In different RPGs, there are different approaches to how open-ended the rules should be. Some RPGs prefer very definite rules, in which there is little left to interpretation, while other games have naturalistic rules that rely heavily on interpretation to make sense.
A naturalistic rule is something like, “this spell allows you to create a ball of fire. It burns hotly and brightly, and can be used to attack people, light areas, or set things on fire.” A definite rule would be something like, “this spell deals 1d10 fire damage to an enemy creature within 6 squares.” The naturalistic rule describes what the spell does in the fiction, while the definite rule describes the mechanical impact of the spell. Most RPGs employ both of these approaches to various degrees, but tend to lean more towards definite than naturalistic. The standard approach is to give fairly clear mechanical rules, along with some naturalistic information on the side. This approach leads to the often-seen “fluff” vs “crunch” distinction.
Different amounts of reliance on naturalistic approaches have advantages and disadvantages. More naturalistic rules allow for a rich play experience where the rules feel intertwined with a fictional world. Naturalistic rules are also by their nature more open-ended, fitting well with a free-wheeling improvisatory style. Definite rules are better suited for establishing an inherently interesting and dynamic game, one that can be fun to play regardless of the fiction of the world, and one that can be fun for optimizing and strategizing within the rules of the game (as opposed to strategizing within the fiction of the game).
As I stated earlier, the standard approach is to give naturalistic descriptions alongside definite mechanical descriptions. Although this approach is standard, designers often fail to clearly delineate their naturalistic rules from their mechanical rules. This leads to arguments over which elements are mere “fluff,” able to be changed at will, and which are “crunch,” a critical part of the design. Good game designers should strive to be as clear as possible about what their intentions were in designing the abilities, so that players and GMs understand the assumptions behind the game and can either a) run it as intended or b) understand what effects their alterations are likely to have.
A big part of this problem is vague wording within parts of the rules that look like they’re giving precise mechanical description. A common mistake is the idea that clear wording is against the spirit of naturalistic description. This view holds that as soon as a rule is made clear, room for interpretation goes away, so the way to increase naturalistic interpretation is to leave the rule vague.
The problem with this is that it does not typically lead to the desired end – an open-ended, ruling-oriented playstyle. Instead, it leads to arguments over what the rules really mean. When you give mechanical descriptions, you are implicitly saying that there is a definite rule in place, and when people read it, they naturally come to a conclusion as to how the rule is intended to be played, and they reasonably assume that the designer’s intention was to codify this as the official way to play. If you don’t want people to read a rule like this, you should explicitly state where the rule is intended to be left open.
A good example of a problematically vague rule is 5th edition d&d’s blindsight:
A creature with blindsight can perceive its surroundings without relying on sight, within a specific radius. Creatures without eyes, such as oozes, and creatures with echolocation or heightened senses, such as bats and true dragons, have this sense.
The problem with this is that it is unclear if this is meant to be interpreted naturalistically, as an interpreted element within the fiction, or if it is meant to be a clear mechanical definition. The use of keyword terms like “blindsight 60 ft” within monster statblocks makes it seem like it’s supposed to be a mechanical rule with a clear definition, and should not require additional interpretation to make sense. However, when you dig into the rule itself, you realize that it is quite unclear what blindsight actually does in mechanical terms. I’ve seen some people argue that it amounts to an “auto-detect” ability, meaning that stealth is simply ineffective against blindsight no matter what. If you think that it is intended to be a definite rule, that is probably the most reasonable interpretation, since otherwise it doesn’t really seem to do anything reliable. However, if it is intended to be a naturalistic rule, then one can assume that it is deliberately left up to each GM to decide how one might go about bypassing a particular creature’s Blindsight, and that this level of on-the-fly interpretation is what was intended.
An example of how to do this right is the way the designers of the game 13th Age did it. Here is an excerpt from a Rogue ability called “Swashbuckle”:
Once per battle as a quick action, you can spend your momentum to pull off a daring stunt the likes of which others could scarcely conceive. You may make an attack as part of the stunt, but you’ll need to roll normally for the attack. This is an improvisational talent.
As a swashbuckler, you do not need a difficult skill check to pull the stunt off.
This description makes it very clear that the ability is intended to be improvisational, literally saying “this is an improvisational talent” (elsewhere in the rules they mention that certain talents are improvisational, and that players should only take them if they are comfortable working with their GM to figure out what they actually do). The ability also gives a nice piece of guidance for how the GM should approach the stunt, saying that the swashbuckler may be required to roll a skill check to pull the stunt off, but it shouldn’t be a difficult skill check. This is setting some parameters within which they expect the GM to operate, communicating their intentions without totally restricting the GM’s freedom.
If this approach was applied to the blindsight rule, it might add something like this to the rule after the initial section:
“Each creature with blindsight is different. Some rely on smell, others on hearing, and others on uncanny supernatural senses. Typically, it is difficult if not impossible to use stealth within a creature’s blindsight radius, but if appropriate precautions are made, it may be possible to confuse or cleverly avoid the creature’s senses.”
This extra bit of text would make it much more clear that it is intended to be a naturalistic rule, and that the GM should expect to interpret it in a way that works for their table. It lays out some parameters for how blindsight is intended to function: it makes stealth difficult, but not necessarily impossible for clever or well-prepared parties.
Of course, this might not be what the designers wanted. Maybe they did want it to be a mechanical auto-detect ability. If that’s what they wanted, they should have just said that. Either way, whatever they were thinking, there is a better way of saying it. Even if they wanted it to be up the GM whether it was an auto-detect ability or not, they could say: “Some GMs will prefer to rule that blindsight automatically detects creatures within range, while others prefer to allow clever play to bypass blindsight.” This at least makes it clear that there is no single assumed interpretation.
The best way to get open-ended rules without being vague is to do what 13th age does: be very clear about what elements are meant to be open-ended, and provide some parameters within which GMs should feel free to work. Vagueness makes the GM’s job more difficult than it has to be, and is a poor way of communicating the assumptions behind the game to the players and the GM.
Intentional vagueness, or “strategic ambiguity,” is fundamentally sloppy. Game designers should strive to state clearly what their intention is, and if their intention is to leave something open, they should simply say so within the rules. There is no need for providing fodder for disputes and rules-lawyering by writing vague rules. Vague and open-ended are not the same thing. Open-ended is often good, and is a perfectly fine design goal, but vagueness is almost always bad and should be avoided wherever possible.