No one RPG can do it all, and each game does some things better and other things worse.
Rule Zero and the Oberoni Fallacy[]
At the heart of every role-playing game is a rule, which consists of doing what works. In many games, this is known as Rule Zero. Rule Zero basically states that the arbiter of an action, typically the gamemaster, can break any rule if it serves to improve play. This is an important principle, because the rules can not cover everything.
The Oberoni Fallacy[]
As formulated by Oberoni on the wizards.com boards, circa 2003, the fallacy is this:
There is no inconsistency/loophole/mechanics issue with Rule X, because you can always Rule 0 the inconsistency/loophole/mechanics issue.
This is considered a fallacy because you obviously cannot fix a problem that does not exist. The existence of Rule Zero in no way diminishes the places where a rule system simply does not do the job.
Tension in Design[]
There is a natural tension in game design between various goals, such as ease of play versus level of detail, or simulation of a genre versus simulation of a consistent world. One approach is to develop a philosophy of game play.
The Problem of Simulation[]
The problem of simulation is basically as follows: if you are creating a game based on an action genre, do you create a game that results in a narrative like those of the media it is based on, or do you accept the reality as presented in the media and strive for consistency? Just as an example, if you were creating a Star Wars roleplaying game, do you want a game that leads to the Emperor being reliably defeated by the power of redemption, or do you want to be able to stat his saber prowess relative to Yoda? As an another example, is it really, truly possible to kill James Bond in a James Bond role-playing game?
One means of solving the problem of simulation is by deciding on a degree to which previously established canon is an accepted reality and establishing such in the setting section or the worldbuilding section of your RPG. Whether said canon is in line with other narratives or does its own thing, it is not advisable to try to do it all at once. Another means of solving the problem of simulation is by not engaging with the established realities at all, instead engaging with a different place inside the same fictional universe at the time of the other events. Limiting factors here include, but are not limited to lessened player agency and a demand for a setting big enough to have multiple narratives at once.
Even in spite of conceivable means of addressing the problem of simulation, this limitation has no uniform, end-all-be-all solution. It is a demonstrable example of mutually exclusive design goals.
Hierararchy of Resolution[]
In general, the best way to resolve an action is, from best to worst:
- Using the rules as written, getting an appropriate result.
- Using the rules as written, with the smallest possible change, to get an appropriate result. For instance, inventing a situational modifier.
- Using the rules as written, interpreting the results in a general way to get an appropriate result.
- Devising a way to resolve an action by analogy. For instance, if two flying characters engage in combat, you might look at the dogfighting rules and try to figure out if there is a trait that can replace the piloting skill in that situation.
- Using a general mechanic. For instance, if the game uses difficulty numbers, you might assign what you believe to be a fair difficulty number for a task and choose an appropriate skill to resolve the action.
- Decide the outcome based on meta-game priorities. You might decide whether one outcome is always preferable, an element of chance if preferable, or that the outcome depends on the dramatic potential of the situation.
- Fiat. An arbitrary choice is better than no choice at all.