There are a dozen ways to slice and dice types of actions and types of resolutions. If the action is either obviously going to succeed, or failure is not interesting, the resolution should simply be the GM saying, "OK, it works." If there is no way the action can succeed (e.g., the character attempts to leap across a forty foot gap with normal human abilities), the GM is well within his rights to give the player a dirty look and say, "Come on, play seriously." But most of the tasks in an RPG fall in the big grey area between, in which both success and failure can lead to interesting stories, and it is not clear which path our hero will go down.
You can't always get what you want
But if you try sometimes...
Every RPG comes with a core mechanic for resolving actions. Most of them use dice. Some of them use other randomizers, such as cards (or, in the case of Dread, a Jenga tower). Diceless RPGs generally use some kind of currency in which the player can "buy" a success now, at the risk of possibly not being able to buy a success later (or some other trade-off). In story games, it frequently comes down to who has narrative control of the scene, and each game has a different way of passing that control around.
The point of the resolution is to bring together all of the various elements of the scene and determine from those whether or not the character is successful. You can separate those elements into some broad categories: environment, character ability, player ability, opponent ability, story concerns, and pure dumb luck.
- Environment encompasses all the things external to the character that might affect the action. In crudest terms, it includes such things as lighting, whether the floor is slippery, and the incessant chittering of the rats in the walls. For certain actions, though, it can also include factors such as whether your true love is watching you, the orders from your superior, and your phobia of chickens being triggered by the duchess' feather boa.
- Character ability is all the things on the character sheet that feed into the action. For most systems, that will be your stat and skill. Some advantages or disadvantages may come into play, as well as "bennie" mechanics such as action points or fate points. It also includes items you bring to bear, wounds you have suffered, and fluff concerns like having had a bad experience with dogs. In pretty much every RPG, this area is thoroughly covered.
- Player ability mostly represents the skill the player brings to the game, in terms of tactics (for combat situations), charisma (for social situations), logic, knowledge of the game system, and, more often than we'd care to admit, pure cussedness. For a player, the ideal solution is to be able to argue the GM away from "make a roll" and into "ok, it works." More often, you are working to maximize your bonuses and minimize your penalties. Note also that player ability covers cheating.
- Opponent ability is essentially the same as character ability, but represents all the attributes of the person (or thing) the character is working against. This could be the armor class of a combatant, the stubbornness of a haggling merchant, the obscurity of a secret door, or the intricacy of an arcane working. In most cases, the GM boils all of this down into a single target number based on the desired difficulty, with only passing regard for all the individual factors involved. In some cases, most obviously in the case of defense in combat, the rules will specify how those factors interact.
- Story concerns are not always at the top of everyone's list. However, in many systems, and even in many games using other systems, whether or not the action is important to the story matters in how it is resolved. That may be a simple matter of whether or not "bennies" or other bonuses come into play. It is one of the main factors in determining whether an action actually needs formal resolution. I also have to admit that I find systems that don't honor story concerns to be extremely frustrating.
- Pure dumb luck should always be a factor, IMNSHO. It's just more fun that way.
There are two of the factors above that require being called out a little more specifically, because they strongly influence gaming style. Different gamers put very different emphasis on player ability, and it's not always related to that player's ability (i.e., it's not always the more competent players voting for more focus on player ability). Different gamers also are interested in very different levels of interaction with the story.
Captaincy is 90% luck and 10% skill. But don't try it without that 10%.
One of the hallmarks of the OSR movement is the emphasis on player skill. If an RPG is a game, then success should be measured by how well one plays the game. Part of the point is matching wits with the GM and overcoming his obstacles with ingenuity and daring rather than bonuses and counting squares on a battle mat. Of course, strangely enough, the story game movement is also strongly in favor of player skill. It is simply that, in that case, player skill is measured by how well you craft the story around the characters, rather than how well you outwit the challenges.
Personally, I prefer to downplay player skill at the individual action level. First, I believe that if you are playing a role, then your success or failure should be determined by the qualities inherent in the role. Second, I believe that emphasizing player skill leads to players working to manipulate the system and/or GM rather than working to make the game as a whole better. Again, these are my opinion, and not presented as facts.
Of course, it is impossible to eliminate player ability. At the most basic level, it includes the rules mastery necessary to understand and apply the elements of the character sheet. It also involves the perception of the various factors at work, the weighing of such factors to determine a strong course of action, and the appreciation of ways to adjust those factors. And, player ability will often make a tremendous difference one level up from the individual action, when plotting out strategies for successfully navigating a scene. A level up from that, when determining what general arc to take through an adventure, should be almost entirely driven by player skill, with little to no impact from the character's assets or defects.
The computer can't tell you the emotional story.
Bending the results of an action to fit the story is definitely a matter of taste. Some gamers see it as terribly unrealistic, as those of us in real life don't get to cheat fate. Some gamers see it as terribly obvious, since a role playing game is all about telling, or at least discovering, the story of the characters. Allowing the blind roll of the dice to screw up the character arc is just a recipe for dissatisfaction.
As you might have guessed from some of my previous posts, I strongly lean towards the latter interpretation. I am interested in acting out our own personal novel or movie. As such, I want to feel the story beats and get the satisfying climax. However, I'm not actually interested in most story games. I'm not sitting down to an RPG looking to write collaborative fiction. If I want to do that, I'll just chuck the system and go free-form. What I want, as a player, is to inhabit the role I've chosen and experience the world through his eyes, in much the same way that a good book lets you experience the world through the protagonist's eyes. But, unlike in a book, I want to control all the choices made, and see where they lead.
One of the most common arguments against letting the story drive the action is that it kills the tension. PCs, by their vary nature, would have to be granted some sort of plot immunity if the story is king. When you know the PC can't die, indeed ultimately can't lose, there's no actual threat any more. Personally, I find this attitude bizarre, as these same people enjoy comic books and TV shows in which you know that everything's not only going to turn out alright in the end, but will mostly end with everything in exactly the same state it was at the beginning of the story arc. If you can feel a thrill when Spiderman or Captain Kirk is in danger, why can't you feel that same thrill when your PC is in danger?
I feel myself building to a long rant on this subject, so I will cut myself off. I'll just end by saying that, in any conflict, some actions are going to be more important than others. That importance is wholly derived from story concerns. Treating critical actions with the same mechanics as trivial ones is unlikely to create a satisfying experience. You will score a spectacular critical hit on a random guard, but totally fail to knock the remote control out of the mastermind's hand, dooming us all.