Stakes, Agency, Consequences

This post comes out of watching a lot of The Magicians (haven’t read the books) and listening to a lot of Ken and Robin Talk About Stuff. Together, these have me thinking a lot about the narrative loop of stakes, agency, and consequences. I want to talk about these, and how we can create richer, more varied narratives.

Stakes to the Heart

I’m pretty sure Robin Laws gave me (not, you know, me… all of his listeners) the formulation “stakes are the thing we hope will happen and the thing we fear will happen.” This is a great functional definition, since it means we can talk about the stakes of most conflicts by defining just two things. Once you’ve outlined the stakes for multiple conflicts, you might look back and realize you’ve written “the protagonists kill all of their opponents or the protagonists die” for every single one. Some variation might be nice, don’t you think?

This is some of the oldest GMing advice in the book: change up the goals of your fights. Not everything should come down to a fight, though; social interaction and exploration are explicit pillars of 5e gameplay. So let’s look at exploration stakes, starting in the form “we find a location in the wilderness or our supplies run low and we have to turn back.”

Now, having a ranger, or anyone with a strong bonus in Survival, in the party makes the failure state almost impossible. If the audience has no concern whatsoever that “the thing they fear” could happen, there’s no tension. Therefore, if running out of supplies is no danger, you need to introduce an alternate danger. Maybe it’s “the power of the Dark Forest (or the One Ring) takes hold of our heroes’ minds.” If you have a group of dramatic roleplayers, maybe it’s “tempers flare within the party and they split up or refuse to cooperate.”

There’s nothing wrong with deciding that you don’t have any interesting tension for a particular stretch of time. One quick narration later, you hurry along to the part that has compelling stakes again, rather than losing the players’ attention. Maybe you give the party ranger a minute to, in a sense, dunk on the challenges posed by the environment: how they make camp each night, how they use their spells to navigate and forage, how they keep watch for dangers. (It’s probably better if you split this up among multiple party members, so that everyone feels gets to contribute.)

Okay, but starving is far from the only danger of an overland journey, even one where you never enter combat. A rickety bridge across the chasm might offer stakes like “we all make it across or someone falls, or else we have to spend time and take further risks to build a new bridge.” Maybe you have spellcasters in the party who can solve this exploration problem with magic: fly or dimension door or wall of stone. Spell use in D&D is limited, though. Solve a lot of problems quickly with a spell, and you might run out. To put that another way, you resolve the stakes quickly by accepting a (probably minor) consequence: “some of my power is gone for the day.”


Grade-A Stakes

One of the great ways to vary your stakes and increase tension is, paradoxically, to put an NPC rather than a PC in the crosshairs. Superficially, it seems like PCs will care less if the fear is a bad thing happening to someone else. If your PCs really don’t give a rat’s ass about anyone but themselves, then yeah, you might be right. But it’s not that hard to get the PCs to care about NPCs, if you start by making the NPCs care about them: as a cheering section, as a source of funding, as a source of legitimate authority. The game-design shorthand for that idea is making NPCs a stat on your character sheet. I mean, who wants to lose stats?

Once you’ve been playing games for awhile, you start to realize that “my character lives or my character dies and I roll up a new one” defangs a lot of conflicts. If the GM really put a PC in the crosshairs and threw everything at them, we immediately get to “the game isn’t fun because obviously the GM can always just win.” The game functions at all because of a social contract that the GM won’t do that: that challenges too great to be fought or reasoned with can be escaped.

(PS. It’s possible to violate this contract properly! I’ve had two DMs do it to me in the last year! But you have to know what you’re doing and communicate with your players.)

There’s no similar promise of fairness around the lives of NPCs. The GM is always allowed to murder their own characters. Okay, sure, you don’t want to kill all of a player’s Contacts that they spent character points on during the first session. Still, you can threaten NPCs that the PCs care about with all kinds of bad stuff. If the PCs fail and you follow through with the stakes, you haven’t ruined anyone’s game. They can keep playing! The important thing is that, if they’re experienced gamers, they recognize what I’ve just explained, and the tension gets a lot more visceral.

To use a recent example, in Altera Awakens, the Crown Prince (now King) has been on-camera for both of the first two events. The first event opened with big assassination plot against him. Now, he’s the ruler of a kingdom that a lot of the PCs have super strong reasons to hate. So you’d think we wouldn’t care… except that in the first couple of hours of the event, the NPC established that while he might be thoughtlessly rude and awful in a lot of ways, he was a least-worst case. The kind of crown prince you can work with, as a foreign national… maybe even someone you could manipulate. With the rest of the setting exploding into chaos, maintaining the stability of even our hated enemies looks like enlightened self-interest. I don’t know that this situation worked for everyone, but it worked for enough of us to get the scene rolling. From talking to the Altera Awakens staff after the game, I know for sure that they were prepared to roll with the Crown Prince dying (and presumably staying dead) in the first event.

Think about what you’re threatening. Make sure it’s something that, if the PCs lose it, the game can go on. Give them things that you plan to threaten. Once you establish stakes, always follow through. Killing PCs is often low-stakes in D&D, because raise dead, so don’t feel too bad about diverting deadly consequences onto things the PCs can’t easily recover. (Check out Kainenchen’s guest post on this topic in Tribality!) In other games, killing PCs might be higher-stakes business, but are you imposing an excessive cost on the player, or ending their opportunity to play through a lot of other compelling stories?


It Stakes a Village

There are exceptions to following through on stakes, too: mercy and relief have their place at the end of a narrative, when you’ve established something else that gives a reason for the hammer not to land. That’s the fundamental principle of Leverage‘s flashbacks. We see the hammer arcing downward, then we see why our heroes’ cleverness means that it’s landing on the villains rather than the heroes or the people they’re helping. We feel the stakes and the tension, even though every. Single. Episode. involves exactly that reversal. The question of “how are they going to get out of this one?” takes somewhere more than 5 seasons to get old. (I assume a lot more. John Rogers, if you’re reading this, I’m another of the people praying for a Leverage revival.)

The other great way I see established stakes get violated is when the hero(es) has repeatedly tried to show compassion to an antagonist, only to be coldly rebuffed again and again. In books (Lord of the Rings), movies (Black Panther), and TV shows (Steven Universe), scenes that end in compassion without reconciliation at the time are functionally promising a bigger, more surprising payoff later.

For example, Tolkien’s overarching stakes for Frodo and Sam are something like “Frodo destroys the Ring or falls under its sway or else Gollum gets it back or else Sam has to sacrifice himself in Frodo’s place or else Sauron gets the Ring back or else the Gondorians take the Ring.” This is part of why Sam matters: to give Frodo another way to fail. He’s a substitute target that we want to see survive even more than Frodo, because he’s all the more innocent. (He’s also our practical side that distrusts Gollum. Characters can be, and express, more than one thing.) All of the other bad outcomes are just intermediate steps for “Sauron gets the Ring,” of course.

Each major event of their quest focuses on a different one of the bad outcomes. The resolution to the whole conflict is that Frodo falls under the sway of the Ring and Middle-Earth will fall under Sauron’s control. At the last moment, Frodo’s earlier, repeated compassion pays off, because he spared Gollum’s life. We had gotten a partial payoff for that earlier on, as Gollum halfway reverted to Sméagol, but that all went sideways. At the last, the punishing weight of the established stakes falls on Gollum, not out of any altruism inspired by Frodo’s compassion, but simply because he is still alive. Frodo did fail, in the end, but his earlier successes were enough to salvage the situation.


Add Agency

If the stakes are the boundaries of expected outcomes, the exercise of agency decides which one happens. We chiefly care about the decisions the protagonists make, so while their actions aren’t all that matters, they’re the main thing. This doesn’t mean we want the protagonists to get what they want. Many stories involve the audience wanting the protagonist to change what they want, or villain protagonists where we want them to be the agents of their own destruction.

The presence and exercise of agency doesn’t mean that the protagonists succeed. It’s completely within bounds for efforts to backfire because of insufficient information, force, willpower, or any other factor. The antagonists should score some wins, creating a new situation that advances the conflict and increasing the tension.

In gaming, the important part is figuring out what players can do to affect a situation. In fantasy adventure gaming, “stab the thing” is among the most common answers. Combat is exciting: built-in stakes, decisive outcomes, and clear applications of the tools on your character sheet. The obvious variation – the one we’ve all talked about for years – is just changing the thing the players can do to “seize the thing” or “destroy the thing” or “survive and escape the thing.”

The problem with combat as a primary expression of agency in adventure gaming is that because it’s decisive and there’s an expectation of a fair chance of victory, we’ve taught PCs to do everything they can to bring all conflicts to a fight. A dead enemy causes no more problems, and a fight can only end in a TPK or some irrelevant consequences that you can recover in approximately one long rest. Two at the outside. But I’m getting ahead of myself; let’s finish talking about agency first.


Noncombat Agency

Nonviolent solutions to problems can be incredibly satisfying, because they take a greater expression of creativity. To use yet another Altera Awakens example, Kainenchen’s character overheard a Duke in a series of conversations. The NPC was clearly up to no good, but way out of her reach for a head-on political conflict. Her character doesn’t any weapon skills or damage-dealing magic at all, and even if she had, a combative solution would have been socially untenable: a Duke in his own domain, with guards close at hand. Her exercise of agency was nothing more than a few words in the right ears about the Duke’s perfidy, but this was all the solving the problem needed, and the Duke’s whole plan unraveled.

Ideally, the GM doesn’t straight-out tell the players what they need to do to bring about what they hope rather than what they fear. The GM communicates the problem, and the players use tools at their disposal to improve their understanding of the situation. Developing their own plan and putting it into practice (in the face of the antagonists’ efforts) is the quintessential expression of agency in gaming.

It’s okay to present problems that the players can’t affect yet. Classically, the core conflict of a long narrative arc gets introduced well before the protagonists can solve it directly. Instead, they have to break it down into a lot of smaller problems that they can solve, or search for new tools or information that puts them on the path of smaller problems.

The Magicians presents a lot of problems this way, especially their major villains: the Beast, Reynard, et al. Despite the incredible powers at the main characters’ disposal, they’re so outgunned by the opposition (at first, anyway) that they can barely imagine how to fight them. But the setting is big, with room to run and hide even from gods while you try to come up with a plan. The solutions they come up with are… usually awful, but you can’t say that they don’t develop and implement plans!


Something of Consequence

Since I’m working with a definition of stakes as “what you hope and what you fear,” consequences are all of the yes-and, yes-but, no-but, or no-and elements that follow on that. The fallout of two different conflicts can vary wildly even if they both had the same stakes and agents. I would go so far as to suggest that this step of the narrative loop is where the majority of the emergent properties of narrative live.

The first thought that sparked this whole post came from how The Magicians emphasizes consequences from every single thing the characters do. I won’t dig into a lot of specifics because I don’t want to spoil the show’s many, many surprises. That may be a future post. Anyway, if you haven’t seen it, what you need to know is that the protagonists never resolve anything cleanly. There are always complications and dangerous loose ends. At best, the conflict costs them some kind of magical asset they have developed or acquired along the way.

It’s a show where the characters are hugely reluctant to engage their enemies in a fight. For one thing, there’s no promise at all of a fair fight in this setting. In fact, they almost never find one unless they’re fighting each other. For another, healing effects work very poorly, while curses work really well and are hard to break. If there are no other negative consequences on the table, you can count on one of the characters to display a horrible failure of judgment or another devastating character flaw.

Now, The Magicians‘s model is far from the only one. Let’s look at two other takes on consequences: superheroes and gritty fantasy. Superheroes pointedly shrug off most physical consequences, and (especially before you get to the darker comics) demonstrate their superpowers by rejecting dilemmas and securing the seemingly-impossible best outcome. I’m thinking of all the villains that hit Superman or Batman with “well, are you going to save those innocents or try to catch me?” Most of the time, Superman uses his nigh-infinite speed or Batman uses his preparedness and tools and does both.

Gritty gaming, including Warhammer Fantasy and Call of Cthulhu, are all about their dire and lasting consequences. They’re games of death spirals: you might be able to remove some consequences with a lot of effort, but mostly every encounter ends in things getting gradually worse for you until you die and/or become a Chaos-corrupted bug (or whatever).

My point, then, is that consequences are primary markers for a game’s tone and many genre elements. The Magicians never quite seems to become a death spiral; instead, the characters are always staying on the move and shifting their tactics and alignments. Things aren’t ever really getting better or worse, overall; it’s more like the narrative version of a dolly zoom. They do interesting (but unbelievably dark, this show is content warnings for days, I am not fucking kidding) things with it, so changing-without-changing isn’t a bad thing.



I would like to see every kind of game, but especially D&D, expand their conceptions of stakes, agency, and consequences. I have plenty more I want to say about this in future posts, getting into specific game elements that help with that and more ways to think about consequences in D&D. Chronicles of Darkness 2e also deserves its own post, because its whole XP system hangs on consequences that follow conflict. I’ll end by linking some of my previous posts that have circled around a lot of these ideas, particularly in terms of consequences.

Wounds | Alternate Threat Tracks | Doom Tracks | Multiple Damage Tracks

Leave a comment

Your email address will not be published. Required fields are marked *