Game worlds are a subset of all kinds of worlds, defined by a specific quality that other worlds lack. There is a problem to be solved, an area to be explored, a reward to be earned or a contest to be won. There is a kind of pressure when playing in a game, and a sense of risk. There is change, death, rebirth and a state of flux.
So in a sense the difference between a game world and a virtual world is one of motion. Game worlds are built for movement.
What ‘World’ Means
Let’s be clear on what the word world means (in this context at least). A world (or reality, ideality, possibility space, whatever your preferred term happens to be) is an environment bounded by rules in which users can take actions. That may apply in the real world (where we are bounded by physics), to a culture, to an industry, etc.
For practical purposes the worlds that we encounter in daily life are infinite, but the worlds that we encounter in games are finite. Game worlds in particular restrict user actions. They deliberately enclose the player and limit what he is allowed to do because those limits also give the player levers to win, and it is within the balance between actions and rules (i.e. loops and the game dynamics that result) that success in the game becomes meaningful. Football with no rules is a much less interesting game than with.
Motion
The upshot of being restrictive is that the player has significant ability to cause change. From his perspective he can alter the world, but the world also reacts to his actions. As he traverses through a first person shooter, for example, his position in the world changes and the levels that he is permitted to play in also change. He acquires new weapons that extend his doll in new directions and enable him to combat enemies that he could not previously have defeated. His agency matters.
Hand in hand with his agency is the world’s pressure. If he does not pick up the gun, the world will destroy either him or the thing that he cares about. If he doesn’t take a move during his turn, twist that tetromino or fails to collect his crops on time then there are consequences. There may be failure or forfeit if he does not decide whether the raise or fold. His army may be destroyed if he just lets the clock run out.
The tug of war between agency and pressure within a finite world is how motion becomes apparent. Contrast this with a virtual chatroom where you can send gifts to friends and talk to them or not with no consequence and you see lots of agency but no pressure. Conversely an experiment in evolutionary dynamics might show its creator the consequences of a mathematical model for life forms where the pressure of survival is paramount. Yet there is no agency.
What about flight simulators then? They have significant user agency and pressure. Are they game worlds? Sure they are, to a point.
The goal of simulator design is to try and be as close to real world experience as possible. They try to be infinite, and so they lose their sense of the finite. The levers for winning become less obvious, and so does the sense of strategy and permutations. While a serious simulation is certainly gamelike, the sense of mastery and the goal of using it are very different from the pleasure of being in a game world as a result.
Dead Worlds
When motion is gone then game worlds often feel dead. A tennis court pre-match is an interesting world, but post-match it is just a patch of grass. Liberty City unexplored by Niko Bellic is an exciting world, but when all the missions are complete and the side tasks are achieved, it’s just a dead world with no purpose.
Without some kind of rejuvenation, game worlds die. A Chess board has a complete state and can be reborn by setting the pieces back to their starting positions for another game, which is a kind of rejuvenation. World of Warcraft rejuvenates by pushing new kinds of content for high level players to get their teeth into and play some more.
If you want your game to keep being played, it needs to either reset or support new content. And you need a plan for that content. In the old days publishers used to do this with franchises, but single franchise publishers have to be more active than that. Releasing a new game every year or two to meet this demand is increasingly not good enough.
Whatever you do, be careful that you’re not creating a world that will die. If there’s no reason to keep playing within it then nobody will.
Simulators are no so void as games. I've been thinking about this a bit lately due to one of your previous posts (don't remember which) but I observed an argument on the internet once between users of Celestia and players of Orbiter. I'll summarize it like this:
Celestia user: "Orbiter pretends to have physics, but they're not really physics, it's newtonian physics and they're even inaccurate cause of the limits of computer simulation." (it isn't practical for Orbiter sim to perfectly model solar system scale flight physics, in fact it's very difficult, so it fails to do so perfectly)
Orbiter player: "I don't care, I like orbiter"
This hints at the difference, one that Celestia user in question doesn't see and the Orbiter player can't articulate. Orbiter is constrained by rules.
Celestia lets you look at things and look at planets and such. It's a planetarium.
Orbiter restricts you to a vehicle with finite fuel, thrust, a strict set of actions governed by rules. It's a simulator.
Celestia lacks this restriction, so the Celestia user in question doesn't see why he'd want it, while the Orbiter player is bored without it.
While it's absolutely the case that Orbiter doesn't have designated "win" states and provided goals, it does allow for death, and the ability to (with planning and work) achieve one state from another.
All sims I've encountered fulfill this, and this is enough for the player to make up a goal and achieve a win in a somewhat deformalized way. His win isn't validated by the machine, but it is a win nevertheless. I've had a multiple great experiences playing X-plane, Orbiter and Jane's F-15E sim and they 100% felt like wins, some very thrilling and stressful.
It's certainly clear that some people, often programmers want to treat everything in a simulationist manner (though these days tower defense seems to be The Thing To Do) and this is usually not appropriate.
But I think it's worth pointing out that simulations do work as games because the people that play them fill in the blanks in terms of constraints, challenges, goals and rules. Like action figures.
Posted by: Joe Cooper | 28 August 2011 at 02:48 PM
The problem with game worlds that reset is that you risk loosing a very large portion of your players each time you reset. So although harder to design and maintain, persistent game worlds are usually more engaging and create more value both for the players and the game world developer.
Posted by: Alexisbonte | 06 September 2011 at 08:56 AM
Alexis,
Agreed. The main issue is that not all kinds of world can be sustained in that way. There is no such thing as World of Chesscraft :)
Posted by: Tadhg | 06 September 2011 at 11:35 AM