What computer games can and can't do
Presented at the Digital Arts and Culture conference in Bergen, August 2nd-4th 2000.
The title of this paper is the grand "what computer games can and can't do". While I am unfortunately unable to give you a full answer to the question in my title, it is meant more literally than you may think. At the same time, the preparation of this presentation caused me to doubt several things that I hadn't really suspected I would doubt. I'd like to share those doubts with you.
This paper has three main points:
Where we are
First, I should like to describe were we are, so to speak. The computer game is somewhere between 39 and 42 years old. Movies, by comparison, are 100 years old. So computer games are younger than movies, but not that much younger.
So the computer game has a fair amount of history. While I do not think the computer game has had it's first Shakespeare - in itself a bad comparison since computer games are mostly built by groups - I do think it has had it's first, say, Dashiel Hammett. To think in music, computer games probably haven't had their first Bach, but probably the first Robert Johnson. As a phenomenon, it is probably the most mature of all the things we discuss in a conference like this. I consider it more mature than hypertext fiction, much more elaborated and differentiated than, say web art, or especially virtual reality.
But we need a separate theory of games. We need a theory that isn't just interactive bits and pieces tacked on to narratology or dramaturgy. We lack a theoretical understanding of what games are and can, and how they relate to the narrative media such as the novel or the movie. We lack the tools to evaluate and place a computer game both historically and in relation to other games.
Computer games did not appear in a void, rather computer games is a form that borrows from something else; games. This poses an interesting situation. Presented with electronic texts or hypertext fictions, the literary theorist can look to his/her standard vocabulary and reappend it to the new phenomena. And thus, presented with a computer game, it should be possible to simply look at the aestethic vocabulary we were already using for non-electronic games. However, it doesn't exist.
But this doesn't mean that there is no discourse on games, because there is. Game developers have their books, web sites, etc... where the aestethic questions are discussed; balancing weapons, graphics, backstory, replay value, level design etc... The way we talk in a place like this is not supposed to be a DIY manual, but it is in many ways useful to consider the actual problems and discussion that you face when creating a game. The academic discourse and the discourse of the game development community have to be more closely related.
I think it is safe to say that the humanities have completely ignored games and focused on, privileged if you will, narratives and fixed sequences. Games have been ignored, to a large extent, I think, because they are sending the wrong signals. To an educated person, literally alien signals of low culture, fun and insignificance.
Frankly, chopping of a monster's head doesn't quite imply the same kind of class that you get by passing subtle references to Rilke. And consequently, one of the recurring ideas for the past 15-20 years has been to create computer games with deeper themes. Most everybody, including people like Brenda Laurel, Janet Murray, several game designers I've worked with, Chris Crawford, myself, most everybody longs for computer games that are not just about monsters, but carry the themes from stories. Love, ambition, intrigue and so on.
There are two traditional ways of trying to make more meaningful games. One is the "interactive movie", where you are presented with simple choices at the end of a video clip. The other is the adventure-game style (as in Myst), where you play a puzzle game, but get rewarded pieces of a narrative as you solve the puzzles.
A new utopia
But neither of this is very satisfying to the computer game connoisseur. A simply doesn't work, and B is quite uninteresting in replay value. We need something more dynamic, something happening, replayable. I'd like to present to you with a new utopia to that end. This is a 4-point program for the creation of a meaningful computer game that is also fun to play:
This is largely based on some recurring problems in interactive fiction and some theoretical observations.
One of the traits of narration is that narration is about something that happened at some other time. This is the whole story/discourse dichotomy, which is something you want to avoid in any real-time interactive product. (There is no way the user can interact with your story while it is being narrated.) 3 and 4 are related, but 3 is one of the problems from point & click adventures like Myst - that for no apparent reason some things can be manipulated and some can't. And finally, 4, that the game must be sufficiently flexible to actually allow the flow, to develop freely.
Unfortunately, during the preparation of this paper, I had a kind of crisis, where I suddenly realised that I did not know why this should be done. I don't see any really compelling aestethic, or for that matter, commercial, reason for working towards this utopia. I think that computer games today are really good. And in the end, I came to the simple Edmund Hillary-inspired conclusion: Because it's there. Or rather, because it isn't there. Yet.
But to get there, it would be nice to have a clearer idea of what a game is. My background is from literature, but when I have created games commercially, it was very obvious that we were not discussing plot, character, narrators and so on. We were rather discussing interface, level design, gameplay, play mechanics. Because these are the kind of issues you have to think about.
What is a game
What is a game? This is a really good question. I'll risk my neck by coming up with a simple two-point definition.
You may begin to wonder how children's games fit in, the answer is that they don't. Because I'd like to make a distinction here, one that I think Chris Crawford at least hints at in his book on computer game design.
It is possible to make a distinction between formal games - rule based games, and children's games which are more oriented to pretending to be something else. When I talk about games here, I'm strictly referring to formal games, as computer games fall in this category.
I would have preferred to make this presentation in a Scandinavian language. In Danish, Swedish and Norwegian, we have a unique distinction between "spil" and "leg". The Roman historian Tacitus describes the Germanic people as being obsessed with games, and this is perhaps the reason why we have this clear distinction.
This definition explain the difference between games and laws of traffic (traffic is not assigned another status). Between formal games and play (play is not based on formal and pre-defined rules, but their rules are rather under constant negotiation), between games and narratives (narratives do not evaluate the reader, and are hardly formally defined).
And this is the important thing. Games have rules that must be sufficiently well-defined that you do not argue about them every time you play (in a board or card game), or sufficiently clear that they can be implemented in a computer program.
The upshot of this is that games are pretty unique compared to the other cultural things we study. Even though the structuralists tried to describe many things as being based on very simple underlying structures, we know now that it probably isn't true. And I do think that computers has helped us realise this by showing how hard it is simulate and formalise stories.
This means that games belong to a formal/algorithmic domain, whereas stories belong in the interpretative domain. Games have to have formally defined rules to be games, and stories, being based on interpretation, are not formally defined.
The definitive proof of this is the fact you can create a world-beating computer chess program. And this has already been done. But you cannot create a world-beating computer program that writes stories. Even the most advanced story-writing programs - which would probably be Selmer Bringsjords BRUTUS right now, are no match for human authors.
Another way to say the same is that reading or writing a story requires a large amount of contextual knowledge. And, probably, a body of some kind, whereas the gameplay of a game is in itself a formal construct, and can thus be played by an algorithm.
Humans, mammals and computers
I'd then like to step back a bit and think this in a larger perspective. As I said, the definitive proof that games belong to some other category than literature, movies, cooking, theatre, ballet or whatever is that games can be played not only on, but by computers. I made a table and it looks like this:
Computers only handle formal games, animals only handle children's games.
I should note that some formal games based on imperfect knowledge, such as card games and especially Poker are formal in their structure, but actually playing revolves a lot around interpreting the signals of the other players. I.e. only humans can do it. Additionally, quiz games require world knowledge and natural language capabilities.
At the same time, hypertext fiction could be described as something of a hybrid since it borrows froms narrative, but also from the formal structure of games in regard to how you can traverse the text, but without becoming games - there are no criteria for success. This is perhaps akin to some simulation games such as the Sims or Sim City. You have formal rules for development, which you can then use for your own ends.
I should also note that few games are completely abstract, but that games often have themes that wear off in use. A text from 625, possibly the earliest written reference to chess, accounts for the peaceful reign of the Indian ruler King Sriharsha, and says that in his peaceful time "only from the gaming-board can one learn how to draw up an army". We don't quite use chess like that anymore.
Resistance to meaning
This difference between games and stories show up in the themes that are handled by each. A brief, informal list could be this one:
For example, almost all computer games are focused on navigating in space, while in stories, space is only included as far as it is interesting, or means something within the story.
And take a thing like random deaths. Stories like to focus on things beyond our control, ascribing meaning to things like disease and death. (Like the movie Love Story.) Games are all about getting better. A bad game is one where you die without being able to prevent it.
For a game to be a game, you have to have a definition of rules and of progression. Definitions that in board games must be at least sufficiently well-defined as not be something you argue about every time you play. In computer games, they must be possible to implement by programming.
From this follows that games resist being assigned the themes we love from stories; life, death, ambition, language, because all of these resist being reduced to formula.
This is why it is so hard to create the utopian game I described above.
This is in fact the beginning of my Ph.D. project. And one of the things I plan to work on is a general model of games as such. I.e. I want to create a theoretical framework that can tell us A) what is a game, B) what is not a game, C) give us the terminology and distinctions needed to describe historical developments in games.
I think I've at least attempted do define games in relation to other phenomena.
I said that we should try to learn from the game development community. So if we try to pick up terms like gameplay, interface design and level design, our perspective changes a lot.
In the newer literature on game development, there are two main points that are repeated over and over.
1) Before you create games, play a lot of games.
2) The second recurring point is this: As a game developer, never forget that your single most important objective is to create games that are fun. This should be obvious, but too many people think in cut-scenes, backstories, graphics and ignore the fun. There are some great phrases for describing this problem. In games where the seemingly only point of your playing is to do the right thing to advance a very linear plot, one may say that "The game designer is having all the fun". If you play an incredibly advanced simulation, with thousands of parameters where nevertheless, what you do doesnt really make much of a difference, one can say that "The computer is having all the fun". Fun is also important in regards to realism. In modern 3d shooters, you can, for some reason, jump and change directions in mid-air. This is of course impossible in the real world. But as a game designer, the only important question is: "Is this more fun?". So you can change directions in mid-air. This works, and its fun.
Ive talked a lot about what is "formal". Games have a formal part, but they have to be fun. Fun is not formally defined, and what this means is that the creating of the formal structure that is a game, is not in itself a formal process. Rather, it involves aestethic choices, intuitions and so on, because everything is relevant to the experience of the player. Creating a game involves all senses.
But gameplay is the central thing that makes or breaks a game. Marc Saltzman:
All the glitz and glitter poured into games these days, such as expensive art, animation, real actors, or the best musicians, cannot cover up for poor gameplay.
A tentative historical view of computer games
To further pick up on what we can do by not starting from a traditional theory like narratology, I'll sketch a possible answer to a previously asked question: How are computer games different from non-computer games?
If you think of computer games historically, they can be described as introducing four main characteristics compared to the non-computerised game. Time, automation/complexity, replay, and levels. The first two are pretty obvious, and more or less quantitative. The last two are more subtle.
Time: In regards to previous single player pastimes, the big change with computer games is that the computer can keep pace. You do get pace in physical games, but this must be enforced by gravity, inertia of a ball and so on. With computer games, you can create real-time games without having to rely on laws of physics. (As in the case of being able to turn in mid-air.) This opens up new possibilities; action games, speed, and is basically what makes the first person shooter possible.
Automation/complexity: Computer games also automate the rules of games. And this means that you can create much more complex games. You can have real time games with thousands of units being moved on each side. This leads to the real time strategy genre.
Replay: The option of returning to the same game, both in the sense that you can go back to exactly the same challenge, play a level of the same game again, and in that you sometimes can save your position. You can play level 3 of Quake II or all the old arcade games again, being presented with an identical challenge. This is not prevalent in other games. The single-player card games that you can replay have a large amount of chance in them. Some puzzles possibly share this trait.
Levels. The early games like Spacewar or the later Pong were conceptually really close to a lot of mechanical entertainments already existing at fun fairs etc.. But I think the video game revolution really starts with Space Invaders, where suddenly you have a one-person quest; you can get better & defeat the aliens, you can move to new levels. This does not correspond to previous single player games. In the earliest forms, the gameplay simply got harder at each level, later games like PacMan introduced new colours for each level, later on came new graphics and level designs.
As a fifth point I might add narration. Narration does occur in some board games like Monopoly, where you pick up a card like "Pay taxes", "You have won the beauty pageant" and so on. But it occurs much more frequently in computer games, mostly in between levels.
What I am proposing is that in the grand scale of things, the most conceptually radical thing introduced by the computer game is introduced in the single player game. Replaying of exactly the same thing; level progressions in the game.
It was recently remarked to me that the popular FPS games today such as Quake have a "sport"-quality. I'm not sure how to define sport, but I think they are pretty abstract physical games. The rise of the multi player computer game, that we today consider "the new thing", means that the computer game to some extent is going full circle and partly returning to its roots in the non-electronic by dropping replay and level progression. You cant replay an identical multi player game, as your opponents will react differently. And level progression is not a central part of the popular multi player games today.
In closing: What ludology?
If you look at literature, we may be talking of many different things at different times, but we do have a terminology of story/discourse, of time, of narrators and so on. Something interesting & probably true has been said of literature, and having such a background to work from is incredibly useful. Lets do the same thing for games.
The only question to me is what form the theory must take.
We need a ludology, but what ludology?
Chris Crawford: The Art of Computer Game Design. 1982.
Gonzalo Frasca: LUDOLOGY MEETS NARRATOLOGY: Similitude and differences between
(video)games and narrative. 1999
David Parlett: The Oxford History of Board Games. Oxford University Press 1999.
Marc Saltzman (ed.): Game Design. Secrets of the Sages. BradyGames 1999.
Tacitus: Germania. Translated by Thomas Gordon.
What is marvellous, playing at dice is one of their most serious employments; and even sober, they are gamesters: nay, so desperately do they venture upon the chance of winning or losing, that when their whole substance is played away, they stake their liberty and their persons upon one and the last throw. The loser goes calmly into voluntary bondage. However younger he be, however stronger, he tamely suffers himself to be bound and sold by the winner.
Acknowledgements & Notes
Note updated May 10th 2003: I originally attributed the term ludology to Gonzalo Frasca, but Lars Konzack has later pointed out that he had told me of the term at approximately the same time that I read Frasca's article (1999). Searches on USENET puts the first usage at 1996, and the earliest reference I have found is in a 1982 article by Mihaly Csikszentmihalyi: "DOES BEING HUMAN MATTER - ON SOME INTERPRETIVE PROBLEMS OF COMPARATIVE LUDOLOGY". Behavioral and Brain Sciences, Volume 5, nr. 1. 1982.
Thanks to Mikkel Cauchi and Peter Lyster, students at the IT University, for discussions on sports, and the suggestion of replay as being unique to the single player game.
After my paper presentation, one of the things discussed was the question of
games and narrativity.