On Narrative Magic

My posts on magic, especially the one on flexible magic, tend to be quite popular, despite them not being on topic. But whenever I write them, a little voice in the back of my head objects.  It says something like this:

"But that's not what they want to use magic for.  You don't understand what they're trying to do with magic."

A lot of people who advocate for flexible magic are not generally advocating for it because they want to have infinitely flexible problem solving power, allowing them just invest in a single stat that will solve all of their problems.  Instead, they speak of adding greater flavor to their stories.  They want to run a game that feels occult or like a fairy-tale, and they feel like static magic doesn't offer them this opportunity, while flexible magic does.  They see my arguments about problem solving as moving in the wrong direction.  I actually address this in my post on Flexible Magic, but I thought I might address it in greater detail here.

First, before we get into the rest of this, I want to hedge that I disagree with the notion that static magic can't feel like a fairy tale or like occult horror.  For example, there's a Pyramid article with Path/Book Magic that's explicitly geared towards fairly tale tropes, and Path/Book magic in general is very occult. 

But these still create very solution-oriented magic that's repeatable, when that isn't what they want.  What they want is something that allows them to express their immediate creativity, to create magic that feels unique to the moment or the story. They don't want "I cast sleep.  I cast sleep again." They want this more dynamic thing where they might cast sleep twice, but the nature of the sleep spell, and how it interacts with the setting and the world changes.  The first time it might be a quick, light slumber spell to help someone sleep while their dreams are safeguarded from ghosts, while the second time it might be a dread curse (that also preserves the target in their current age and health) that keeps them asleep until the spell is broken.  Static systems struggle with this sort of nuance and detail, so they naturally reach for a dynamic system.

But I think the problem here isn't actually the dynamic system, it's the mechanical dissonance. They're trying to use a narrativist mechanic in a gamist, solution-oriented game system, and the net result of that inevitably creates the sort of problems that I complain about in  my flexible magic post, where people realize that magic is the One Stat to Rule Them All, and just rolls magic until the problem goes away.  There is, however, a solution to this that allows the creative player to have their creativity without breaking the system they're working with.

A Quick Primer on an Outdated RPG Theory

Before I dive into this explanation, it's an outdated and contentious concept from the Forge called "GNS theory." I don't think they use it anymore, and most people don't even understand it, or use it wrong, and doubtless this will generate some comments from a certain crowd of people, but the reason I want to use it is the same reason the Forge created it in the first place: to highlight different ways to think about gaming, and thus more clearly differentiate what they were trying to do at the time from how other games worked.  I don't want to argue that this theory is "correct," as I have some problems with it. It's certainly not the One Great Truth of RPGs, it's just a useful model to explain a single, critical concept.

Traditional-Gaming: Solution-Oriented Mechanics

I was tempted to call this "Adversarial gaming" but that might distract from my core point.  Even so, thinking about it this way helps us understand the specifics of how the typical RPG works.  

• The game master presents a problem.
• The player character proposes a solution
• The character's stats determines their chances of success with the proposed solution
• The dice represent the role of luck in determining the final outcome.
• A successful solution rewards the players while a failed solution punishes the players.

In this traditional model of gaming, the game-master represents the opposition to the players, the world that the players seek to defeat.  The players have an objective and the GM outlines opposition to that objective, be it an enemy NPC or a locked door.  The player seeks to acquire traits and abilities on their character sheet that help them overcome this opposition, improving their chances.  People say you "can't win RPGs," but you kind of can, especially in a traditional RPG model.  A group of heroes who go through a dungeon and are defeated "lose," and a group of heroes that defeat the big bad at the end and take the treasure "win."

If a character faces a locked door, they may want to get past it; perhaps it has a nice treasure on the other side.  They might be able to pick the lock or bash the door down or cast a spell to open the lock. The player will choose the solution that their character is most optimized for.  A barbarian is unlikely to try to pick the lock, and the thief is unlikely to try to bash the door down.  If they choose a "sub-optimal" approach, they risk failing to open the door, and thus lose the treasure on the other side.  The consequence of that might be that they don't get as much gear or experience, which means they'll have a harder time later on.  Thus, they're highly motivated to "succeed" and make the right choice and thus successfully open the lock.  Opening the lock is interesting and fun, failing to open the lock is punishing and not fun.

Nouveau-Gaming: Narrative-Controlling Mechanics

One of the purposes of the Forge and their GNS theory was to clearly articulate a different way of thinking about gaming. This differed sort of gaming might be thought of as "cooperative gaming" in contrast the the "adversarial gaming," I sometimes also see it called "GM-less gaming" as if you use this approach, someone representing the world becomes less important. This variation of gaming looks more like this:

• Someone presents a situation
• A player proposes a complication or additional context
• Other players also propose additional complications or context
• The dice (or other resolution mechanic) determines which version of complication and context gets added
• Successfully adding a complication is more of an artistic choice than it is a reward.

This model of "resolution" more resembles negotiating a shared narrative. For example, our heroes venture through a dungeon, and come across a locked door.  What's on the other side? I dunno.  What do you want to be on the other side.  You could say treasure, but maybe it's a big monster, locked behind this door to keep the world safe from it. Which is more interesting?  Perhaps failing to open the door is the best result, because the world will be safe, then.  But on the other hand, it might be way more interesting if the monster got out!  So perhaps we do succeed at opening the door, and perhaps casting a spell is the only way to do it, even though we're not good at it, so we cast the spell and open the door, but only partially! So the monster doesn't get out, but its influence does!  And it possesses my character, but only partially! So now I'm cursed with a dark power!

In this model, "reward" is more the chance to shape the story in the way you'd like.  There's far less winner/loser here, because everyone can win if everyone likes the proposed elements.  Players can negotiate with one another and come to an agreement, removing the need to roll dice or spend resources.  This sort of game gets strange, because you quickly realize you're really just jamming a story with your friends and the mechanics serve more as inspiration; resolution mechanics like flavorful tarot cards that provide interesting inspiration or random result tables become more useful to such a game than resource management or dice rolling.

Narrative Magic vs Solution Magic

So, like, who cares? Some hipsters play weird games. What does this have to do with GURPS? Well, the core point here is that these two approaches can be antithetical to one another. If you're in a purely narrativistic system, then the solution-based mechanics become redundant: players don't actually care if they succeed or fail at their lockpicking roll, because failure can be as interesting as success, but if you make it so they do care, the ability to just add whatever details they want undermines a lot of the tension (if treasure always makes your character better, and being better matters, and you get to decide what's behind the door, you'll always choose for as much treasure as you can possibly carry, and never for a slumbering Cthulhu).

And this is where we start to see our problem with these more "narrative" magic systems impinging on solution-oriented gaming.  If you're playing the first game, and you have magic that can "do anything you want" then you want that magic to unlock the door to get the treasure.  And also to verify that there is treasure on the other side, and not a monster, because if there is treasure, you get rich and can buy more magic, but if there's a monster, you get dead, and you don't get more magic.  Thus, the solution-oriented system incentivizes you to remove complications rather than add them, and a narrative system needs you to add complications to keep it interesting.

Hence, if you're playing a game like GURPS, no narrative magic for you.  At least, not without some core, critical elements and an understanding of what narrative magic is for.  

First, if the magic doesn't directly solve problems, it can't really impact the "solution-oriented gameplay" of the traditional game.  Games are often built around core restrictions, like you can't teleport directly to the end of the dungeon in a dungeon crawling game, wish your enemy dead in a martial art tournament, or look back in time to see the crime in a whodunit, because these things would invalidate what the core gameplay is about. Whatever your magic is, as long as you can't "creatively" violate whatever your core gameplay mechanics are, it should be alright.

Second, if it does directly solve problems, it should itself have some limitations on it.  It's probably best to make it rare. As we'll see, the process is likely involved and can distract from the core gameplay if it's called upon too often.  Furthermore, if we allow it to directly impact gameplay, having a single "ace up your sleeve" means people are unlikely to abuse it: you could magic that door open, but maybe you'll need it later on!  You can also build some interesting gameplay elements into it: perhaps the character needs to collect particular resources to use it, or the nature and shape of what you can do with it depends heavily on the stars or some other external factor ("I cannot open the door, for now is the time of secrets, but I can know what is on the other side, and if there is another way into the room.").

Third, it should involve negotiation with the GM, and it should always involve adding complications to the story.  What we're ultimately trying to do is inject some of that narratavism into our solution-oriented game.  We want to add that negotiation-oriented gameplay where the players join the GM to come up with a shared world (or at least a shared moment) together.  That means the player should not be unilaterally declaring the effects of their magic, but together, as a group.  This means that the GM needs to have some say in what happens, that the net result should be both a positive, a solution, and a drawback, a complication.  Perhaps the wizard can open the door with a spell, but it will cost them something dear, or while there will be treasure, it will unleash whatever evil is beyond the door.

Finally, I wouldn't use this to replace repeatable, predictable, solution-based magic if you still wanted magic to be a standard, possible solution to your game mechanics. The point of this "narrative" magic should be to add an opportunity for creative complications, not to be how you solve problems.  In fact, it should not be how you solve problems, at least not a core means of solving problems. You should be using this magic to do big, narrative things, and use an unlocking spell to unlock doors.  This allows you to retain game-style "series of interesting choices" mechanics with your magic, with wizards counting their magic points or spell slots and rolling for success and eagerly acquiring new spells, while having an option for some high level, complication-adding magic or magical components.

 A Model of Narrative Magic for GURPS

It might be tempting to dismiss this as a possibility for for GURPS, but in fact, GURPS has several possible models for implementing exactly this sort of magic.

The first is Divine Favor, as I alluded to in my post on Flexible Magic.  It has repeatable magic in the form of learned prayers, which have strict rules that you have to follow, and beyond that, invoking God is difficult and slow and has rules that prevent you from doing it too often, and the results are ultimately up to the GM, and thus inherently involve negotiation (even the roll to perform this form of "magic" is a petition roll!).

Words of Power are similar.  They cost a staggering amount of fatigue, which means you're rarely use it (and regret it when you do), and the GM, not you, determines what happens, though this means the player has less of a negotiating position.

Finally, there's a go-to mechanic in GURPS for whenever you want direct player input on a scene: the scene-editing mechanics of Serendipity. Serendipity is inherently limited (once per session) and is inherently based on negotiation (the GM ultimately decides what happens, but the player can ask).  We can expand this to make it even more of a negotiation process.  We can add Wishing +100% to explicitly allow the player to have a direct say in what happens.  It's not "I use Serendipity, can the princess fall asleep?" it's "I use Serendipity so the princess falls asleep."  Given the scale and scope of what we're talking about, some level of Cosmic (+50%?) is probably applicable: you can do a lot more than what Serendipity suggests: this is not a coincidence, but an explicit and powerful scene-edit.  We need to make sure that this is a negotiation, though, and not just the player enforcing his will, so we can add a Nuisance Effect -10%, something like "There's always a consequence or price to pay" that the GM will determine, but the rule description should include some specific language about negotiating back and forth.  It's obviously Magical -10%, and probably involves an IQ Roll -10%, likely some sort of magical skill.  The skill roll is not explicitly necessary, but it's likely expected, and it would encourage the player to monkey with modifiers; however, I would suggest writing the rules so as only a critical failure outright invalidates player intention: on a critical success you can escape (or diminish) the nuisance effect; on a failure, you get a much worse side effect in addition to the normal side effect, but the basic application still works.  On a critical failure, you get the drawback, but not your intended effect.  We could add some additional costs, if we wanted; it would be in genre for it to have an extensive fatigue cost!

The final cost of the trait would be 33 points.

The idea here would be that the wizard can do anything once per session, but there's always some cost, and he has to negotiate with the GM as to what he can do and what the drawbacks will be. The drawback is the primary check on this power: You can wish your enemy dead, but you risk the GM responding that your enemy became undead.  You can wish the door unlocked, but the GM can suggest that this will unleash a terrible evil.  You won't fail at your intention, but you might unleash something really bad.  Your magic is unpredictable and it tends to complicate things. But you have the room to define whatever magic you want.  Indeed, the back-and-forth negotiation will encourage it.  This complexity, and the single use, makes it unlikely that the wizard will use it often.  The wizard becomes  a Gandalf, who mutters about how his dark power could solve the problem "but at what cost?" and actually backs that up!  When he does finally use it, the results will be wondrous but terrible, and will result in the player and GM collaborating on a scene edit rather than just solving the problem with a die roll. It's also not going to break the bank for a player character.  You could easily get this as a power-up for a DF Wizard, if you wanted.

You could create more of a system around it.  Another nuisance effect, Environmental, or some Aspect modifier could limit the flavor of the magic, or make it dependent on the state of the world, further reducing its cost and focusing the player's inspiration.  Instead of being "I can do whatever I want, but with some consequence" it becomes "I can do whatever I want in this arena, with this theme, with the results shaped by invisible variables."  That should be more than occult enough for most people.

Addendum: Not MY Flexible Magic

So at the beginning, I commented that "a lot" of people use their flexible magic for these purposes.  Doubtless, some astute reader has an objection to raise: actually, they use flexible magic for this other reason!  This is true: people use flexible magic for lots of reasons. I could have mentioned this at the beginning, but this would deprive such a reader of the pleasure of seeing that they're so right that I've already pre-emptively agreed with them once they got to the end!  So let's go over some of these reasons and discuss how I feel about them.

Flexible Magic is Easier

Rather than memorize a bunch of specific spell rules and a ton of modifiers, a flexible magic system can allow someone to roll some dice, and the GM and he can just figure it out together.  It's much easier to learn and to GM.

The problem is this still runs into all the sorts of problems I highlight above and in my Flexible Magic post.  In fact, what I've noticed is that most systems that introduce a flexible magic system in their first edition invariably complicate it and limit it in subsequent editions until it starts to look like a static magic system for this very reason.

I think if you want an "easy system" I think you should still consider taking my advice above and understanding what the limitations of it should be. You can have that simple-to-adjudicate system, but not if you let players use it to run roughshod over the core conflicts of your game mechanics.

Flexible Magic as a Design System

I've seen a lot of people use Flexible Magic less with the original intent of "You can create anything you want on the fly" and more as a means to create their own repeatable, static spells.  This is more-or-less the direction Mage: the Awakening went, and I would argue it's a lot of the appeal behind RPM, which has more thorough guidelines for static spell creation than GURPS Magic has, but less tedious than Sorcery.  I find RPM is too exhausting it is spell creation to really allow "Flexible Magic is Easy" that tends to be the core draw, but it works fine as a design system.

I think people using it as a design system can simply carry on doing so.  I don't really see a problem with it.  They should be aware of what impact the spells they allow has on the mechanics they set up, but that's always true.  They don't need the flexible magic system I describe above, though they could use it as well as game with a more explicitly static spell system.

Flexible Magic is for NPCs

I've done this myself: if I have a fantasy setting with mysterious magic, then swamp-witches and mountain-wizards can do whatever I need them to do.  Their magic is mysterious and hard to anticipate, but follows a certain fairy-tale logic.  The players can't predict exactly what their spells will do, because no spells are written down anywhere. Magic worked like this in Cherry Blossom Rain, for example.

I think this sort of magic is fine.  It has a lot going for it, especially in that it makes the enemies less predictable.  If your magic is too unpredictable, if players get the sense that wizards can do literally anything, they may become frustrated and feel powerless.  For example, a "locked room" murder mystery is a lot less interesting in a setting where wizards can do literally anything, because then the answer is always (or could always be) "A wizard did it."  But if you take that caveat, keep your magic limited, and more-or-less repeatable, and define things that need to be defined so players can know at least some of the rules, I think you'll be fine.

By definition, NPCs don't need to pay point costs for things unless you're being a stickler on Ally or Patron values, and they'll only do whatever the GM wants them to do, so sure, the swamp-witch could solve the problem with a wave of her hand, but she doesn't, because reasons, and so the players are allowed to shine as they solve the problem.  So I don't think NPCs need a strict system like the above.