily seen due to high question volume) routinely deletes or disambiguates tags with multiple meanings. Arqade also follows this policy, and to my knowledge has no tags that can be used on different questions with substantially different meanings. – murgatroid99 Dec 3 '14 at 0:53
  • @murgatroid99 Yes, it's not acceptable on most Stack Exchange sites. We are not Most Stack Exchange Sites, we're this one, and can make our own policy decisions, based on what works well uniquely here. On RPG.SE, it works extremely well. There's precedent both ways. – doppelgreener Dec 3 '14 at 0:54
  • @murgatroid99 Re adding tags, it sets a precedent we could add those tags, but does not mean they will necessarily be added (ever). On RPG.SE, we employ the usual policy that tags don't need to be added until someone finds it useful to add them. They're not generally useful to have unless there's a lot of questions on each topic. But if we do have that many, it's going to be pretty useful to have a tag about it to help search for that topic separate from the droves of other MTG questions. Note RPG.SE only has tags for classes that get a lot of questions, but there are many tagless classes. – doppelgreener Dec 3 '14 at 0:57
  • You are right that we can make our own policy, but there is often a reason why the majority makes the choice they do, and we shouldn't just blithely disregard that. Since you don't think we should immediately add tags for all of the things I mentioned, can you explain what makes the tags you added different? For example, why did you add the [heroic] tag to 2 of the 14 questions returned when searching for the word "heroic" in [magic-the-gathering] questions, and what value does the tag add when every single question there would have the tag? – murgatroid99 Dec 3 '14 at 1:10
  • @murgatroid99 I added two because tags on just one question get auto-destroyed. I did not add it to all fourteen because (a) I was testing the waters, and (b) the front page had just been bombarded by bestow being added to 18 questions, I wasn't about to do that again immediately. I added that tag in particular because we are recently getting a lot of questions on the topic, one of which was active and brand new, and so it adds immediate value. And to "every single question in the search [for the plain keyword] would have that tag", that will naturally be the case for some tags. – doppelgreener Dec 3 '14 at 1:19
  • I understand. In that case, what waters were you testing that were not already probed by the additions of the [bestow] and [protection] tags (the latter of which has been reversed, by the way)? – murgatroid99 Dec 3 '14 at 1:31
  • @murgatroid99 Dude, I added two instances of two tags - protection and heroic. What is the point to this grilling? – doppelgreener Dec 3 '14 at 1:42
  • You suggested creating some of these tags but not others based on whether they would be useful. I was trying to understand why you thought that the tags you did create were useful. Also, regarding your recent edit about other sites' policies, I think you're fighting a straw man. I didn't say that we should choose this policy because other sites do; there's just no good reason to expect that the reasons they had for making that decision are any less valid for us. – murgatroid99 Dec 3 '14 at 2:00
  • @murgatroid99 I don't know those reasons. I understand Stack Overflow's choice at least. The question mentions lessons from Arqade, but doesn't explain what those are. Meanwhile I use a site that has come to different conclusion. And man, all I was trying to do was help the site. That's it. I found those tags useful because we have many questions about it. – doppelgreener Dec 3 '14 at 2:11
  • I'm sorry for responding so harshly. I'm just trying to help the site too. I think this is a bad idea, and a bad precedent to set. Honestly, the ambiguous tag issue isn't even that relevant to our particular situation, so I won't belabor the point. The primary reason for avoiding them, as far as I've seen, is the one you mention in your answer: the tag alone doesn't represent a cohesive topic or area of expertise. The lesson Arqade learned is to avoid that issue by prefixing the tag with the game name, which is why we have [tf2-spy] and [diablo-3-monk]. – murgatroid99 Dec 3 '14 at 2:23
  • So why then is [magic-x] or [mtg-x] a bad idea? I would be happy with those too, honestly. As I've mentioned I'm not satisfied with just searching for the word, it isn't reliable on account of the false positives it turns up. – doppelgreener Dec 3 '14 at 2:26
  • @murgatroid99 Now that I understand the context better, I've heavily revised this answer. – doppelgreener Dec 3 '14 at 2:47
  • Do you think tagging would make search consistent enough to be worth it? We'd have a ton of retagging old questions, and probably plenty on new ones too. And then there's the issue of knowing which mechanics have tag names and which don't... – Cascabel Dec 3 '14 at 15:29
  • 2
    @doppelgreener I guess then it would be helpful to discuss, maybe in another question, what the best tagging scheme is at can find, so that we can judge the idea based on a real notion of how things would look, and how hard it'd be to accomplish. – Cascabel Dec 4 '14 at 1:03
0

(not a proper answer, but to long for a comment)

Normally I would say no to such a proposal, but in this case means to and end reasoning compels me to support such a feature.

I feel that we have too many MTG questions that really should be marked as duplicates of other questions, but because they mention different keywords or cite different specific cards they are often not marked as such despite having the same or nearly identical answers. How many answers do we have to copy the same rules text snippet into before we admit we have a problem?

| |
  • I'd like a better way to deal with this too, but this just seems to be part of resolving MTG questions, unfortunately. We've got 1298 MTG questions, which is a lot, but not that many considering the number of MTG mechanics. A lot of them ask about the same mechanics as others, which means we have to revisit those mechanics over and over, but we haven't been asked before about those particular mechanics in that particular combination in that particular gameplay situation and so on. It's not ideal, and I wonder if there's more we could do about it. – doppelgreener Dec 4 '14 at 22:36
  • 2
    If questions are about different keywords, this proposal won't help find duplicates. Plus, questions are not duplicates just because they have the same answer. – murgatroid99 Dec 4 '14 at 23:05
  • @murgatroid99 for keywords, sure this will not significantly help find dups, but with things like priority, turns phases, and the stack I am confident it will help significantly. – Colin D Dec 4 '14 at 23:37
  • @murgatroid99 for dupe answers, sure it is not always the case. If 2 questions ask the same thing but with different specific context I would make an argument for them being dupes. Maybe this is an argument for editing out specifics that are not crucial to the question (specific creature or ability text for example) and replacing them with generic terms, idk. – Colin D Dec 4 '14 at 23:45
  • In regards to your latest comment, you might want to check out the discussion on this question. – murgatroid99 Dec 4 '14 at 23:52
  • Honestly, I am unconvinced that having more tags will actually make it easier to find duplicates, even if we somehow have every single question tagged correctly. – murgatroid99 Dec 4 '14 at 23:58
0

I think some tags could add value. These would be tags in the form of mtg-something, because they only have meaning for magic.

These tags would be useful mainly to avoid false positives when searching. Yes, it's true that a question about protection will probably say protection in the question. But a lot other questions could say that too, even if they really are not about protection.

These tags would be much more useful for topics other that protection. A question about state based actios, for example, is unlikely to say state based actions in the title or question, because new players don't know these terms, even if the question is about that. And even if the answers has the words state based action, searching for that could result in a lot of false positives, because people could be citing rules that say it even if the question is not really related to that topic.

I remember how difficult have been sometimes to search for rulings in some forums about magic. For example, if I wanted to search for rulings about events and replacement effects for those events, that is really hard. I usually started searching for references to rule numbers, or specific cards that I know could have an interaction, and that would give lots and lots of false positives that I needed to weed out.

| |

You must log in to answer this question.

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