Friday, January 15, 2010

Simulationism and other claptrap

I object to the terms "simulationism" and "simulationist". Simulation is not a belief system, nor is it a philosophy. Simulation is a tool which uses as its main mechanics modeling and abstraction. The goal of that tool is given situation X, produce result Y, where Y could plausibly happen in the real world, if the differences between setting Z and the real world are factored in. Modeling is an attempt to produce objects, forces, and people which mimic real objects and people, given the differences in setting, while abstraction is an attempt to lessen the data required to produce the fastest possible result with the least effort.

Thus simulation is always a compromise between modeling and abstraction. You can think of it in terms of digital sound sampling - the more samples you take over the duration of a given sound, the closer that sound is to the actual analog sound wave, but try to force too many samples for your playback system, and the system breaks down under the data load. Abstraction reduces the samples processed by the system to match the data capacity at the cost of some loss of fidelity.

So, you can easily see this with weapons in games, which is one reason why it's always brought up - the other reason being most roleplayers think they know a lot about guns, whether or not they've ever fired one. If a Smith and Wesson Bodyguard .38 special snub nosed revolver is mechanically different from a Colt Cobra .38 special snub nosed revolver, you have a very high ratio of modeling to abstraction. If a .38 revolver is mechanically different from a .45 revolver, but you can't tell the Colts from the S&Ws, you have a moderately high ratio of modeling to abstraction. If a heavy revolver is mechanically different from a light revolver, but you can't tell a .38 from a .45, you have a moderately high ratio of abstraction to modeling. If you can't tell a revolver from an automatic, then you have a high ratio of abstraction to modeling.

What the best ratio is depends on the game's mechanics - the data throughput capacity - the desires of the participants for fidelity - the listener's ears - and the particular aspect of the game in question - some people like lots of combat, others like lots of social interaction, still others like investigation, etc. Your ratio does not have to remain consistent throughout every aspect of the game.

-clash

4 comments:

  1. That's good stuff, clash. I am reminded of one of my favorite Operations Research chestnuts: "All models are wrong; however, some models are _useful_."

    ReplyDelete
  2. A good angle on the 'great debate'...

    Earlier you mentioned little response to setting articles as opposed to system articles.

    I suggest that this post could be fruitfully adapted to setting theory. That would interest me, for sure :)

    ReplyDelete
  3. @ Rich:

    Yep! I know that one! :D

    @ Vincent:

    But it scares *me*! :O

    -clash

    ReplyDelete
  4. Well said. I'm all for "given situation X, produce result Y, where Y could plausibly happen in the real world, if the differences between setting Z and the real world are factored in." with as much abstraction as possible.-Mike

    ReplyDelete