6

Well this is actually a continuation of a rant I had back in July, but the content here is a little different and more specific, so please disregard the previous thread :)

Many tags currently have different meaning in different games. I really think we should try and avoid using tags like these.

There are tons of them. For example (and I've only used tags with more than one appearance):

  • Profession names, such as or or etc. are used in gazillion games and can refer to completely different things in each one.
  • Likewise, names of buildings or places such as or .
  • Other concepts that mean different things in different games like or or .

So what's my rationale? I'm going to basically copy-paste from my previous post:

Some users assert that tags are helpful for searching, and I believe this is just not their intended purpose: tags are helpful for filtering (and for gathering statistics). There are some discussions about tagging in meta that seem to agree. I really liked one particular sentence:

Tags connect experts with questions they will be able to answer

In Stack Overflow, one can be an expert in Java or experienced with strings. But there's no way to be an expert on "priest" (since they appear in so many games) and I doubt anyone would consider herself an expert on "energy", let alone filter on it.

On the other hand, of course, it could be asked what's the big deal, what harm is there in a few excess tags? Well I confess it isn't a big deal, but it's a fact SO has over time increased its tag-creation threshold. The more tags there are the more noise and needless complexity we add, in my opinion.

  • I see what you did there, link to a post, tell us to disregard it and then quote it! Your confusion roll is successful, good sir. – badp Jan 25 '11 at 23:20
  • In my experience, most WoW questions that ask, say about "rogues" or "Paladins" make that fact obvious from the title: "What are the best food buffs for Rogues?" "How do I tank as a Paladin?" Hasn't the idea always been include all salient points in either the title or the tags, but above all be clear? – Raven Dreamer Jan 26 '11 at 7:04
3

I don't see a big problem with those tags, as the only use for them I can think of is searching within a game-tag. So if I'm searching for WoW rogues, I'll use and .

I don't much like the fact that the tags have different meanings in different games, but I'm pretty sure noone uses those tags (on their own) for browsing or favorites anyway.

If we argue that the tags are not for assisting searches and that real tags would have to be useful for filtering/favoriting on their own, I think we would have to remove 90% of the non-game tags here. I don't think that makes much sense.

I'm only using the game tags, I could see a use for genre tags if we applied them consistently, but personally I find no use for any other tag. I think they might be more useful when we have a bigger volume of questions, but we really have to get some kind of order in our tags for that.

I think there is really not a substantial difference between game-specific tags like , semi-general tags (same name but different implementations) like and general tags like or [talk:tips] in how they are used. I can't imagine a case where I would filter only on on of those tags, I would only use them to search within a game, not across multiple games.

For that reason I think it is perfectly acceptable that some tags have different meanings in different games. They might not be useful on their own, but nobody will use them like that anyway, and if you use them together with a game tag they are again useful.

But there is some need for cleanup in our tags for sure, but I don't think this is an important criteria for deciding about the usefulness of a tag.

  • 1
    "I'm pretty sure no one uses those tags for browsing or favorites anyway" - that's precisely my point. "If we argue that the tags are not for assisting searches ... I think we would have to remove 90% of the non-game tags here." - I'm not so sure. Some tags describe genres, some tags describe the nature of the question (tips / strategy / plot / troubleshooting etc.), and some tags just mean the same thing across many different games (achievements / bugs / secret) - I have no problem with any of those. I agree it does mean avoiding the use of many of the existing tags, though. – Oak Jan 26 '11 at 10:01
  • @Oak I've (hopefully) clarified my point somewhat. – Mad Scientist Jan 26 '11 at 10:57
  • Regarding your observation that you can't image a case when someone would filter only on these tags - me neither. The official SO stance is that tags need to be able to stand (or more less) on their own. – Oak Jan 26 '11 at 13:57
  • @Oak The tags we are discussing are not really meta tags like mentioned in the blog entry. They just have different meanings in different games. I don't think this blog entry really applies to this situation. – Mad Scientist Jan 26 '11 at 14:50
  • @Fabian to be honest I don't think it applies either :) I definitely wouldn't call them meta tags, but I do think it carries some resemblance. Also to stress my point I do not propose burning all these tags down, just avoiding creating new ones like these. – Oak Jan 26 '11 at 14:53
6

There's two different classes of these tags that we deal with. I believe we should pay heed to this, because I don't think it is remotely wise to avoid a tag just because its meaning varies between games. Take nothing of this to address fully game-specific tags where the tag represents something that's undeniably tied to a single game or franchise.

  1. Semi-dependents, or good generics. A good example we have is , and one that I'd like to become a similarly useful example would be . You bring up the "Death of Meta tags" article. Semi-dependent tags will have different meanings when associated with different games, but will still have a mostly universal meaning to gamers when it is alone - even if it will never get a question that it is the only tag for. I use [strings] on Stack Overflow as an example all the time because it's the same principle. The implementation differs between languages, but the core essence of what it means will always be understood by the target audience - no one is going to confuse these for quartet musicians. I address a lot of this in a different Meta question. This is a little different, though, so you can disregard that thread and just bother with the following quote. ♪

    The plague of duplicate questions that differ simply by choice of words is testament to this foul shortcoming of text search: you simply cannot find what isn't there. This same shortcoming applies to tags, but the thing here is that tags can represent the specific as well as the general. So you can ask a question about weapons without ever saying "weapons". In a multi-question theoretical example, suppose we got questions about the game Sora, one question about the Pilebunker, one about the Flamethrower, and one about the Bullet. They really don't need to share any words besides the name of the game and maybe "damage" in their question bodies, but the presence of a unifying tag lets me group these together in a way that they should since they're all about the same content: weapons. You can accomplish this categorization without needing to alter the word choice of the individual questions, because tags are independent of the author's expression of the problem.

  2. Dependents, or bad generics. is actually a pretty good example. In StarCraft and its successor, a Bunker is a building that Terran Infantry can utilize to stay safe from damage. In Resonance of Fate, a bunker is a wall that will block projectiles from a long range but can be fired through when adjacent to it. Though they share the term because of a matching dictionary definition, this knowledge will not help identify what the term means if you are unfamiliar with the game. This is because these don't exist as independent video game constructs, and fully depend on the context of the game to define it within the scope of what we care about.

A good acid test is to think of the tag and consequently how likely it is that you can pinpoint very closely what it'll be in any given applicable game whether you know it or not with its universal meaning. Or, if you can write a tag wiki description that isn't just "This means things so completely different between so many games", that helps.

  • would fail because Team Fortress 2, Disgaea, and StarCraft all have very divergent meanings that extend well past the universal meaning, just to name a few.
  • would succeed as the difference in implementation between the oh-so-many games with minimaps doesn't change its universal meaning to gamers as a small and dynamic map of the current area.
  • We shouldn't even begin to describe how bad would fail.
  • There are many different implementations of in games, but they all surround a core concept that is well understood by gamers as some sort of objective that must be cleared for some manner of reward. Terribly vague when it's alone, but successful semi-dependent tags tend to be like that.

Using this, you can help sift out the junk dependents while keeping the useful semi-dependents. I'm still one of the advocates of actually using tags in searches (as highlighted in the Meta thread I told you disregard), but my advice here is actually geared towards the utility of tagging for categorization and filtering.

  • What about the tags warrior, mage, paladin, rogue, ...? I would consider those useful, but the implementation in different games vary a lot. – Mad Scientist Jan 26 '11 at 15:11
  • @Fabian Rogue happens to be the name of a game. Mage, if I'm not mistaken, is also such, but that might just be a P&P game. – Grace Note Jan 26 '11 at 15:15
  • Paladin is its own special case, which I did argue for in the past but I'm not as comfortable with anymore. As far as universal meanings for gaming, all a Paladin means is a holy knight. In games like Ogre Battle or World of Warcraft, it's a job class. But it's also not uncommon to see its use outside of job class. For example, in one game there's a militant order known as the Paladins (I think it was Starsiege), and a Paladin is also a high rank in Brotherhood of Steel from Fallout. To me, the lack of unifying video game construct weakens its utility for categorization. – Grace Note Jan 26 '11 at 15:27
  • In the end, it's tricky. Games like Ogre Battle and Etrian Odyssey, we really don't need as much of a deep need for class-specific categorization of questions. But for something as deep as World of Warcraft, that kind of categorization can really change things. It's something to be careful of, though, because practices that promote bad dependent tags just leave windows open as examples for more bad dependent tags to be created. – Grace Note Jan 26 '11 at 15:32
  • 1
    It's been worse in the past, but let's take a gander through the recent list of new tags and pick out some of the horrors. We have attachments, assassins, anti-tank, and spotting just to name 4. – Grace Note Jan 26 '11 at 15:36
  • So if we want to start elimintating bad tags, we (as non-moderators) should just remove the offending tag from all questions and deletion then happens by magic, right? – Mad Scientist Jan 26 '11 at 16:32
  • @Fabian Actually, that's all moderators can do, too, hehe~ By the by, I'm not saying that your example tags are as bad as these (except that they do match assassins). It's not exactly a clear cut solution that we have yet, but there's some methods we need to dwell on for keeping useful semi-dependents while discouraging the dependents. The latter bullet points from Oak's are far more egregious and less useful, but job classes sit in an unfortunately poor position that leans too much towards dependents to be honest. – Grace Note Jan 26 '11 at 16:35

You must log in to answer this question.

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