An example where granular resolution based on setting => situation didn't work

[pemerton]

There's a one sentence rule that says "in this case, make something up." This rule is explicit in one rulebook and implicit in another. The rule itself is vague- "do something" isn't very specific. But I think I see your point here. However, I find it very difficult to understand how not being explicitly told that the GM needs to do something in a particular instance, especially in the frame of NPC reactions, is a failure of campaign design.

I'm not sure which rulebook you're saying has an "implicit" twist rule.

There is no implicit rule in RM that, if a PC fails (say) a Perception check, the proper narration of a consequence is that their enemy shows up. In fact, that would be considered pretty outrageous, because it breaks the nexus between then at-the-table process of resolution and the in-the-fiction process of the character looking around.

I also think that there is no such implicit rule in classic D&D, nor in 3E D&D, and I've never seen it suggested that it was an implicit rule in 5e D&D. It is a rule, implicitly stated, in 4e D&D - the example which reveals it but doesn't expressly state it is the skill challenge example in the Rules Compendium (where a failed knowledge check, in relation to an abandoned building, has the consequence that an enemy NPC turns up).

I posted this back in 2012:

When it comes to out-of-combat resolution, the main requirement is to explain to GMs how to resolve the failed checks that will inevitably follow upon players making checks in which their PCs have poor bonuses. Burning Wheel does an excellent job of this. D&D, to date, has done a terrible job. Judging from posts I read around here, the default narration for the dwarf fighter attempting and failing the Diplomacy check is "You open your mouth and spray your spit over the mayor - sucks to dump CHA, I guess!" - and then people complain that their players won't use anything but their biggest numbers!

If the fighter fails the Diplomacy check, then there are any number of ways of narrating that failure without making the PC look like a fool - from "The mayor listens briefly, but then excuses herself to go off to the next meeting" to "Of course the mayor would love to help you, but she swore an oath to her late brother that she would never do XYZ" to "As you begin your address, rain starts to fall, and the mayor's entourage usher her back into the city hall before you can get your point across".

This example, and ones like, it - which is an example of the "twist" approach found in Torchbearer, although back then I was referring to other RPGs that use similar approaches like HeroWars and Burning Wheel - produce responses like "Why does my failed attempt at diplomacy cause it to rain?"

And still do. For instance, you can see references to "Schroedinger's encounter" or "Schroedinger' secret door" very often in posts made in recent threads on these boards.

 

[Baron Opal II]

I'm not sure which rulebook you're saying has an "implicit" twist rule.

I think they all do. It is a possible response to a failed roll. Not every one, but it is a response depending on the situation.

In all of your threads, you contrast two overall schemes: Map-and-Key vs. Narrative. A strength of the Narrative system type is that they are more specific in GM responses when PC characters fail / near miss / "Yes, but", &c. Here is a scheme to turn a failure into a story point. Here is a scheme to allow PC access to meta-currency based on their needs and behaviors. Here is a way to have gradations of success to allow you to introduce interesting narrative complications.

Map-and-Key tend to not have these specific schemes. They are more focused on abilities and rules to overcome challenges. The narrative grows not through deliberate or opportunistic beats but through clever use of mechanics and the vagaries of the dice. In my opinion, the best of these map-and-key games have emergent properties that can be discovered that facilitate play but don't overwhelm it. The difference between an emergent property and an exploit, if you will.

Having defined, or granular if you wish, abilities is not a detriment in of itself. The process of learning how to GM with granular rules could be instructed better, but I think determining how a group of NPCs will react is very close to the process of deciding how a group of PCs would act. Determining how may casters are in a group of NPCs is something the GM will have to derive from what the rules have mentioned and the aspects of worldbuilding / setting construction that apply.

 

[pemerton]

I think they all do. It is a possible response to a failed roll. Not every one, but it is a response depending on the situation.

<snip>

Map-and-Key tend to not have these specific schemes. They are more focused on abilities and rules to overcome challenges. The narrative grows not through deliberate or opportunistic beats but through clever use of mechanics and the vagaries of the dice. In my opinion, the best of these map-and-key games have emergent properties that can be discovered that facilitate play but don't overwhelm it. The difference between an emergent property and an exploit, if you will.

The passages on either side of my <snip> seem to me to be in tension.

I agree that setting => situation play, of which map-and-key is definitely a paradigm, does not use twists. But I think that contradicts the suggestion that (say) Moldvay Basic has an implicit twists rule. I don't think it does. I think the first RPG that I can think of that even puts something like twists on the table is Prince Valiant. And it was still very unusual. Over the Edge would obviously benefit from twists, but in the foreword to the 20th anniversary edition Jonathan Tweet notices that they're missing and observes that if he were rewriting he'd put them in.

 

[Baron Opal II]

I agree that setting => situation play, of which map-and-key is definitely a paradigm, does not use twists.

What? It absolutely does have twists. Sometimes caused by die rolls, sometimes pre-plotted, sometimes player generated. It is a basic aspect of narrative flow. No agreement here, apparently.

Do Messrs. Baker, Olavsrud, or Crane have any kind of acting background?

 

[pemerton]

What? It absolutely does have twists. Sometimes caused by die rolls, sometimes pre-plotted, sometimes player generated. It is a basic aspect of narrative flow. No agreement here, apparently.

I am using Twists in a (semi-)technical sense, to refer to the Torchbearer method, which is a version of the BW, DitV, Prince Valiant, etc method, in which what is narrated as a consequence need not flow, in ingame causal terms, from what the PC attempted. A slightly more generic term is "no whiffing" (Ron Edwards) or "fail forward", but the latter is now more often used not in its original sense but rather to describe a particular technique in railroaded play, so I'm hesitant to use it.

An example of such a Twist, that you brought into the thread, was my decision that a failed attempt at entering Megloss's house would mean that (as a Twist) Megloss turned up.

This is not a thing that is even hinted at in any D&D book I'm aware of other than the 4e Rules Compendium, by any Rolemaster book (which is the system mentioned in the OP), by Hero as best I'm familiar with it, etc.

As I already mentioned, the first time on these boards I suggested, as a possible consequence for failure (ie a Twist) in the context of a Diplomacy check, is that rain starts falling and the crowd disperses rather than listening to the PC, the response I got was "Why would my failed attempt at Diplomacy cause it to start raining?"

 

[AbdulAlhazred]

What? It absolutely does have twists. Sometimes caused by die rolls, sometimes pre-plotted, sometimes player generated. It is a basic aspect of narrative flow. No agreement here, apparently.

Do Messrs. Baker, Olavsrud, or Crane have any kind of acting background?

Yeah, a GM can introduce a plot twist, so to speak, in any RPG (well, trad ones at least). There are no rules covering the technique of introducing a plot twist as a response to a character failing at some task though, or as a form of currency (this is possible in FitD where a player can ask for a Devil's Bargain, an extra d6 in her pool in return for letting the GM describe some sort of complication which can be unrelated to the current situation). So in the BitD Devil's Bargain case, for instance, once my character was in a fight and really needed another die, and didn't have any resources left to get one, so I got a Devil's Bargain. As I glanced to my right, I saw, in the window of the adjoining abbatoir, the face of a girl from my character's home country, and my demon-sword expressed a terrible thirst for her blood! I got my d6, but now I got new problems...

 

[clearstream]

Yeah, a GM can introduce a plot twist, so to speak, in any RPG (well, trad ones at least). There are no rules covering the technique of introducing a plot twist as a response to a character failing at some task though,

Aside from those in the DMG. There should be consequences and there is a method to produce twists of varying severity. The parts on consequences are fairly well constructed. The parts on nuanced outcomes aren't. I think objectives with twists such as Diplomacy causing rain is that even DM is bound to narrate in a way that the group see as legitimate against their fictional position. And so there needed to be more in place for folk to accept that twist as legit. (Which doesn't mean it wasn't legit... only that folk couldn't see how it was!)

or as a form of currency (this is possible in FitD where a player can ask for a Devil's Bargain, an extra d6 in her pool in return for letting the GM describe some sort of complication which can be unrelated to the current situation). So in the BitD Devil's Bargain case, for instance, once my character was in a fight and really needed another die, and didn't have any resources left to get one, so I got a Devil's Bargain. As I glanced to my right, I saw, in the window of the adjoining abbatoir, the face of a girl from my character's home country, and my demon-sword expressed a terrible thirst for her blood! I got my d6, but now I got new problems...

There's no currency for twists in 5e that I can think of. Devil's Bargain is actually an interesting rule. If you set about designing a system for yielding twists from a fortune mechanic and saying what those twists are, and you want to avoid assumptions about who should say that, then you quickly notice design space for a mechanic that asks players to choose a twist.

Lately, I experimented with that for 5e. It was interesting, because I didn't say - hey, I want to get Devil's Bargain into 5e. I said - hey, I want to have a system that better says what consequences may include. Of course we already have the situation and what the player describes doing, but I wanted to go further than that. I came up with a method where player rolls 2d10 and if they need to use both dice to succeed, then they choose to narrate a twist or deal with failure. I immediately realised I'd recreated a form of Devil's Bargain. My version had some limitations so I went with another approach. It's also interesting to compare Devil's Bargain with some design choices in TB.

The limitation that most concerned me relates to the Czege Principle. With Devil's Bargain, I'm not sure if the bargain will get me what I want. It's just one die and I might still fail. For me that puts the bargain on fair-footing. With my take, the bargain guarantees success. I felt that put it on doubtful-footing. Possible design directions to explore from there are obvious of course.

 

[pemerton]

Aside from those in the DMG. There should be consequences and there is a method to produce twists of varying severity. The parts on consequences are fairly well constructed. The parts on nuanced outcomes aren't. I think objectives with twists such as Diplomacy causing rain is that even DM is bound to narrate in a way that the group see as legitimate against their fictional position. And so there needed to be more in place for folk to accept that twist as legit. (Which doesn't mean it wasn't legit... only that folk couldn't see how it was!)

The reason for objections in the posts is not the one you posit.

It's very straightforward: the objector said (to the effect of) Why would my poor public speaking cause it to rain? That is: the objector wants the consequences that are narrated at the table, which are caused by the failing of the check at the table, to be - in the fiction - things caused by the actions of the character. And the actions of the character, very obviously, don't cause rain to fall.

Another way to put it is this: at the objector's table there is no need for the AW principle, "Make your move, but misdirect." The purpose of that principle, in AW, is to remind the GM that when they say things because of what happens at the table, they need to introduce an appropriate in-fiction cause. There is no need for that principle if every consequence that is narrated at the table is caused by the actions of the character in the fiction, as in this case the in-fiction causation is always obvious without needing to be expressly elucidated.

 

[clearstream]

The reason for objections in the posts is not the one you posit.

It's very straightforward: the objector said (to the effect of) Why would my poor public speaking cause it to rain? That is: the objector wants the consequences that are narrated at the table, which are caused by the failing of the check at the table, to be - in the fiction - things caused by the actions of the character. And the actions of the character, very obviously, don't cause rain to fall.

Another way to put it is this: at the objector's table there is no need for the AW principle, "Make your move, but misdirect." The purpose of that principle, in AW, is to remind the GM that when they say things because of what happens at the table, they need to introduce an appropriate in-fiction cause. There is no need for that principle if every consequence that is narrated at the table is caused by the actions of the character in the fiction, as in this case the in-fiction causation is always obvious without needing to be expressly elucidated.

That reveals a few interesting differences in modes. One motive I have for positing what I do is a conversation I had with another poster (@iserith IIRC) relating to game text they had created outlining possible consequences of D&D ability checks. One example was of aggressive, noisy birds as a consequence for failing a climb check.

How do we know in a principled way what consequences are possible? At the time, my concern was that some of the suggested twists might not be well legitimated against the fiction of situation + what player described. As I understood the climb situation, the stakes in view might well not include birds. The poster's text I subsequently accepted as demonstrating how meaningful consequences could be made broader than failing the actions of the character. Birds could very well be included in a situation and it can then be legitimated as a consequence.

Additionally, In DMG 5e a roll is made only when there are meaningful consequences. As described in the supplemental text, the birds presented a meaningful consequence (noisy and perhaps dangerous.) Raining should be in view and meaningful for it to be legitimated, else it is window-dressing... the meaningful result in your example remains that the attempt to persuade fails. (Although I think there is also a social consequence at the table, of the dwarf not being characterised in a tropey way, which is a strength of the approach you advocated.) Following DMG 5e consequences resolution, performance need not be all that is at stake: failures in performance often only matter in DMG 5e in view of other consequences.

Another difference is that D&D traditionally drives a forwards-going narrative. In D&D, it's normally less acceptable to retroactively add birds to a cliff just because a character fails their climb check. Those consequences need to be in view up front. That's not true of game modes that take a result and can apply it retroactively. These differences are about principles of legitimation.

 

[pemerton]

@clearstream

I'm not a party to your discussion with iserith and others. But I was a party to the discussion about the Diplomacy example. Here is the reply in question (and here's a hyperlink - I'm not using the quote function as a poster from a 10 year old discussion doesn't need to be notified of this particular conversation):

I would be pretty pissed off if my dwarf's low diplomacy roll made it rain. I'm playing D&D, not Toon. I don't think the consequences should be unrelated to the cause.

That's not ambiguous. It is exactly as I posted above. It has nothing to do with "legitimacy" given the fiction, and everything to do with a principle about how in fiction causation and fiction introduced as part of consequence narration should be correlated.

How do we know in a principled way what consequences are possible? At the time, my concern was that some of the suggested twists might not be well legitimated against the fiction of situation + what player described.

<snip>

Raining should be in view and meaningful for it to be legitimated, else it is window-dressing... the meaningful result in your example remains that the attempt to persuade fails.

<snip>

Another difference is that D&D traditionally drives a forwards-going narrative. In D&D, it's normally less acceptable to retroactively add birds to a cliff just because a character fails their climb check. Those consequences need to be in view up front.

Three things:

* It is not retroactive to introduce it starts to rain - that is purely forward looking. It's just that instead of a random weather check, the GM has responded to the failed Diplomacy check.

* If you read the post I quoted, you will see that there is no complaint that (eg) the GM didn't telegraph with cloudy skies.

* Not every RPG follows the AW principle of hard moves follow on soft moves. In AW, narrating the birds in the tree, or the clouds in the sky, creates particular opportunities for subsequent hard moves, including if the player hands an opportunity on a silver platter. No such rule applies in 4e D&D, in part because there are no "silver platters" in 4e D&D: the GM does not get the opportunity to "follow through" in the AW fashion if a soft move is made and then ignored. What 4e D&D has instead is scene-based resolution in the form of a skill challenge, and narrating that it starts to rain is permissible.

Whether it's well-judged is a further question, but relevant considerations there would include broader questions of them, is the speaker a cleric of Melora (probably doesn't get arbitrarily rained on), etc. Not whether or not the GM had narrated some clouds.

EDIT: A further set of considerations:

* In AD&D, and I suspect in 5e D&D, narrating birds on the cliff at the opening of the climb check is an invitation to the players to describe how their PCs chase away the birds or put on bird-proof gear or something similar that engages with the granularity of the situation - and all that stuff is low- or no-stakes. So in effect, every time a possible consequence is flagged it invites the players to deal with it. Narrate clouds and they bring their umbrellas. As a result the space for failure narration is winnowed down more and more. In another recent thread @Manbearcat called this sort of thing, when initiated by the players ("Are there birds?" "Are there clouds?" etc) a "conversation trap" used by players to manipulate obstacle ratings.

* In AW, on the other hand, chasing away the birds is just another action, and either triggers a GM soft move or - if the circumstances are right - is Acting Under Fire and hence triggers a player side move. So it's not a way of "bypassing" challenge by hedging the GM in with more and more low-stakes fiction.

* 4e D&D would permit a "weather watching" attempt as part of a skill challenge to give a successful oration outdoors - a check on Nature, say, to affirm that no rain is expected - and depending on framing it might be a primary or a secondary check. This is different from both the above as a technique, but it locates the attempt within the resolution process, so it just feeds through to the outcome without any disruption or bogging down or "conversation traps".