Agency

(A lack of) Agency

Peter Parker made a choice when he witnessed the thief stealing from his wrestling manager. A player, acting as Jack, saves or harvests Little Sisters in BioShock. Agency reflects the desires and values of the character or player upon and within the fictional world they inhabit.

I've discussed matters of agency in video games twice now, but haven't really discussed the concept itself in-depth. So I decided to take a bit of a closer look at two kinds of agency present in video games, both character agency and player agency.

To make sure we're all on the same page, however, let's first come up with a simple definition. In this case, agency is the illusion of free will. It is the idea that the player or a character is acting in accordance to their own directive and choice, marching to the beat of their own drum.

While there are a lot of other more complicated discussions on agency within fiction and entertainment, this should give us a good, basic place to start.

Character Agency

Fiction relies on characters that the audience can, in some way, identify with. The easiest way to manage this is usually the same as in real life, by making a character funny or amusing. Making your protagonist a bit of an underdog typically helps as well, as everyone likes to see the character(s) accomplish something. A sense of growth and development, of conquering some personal trial.

All of this can shatter the second the audience no longer finds the protagonist(s) to be believable. Once the character does something that breaks the audience's understanding of the character's established personality or logic, then the audience is abruptly wrenched from the film, book, television show, or even video game. "That doesn't make sense," they may exclaim, watching as the characters decide the best option is to split up inside of the horrifying cavern. Or perhaps the girl that gets all A's in class sees the door is ajar and thinks "That's funny, I know I just locked it before bed," yet she simply closes the door and shrugs it off, proceeding on her predetermined demise.

Even antagonists are not immune, however. Look at the rogues gallery of James Bond villains and how many chose to leave him in an escapable situation rather than pulling the trigger themselves. These elements do not often reflect the actions of rational people, and are instead cheap gimmicks of suspense on the part of the screenwriter. Even if we were capable of forgiving such transgressions, the many questions of "why" would certainly come about in time.

Contrast this with Frodo Baggins in The Fellowship of the Ring. A simple Hobbit that enjoys tales of adventure, sure, but once he is informed of the true, dark nature of his uncle Bilbo's ring, he panics. He's been given what amounts to having the launch codes to the world's nuclear arsenal. Attempting to reject that power, he tries to shove it into the hands of Gandalf, the most powerful and wise person he knows. Fortunately for Middle Earth, Gandalf restricts Frodo's options. As with his uncle's adventure, the decision was made for him. The Bagginses are given a charge — a mission, should they choose to accept it.

Of course Gandalf makes it clear that not accepting is not much of an option. There are no others who can make the journey. With little choice, Frodo ventures to Rivendell. He is reluctant, but chooses to go forth and begin his adventure. The actions make sense, and we can accept the bravery because we've seen Frodo consider other options. As a result, when he makes the choice to venture to Mt. Doom, we can believe he is brave enough, despite his fear, to do such a thing. His decision is consistent with his earlier choices and actions in the film.

The illusion is that Frodo is acting of his own free will, when in truth he is merely a pawn in the game J.R.R. Tolkien is playing. It doesn't feel like that, no, but it is true. If Frodo did not behave as he did, however, if he didn't have to face reluctance and fear, we'd never buy into the story that Tolkien was trying to sell. Instead of being a story about struggles and strife, it would have been about guys fighting stuff in places with trees.

Player Agency

Have you ever been exploring the various corners of a map when, suddenly, you crash into an invisible wall? The road ahead continues onward, but the game prevents you from proceeding without giving any reason why. This is often labeled as being "immersion breaking," primarily because the player's sense of agency is broken. The player suddenly becomes aware that they are not truly free in this world, that the freedoms apparent aren't as they seem, and that they're simply following the breadcrumbs and restrictions left by a developer.

A player can never be given absolute freedom in a game — that'd be impossible to code — but a good design team will be able to craft a world that hides this fact. There is an illusion of free will as the player uses the tools and follows the rules provided to accomplish predetermined goals.

It sounds like a difficult task, but it actually varies on the complexity of the game. In the original Super Mario Bros., having to travel to the right is simply one of the laws of the game world. As you can only travel side to side to begin with, you're willing to accept this further limitation. The world makes up for it by allowing you to leap and find a variety of surfaces to walk or run on. In addition, the player can travel down pipes or up vines and into the clouds. While the player is still essentially following the same rule (progress must be made towards the right, not to the left), it feels like there are a variety of options within that ruleset. The player is not simply restricted to holding down the D-Pad, moving in a single direction the entire time.

Modern games, however, tend to become more complex and, as a result, can find it harder to construct that sense of agency. In the recently released Batman: Arkham Origins, there are a series of buildings that, despite appearances, the player cannot grapple onto. This is to prevent the player from accessing story content too early, but similarly breaks the sense of freedom. The illusion of freedom and choice is immediately broken. The player thought they could grapple to any ledge or surface in order to navigate the world, and instead they discover only surfaces deemed capable of being grappled. The player is now removed from the game world and sees it for what it is: an artificial construct designed to constrain the player to behave a certain way.

The original Arkham Asylum hid these points much better. Anything that could not be grappled onto either had spikes jutting out, preventing a grapple, or the surface was simply crumbling and too frail to allow it. The game communicated this reasoning. Even if the player recognized these gateways of progression, there was a clear and visible reason that made sense within the game world. In Arkham Origins, the roof of one building can look no different than another, but Batman cannot grapple simply because the designers do not wish for him to do so.

Dishonored, on the other hand, allows the player to travel nearly anywhere with the Blink ability. Typically any surface that does not permit Blink is too steep or simply not a sturdy surface. This ability could have shown the holes in the game's sense of freedom, but instead that range of movement reinforces that freedom. The player suddenly has more options available throughout the game world, and any surface that cannot be Blinked onto is typically insignificant. There is always a path available, and there are always many different ways to skin a rat, so to speak.

Every power and ability in Dishonored works like this. Each power or weapon, every upgrade, increases the ability for the player to approach a situation in a different manner. Just as Super Mario Bros. kept the rules of movement simple (you can only progress to the right, never to the left), Dishonored too keeps rules simple. You must defeat your target, lethally or non-lethally. The game often provides special tools to accomplish these goals, but the player does not have to make use of them. They have plenty of other tools at their disposal to accomplish the same goals.

What makes Dishonored a truly special game is that the player's agency extends to the narrative. In many games, player agency can potentially conflict with character agency. For example, it does not make sense for the Master Chief to kill his fellow humans in the Halo series (and if you do, well, let's just say killing Captain Keyes in the opening of Combat Evolved is a bad idea).

This is why a lot of games implement silent protagonists or allow the player to customize their own. In Dishonored the player acts as puppeteer to Corvo. Despite having a face and a name, the player has total control over Corvo's actions. If the player kills ruthlessly, the game world will reflect their habits by flooding the streets with zombie-like Weepers and hordes of hungry rats. If the player spares the majority of guards and hostiles they encounter, then the streets will be relatively clean (for a plague-ridden city). Each option has merits and consequences in relation to gameplay, despite effecting the narrative.

Speaking of, if the player is a cold-blooded killer, then the heiress Emily's personality will be affected. Her demeanor and her drawings will be more hostile. Characters will betray you that might otherwise not have done so. The final conflict will take a very different shape.

Visible changes based on your behavior are a necessity to providing a sense of narrative agency to the player. It's not enough that their choice results in, say, red, blue, or green explosions. Telltale's The Walking Dead is a great example of choice and consequence. Even if the overall story is linear and protagonist Lee has a predetermined personality, the game manages to present difficult choices with apparent consequence. The player knows going in that a decision will have an immediate or gradual consequence, and the game is sure to remind players of that. So while the game is certainly linear, it feels as if the player's choices and decisions resonate.

Which is what agency is all about. What makes video games different from other media is that you not only have narrative agency, but you also have mechanical agency. While narrative is not a necessity or can cross over with character agency, mechanical agency is the key to a game that keeps the player coming back. Narrative agency can help out, but if the mechanical strategies, tactics and interactions are the same every time, then the desire to return to that game world diminishes.

Conclusion

The key to enjoying a good fictional product, be it video game or film, television or literature, is that the audience can suspend their disbelief and accept the illusion of the world presented. Characters and their decisions must make sense. Game mechanics must adhere to their own rules in a logical manner. There must be visible feedback and consequences of any decision, even if it is a simple success (press the jump button, and Mario leaps into the air).

Without these elements, a story is just a series of events with no emotional purpose or connection. A video game would simply become a session of Simon. Neither of these truly engage the emotional minds of the audience, and without that engagement, well, what's the point?

Comments

Yay, one of my favorite topics! While I don't quite agree with everything you wrote, I think making the distinction between player and character agency is important, so I'm glad you wrote this.

I await the disagreement.

You're clearly not a hardcore Simon player.

Okay, I looked on youtube, but didn't find what I was looking for,
am I the only one who gets really intense and focused when playing Simon? I get pretty emotional.

As a result, when he makes the choice to venture to Mt. Doom, we can believe he is brave enough, despite his fear, to do such a thing. His decision is consistent with his earlier choices and actions in the film.

You know, a fair number of authors talk about how they don't really know what their characters are going to do when they start out writing something. Finding out, often, is one of the real rewards for creating the story.

In Tolkein's case, the books feel fairly carefully planned, so Frodo's apparent agency may indeed be all illusion, but that's not the universal case. Sometimes, the author is as surprised by what transpires as you are.

I thought of Symon, but that's neither here nor there.

Reading closely, I actually agree with the main points you're making. The balance of mechanical agency between the perceived possibilities and the actual systems is important.

I have a slight quibble with equating agency with free will. It's a common connection, and probably the right simplification for your explanations. My personal feeling is that the usual conversations about agency import too much from free will, leading some to imply that a game has to be a complete open-world sandbox to have enough agency. You do a good job of addressing this in the article, though, so it's just a quibble.

I also want to emphasize, though you touch on it, that a very linear game can still have both character and mechanical agency. Portal 2 and Kentucky Route Zero each have both in spades, though via very different approaches. Valve is especially good at constructing scenarios where the player's goals and the characters goals so closely align that you don't notice the limits of the space as much.

Malor wrote:

In Tolkein's case, the books feel fairly carefully planned, so Frodo's apparent agency may indeed be all illusion, but that's not the universal case. Sometimes, the author is as surprised by what transpires as you are.

Simplifying a bit, Tolkien wrote straight through and then went back and revised it. He wasn't quite sure who Strider was when he first appeared, for example; there's a definite sense of discovery in some of the drafts. There's probably a lot to dig up in the intersection between Tolkien's writing process and character agency.

I think the basic structure that lends the truth, not the illusion, of player agency is the contract of choice. The designer lays down at least two true choices to the player, the player then chooses one of them, and then the designer, within the inherent understanding of their contract, proceeds to tell the player what the choice means.

SMB works because it communicates the choices it gives the players clearly, and it follows through on the expected results of those choices consistently. Any deviations from consistent follow-through constitute a conscious decision from the designer to subvert the rules, but in a manner that is clearly rewarding to the player (and thus does not erode trust), and is then repeated in a number of ways, which expands the rules rather than merely introduces a series of exceptions.

"Invisible walls" violate the contract between designer and player in a world that otherwise simulates reality, because they are often encountered well within a time frame that exceeds establishment of the contract, and are often positioned to benefit the designer's limitations rather than excite player engagement and curiosity.

If the designer were a parent and the player his charge, the invisible wall constitutes not only a "because I said so" rationale for thwarting the child, but also an arbitrary one that erodes previously agreed-upon rules. It violates very basic human interactions regarding trust and consistency.

It was actually one of the articles you posted in Draw Your Sword that helped me finalize what I wanted to put down in here. I used Dishonored because it not only has choice of play-style, but it's a fairly linear game as well. You're right, when most people think "freedom" in games they think of open-world sandboxes, but there's much more to freedom and agency than that.

I considered bringing up Halo, but decided it wouldn't be as good of an example. However, I strongly believe one of the reasons the game is so good is because they know how to handle the 2-weapon system. It works in terms of forcing resourcefulness out of the player, but it also allows the player to choose which weapons are most effective for their playstyle. It goes against the grain of most shooters, but Plasma Pistol + Magnum is actually a really good combination. The Magnum takes care of Grunts and Jackals with quick headshots, and the Plasma Pistol wipes away enemy shields and thus allowing you to fire away a couple rounds to the head. Yet this isn't the only way to play. I happen to prefer the Needler over other Covenant weapons, for example, though the trade-off is more limited ammo.

But that still feeds into visible consequences and results. Each weapon behaves differently, has advantages and drawbacks. Therefore the choices matter, or at least they feel like they do, and the player now considers which weapons they want to pick up off the ground.

Most other shooters these days don't manage the 2-weapon system well, and it's because they don't understand why it worked. I'd say the only other game where it makes sense is Call of Duty, because authenticity.

“What I did I did without choice. In the name of peace and sanity.”

(Sorry, I'm a little giddy about Dr. Who right now.)

Watched some Let's Plays of Stanley Parable this weekend. Now there's a game that respects player choice.

wordsmythe wrote:

Watched some Let's Plays of Stanley Parable this weekend. Now there's a game that respects player choice.

The Stanley Parable is very aware of agency and it's limitations.

The game's ability to respond to the player provoking it contributes greatly to the player's sense of agency. I think it's because the game exceeds our expectations. We think we know what games will respond to, so we can be surprised when the machine responds to something that makes logical (or emotional!) sense but isn't something we're used to games noticing and responding to.

Looking at it another way, the game judges all of the player's inputs but can only respond to a fraction of them. We're continually looking for the game's approval, since that's the only way to progress the system, or even interact with it at all. So when the game judges us for something unexpected, we get it's approval in the form of more interaction. Playing a game is seeking attention from the system.

That's why something like Stanley Parable, that pays attention to all the little things the player barely thinks of, or Minecraft, where every block encodes a tiny interaction with the system, feels more agency-ful. There's more system there to get attention from via interaction.

Gremlin wrote:
wordsmythe wrote:

Watched some Let's Plays of Stanley Parable this weekend. Now there's a game that respects player choice.

The Stanley Parable is very aware of agency and it's limitations.

The game's ability to respond to the player provoking it contributes greatly to the player's sense of agency. I think it's because the game exceeds our expectations. We think we know what games will respond to, so we can be surprised when the machine responds to something that makes logical (or emotional!) sense but isn't something we're used to games noticing and responding to.

Looking at it another way, the game judges all of the player's inputs but can only respond to a fraction of them. We're continually looking for the game's approval, since that's the only way to progress the system, or even interact with it at all. So when the game judges us for something unexpected, we get it's approval in the form of more interaction. Playing a game is seeking attention from the system.

That's why something like Stanley Parable, that pays attention to all the little things the player barely thinks of, or Minecraft, where every block encodes a tiny interaction with the system, feels more agency-ful. There's more system there to get attention from via interaction.

I think both Minecraft and Stanley are also good examples of the way the narrative track grows shorter as the possibility space grows wider. If you play either just to follow the narrative to its conclusion (especially if you follow the "right" path), you get a vastly contracted experience compared to what the game allows.

I think there are different ideas here of how to create and construct a narrative and what the purpose is supposed to be. If we consider Street Fighter as a game that allows you to craft narratives through personal experiences, we see that the game is carefully crafted in order to make specific kinds of narratives possible - like a narrative lego set. A branching narrative comes closer and closer to this end of the spectrum by necessity as it increases in complexity and scale. At the point where it allows the player every plausible action within a real context, it becomes very sim-like. Hence, romance VNs of a very complex narrative structure are often also called "dating sims."

The ultimate dating sim Love Plus is more of a sim than it is a crafted story, though it does have arcs and broad narratives for each possible girlfriend.

In the sense that a lego set of narrative possibilities can approximate life experience, the narrative experience isn't necessarily shorter. Indeed, in the case of MMOs and other "life" games that provide means to simulate life activities (hanging out, chewing the fat with friends), the narrative length and personal stories players craft from the possibilities within the game take over their lives as the game and their life activities integrate on fundamental levels.

I think there are different ideas here of how to create and construct a narrative and what the purpose is supposed to be. If we consider Street Fighter as a game that allows you to craft narratives through personal experiences, we see that the game is carefully crafted in order to make specific kinds of narratives possible - like a narrative lego set.

And this sort of emergent narrative is certainly valuable, but it's also a different sort of narrative (especially in terms of what it takes from the developers) than the pre-crafted stuff, where lots of canned material is required in order to give a response to the player-character that recognizes previous activities and choices, rather than only responding to the current choice/action.

Part of the reason that scripted narrative elements stay around is that we're still figuring out how to recognize, talk about and critique emergent narrative. The screen writers guild (Writers Guild of America) gives awards for game writing, but submissions have to be in linear document format (I think they request .doc). That's all well and good for a Final Fantasy script, but I struggle (and I'm not alone in this) to conceive of a way to convey the genius of Stanley in that format. It's work like Aarseth's that helps lay the groundwork for engaging these new forms of storytelling in a critical way.

Huh, I knew you had to be registered with the Screenwriter's Guild, but was not aware of the format having to be in a linear format. Considering how much dialogue there is from random folk in the environment in a lot of games and all the other non-linear information, that just seems...to ignore some of what makes writing for games both unique and uniquely difficult.

However, there certainly is a difference between telling a story and creating them. There are moments in Morrowind and D&D my friends and I share, but they are very different from discussing a story story. It's more like retelling a story of our own life. These are very different, and do not always present a feeling of agency.

In fact, Far Cry 3's "emergent gameplay" irritated me at times and also felt as if it broke down the system. As often as it can be amazing, it can also be horrible or pull the player out of their experience.

To be clear, I think the Guild judges understand the difference (or at least that there is a difference), but they're not equipped to receive and judge submissions in a nonlinear format. They do accept subsections, though, rather than insisting on all the writing that was done for a game.

I have personally dabbled with things like converting a game's writing and narrative design into a Twine or similar format. That was actually how my thesis project first started out.

I wonder how this maps on to strategy games, where you tend to control an organisation rather than just one character.

With XCOM, you have very little narrative agency -- the story will hit the same beats as long as you complete your objectives -- but a broad mechanical agency. Like playing chess, though, it often though it feels like trying to divine the "right" choice that will net you the strategic/tactical advantage.

As an aside, the notion that mechanics must appear clear to the player is a good point, but as Sid Meier has noted re proability rolls with Civ combat, there's a difference between what is "fair" statistically and what feels fair.

Some of the more engaging game narratives for me -- the watercooler stories -- are the ones that come out of strategy games like Civ or XCOM, where the outcome of the games may be the same but the means I took to get there differed each time.

Felix Threepaper wrote:

Some of the more engaging game narratives for me -- the watercooler stories -- are the ones that come out of strategy games like Civ or XCOM, where the outcome of the games may be the same but the means I took to get there differed each time.

Crusader Kings is calling your name.

Felix Threepaper wrote:

I wonder how this maps on to strategy games, where you tend to control an organisation rather than just one character.

With XCOM, you have very little narrative agency -- the story will hit the same beats as long as you complete your objectives -- but a broad mechanical agency. Like playing chess, though, it often though it feels like trying to divine the "right" choice that will net you the strategic/tactical advantage.

I think it applies. There's a certain story structure to every Civ game, because the rules set up certain emergent constraints. Civ is fine tuned to result in a dramatic arc in most games, with the rising action of exploration leading to an end-game climax. Chess is similar. They each have a kind of narrative agency encoded in the systems; call it a social or economic history as opposed to the Great Man narrative of a single-person story.

XCOM has a certain progression structure because you have to work your way through certain checkpoints to unlock the next act. You do so emergently, so it's not strictly linear. I'd like to see a game take a similar approach, where your actions in the tactical levels affect what course you take at the strategic level, either with a branching tech tree or some other emergent narrative progression.

Felix Threepaper wrote:

As an aside, the notion that mechanics must appear clear to the player is a good point, but as Sid Meier has noted re proability rolls with Civ combat, there's a difference between what is "fair" statistically and what feels fair.

I'd actually make the distinction that the mechanics don't need to be clear, the player just needs a clear route to play with them. That's where the games that over-explain their mechanics go wrong: the best way to learn is by playing around with the different possibilities and figuring out what works for yourself. The game should give clear signposts as to where to look, but the player is the only one who can actually close that gap.

That is a good point. I don't play all that many strategy games, and when I do they're typically in the form of Final Fantasy Tactics (X-Com is on my PS+ pile, though). However, perhaps strategy is where agency matters the most, it just isn't quite so disguised.

Going back to Final Fantasy Tactics, I can sit there for a long time trying to figure out what to do with a single character, measuring the probability of certain abilities working, which foes I should strike at, and what the potential consequences of my decision could be. Feedback and response is typically immediate. You are rewarded for 3/4 characters being buffed by the Haste spell, or your attempt to slay one character opened you up to a Break Armor from a Knight above.

A strategy game is more about mechanics than a traditional narrative, certainly, but if those mechanics are overly simplified or are too obtuse, then the player will not feel a proper sense of command and thus agency. The game must also respond in an intelligent and sensible fashion. Operation Darkness was a supernatural WWII tactical RPG for the 360 in its earlier years that I enjoyed the concept of, but the execution was lacking in a lot of options found in a game like Final Fantasy Tactics, and also would have been too easy if they didn't overcompensate with surprise reinforcements multiple times in one battle, and thus instead became too difficult.

I imagine Master of Orion 3 could make an excellent case study as to how agency can be ruined in a strategy game. My understanding is you either have the computer do too much work for you, or not enough and are forced to learn obtuse systems. A proper strategy game is complex enough to provide the player a series of options, sure, but can also be understood easily with a steady learning curve.

This could be one of the reasons Brutal Legend wasn't viewed positively. Once you hit the strategy segments, they were really quite easy. Spawn as many units as you could, send them all to a geyser, use team-ups and shred on your guitar as often as possible, rinse and repeat. The actual strategy part is simple. It's all about being in the thick of it. But the design of the game just doesn't flow well with a lot of people.

ccesarano wrote:

I imagine Master of Orion 3 could make an excellent case study as to how agency can be ruined in a strategy game. My understanding is you either have the computer do too much work for you, or not enough and are forced to learn obtuse systems. A proper strategy game is complex enough to provide the player a series of options, sure, but can also be understood easily with a steady learning curve.

Master of Orion 3 had two problems: first, it was incomprehensible. Textbook case of the Tale-Spin effect, where there's a massively complex system that the player has no real way to interact with or learn about. The planets had a really cool multidimensional environmental model, for example, where different races prefer different conditions. But there was little documentation and it wasn't really clear what the player could do about it. The second problem was that parts of the game didn't work at all, but because of the first problem this wasn't immediately obvious.

Strategy games, because they rely so heavily on the emergent systems to work at all, are more than usually contingent on mechanical agency. That goes for sims, too. That's one reason why I personally prefer SimCity over Impression's City Builders, because games like Pharaoh, much as I love them, often end up as puzzle games instead of the sims I want because the progression was made hard enough that there isn't as much room for player expression in the solution.

Gremlin wrote:

Crusader Kings is calling your name.

I hear it, or rather CK2. It's sitting in my Steam library, complaining that I haven't started it up since I got it on sale!

Gremlin wrote:

I think it applies. There's a certain story structure to every Civ game, because the rules set up certain emergent constraints. Civ is fine tuned to result in a dramatic arc in most games, with the rising action of exploration leading to an end-game climax. Chess is similar. They each have a kind of narrative agency encoded in the systems; call it a social or economic history as opposed to the Great Man narrative of a single-person story.

I like your point here and don't have much else to add.

ccesarano wrote:

This could be one of the reasons Brutal Legend wasn't viewed positively. Once you hit the strategy segments, they were really quite easy.

Brutal Legend was hard done by.