Because they create a world into which a player can project, video games often incorporate some level of simulation. Constraints such as gravity or real time regularly play a part in games, as do attempts to use procedural algorithms for a variety of things from rag doll physics and enemy behaviour to terrain generation.
However simulation is not actually the point of games. The reason that we use simulation in games is not to create a real world, but instead a deliberately simplified impression of one. In games it is generally better to fake it rather than make it.
The Simulation Trap
Some game makers (often, but not restricted to, programmers) are simulationists. Simulationism is different from both tetrism and narrativism, in that the simulationist believes that accuracy and complexity are the things that matter most. To a simulationist, a game is an opportunity to explore permutations, and both story and gameplay need to work within the boundary of something approaching reality.
Simulationists relish the engineering challenge that complexity brings in figuring out the most elegant sets of algorithms and procedures to reap the greatest benefits. They are the people who think that trees which grow quasi-naturally are important, or that shadow effects have to look totally realistic, or that bullet-through-paper issues are a significant problem in shooters. Sometimes that kind of obsession is exactly the right attitude to have. Sometimes it’s disastrous.
For example, racing games like Gran Turismo benefit greatly from physics that have enough depth to permit car customisations to be felt by the player. Complex management games such as Sim City 4 rely on permutations of activity to provide many hours of absorbing fun. Even games like Nintendogs use some level of simulated behaviours to convince younger players that the virtual dog in their DS is actually their friend.
All of these kinds of simulation are good because the effects are about the player, and directly affected by the player’s actions. They are also within the context of what the player expects the game to do, and so they form a part of the marketing story.
However just how realistic does a blood splatter need to be? Does an arcade-style racing game like Burnout really need 25 levels of engine tuning? How subtle do a character’s aging routines need to get? Does Nintendogs need realistic gravity? The simulationist will always answer in the affirmative.
Simulationists believe that the reason for playing a game is to immerse yourself in a world (this is true) and the best way to do that is to make that world as real as possible (this is not true). Simulation run amok tends to result in dead-end development, unnecessary features, and huge amounts of wasted effort.
Misdirected simulation produces Uncanny Valley-afflicted cut scenes, confusing gameplay, overkill amounts of rule complexity, complicated game controls and endless cruft. Most damagingly of all, simulation is one of the biggest causes of opacity. The simulationist doesn’t realise that all his hard work is invisible. Invisible work is wasted work, and the only real effect that inordinate simulation has is to slow the game engine down with unnecessary overhead.
What Players Actually Notice
To a lot of players there is simply no difference between a very complicated system, and a system that is merely pretending to be very complicated. Consider FarmVille.
FarmVille constructs the impression of a farm in which you can grow vegetables and keep animals. The various objects in the game only vaguely interrelate with each other, providing a steady stream of things for the player to click on and do within a set of rudimentary game rules. The controls are simple, the rules concerning time and energy arbitrary, and the game is only really giving the player a farm in a cosmetic sense.
FarmVille’s veneer of simulation may be about a millimetre thick, but its players don’t care. It turns out that a FarmVille player wants meaningful actions like planting and harvesting to be direct, and is far less concerned about whether the farm has any veracity. It empowers players to get into a mild fantasy of keeping their own patch of land and growing their own tomatoes, and it works largely because it’s in no way realistic.
Another example: For a long time the argument around first person shooters was whether a joypad could really convey enough accuracy and control compared to the classic mouse-and-keyboard combination. While big first person shooter makers pursued a path of endless refinement and resisted ‘impure’ console versions, console shooters exploded in popularity, and some of their developers became heroes.
Even though console shooters are definitively less accurate and realistic than their PC counterparts, players don’t care. They don’t care that the Master Chief from Halo can jump three times the height of Michael Jordan, nor that the troops in Call of Duty can run permanently crouched. What they care about is the game dynamic, whether the results of the game generate enough little and big wins, and really whether they are having fun or not.
Players are unable to notice all of the rich detail of a simulation because of lensing. Lensing is a shorthand term for the amount of information that a player can reasonably be expected to absorb from a game, and the amount of actions that they can meaningfully take. Lensing is one of the five major constraints on all video games (more on this in a future post).
When perception is overloaded, the brain starts to filter and prioritise incoming information according to threat levels, and so it just ignores irrelevant detail. Since most games generate some level of tension for the player, their senses are almost always on high alert for threats, aid packs and other useful game items. The subtle detail of the simulation passes them by.
A really great game designer has to know when too much information is simply not worth including, and instead focus on delivering along lines that the player will notice. There is not much point in creating realistic plant life growth patterns in a game where the objective is to shoot space orcs, and there are often many technical reasons as to why including such elements impedes game engine performance in other areas.
Where simulation goes off the rails entirely is when it starts trying to make gameplay less abstract and more like real life. Since games are technologically more complex than ever, the simulationist thinks this means that goals and actions should also become more complicated. Choices should have more consequences, behaviour of AI characters should be more refined, and players should perform less abstract and more real actions. This would then lead to more immersion. ‘For Every Choice, A Consequence’ was the tag line of the original Fable, for example, conveying the idea that every action that the player took would in some way flow through the rest of the game.
In my opinion, this kind of simulationism is a cardinal sin of game design.
What results is games which are opaque: Every choice may indeed have a consequence, but most of the time the player is unable to perceive the line of cause and effect in an opaque game. So they are unable to tell whether a loop has closed through their actions, and cannot know whether what they have done in the game matters. Complex goals rob the player of the knowledge that they are winning.
Opacity leads to feelings of randomness and frustration. How many times must I teach the virtual creature in Black and White to do a task before he gets it? How often do I have to do good or bad deeds in Knights of the Old Republic to become a Jedi or a Sith? What is the correct conversation path to get Yuffie from Final Fantasy VII to join your team? All of these are examples of opacity. The game becomes vague, uncertain and hard to comprehend. The player starts to filter or ignore peripheral simulation elements as much as possible, focusing their effort on the causes and effects that they can perceive.
In Fable, for example, the extent of the simulation may well be vast but the combat is actually very easy to master and so all that the player really does is hit stuff and visit towns. There is a lot of detail in the behaviour of the environment, the various characters, permutations of appearance, class, powers and abilities. In fact, in theory the player can get married, own property, grow a reputation as being good or evil, and otherwise engage with the environment.
In practise, many players realise very early on that there is an optimal way to play the game: Focus on the magic class and ignore the consequence that using magic may prematurely age your doll because it has no tangible effect. Ignore most of the game and focus on using magic to battle our way through the missions until you get bored or complete the game. That’s it, and everything else is left by the wayside because it doesn’t matter nor feel relevant to winning.
Fable scored highly among reviewers and sold well, as did its sequels, because it garnered many fans for its wise-ass humour and relatively entertaining adventure, but the extra simulation layers within the game usually go unnoticed. Lionhead might as well have not bothered and simply faked it instead.
Pursuing clarity, the opposite of opacity, means knowing when the simulation elements in the game are causing more trouble than they are worth, and scaling them back. Achieving clarity is one of the hardest things to do in game development because it usually means ditching many of the pet projects of simulation in favour of something cruder, and fighting with the simulationists in the team on behalf of the users.
Perceptible changes are the ones that players care about the most. We rarely encounter straightforward win/lose choices in our daily lives, and the levers with which we make those choices are similarly rarely obvious. It is at the heart of games that they simplify wins and the means of action into an elegant loop, so that an abstract action (like pressing a button on a controller) has a predictable outcome. Games are played because they are empowering, simplified and fairer than life.
While the simulationist is obsessed with providing fidelity, the player is actually looking for a simulacrum that gets their imagination in gear. They want ideality, not reality. An ideal world is an empowering world, in which the player’s actions have effect, and the wins that they attain from that world are clear to understand. Ideality merely gives the passing impression that it is reality.
Guitar Hero works as a game because it turns the complicated skill of playing guitar into a series of abstract actions that can be easily understood, and therefore won. The actions that the player takes perceptibly effect the visual, numeric and auditory feedback of the game directly, so there is no opacity.
Zelda works as a game because every thing that the player does is substantial, and each adventure is constructed to let the player play in an exciting way. A new and significant action (such as the Ocarina) is the focus of each game, and is both tangible and extensible.
Even in more long-winded games like The Sims, the player needs to be able to see consequences, or else they are not consequences at all. Whether creative, immediate, tactical or strategic, all games must feel clear. Otherwise they are inherently unfair.
Backdrop, and the player’s ability to infer numina, can be created using simulacra rather than simulation with zero ramifications. Like a stage play, a game can ask a player to make the imaginative leap to see beyond the edges of what he actually sees, and if done well this enchants the player.
In the Grand Theft Auto series, the presence of other people on the streets, and their various snippets of dialogue, does a great deal to enchant the player. It adds a layer of richness to an experience that would otherwise be pretty sterile, and the game is consequently more thaumatic.
It appears to simulate a living city, but what’s actually happening is that the game is using a cylindrical area around the player and adding or removing people and cars just out of sight. When the player moves on from one street to the next, the cars that he left on the previous street are removed from the game, as are the people, and new ones are created instead. And the cars and people are all constructed from easily-identified types.
The sharp-eyed player will eventually notice this, but actually it doesn’t matter. An ideality is not trying to fake a reality whole and complete, it is merely trying to create the impression such that the player’s imagination can fill in the gap.
He feels like he is in a city, and he has tasks to perform and things to do. Even if the same faked human in the city walks past him 25 times and says much the same thing, it doesn’t matter. The game asks the player to suspend his disbelief a little, to understand what is meant in the environment rather than what is actually shown, and the player is fine with that because he is playing and actively trying to get things done.
Part of the art of making games, and understanding what games are, is realising that worldmaking is not reality-making. You are allowed to fake it, to give the impression of a simulation rather than actually making one.
To the player there is no difference, and to the art of the game likewise, and games are always about the player when all is said and done. So stop worrying about tree-growing algorithms, and trying to summon Techthulhu, and focus on the things that matter instead.