GM fiat - an illustration

This thread is about identifying and analysing GM fiat as a way of establishing the shared fiction in RPGing. My experience is that, often, many cases of GM fiat go unnoticed, or at least unremarked upon. Here's one example, that I posted in a recent thread:

pemerton said:

Rolemaster is a very rules-heavy RPG. If I declare that my PC goes to the market hoping to bump into my long-lost sister; or hoping to find an angel feather for sale; what is the chances of success? How do I (as a player) even know that there is a market place for my PC to visit, or that my PC has a long-lost sister, or that angel feathers exist in the (imaginary) world? That is all up to GM fiat.

Click to expand...

In this post I want to present another example, and say a bit about it. My example is the Alarm spell, from D&D 5e:


On its fact, this spell looks like something that a player could use to help control the risk environment for their PC. But on closer analysis, it turns almost entirely on GM decision-making that is significantly unconstrained.

For instance,


All of this depends on GM decision-making. That decision-making is largely unconstrained, except by some pretty loose notions of "fair play". By choosing to use the spell, does a player actually affect the risk to their position in the game? Does this happen in any way other than by invoking the GM's notion of "fair play"? Perhaps if the GM is relying on a very precise timeline for introducing threats, the 1 minute and/or 8 hour issue might be obviated. But that still leaves the other issues.

Here is a superficially similar spell from a different game - Torchbearer 2e's Aetherial Premonition:


The fiction of this spell is very much the same as that of the D&D Alarm spell. But the gameplay is different:


It's possible, in TB2e, for a wily intruder to avoid the alarm, but that would be a narration adopted after the camp event roll is made and an unhappy event results despite the bonus. And it is possible for the watch to be too distracted or drowsy or whatever to effectively respond, despite the alarm; but again, that would be a narration adopted after their test to avert disaster fails, notwithstanding the +1D bonus.

The GM is not at liberty just to narrate things in such a way that the spell makes no difference.

Some RPGers might prefer the GM fiat-free Torchbearer 2e approach; others might prefer the approach of the Alarm spell, which puts some parameters around the GM's narration (eg the GM can't just narrate someone wandering into the warded area 4 hours after the spell is cast without also narrating that the alarm is triggered) but otherwise leaves the GM free to introduce a threat, or not, that does or does not trigger the alarm, as they see fit.

But I think the difference between the two approaches is clear.

I'm a little dense. What's the point here?

As always, this fails to consider limitations on what the GM might "narrate" that live outside any individual act of resolution. The states in which the GM might or might not "introduce a threat" are not necessarily unknowable, unknown, nor generally equivalent outside of the alarm spell. Neatly dividing the world between "a die roll" and " the web of interacting consequences being simulated in the GM's brain" and calling the latter fiat is way too reductive.

There's frankly more difference in how two GMs with differing understanding of their roles might determine what happens next, and the relative agency of players in their games. An alarm down might be an invitation to trigger it, a request to have problems happen or it might be a carefully calibrated precaution that renders surprise impossible. Unless that +1 is capable of overwhelming the RNG on that die roll, it cannot be the same decision as the latter case, and unless the camp roll doesn't happen every time, it cannot be the former.

The space being called "fiat" here is vast and almost certainly subject to more constraints than the name and comparison suggests.

GM Fiat is definitely a thing in most games. At least most games that include Rule 0.

GM fiat always exists to some point or another, it is like the argument about metacurrencies in games, and then looking at hit points, they have always been there to some degree.

Pedantic said:

As always, this fails to consider limitations on what the GM might "narrate" that live outside any individual act of resolution. The states in which the GM might or might not "introduce a threat" are not necessarily unknowable, unknown, nor generally equivalent outside of the alarm spell. Neatly dividing the world between "a die roll" and " the web of interacting consequences being simulated in the GM's brain" and calling the latter fiat is way too reductive.

There's frankly more difference in how two GMs with differing understanding of their roles might determine what happens next, and the relative agency of players in their games. An alarm down might be an invitation to trigger it, a request to have problems happen or it might be a carefully calibrated precaution that renders surprise impossible. Unless that +1 is capable of overwhelming the RNG on that die roll, it cannot be the same decision as the latter case, and unless the camp roll doesn't happen every time, it cannot be the former.

The space being called "fiat" here is vast and almost certainly subject to more constraints than the name and comparison suggests.

Click to expand...

To me this sounds like another iteration of "system doesn't matter". Except it does! Even if you assert that whatever the GM is doing is the most significant thing, you can STILL discuss the differences between the examples of rules given in the OP, right?

And I find it hard to argue that the more story-centered focus of the TB2e version has no impact on the question of where the authority to actually resolve these things rests.

As for the assertion that it doesn't matter because no true ScottsGM will ever do X, Y, or Z... Let's just agree that we differ on that, as I've played under some fine GMs who absolutely 'managed' things by exactly the sorts of techniques @pemerton suggests. Again, good GMs, top notch, I am not criticizing those techniques, just noting they're not some hypothetical spherical cow.

AbdulAlhazred said:

To me this sounds like another iteration of "system doesn't matter". Except it does! Even if you assert that whatever the GM is doing is the most significant thing, you can STILL discuss the differences between the examples of rules given in the OP, right?

Click to expand...

No, precisely the opposite. My point was that the fiat-less mechanic demonstrated cannot do some of the things the alarm spell as presented could. I am absolutely not making the system doesn't matter point, I'm saying this fiat/no-fiat divide isn't a particularly incisive divide (or at least, it's insufficient by itself).

AbdulAlhazred said:

And I find it hard to argue that the more story-centered focus of the TB2e version has no impact on the question of where the authority to actually resolve these things rests.

As for the assertion that it doesn't matter because no true ScottsGM will ever do X, Y, or Z... Let's just agree that we differ on that, as I've played under some fine GMs who absolutely 'managed' things by exactly the sorts of techniques @pemerton suggests. Again, good GMs, top notch, I am not criticizing those techniques, just noting they're not some hypothetical spherical cow.

Click to expand...

I'm not making that argument. I pointed out two specific cases that are not possible with the mechanic demonstrated and both exist in the realm of "fiat" suggested, but are completely different in play. My point was that the realm of "fiat" thus described is not monolithic, nor particularly useful as a category division. The two cases I presented were these:

1. The alarm spell is a flag presented to the GM, announcing "I am interested in being attacked at night" and puts a request or obligation for them to facilitate that happening. In this hypothetical game, the constraints the mechanic puts on the GM are about what kind of thing they should present to the players.
2. The alarm spell is part of a larger chain of preparations, that based on established circumstance, cuts off the possibility of any ambush occurring. In this game, the GM is constrained by some other set of expectations about the fiction or chains of pre-established events/causality.

My point was that these are so different that lumping together as two examples of the same thing is too reductive to be useful. Imagine a player expecting the latter case in the first game, or vice versa. They would be getting literally the opposite of what they expected/desired when the mechanic was invoked. They aren't identical cases, and require different actions from both players and GMs, and lumping them into the same category on the basis that GM and/or have roughly equivalent "power" in both vs. the TB example leaves too much out.

My response as GM reading both spells from the two different systems is..."challenge accepted!"

The OP has two very different examples of DM fiat. One is story, one is rules. If you go to the market looking for your long lost sister, and she doesn’t live in that city, DM Fiat, you don’t need to roll, she’s not there. Same with the Feather, if it’s not there, it’s not there. Not everything has a chance of success, unless you have quantum mechanics game rules. But if you’ve tracked the sister or feather to this market, maybe GM Fiat makes up the odds on a roll, for bumping into her or finding the feather or maybe for story or fun reasons yes she’s there or no she’s not. GM Fiat shouldn’t be a dick about withholding what they’ve worked for nor should it just give away stuff w/o work. Maybe it takes a couple Tuesday market visits to find the sister, you don’t know the odds, sure, but it’s doable.

Edit, Fiat or not, if there’s a roll, I always reveal the target for success, not always before they decide to roll, but always after there’s no backing out, that’s the drama of the game. You can Fiat less bad failures, or less good (oh but then) successes, but you can’t Fiat meaningless roles or the game is broken.

The second issue is a long long rulesy thing. On rules, my take on GM Fiat is that you help the players do what they want to do, and round off the corners of the actual rules when necessary. For sure, sometimes my ruling is, “Dude, you clearly don’t know what “and” means, maybe you’d like to take that action back”. Sometimes though, close enough, works. Sometimes, “I don’t think that’s how that spell works, what’s your reasoning as to why it does”. And never is it a recourse to the internet to see how anything has been litigated before. Come up with a plausible reason, great, ok. And later I’ll look it up and maybe explain why, no, can’t do that again cause “exploitable in xyz way”…or maybe I don’t and see if they figure out the exploit and let them do it once.

Edit, but I see you want edge cases, and some rulesyness. Then it becomes about was the intention good, but technically flawed, like did they want to block the hallway with an alarm on a VTT but set it two fat too far to the left, dick move to squeeze a shapeshifter through, but has the first watch not slept in two days? Ok to GM fiat some odds of falling asleep.

GM Fiat is there to move the game along. It’s, sure, about trust, but trust in that whatever the GM is ruling is about making the game more fun, or more interesting, and keeping it moving, and yeah, won’t always agree, but trust that their brain a heart is in the right place. If you’re a GM and that’s not your motivation, stop being one.

I always thought that an "illustration" was meant to simplify data and information.