I don't think your first line here is, in fact, true, at least the same way rules are. Most of that tends to tell you about output result in broad strokes, not likelihood or process, at least in any way that will reliably mean the same thing to everyone involved.
nods For instance, a narrative example (rather than an action-by-action play) of a dragon soaring through the sky and occasionally unleashing a gout of flame is unlikely to allow one to deduce how quickly it flies, how often it can breathe fire, or the area of effect or the precise lethality of its flame breath. Arrows and spears may bounce off its scales, but is it merely
unlikely or literally impossible for a mundane projectile to have a meaningful impact on it? And so forth.
If one is running a system that abhors the very notion of combat as a subsystem and prefers to reduce the entire scene to a single die roll or two based on preferred drama, estimated task difficulty, and some narrative metacurrency -- such details may not be important.
But if not, and if players are expected to make more granular, tactical decisions; well, it's best if both the players and the GM have consistent visions of how such things work. It's unlikely to be a good time if there are frequent arguments as to how likely it is that an AoE spell would be able to catch certain targets while being positioned as to spare others, or whether it
should be harder to cast a spell while struggling to avoid being grabbed by some tentacled horror, or whether wounds should have lasting effects, or who should have the right to act first in a situation where turn order matters -- never mind more extreme situations, such as if a player seems to want this character to have movie-protagonist plot armor and competencies while a GM wants to run a more grounded game.