Worlds of Design: The Tyranny and Freedom of Player Agency

“Player agency” refers to the player being allowed by the game to do things in the game that have real consequences to the long-term course, and especially the result, not just for succeeding or failing. Some campaigns offer a lot, some only a little. Are players just following the script or do they have the opportunity to make decisions that cause their long-term results to be significantly different from another player’s?

Picture courtesy of Pixabay.

​

“I play games to specifically be an agent in the universe that does effect things. I watch netflix or read books to be an observer. I have to be in control of something.” Kaze Kai​

​

The subject of player agency is a controversial topic in game design. We have “rules emergent” games which are “open” versus “progressive” games which are “closed”; or “sandbox” which is open versus “linear” which is closed. The first of each pair can also lead to strong player agency, the second almost never does. I'll add a third one: games, which are open, versus puzzles, which are closed, because in a pure puzzle you must follow the solution devised by the designer.

Player agency is important because many long-time gamers want control, want agency, yet many game and adventure makers want control themselves, and take it away from players. It’s the difference between, say, Candyland or Snakes & Ladders(no agency), and games like Diplomacy and Carcassonne. For adults, Tic-Tac-Toe has no practical player agency, as it is a puzzle that is always a draw when well-played.

When a GM runs a particular adventure for several groups, do the results tend to be the same for each group (beyond whether they succeed or fail), or do the results tend to be “all over the map”? If the former, it leans toward being a linear adventure, while if the latter, it’s more “sandbox.”

Books can help us understand this. Most novels have no “reader agency”; the reader is “just along for the ride." Films offer no viewer agency. On the other hand, “Fighting Fantasy” and similar “you are there and you make the choices” books, where you choose what to do next from among about three possible actions, gives the reader-player agency over the short term. (Dark Mirror’s Bandersnatch is a more recent example.) Though in the end, if the player succeeds, there may be only one kind of success. Video games usually let players influence the small-scale/short term stuff a little, but not the large scale.

In between broad player agency and no player agency can be found games with false impressions of player agency, which you can recreate in an RPG adventure just as well as in a standalone game. The Walking Dead video game was often praised for the choices the player had to make, but in the end it all comes out the same way no matter what the player does (see this reference for a diagram of all the choices). Mass Effect is another game highly touted for player choices that ended up in the same place despite their decisions.

Full player agency creates story branches that don’t come back to the same place; the player’s choices just continues to branch. The reason this is rare in video games is because more choices and branches means more development, which costs money. In tabletop RPGs, a good GM can provide whatever branching is needed, on the fly if necessary.

The one place where player agency is seldom in question is in competitive tabletop games, especially wargames. Even there, many of the old SPI games more or less forced players to follow history. And many Eurostyle “games” are more puzzles than games, hence players must follow one of several solutions (“paths to victory”).

Why would a designer not provide Agency? I don’t understand it emotionally myself, but I can understand it intellectually. Some game designers are frustrated storytellers (or puzzle-makers) who have chosen not to use traditional forms such as novels, film, plays, oral storytelling. They want to provide “experiences." But in order to do so in a medium not as suited for it, they must introduce limitations on players in order to retain control of the narrative.

Only games (as opposed to novels or films) offer the choice of agency or not. There’s nothing wrong with a “lack of agency”, if that’s what players expect – as in a typical film or novel. I am not saying it’s wrong, just that most highly experienced game players don’t like lack of player agency.

I recommend you ask yourself a general question: “am I imposing my ideas and notions on the game, or allowing the players to use theirs?” Part of that answer is relevant to player agency. What you want the answer to be is up to you.

This article was contributed by Lewis Pulsipher (lewpuls) as part of EN World's Columnist (ENWC) program. Lew was Contributing Editor to Dragon, White Dwarf, and Space Gamer magazines and contributed monsters to TSR's original Fiend Folio, including the Elemental Princes of Evil, denzelian, and poltergeist. You can follow Lew on his web site and his Udemy course landing page. If you enjoy the daily news and articles from EN World, please consider contributing to our Patreon!

 

[Jay Verkuilen]

Trust me, there are puzzles that have more than one solution. But they tend to be more complex.

Yes, definitely. I think the skill challenge is a better replacement for a straight out puzzle, or else the DM should allow for different methods of some sort to achieve the goal, with the puzzle being one obstacle. The puzzle could be one way, but there may be others, such as a sequence of athletic feats, or a complicated lock. That gives players agency but still challenges them. It can also be for interesting RP if the character trying to solve things by wits gets in a race with the one trying to solve it by muscle. That's totally genre-appropriate in a more swords and sorcery style.

 

[dragoner]

Many of the rules that got added were simply put in place to make things more fair, clearer, give players more opportunities, and so on. They weren't done with bad motives in mind, but things got out of hand.

It became having a rule for everything, simultaneously promoting a rules lawyer type of play, and making it more difficult for the GM to get to the point of rules mastery; plus a creep towards so many rules, that some didn't work as well, or were not play tested. Then there was the one player who wanted to use the additional rules, making the GM have to put in more work for that one player, often without much interest from the rest of the group, for example Psionics from AD&D or Cybernetics from Traveller.

Unquestionably, but game designers can't write games under the assumption that you first need to find a kicking DM.

The admonition for writers is to write for yourself, write what you know, it is true for games too. Some writers do capture lightning in a bottle, that's why they get the following they do.

 

[Malrex]

Yes - I like sandboxing and with proactive players it's great, but I have also seen players respond with blank stares to "So what do you want to do?" - some players want and need a clear path in front of them, while others love bouncing around instigating new stuff. The latter can be disruptive in a more linear campaign.

I think that's fixed with an array of adventure hooks--some can be a clear path, some can just be vague rumors. But I agree, it depends on what sort of campaign is being run.

 

[Malrex]

There's a severe category error in this thread. There are more ways to play than hex crawls and railroads. The idea that one needs to railroads in order to have a game that is compelling on a narrative level is misguided. One simple way is to require players play characters with hopes, dreams, and goals and then present honest defined antagonism and playing to find out what happens.

I think there is a lot of truth to this. However, I would never 'require' a player to do anything--I would read the player during play and if the character has goals or hidden agendas, then I add wood to that creative fire. This is how my group plays...it's great! Sometimes it can lead to inner party turmoil as goals conflict--but this usually takes a very looooong time to come to a head (a year or 2 in our case).

I was also going to add about the article...there is also the tool of the 'timeline'. In a way, a good timeline can almost combine a linear and sandbox playstyle. Key word--almost. The timeline MUST be loose enough to allow character agency--i.e. they can ignore it or partake in it if they want. The actions of the characters can change the timeline to slow things down or speed things up, or the timeline provides an ever changing environment that characters make choices to adapt. For linear, the timeline can cover that overarching plot and be used as tool to get a party 'back on track' of the story----I'm guessing about this as I'm a sandbox DM, but would assume it would work.

 

[Jay Verkuilen]

I think that's fixed with an array of adventure hooks--some can be a clear path, some can just be vague rumors. But I agree, it depends on what sort of campaign is being run.

For some players more choices that "left or right down the hallway" is about as much as they can/want to handle. Three or four plot hooks is too much for them. I've seen choice cycling, where they start looking at option A, then B, then C, then back to A. This can happen for a variety of reasons, not wanting to be the guy/gal who makes the choice and offends the other players being one, but there are others. For a group like this, choices need to be pretty clear and laid out, for instance, A -> B or C (choose which order to do) -> D.

I've come to make use of this kind of structure in adventure designs a lot. I'll start in one area, say A, with the eventual goal to get to a particular other area, say G, but give a choice among different paths. It's kind of a "subway map" style dungeon.

A: Starting point
B: Plant monsters in a chasm
C: Ghouls
D: Troll
E: Phase spiders
F: Attack by soldiers of their main adversaries
G: Goal

Valid paths:

A -> C -> F -> G
A -> D -> F -> G
A -> B -> E -> G
etc.

I just chart these out on a piece of paper and budget XP accordingly so they're suitably beat up by the time they get to G, though it depends on the choices they made. This allows for player agency, but still gives me the DM enough structure to be well-prepared. I think it also helps to let exploration- or skill-oriented characters really benefit the group by letting the PCs make good choices or even bypass some of the encounters. Too often in a very narratively structured game they don't really get a chance to be useful.

Of course, the same kind of structure can be used for adventure design. Yeah, the total hex crawl or else grognard will hate this as still not providing enough player agency, but at some point I just stop listening to them.

 

[Jay Verkuilen]

It became having a rule for everything, simultaneously promoting a rules lawyer type of play, and making it more difficult for the GM to get to the point of rules mastery; plus a creep towards so many rules, that some didn't work as well, or were not play tested.

Yes, and of course game companies do need things to sell. I totally understand the desire to provide a very comprehensive set of rules, but, like anything else, it can go too far and become abused. Not enough rules? It's left to the DM and DMs vary. Too many rules? The rules lawyers and power gamers have a field day and everyone else just gets bored.

The admonition for writers is to write for yourself, write what you know, it is true for games too. Some writers do capture lightning in a bottle, that's why they get the following they do.

I was referring to the fact that there's so much nostalgia for modules like GDQ or the like and WotC seems to want to try to capture that lightning over and over. Many of those modules really aren't that good. It's like listening to album tracks by classic bands. A lot of those songs weren't that great and the recordings were not good either, but in the context of the times they were important.

 

[GrahamWills]

I have never heard of a puzzle with multiple solutions. Do you have an example?

My personal favorite is the classic puzzle on how to measure the height of a building using a barometer. Think about it for a few minutes, then follow this link:

Esmerel: Elementary Physics

 

[pemerton]

There's a severe category error in this thread. There are more ways to play than hex crawls and railroads. The idea that one needs to railroads in order to have a game that is compelling on a narrative level is misguided. One simple way is to require players play characters with hopes, dreams, and goals and then present honest defined antagonism and playing to find out what happens.

If I had a dollar for every time I've posted this on these boards . . .

Here's a post I made about this back in 2014:

Lewis Pulsipher was a prominent contributor to White Dwarf in its early days. The following quotes are from his article in an early number of White Dwarf (my copy is in Best of White Dwarf vol 1, 1980):

D&D players can be divided into two groups, those who want to play the game as a game and those who want to play it as a fantasy novel . . . The escapists can be divided into those who prefer to be told a story by the referee, in effect, with themselves as protagonists, and those who like a silly, totally unbelievable game. . . In California, for example, this leads to referees who make up more than half of what happens, what is encountered and so on, as the game progresses rather than doing it beforehand. . . . [T]he player is a passive receptor, with little control over what happens. . .

Gary Gygax has made it clear that D&D is a wargame, though the majority of players do not use it as such. . .

The referee [in a skill campaign] must think of himself as a friendly computer with discretion. Referee interference in the game must be reduced as much as possible . . . Effectively, this means that the referee should not make up anything important after an adventure has begun. He should only operate monsters encountered according to logic and, where necessary, dice rolls. . . . Occasionally an adventure will be dull, because players take the wrong turns or check the wrong rooms, while others may be 'milk runs' because the players are lucky. Referees must resist the temptation to manipulate the players by changing the situation. Every time the referee manipulates the game on the basis of his omniscience, he reduces the element of skill. . .

The referee who, for example, schemes to take a magic item away from a player is incompetent. If the player doesn't deserve the item he shouldn't have obtained it in the first place. Don't lie to the players when speaking as referee. If players can't believe what the referee tells them they are case adrift without hope. . . .​

I tried to implement this advice in my early GMing (around 1984). I was not very good at it - I'm not a particular effective "computer with discretion", and my players didn't (and don't) like the occasional dull adventure. But I do like the advice about not manipulating the players. It was around 1986, with original Oriental Adventures, that I started to discover a way of GMing in which the GM would make stuff up on the spot, while still allowing players the scope to make choices which are genuine in their consequences, thereby avoiding the railroading that Pulsipher warns against. (More than 15 years later I discovered that this approach to GMing had been refined and theorised by Ron Edwards and others at The Forge.)

I don't hold it against @lewpuls that his framework for thinking about RPGing doesn't seem to have developed a great deal since the early 1980s. He know what he knows and generally writes about it well.

But as a RPG community I think it is worthwhile to pay at least some attention to developments in RPG design and play over the past 30-odd years. It's not as if the sort of RPGing you (Campbell) have posted about is particularly esoteric.

 

[aramis erak]

This is a misunderstanding of agency:

"The wider the field of choices, the harder to make those choices it becomes for the players. The rule system itself is a constraint on both player and GM agency... and it works because it reduces near-infinite choices to a manageable few."

Agency isn't about number of choices, it's about the ability of players to influence and change the outcome. That can be achieved without offering players myriad choices. But the choices must be important (in so many games, they're not, they're choices for the sake of choice).

At the level Mr. Pulsifer is talking, and that I' pointing out he's ignoring, the difference between agency and choice of action is moot.
The range of meaningful actions must be reduced to a level where the GM can handle it; the ability of players to affect the story to a range of meaningful outcomes also is reduced, but the value of that is, in fact more playability, in that the GM has less to think of and has scaffolding upon which to hang, the actions of the players and judge their impact.

Agency and player choice, at that level, are still a unified function. The feel of agency is enhanced by prior expectation of probable success, and then that success having tangible in-fiction consequences.

When one gets down to the play level, the reduced agency implied by acceptance that the nature of who can define what and who can refute what becomes separated by the question, "Does this roll really matter to the fiction?" - and, largely, they do. The restriction on action set is also a restriction on agency, but it'a also a restriction on player ability to go afield.

If Player A has total agency, no one else has any. If all players have equal agency, none have total agency. While I hesitate to describe group total agency as a zero-sum-game situation, in many ways it is.

If A isn't allowed to make meaningful changes to B, then A's agency is limited. If A is required to use mechanics to affect meaningful changes to B, that's still restricting A's agency... but it does so in a manner that makes it potentially more practical.

Let's look at a non-agency rule by comparison... In Star Trek Adventures, all stress damage is recovered at end of scene. Unless and until a wound is inflicted, the character hit has no lasting change. Stress damage is a non-agency tool for affecting a sum of actions that creates a lasting impairment (at least until getting to sickbay or a doc with a surgical kit)... that impairment is proof of agency, in every way that stress isn't. The decision to shoot in that system also includes options to spend metacurrency to take the shot that hits and push the damage up to wounding... (adding 5 momentum almost always guarantees a wound, and thus imposing a wound-type disadvantage, which has lasting story and mechanical impact.) the rules restricting the "I shoot and kill" level of agency result in a less certain outcome, invoking the game aspect, and making the choices fewer, but thus easier to identify which choices are meaningful, and which are more/less likely to succeed.

Perceived agency versus actual story agency - yes, they are different. But they're so conflated that restricting one axiomatically restricts the other, but also increases the ease of use.

 

[aramis erak]

I have never heard of a puzzle with multiple solutions. Do you have an example? All I can think of is a Gordian Knot puzzle with an intended solution that allows for out of context unintended solutions.

the simplest case to exemplify: Mazes. We think of mazes as a singe solution. But a maze can have two ways through. Or even three.

riddles often have multiple correct answers. EG: What has many teeth, but cannot eat? Comb, zipper, rake.

many "park the car" sliding item puzzles have 2 or more solutions; there is usually one optimum one and one or more longer ones.