First of all, my original post wasn't even intending to classify anything as "not an RPG," or to propose a new definition of "roleplaying." All I was trying to do was to set up a framework for classification, to make these issues easier to talk about. I never intended to say that one way was better than another or that one way "wasn't roleplaying."
But anyway, HowAndWhy, I'm still a little confused here...
Alex319: I think I understand what you're saying now. You're defining a "role-playing game" as "a game such that the players don't have to know any of the rules to start playing, and they can figure it out along the way by describing their actions and letting the DM translate it into game terms."
howandwhy88: Actually I'm defining an RPG as it has been designed and played for the last 35 years, a game where the participants need to succeed in their role in order to win.
But on the other hand, later in the thread...
LostSoul: If I understand this - let's say there's a table in the game rules, a table that says "Modifiers to AC". One of the entries is "a chair or other improvised shield". In that case, using that would mean that I have a good sense of the role that I am playing; that when I fight something out of my league (a mummy, though I've only picked up on that through play) I refer to my knowledge about what will aid me in defending against its attacks.
howandwhy88: Does that make sense, or am I missing something? That a good description, but it leaves out the core element that makes it a roleplaying game versus a simulation game. The rule is on the DM's Screen, not the Players'. The operation of the rule is learned it play. As it is a definer of the role it's an element that must be determined through play. Or proof of proficiency in the role becomes suspect. As you might read in my response to Hussar: any such response in example C. becomes simply the repeating of a stolen answer back to the teacher (or test giver in a Referee's case).
It seems here like you ARE saying that "the players don't know the rules at the outset" is a key element of "roleplaying" under your definition, because you're saying that if the rule is on the DM's screen, it counts as "roleplaying", but if it's on the PC's screen, then it isn't.
Also, here's another thing. You state that having the operation of the rule be "learned through play" is an important aspect of roleplaying. Is the only way to "learn it through play" through trial and error, or are other methods possible? In that example, suppose that before the battle, the PCs were back at home training, and their instructor told them that "these are the things that can help you defend yourself, and these are which ones are more effective" and gave the PCs the "modifiers to AC" list. Any problem with that?
And if you're going to say that the instructor wouldn't have a list with quantitative data (like "X gives +1, Y gives +2") etc., consider the following hypothetical. Suppose that it was a sci-fi game, and the players were officers on a starship, and before their mission they were given a document with the technical specifications of their ship. This would include lots of quantitative data ("Weapon X is N% more powerful than Weapon Y but consumes M% more energy per shot") and of course this data would correspond to the game rule stats of the weapons (if it didn't, the data wouldn't be accurate). Would this break the "roleplaying"?
(To go back to your "telephone repairperson" example: The examiner asks the person how to repair a phone with a particular problem, and even though he had never repaired a phone with that particular problem before, he knows the answer because he read it in a telephone repair manual.)
And anyway, earlier in the thread you say:
Alex319: What "error" are you talking about? And why does the imperfect model interpretation "not count as a game at all"?
howandwhy88: I should have said "gameplay convenience" falls into the same error as game-based design. It's not bad, but it is the element of design that, while nice, causes game to be un-roleplayable in the manner I point out above.
Imperfect model games are based on building DM fiat into the rules. Rule Zero and all that. Whether these are considered games or not is up to you, but without rules in place of "because I say so" an activity doesn't really stand up to the definition of a game.
So you're saying anything that involves "DM fiat" isn't roleplaying. But if a player describes an action that isn't in the rules (which will happen a lot if players don't know what the rules are, as you seem to advocate) then what other way is there to resolve it other than "DM fiat"? Or is the only way a game can qualify as a "roleplaying game" under your definition if the rules cover every action that a player could conceivably describe?
And finally...
For one, there are just so many actions within an RPG in which I could have no actual knowledge in order to role play. I certainly have no idea how to pilot a starship, but my PC does. Does that mean Star Trek is no longer an RPG? I really can't really hack into a computer, so Shadowrun is no longer an RPG. And I'm pretty darn sure that no one reading this can ever cast a spell, so, most of D&D is out as well.
howandwhy88: Those games are RPGs, but you are right that those specific elements of those games are not roleplayed. Why? Because the designers decided they were not definitive of the roles being performed.
So you're saying that a game can still be an RPG even if certain elements of that game are not roleplaying. So suppose that you play 4e as normal, but during out-of-combat scenarios, you have players describe what actions they want to do, then have the DM decide what skill checks are necessary, roll the dice behind the screen, and tell the players if they succeed or not. So in that case, 4e as a whole would still count as an "RPG" even though the combat aspect is still "not roleplaying" according to your definition. Correct?