this isn't really about mechanics. It is about deciding the themes sticking to them while framing the fiction.
What's the process, in D&D, for a PC hoping to meet a friend or ally - and having a chance of doing so - but also having a chance of an enemy turning up instead to rain on their parade?
What's the process, in D&D, for having a king's guards turf a PC out of the castle into the moat that doesn't have a real chance, and likelihood, of escalating into a deadly conflict?
What's the process, in D&D, for a PC attempting to shame a NPC into taking some action, and having the redound upon them, so that they are the one who has to carry the weight of embarrassment?
What's the process, in D&D, for a PC to persuade a skeleton lord to give up their guardianship of their forlorn forest, and instead convert from heathenism and have the bones of them and their followers placed in the PCs reliquary?
What's the process, in D&D, for two characters to argue about whether or not one will mend the armour of the other, with the outcome of the argument not just being chosen by one or other controlling participant, and with the outcome being binding at the table?
I would just want people to articulate why they think it is these mechanics, instead of the fiction and themes they choose that produce this difference. Like that is the part I don't get.
The above examples are intended to provide some illustrations. If all that can be staked and resolved, without the outcome just being GM fiat, is PC death, then it is hard to address other thematic concerns.
If resolution processes are highly technical and mostly involve mathematical decision-making and optimisation without those decisions and processes generating thematically meaningful fiction - ie if they closely resemble AD&D hp-attrition combat - then it is hard to address other thematic concerns.
If the outcomes of declared actions will be decided by the GM primarily by reference to and extrapolation from their idea of the fiction, and if they main way for the players to learn what that is is via low-stakes action declarations that "poke" at the fiction to prompt the GM to reveal it in ways that don't hose the PCs, then it is hard to make high-stakes thematic concerns a regular focus of play.
Perhaps it is about the difference between principles and mechanics. Like I totally get why principles matter for this great deal, and I am way less convinced that the mechanics matter that much.
The two are intimately bound up. For instance, a principle that says "be a fan of the PCs" is not very useful if the main procedure actually available for scene-framing is for the players to declare that their PCs enter a building (or similar complex) mapped and keyed by the GM, to then declare that their PCs enter certain rooms or open certain doors, with the GM then telling the players what they see and who they encounter by reference to the key.
A GM move like "announce badness" is not very workable if there is no process that tells the GM when to do it or not do it, and if there is no process whereby the players can have their PCs reliably respond to the announced badness.
Here's a concrete example
that Edwards gives:
The time to traverse town with super-running is deemed insufficient to arrive at the scene, with reference to distance and actions at the scene, such that the villain's bomb does blow up the city. (The rules for DC Heroes specifically dictate that this be the appropriate way to GM such a scene).
He labels this "simulationism overriding narrativism". What he means by that is that a faithful application of the mechanics - where the goal of the mechanics is to model time and space and to permit extrapolation of events from that modelling, without any regard to emotional or thematic considerations - can and likely will result in outcomes that undermine a focus on premise and theme. If you're using this sort of resolution engine, then "announcing badness" does not contribute to rising action: it just taunts the players!
To elaborate further on the previous paragraph: I have more than once seen the complaint that, in a game like AW or BitD or BW, a player can
make things worse off by declaring an action, because if the action fails then the GM doesn't just say "nothing happens" but rather narrates some adverse consequence. This orientation is precisely a response to the sort of example just given - ie the GM's announcing of badness is just taunting, because the resolution engine does not in any way guarantee the players an actual meaningful capacity to influence that badness or the threatened outcome - it all just depends on how the GM has made decisions about time and space, and how those all combine to deliver a yes or no answer.
Here's Vincent Baker's basic theory of
how to get narrativist RPGing:
After setup, what a game's rules do is control how you resolve one situation into the next. If you're designing a Narrativist game, what you need are rules that create a) rising conflict b) across a moral line c) between fit characters d) according to the authorship of the players. Every new situation should be a step upward in that conflict, toward a climax and resolution. Your rules need to provoke the players, collaboratively, into escalating the conflict, until it can't escalate no more.
Character creation in a Narrativist game might work by creating characters who, in some key way, have nowhere else to go. Life o' Crime, the rpg: create a character who owes somebody more money than he can repay.
Setting in a Narrativist game might work by applying pressure to that key point in the characters. Life o' Crime: there's recession, few jobs, no way up or out, but worse class difference than ever before anywhere. You see wealth but no opportunity.
Situation in a Narrativist game works by increasing the pressure. Life o' Crime: Someone depends on your character to bring home groceries and pay rent. Someone else has just been evicted and is facing homelessness. Someone else asks you if you know where to get drugs. Someone else just got beaten by the authorities. Someone else just got beaten by the guy you owe money to. Someone else offers to cut you in on a job. Someone else wants the whole take for himself. Someone else knew you'd never amount to anything. Someone else can't be trusted. Someone else can be.
System in a Narrativist game works, again, by resolving one situation into the next. Life o' Crime: what do you do? How does it work out for you? Does it a) hurt? b) give you breathing room? c) piss someone else off? d) hurt someone else? and/or e) set you back? How does it increase the pressure? Remember the moral line defined by your Premise, and remember that the players are the authors!
And Color permeates a Narrativist game same as any other. Life o' Crime: is it Thatcher's England? Victoria's England? Shakespeare's England? Bush's US? Hoover's US? Colonial Massachussetts? Mars? The Kingdom of Thringbora? The details change, but the core of character situated in setting - the fit characters locked into conflict defined by a moral line - doesn't.
Many D&D characters are not, by default, "fit". It's not a coincidence that the narrativist AD&D I was GMing in the second half othe 1980s involved thieves, and OA bushi and kensai: AD&D thieves are fit (not unlike Baker's own "Life o' Crime" example) and so are the OA characters, with their family loyalties and tensions and martial arts masters and rivals.
D&D does not make it easy to create pressure via setting and situation, in part because the default setting is one in which the PCs are wanderers or strangers, and in part because the GM often presents the setting or situation with an answer already in mind - the players will have their PCs choose the plothook and take up the quest. The narrativist AD&D I was running eschewed plot hooks and adventures, and used the hooks built into the PCs to create situations out of the setting that allowed the players to choose "across the moral line" (in OA, very roughly, honour vs glory; in the thief game, again pretty roughly, good-hearted roguishness vs genuine villainy).
And perhaps most significantly, the system in D&D doesn't easily establish how things work out, other than hit points lost and gear or spells expended. It is hard to get binding results of breathing room, or social or emotional set-backs, for instance, because (i) the process of action declaration in D&D doesn't make it easy for a player to put this sort of thing at stake, and (ii) a GM in D&D has a lot of liberty to decide how things happen more-or-less independently of how the actions resolve.
As I've posted already, and reiterated in this thread, I've GMed narrativist AD&D. And also narrativist RM. But I wouldn't go back to either system now, because better systems are available. For me, that's Torchbearer and Prince Valiant rather than AD&D (some might opt for Dungeon World instead), and Burning Wheel rather than Rolemaster (The Riddle of Steel would be another option in this general space, I guess).
