Here is the Gamasutra summary of my keynote at the Serious Games Summit at GDC today.
Regarding writing, I will put up the article version of the talk soon.
Skip to content
My name is Jesper Juul, and I am a Ludologist [researcher of the design, meaning, culture, and politics of games]. This is my blog on game research and other important things.
Hmm… Not sure how accurate the Gamasutra precis of your talk is, of course… But if you’re actually claiming that “games without goals” are a good idea, I think you’re offbase.
The Sims has a goal: satisfy your Sim’s needs, and get better stuff for him/her so you can satisfy them better. And a variety of subordinate user-selectable goals: get laid, raise a family, improve your job, build a better house, etc. Lots of these, and you can choose which–but as I’ve argued in regards to Sim City, part of the reason the game works is that even though there is not a single over-arching goal, once you get beyond a certain point, there’s support for a wide variety of user-selectable goals.
Tabletop RPGs are similar; while they share the level (or power-) raising goal with MMOs, beyond a certain point most players are interested in other, user-selectable goals, rather than an increase in pure power. But unless the GM provides them with either one or a variety of potential goals, they will lose interest.
You talk about the flexibility of GTA, but please note that GTA rewards you for pursuing goals other than fulfilling the next mission–e.g., awarding points for stunts. You can choose to avoid the mission structure and instead go after rampages, say–but the system itself supports and rewards a decision to go after one of these alternative goals.
Games are, fundamentally, goal-directed interaction. That doesn’t mean you need to have a single win-condition, or penalize behavior that isn’t directed toward a goal. But ultimately, if there’s nothing that rewards me for doing one thing over or another, or cultivating certain skills, then there’s no point in continuing, and you get something that is, well, pointless.
It’s worth suggesting that designers consider a more flexible and open-ended set of potential goals–but goals remain essential to good game design.
At the rant I asked if “interactive storytelling” was nessecary, and if “drama games” weren’t a better term for interactive spaces about characters and social dynamics ect. Jonathan Blow countered that a game requires a goal and that interactive storyworlds didn’t have that. But as Greg points out, a non-explicit goal induced by the player (rather than an explicit goal deduced, like a puzzle or linear level) are goals non-theless. Serious Simulations and Storyworlds are still games in this sense, though, as Jesper, points out, a new kind of game.
Maybe Chris’ grotesque analogy at the rant, that games are a braindead patient and a new baby is being born down the hall, would be better framed with the braindead patient being a goal driven slave to a machine outside of her control, and that the new baby will enjoy the future freedom to choose her own goals. That, I think, is where higher agency and interactivity lies, and thats why everything I design from here on will allow players to find their own goals in a more playful, non-linear space.
I like the idea of implicit goals, made possible by the game system but created and upheld on the player’s mind – a little bit like Juul’s example at the beginning of Half-Real (the computer racing game lacking programed rules).
That is why I beg to disagree with Greg’s skeptical view on Elektroplancton, which it’s like playing a musical instrument. The reward can be achieving a virtuoso performance or just making some nice sounds (or even just watching the planctons’ behaviour) – it depends on the mood and dedication of the player.
At GDC Juul presented a good example of goal-less play. He talked about “just wanting to see what happened” if he made his Sim eat ridiculous quantities of food. I believe Juul said the oven caught on fire and his Sim burned to death. In many ways, this example is anti-thetical to what Greg states as the goal of “meeting your Sims needs.” Indeed, Juul purposely failed any obvious goals but still received an entertaining outcome.
Juul generalized his ideas by thinking about games as language: They have words, syntax, and semantics which translate into objects, actions between objects, and meaning of objects. A game with a goal is like trying to express a particular idea — everyone ends up coming up with the same basic sentence. But remove the goal and people are free to make whatever whacky sentences they want.
For example, if I give you the words “cat” “hat” “ate” and “sat” and I tell you, “To win, write a sentence about a cat eating a hat” you’re going to write: “The cat ate the hat.” But if I tell you, “Write any sentence you want” you may write, “The cat sat on the hat” or maybe even “The hat ate the cat!” Or you may write the original sentence, “The cat ate the hat.” Given freedom, you can play the original goaled game if you want or you can do so much more!
To address Juul’s challenge of creating a game that makes the player cry because of the actual game rather than through the expressiveness of the gamers themselves, I think you need two elements: 1) Removing goals to minimize the “Rules of Irrelevance” 2) Emotional semantics. A really simple example of this is to have the following game: Press A to murder 100 random people and cure AIDS. Otherwise, Press B. Obviously, if this game had goals, it would be pointless — if your goal was to cure AIDS, then the choice is automatic. If your goal is to NOT directly kill people, then the choice is also obvious. But with no goal and strong semantics the game becomes heartwrenchingly difficult. (Perhaps a better example of this would be a Sophie’s Choice type decision — you can choose to save one of your children at the cost of losing the other. Which child do you save?)
Remove goals, choose powerful semantics, provide ambiguous outcomes to prevent the player from thinking they’ve won, and voila, you’ve got emotional interactive gameplay.
Heres a ripe bit of theory, written up in fairly accesible prose, that I mentioned last Wednesday: http://kingludic.blogspot.com/2006/03/ontological-numbering.html
I think the distinction I make between open-but-bounded and closed structures correlates in an orthogonal but not direct way with the difference between open-goals and explicit ones.
Greg, I am not quite sure how much we disagree.
But: For the Sims 2, I don’t think that Sims 2 has any explicit goals of making your sims happy (at least I don’t see that anywhere), but it’s hard not to try a bit. Yet you can also do all sorts of interesting things involving killing them, scaring them, and so on. The box encourages you to “realize their fears”, not just to “make them happy”.
So the point of the talk was really it can be a good idea to consider whether a game should have goals (as in “this is what you should do”), and that games without goals can be boring if the player does not care about the difference between choice A and B, that GTA allows players to do lots of interesting stuff without following the goal (i.e. it’s part of the design), so if you make a game without a goal, make sure it’s expressive, or at least make sure that player has something to do that they will care about.
Sim’s
sums
isms
skims
Edit…
Ignore all
Add to dictionary