Who owns the word “game”? (A definition of definition)

At the State of Play conference, Eric Zimmerman described MMORPGs as borderline games since they have no (quantifiable) outcomes (you can continue playing). My game definition also describes MMORPGs as borderline games for the very same reason. In his blog, Greg Costikyan disagrees strongly. There are really two issues here: 1) how to define “game”, 2) why is it so important to be a game or not? “Game” is just a word, isn’t it?
Here’s Greg’s conclusion:

[…] from my perspective, if you produce a definition of “game” that excludes things that most people call “games,” either your definition is clearly wrong, or you need to make a strong argument for why the excluded entities aren’t really games. Simply saying “they don’t come to quantifiable outcomes” is circular, since it is saying “This is my definition of games, these don’t fit, they aren’t games, QED”.

Greg is of course right that Zimmerman’s definition is tautological, but all game definitions are: When you attempt to define something like games, your definition will necessarily be designed to match an a priori idea of what should be included or excluded in the definition. Costikyan even points to the fact that his own game definition comes from his own RPG background. So the cornerstone in Costikyan’s argument will be this one:

They certainly fit my definition of “the game,” but it’s clear to me that they should also fit any reasonable definition of “game.”

Which is of course no better than any other argument here: A reasonable definition of games should include MMORGPs because that’s what a reasonable definition of games would include.

So what’s a poor theorist to do? Why, appeal to something external to the definition, of course!

For example: If we had been discussing whether a whale was a fish or not, one side of the argument would state that whales are fish because they swim in the sea; the other side would counter that whales are mammals because they have lungs and give birth rather than lay eggs. The issue is most easily resolved by using evolution as the measure of all things, showing that whales evolutionarily speaking are mammals etc.

But human culture is always stranger than that, so there are a number of things we can appeal to in order to prove that our personal game definition is the right one:

• We can appeal to consensus: Most people will agree that “game” should be defined like I am doing it here.

• We can appeal to consensus of the enlightened: Knowledgeable game scholars agree that games should be defined like I am doing it here.

• We an appeal to empirical factors: Many “game” stores carry RPGs, hence RPGs are games.

Or we can (drum roll) include a sense of history in our definition: I called my game definition “The Classic Game model” because it describes a model for games that has been dominant from 3000 BC to approximately 1970 AD. And during that historical period, nearly all “games” had outcomes. MMORPGs differ from the classic game model in that they do not have outcomes, but they do belong to a broader family of games. This does not give us a simple yes/no result, but it gives us a qualified answer that is open about its premises. So there.


Another thing of interest is that half of the world seems intent on describing their obviously game-like projects as being non-game – “it’s not a game, it’s an interactive narrative”, while the other half of the world is bent on describing their borderline projects as games.

Is the label “game” something to be avoided or something to be sought at all cost? My feeling is that “game” is moving towards hipness – in a few years time, perhaps people will be using the word “game” as indiscriminately as the words “interactive” and “narrative” have been used lately … And we will be victims of our own success.

11 thoughts on “Who owns the word “game”? (A definition of definition)”

  1. Hmm.

    I think Greg’s approach is on the right track for a simple reason–finding these “genetic” similarities between games lets us study and compare things that go together, reflect one another and reach conclusions that would make sense to the average, well, uh, gamer. Otherwise we risk erecting artificial theoretical frameworks to reach equally artificial conclusions. The results of our theory should have some common sensibility in their outcomes, it seems sensible to argue.

    If we sideline some example as borderline (which seems to me to admit that they ought fit the definition, but dangit, they just wont cooperate) then we exclude data from the analysis unnecessarily. I’m with Greg, D&D and The Sims are games whether they have goals or not. Then again, I think a lot of things under the scope of game studies that typically are not. So, I’d more at risk over over-inclusion that the exclusion your definition risks.

  2. half of the world seems intent on describing their obviously game-like projects as being non-game – “it?s not a game, it?s an interactive narrative”
    Is the label “game” something to be avoided or something to be sought at all cost?

    As someone developing what I think you’d call a borderline case — Facade has no explicit rules (you’re not told what you should do or need to do, since there isn’t), but it does have a quantifiable outcome, and valorization of outcomes is certainly possible — I think we used to not call this interactive drama a “game”, but it turns out we tend to describe much of the underlying interactive mechanic as “affinity game” or “therapy game”. You’re correct that “interactive narratives”, particularly ones with non-trivial interactivity (specifically, agency), will often have game-like mechanics under the hood. But since as much of the intent of such experiences is to create a well-formed story (versus the not-well-formed “emergent” narratives of game-games), labelling an interactive narrative a “game” is to be avoided because it’s an incomplete label. And labelling it a “story” is incomplete too. Lately I’ve been advocating that we need a new word for this form, kind of like how the word “movies” was invented for narrative feature films… and no, not “cyberdrama”

  3. I think you were right on track with the appeal to evolution. Why did you stop? A massive game (to borrow a better-than-“MMORPG” term that I first saw in a Penny Arcade newspost) has a pretty blatant and direct lineage from non-massive games.

    Now, here’s an interesting argument for you. Let’s say you play five hours of the new Prince of Persia. (That’s about halfway through it.) Now, at the moment — if we draw only on my immediate experience of the game itself — it’s quite possible that the game will have no victory condition. Perhaps it will suddenly open out into a massive, open universe full of walls to bounce off of, trapeezes to swing on, and so on — with no further explicit goals and no “ending” condition. (Of course, this is not what the game thus far has led me to expect — but, from my limited perspective, it’s still possible.)

    Now, let’s say I choose never to go further in Prince of Persia. At this point I can’t tell whether or not it has an ending condition. If I use the “must have an ending” definition of “game,” how do I know for sure whether PoP is in fact a “game”? And what am I to do with the fact that all my instincts tell me: *obviously* it is a game! The experience so far is enough to convince me that it is.

    Or! Here’s another one. Let’s say I’m playing .Hack — a single-player game set in a “virtual virtual world”, an entirely offline simulation of a massive game.

    Now, hook it up to Xbox live, and let other real-world people chat with in place of the fake-PC NPC’s in .Hack. There should now be essentially no way for me to tell the difference between my experience playing .Hack, and a real massive. There’s no way for me to know that .Hack does in fact have a victory condiiton. Given that the two experiences are effectively identical, why should only one be defined as a “game” simply because it stops?

  4. Interesting point, Jay. Reminds me of Edward de Bono’s L game, where it was possible for a game to go on indefinately, even though a finished condition was possible. Hell, even Tony Hawks Pro Skater 3+ has the possibility of going on forever, if you’re good enough to keep a combo rolling over the time limit.

    Personally, I’m quite fearful of trying to nail a game’s definition down too precisely – much like general Scientific theory, you can never know that your theory won’t be undermined by something unclassifyable, tomorrow. That’s not a reason to not try to, of course, but you see, from Jesper’s original post how much shit flinging goes on as a result of someone’s preferred game type not being included!

    Instead of such properties as “finish states” and “optional safety”, I like to focus on a slightly more pragmatic approach. All this does is to express the game in terms of a human interface device/system, and note the common facets of all games. It may not give a nice wordy definition of games, but it has certainly helped me to think about the common traits of games in a more structured way, and this has led to a very nice modular approach to writing game mechanic design documents, i.e. a common way to define game states and rules, ones’ interactions with them, and feedback from them. Here’s the first draft (although you may have seen it before, but it seemed somewhat fitting): http://www.antifactory.org/archives/000027.html

    I think most people’s problem with it (apart from being badly written :p) is that calling a Game a system is too simplistic a definition – at that point, you can call “Life, The Universe And Everything” a game, and obviously, that’s a metaphysical bridge too far for some people. I still believe it gives some strong pointers to the common elements of the systemic elements of the game as a system, even if it’s basically the same avenue as Human Computer Interaction research.

    Oh, and talking about metaphysical cans of worms, here’s a new article you probably ain’t seen yet, “Why Games Aren’t Interactive”: http://www.antifactory.org/archives/000031.html

  5. Jesper: If your argument is that “games” are those that derive via a definable historical thread from things called “games” in prior history, then MMGs are indeed games. MMGs evolved from MUDs; MUDs were attempts to combine text adventures with tabletop RPGs; I believe you accept a text adventure as a “game”; and tabletop RPGs derive from miniatures gaming (Arneson developed D&D, originally, as a Chainmail variant).

  6. Greg, the point was that there is a “standard game model” that has been dominant historically, and that MMORPGs deviate from that model in that they don’t have final quantifiable outcomes. So MMORPGs do not match the classic way of making games.
    MMROPGs are different from other “games”, but they also share many features with other “games” and are clearly historically related to other games.
    Whether we then want to call MMORPGs “games” or “borderline cases” is in a way of minor importance – it is exclusively a battle of words. But if we can agree that everybody should be clear about what they mean by “games” and why they are choosing to define it in a specific way, I am happy.

  7. Hi guys
    To my mind the definition of ‘game’ (as any definition) should be based on the pragmatics of the subject – how and why do we use it? The answers to these questions can, as hinted in the debate, be found through the perspective of evolutionary psychology. In this perspective games are devices that allow us to practice behaviour and strategies, without risking the otherwise potentially harmful consequences (in this sense games are very much productive, as they allow for the accumulation of experience). As a result games can be said to exist of a framed conflict, in which the player(s) participates. The frame consists of resources to be manipulated and the rules for this manipulation, in order to achieve a certain goal. Simple as that. Whether the game ever ends (by achieving the goal) or not does not matter. There should be a defined goal, but that does not necessarily mean that the goal is ever reached. Without a defined overall goal, as in some MMORPGs, it is not a game but a simulation. But notice that simulations often contain lots of small goals to strive for, as this activity forms the basis of much of our understanding of the dynamics of the world. This half-state of gameness (refered to as borderline games) might be part of the reason for the controversy around this subject.
    Regarding the computer’s ‘affinity’ for mediating games, this comes about because the computer is nothing less than a revolution in game-mediation ? for the first time it is possible to combine real-time rule-governed symbol manipulation with a complex and sophisticated narrative frame in which to participate. The novel can mediate a deep narrative frame because it is pre-designed; the game cannot because it has to allow for instant changes, which puts restrictions on the scope of the resources being manipulated.
    As for the player’s reaction (attachment of the player to the outcome), this does not happen by convention or by being agreed upon. It happens automatically as soon as we invest ourselves emotionally in the game (as in anything), as a consequence of goals being reached or lost. That games without real-world consequences can evoke emotion at all shows our compulsion for make-believe, which is the foundation for emotional reactions to fictional events in general.

  8. MMORPGs are not so unique in this respect if you think of sports. Single tennis and boxing matches and tournaments have very goal or victory condition or whatever you want to call it. Now, they also have rankings (ATP in tennis, several at boxing), which get updated forever, and there’s no victory condition there. Sure, you can be the best at end of the year, or be the one with most weeks on #1 position, but “you can continue playing”, and soon someone passes you.

    I’d equate Ultima V, tennis tournament and your usual time-limited quest in MMORPG as rule-bound… games(?) with goal / victory condition, and then equate ATP scoring system and MMORPG system as gaming milieus/environments (my poor english fails here), INSIDE which gaming takes place.

    I’m not a theorist really, so I’ll let you do the thinking, but answer this: currently, NBA basketball league has annual play-offs and finals as “goals”, but what would NBA be, if – instead of this seasonal cycle – the teams would play only regular matches, every week, around the year and ad infinitum? Would it be quite similar goal-wise to most MMORPGs?

  9. I’m tempted to agree with Greg solely to avoid entagling the meaning of “game” in a culturally founded definition. I’d wager that you don’t need to rely on cultural usage to put forth a strong definition. Rather, we can work out a robust structure for what is an is not a “game” and then employ it under the term game. I’d also have to argue that we would rather have a definition that based an existent objective system rather than a sysnthesis of historical usage. or perhaps I’m simply babbling

    My two cents

  10. How would you make a definition of a cultural thing like games without referring to cultural usage?
    Of course you can make your own definition of anything you want without referring to anybody else … but does it make sense?

  11. If we agree that games are a kind of art, then aren’t we just trying to define a section of art? Or am I being stupid here?

    I guess I’m resigned to the fact the “game” is one of those incredibly loaded terms. Pub-philosophers say “life’s a game” while soccer moms say “life isn’t a game”. I can appreciate the need to define games in terms of their cultural relevances, but as far as using the word “game” in a fixed and well defined context within ludological discussions, I don’t hold up much hope. Instead, I say, harness the power of the adjective, which will help you express exactly what aspect of “game” you mean.

    Then again, I’m forever missing the point…

Comments are closed.