When I was developing parts of the curriculum for the University of Advancing Technology’s Serious Games coursework, there were very few source materials to draw from; ten years on, there still seem to be very few standardized critical frameworks for discussing and analyzing the many different styles and types of play lumped together under the “serious games” label. The following is by no means an attempt to be definitive or establish a standard, but rather to spend some time thinking aloud about the boxes we use to compartmentalize our projects and how we might sort game design approaches more accurately for examination.
I began my material for these courses by identifying five different approaches to simulation, in the context of “serious games” (I continue to use the quotes, not as sarcasm, but as an indication of how dubious I am about what this particular label can tell you about the product itself). These approaches bleed over into games for entertainment, which is a natural occurrence as both, at their core, promise some degree of interaction (and yes, fun.) These approaches are on a gradient, so to speak, from the most sincere and direct attempts at simulation to approaches that try to balance simulation with the abstracted world we associate with video games.
Those games/interactive products which most directly attempt to mirror or mimic reality, I classified as Mimetic. This means “mirror image”; these sims attempt to create the most realistic and authentic experience and may have little actual “game” content.
Mimetic sims do not rely upon visual fidelity to create a sense of simulation; instead they attempt to engage the imagination of the player or user and do not sacrifice realism for attractive graphics or traditional gameplay elements (such as earning a score). The goal is to re-create an authentic experience, for someone who has experienced this before to some degree in the real world.
These sims are made for a small, “hardcore” audience and may include such non-game subject matter as air traffic control, orienteering, mountaineering, traffic management, stock market trading, and hydrology. Often there is a high emphasis on educational content and an assumption that the player is very familiar with the source material, if not in fact already a professional in that field. The core approach of a mimetic simulation is to duplicate the interaction points between the human and the system being simulated with 1:1 fidelity; so a mimetic air traffic control simulation would be real-time and include all the associated duties, perhaps even as far as not including the ability to pause the game (as, alas, we can not pause time in real life).
Though they are produced for niche audiences, mimetic sims are in fact much more common than you would think, especially in the areas of training simulation. On very rare occasion (fueled by massive contracting budgets) you would find a commercial simulator that also attempts 1:1 visual fidelity, but again this is for presentation of information and not aesthetic purposes. One example of this would be a commercial flight simulator - these feature extreme visual fidelity for weather, time of day, lights on the ground, etc. but do not actually simulate the crash of a plane if the sim pilot is careless; on Mythbusters, when they crashed the plane in a full-cockpit simulator, it just skidded and bounced off the ground. Thus the goal of visual mimetic presentation stopped at the point the simulation stopped, rather than continuing into a more game-like presentation of a spectacular crash.
A few examples: http://www.catchingfeatures.com/ is an orienteering simulation; http://www.atc-sim.com/ is a control-tower simulator; https://www.circuitlab.com/ simulates working with digital circuits.
Systemic sims attempt to replicate real-world systems of activity in a way that is approachable and logical; they operate on “common sense” designs that work with our instinct as to how a system should work. As with mimetic sims, visual fidelity is not a requirement; systemic simulation tries to replicate a “big picture” of a known environment. The systemic sim relies on a mix of the player’s sense of how things should work, given a known set of fixed rules, and limited direction on how to interact with that system.
The systems being simulated do not have to actually exist; instead, they just have to operate by known, common-sense rules of how we expect them to operate. Systemic sims place the player in the role of a manager of complex systems, but often hide the actual structure and workings of those systems; in this way they differ from mimetic sims. A systemic sim could be considered one step back from mimetic; for example, you could have a systemic air traffic control system which involves managing the staff of the control tower or multiple airports, rather than the direct air traffic control itself. Systemic simulation often has a “god’s eye” point of view rather than direct interaction with the simulated system’s integral parts.
One of the most interesting (and popular) commercial applications of Systemic simulation can be found in the various "sports manager" games, which simulate owning, coaching or otherwise running a sports franchise in soccer/football, cricket, Formula 1 and even darts. These management sims rely heavily on the idea of what this might be like, rather than a direct 1:1 simulation (for example, you do not do your team's taxes, thankfully).
A few examples: http://en.wikipedia.org/wiki/Championship_Manager_series Championship Manager, perhaps the most successful systemic sim of all time; http://www.citiesinmotion.com/ a fascinating, zen-like simulation of city transporation management; http://www.introversion.co.uk/prisonarchitect/ Prison Architect, an unusual systemic sim that uses cartoon visuals and a less numbers-driven approach than most simulation games.
Adaptive sims attempt to simulate a specific environment and set of behaviors, but with an emphasis on gameplay rather than pure realism. Often of high visual fidelity, these types of sims use mimetic and systemic elements, but focus on gameplay as their core determining structure. The branching-off or blurring of the entertainment vs. “serious” game audience occurs in adaptive sims, and they tend to have budgets that skew towards the entertainment side of the industry (as in larger) and be aimed at a commercial return on investment.
Adaptive sims include “all the good stuff” from an environment, but leave out those things which make mimetic sims frustrating; thus they “adapt” the rigid mimetic rules into a more friendly gameplay environment. However, unlike Systemic sims, they like to put you right in the middle of the action and interact directly with the simulated world, rather than manage its systems.
For instance, in a mimetic F1 racing sim, you would have to make pit stops or run out of fuel; this is rarely a feature in adaptive sims, which don’t often don't even bother to include a fuel gauge on the dashboard. The sense of simulation and authenticity in adaptive sims comes from a combination of our common sense responses to that game environment (as with Systemic sims) and from visually replicating the environment, its elements and using real-world places, names and products. Quite a few commercial titles could be called adaptive simulations, although there are many adaptive sims that would also sit in the hobbyist or even educational markets.
Some examples: http://www.rfactor.net/ rFactor, an auto racing simulator that skews towards driving fidelity rather than commercial flash, and encourages fan-created content; http://www.digitalcombatsimulator.com/en/products/warthog/ DCS A-10C Warthog, an incredibly accurate sim that nears mimetic (especially in the difficulty of controlling the aircraft) but does include enough assists, scenarios and the ability to drop in to the middle of a battle to render this an adaptive sim.
Intrinsic or Inclusive Simulation
Games with intrinsic simulation include realistic elements inside an unrealistic or impossible environment, as a way of bringing authenticity to the experience. These could be considered the inverse of an Adaptive approach, as we know reality is not being addressed yet we have expectations of some degree of “realism” within the context of the game world.
As with systemic sims, common sense is a big part of the design; we’ve never fired a plasma pistol, but it makes sense that they require ammo, sometimes overheat and can be found on the bodies of dead soldiers that we have just witnessed firing said weapons at our person.
Visual and audio fidelity are very important in intrinsic simulation; if an M-16 rifle looks and sounds right, then it becomes an element of reality even when using it to battle zombies and vampires, which (one would hope) the player knows are imaginary.
Physics, gravity and other real-world principles are also intrinsic simulation, and can be effective design elements even in games which are utterly unrealistic otherwise. Mario’s jump in Super Mario, perhaps one of the most discussed and dissected game mechanics of all time, is indeed based (loosely) on a simulation of gravity, although every other element of the game is pure fantasy and bears little or no resemblance to reality.
Some examples: https://kerbalspaceprogram.com/ Kerbal Space Program, an absolute fantasy game that nevertheless includes really solid material and orbital physics simulations; http://www.runestorm.com/viscera Viscera Cleanup Detail, a game where realistic physics and behavior of fluids, flexible materials and fire are simulated to create an incredibly disgusting experience of cleaning up after an alien invasion.
I chose to break out history-based simulations into a unique category (although they often use some or all of the approaches above) due to the specific mindset and audience for those simulations. Historic sims are often mimetic or systemic in nature, but always lean towards authentic reproduction of historical events in an attempt to “replay” history; thus the systems and mechanics are altered by this goal rather than being a generic approach to pure simulation.
These historic sims may be adaptive in design (such as a WWII battle sim) but draw strongly from historical information and attempt to replicate a particular period or event experience for that specific audience. These games often have an educational basis or come with supplemental materials, to prove the authenticity of the simulation and set it apart from a pure entertainment product. Almost the entire catalog of Battlefront (http://www.battlefront.com/) could be seen as Historic in mindset, with varying degrees of mimetic, systemic and adaptive approaches as seems to fit the source material and audience.
What seperates a Historic sim from a purely adaptive or mimetic sim is the idea that in all the game design decisions, history wins - while a mimetic sim might create a completely neutral Battle of Waterloo that can be played from either side and with many outcomes, a historic simulation of that battle will try to replicate the known outcome. Historic sims also include very large amounts of information and historical background, something that may not be present in a mimetic sim (as those assume knowledge of the topic).
Some examples: http://www.matrixgames.com/products/360/details/Birth.of.America.2:.Wars.in.America Birth of America II, a Civil war simulator that includes the timeline from the late 1600s until well into the 1800s. While it has many systemic elements, historic elements abound (for example, you can't use a specific general outside their lifetime); http://www.matrixgames.com/products/472/details/Germany.at.War:.Barbarossa.1941 Germany At WarL Barbarossa 1941 is another game with many systemic elements, but the historical approach overrides many game elements to bring as much historical data and events to the table as possible.
So, what is the purpose of such a fine degree of classification? Many of the core design dilemmas present in “serious games” design and educational/training game creation hinge on what elements of this “game” should be interactive, and what are rendered as didactic or “chunked” (overused buzzword alert!) information for the user to digest. On many occasions I would try to present a design to a client (or simulation problem to a student), and they would insist that some elements needed to be “modeled” or presented as interactive, while others needed to be text or quiz elements that establish the facts you need to know. Being able to classify elements as different approaches to the goal made better sense of those conversations, and helped tremendously in breaking down a complex simulation into development targets - as well as helping to convince the client that the quiz or text was most likely not necessary as long as the simulation was doing its job.