Sign up for news

About a month ago Steve Gaynor posted a nice consideration of storymaking – the ways in which players create or embellish their own stories of playing a game. To me, as an AI/gameplay programmer (what I get paid to do), and as a systems designer (what I often need to do), it’s interesting to me to think about how to build systems that encourage the player to narrativize their experience – to view their play as an exciting story that they have had, at least in part, some hand in creating. This is one key path to creating depth and meaning via game mechanics.

This behavior naturally has a spectrum – from the player turning to a friend and describing their experience in the stream-of-conscious expression that is punctuated by descriptions of explosions and character death, to the player attributing & creating more narrative aspects than are represented in the game itself (say, if you’ve ever been a creatively repressed teenager with too much time on your hands playing a party-based RPG, or anybody that could get past going to the bathroom and other more mundane aspects of The Sims). 

How can we create systems that encourage the player to do more of this, formulating their own stories about their play? Without the resulting mess of narrative & mechanics overwhelming.

Character & Character Creation

It should probably go without saying that stories involve characters, but I remember once a friend of mine, a game designer, turned to me and said he longed for the day a game like Tetris would make someone cry. I think he was joking, but just in case – stories involve people, or anthropomorphized stuff. Like toasters. That is what makes us cry. That, and skinning a knee.

The biggest step towards this goal is the player’s ability to create characters in their story, and the tools they have to do it. The frame that character creation provides is also the introduction the player has to how much ability they will have to tell their own story. It helps to craft these mechanics around a narrative context, to point them in this direction.

An example is the perks in the Fallout series – they take a systemic effect about a character trait that you can select and wrap in a concept related to the story (like the Daddy’s Boy/Girl perk from Fallout 3, which gives you a bonus in your father’s skills of Science and Medicine). More interesting still are the Merits and Flaws of White Wolf’s tabletop Storyteller system. You’re mechanically encouraged to choose flaws, which add penalties to your character through various game systems, but provide additional XP. You can use XP to buy merits, which are mechanical bonuses in similar wrappings.

The difference is cosmetic in a sense, but by couching it in this way, you’re encouraging the player to build up a role for their character using these descriptive terms as a shorthand for the character’s background.

Character Growth

There’s two aspects to character growth. The first is how your own created characters are changing because of your choices. It can be pretty daunting to throw too much character creation on the player up front. One way to manage that is use the same ideas but apply them later on as well, to what the player can earn or change about their character as they play the game.

The second is how other non-player characters, those presented to the player as part of the game’s story, change in their relationships with the player characters. Ideally these characters are not without change, as how else would you define character? There’s no reason you couldn’t go to the same depth with player character traits as you would these NPC traits. Sure, they might use the same stat systems, but the player’s choices & interactions could allow for adjustments in those stats with appropriate relation to the story.

Some party-based RPGs have certainly done this, like relationships modulated through persistent dialog choices (ala Bioware games’ love interests and such). There are a host of existing system paradigms that can be applied to character relationships like mapping relationship states to gameplay states and tracking relationship status changes as achievements/bonuses/penalties. It’s amazing to me that we don’t actually see more of this.

Earned Personalization

I’ll admit, for a long time I abhorred any character customization that had no mechanical impact, like your character’s fashion choices. So easy to give it even the slightest consequence from your choice (like Fable’s clothing & tatoos). But the power that such customization affords the player for investing in their own story is not to be overlooked or undervalued just because of that bias.

On the surface, having many objects in a system that have no mechanical differences, only aesthetic ones, would seem to be a non-systemic storymaking element. Systems do have to be built to support the addition of such, but the system side becomes more interesting when the act of personalization is treated as reward or has a systemic role itself.

In Fable 2 you can give any character a nickname that takes precedence over the name given to them by the game’s creators. Literally tacked on, these names don’t give much power to the player. Were they earned, they would provide a natural inflection point in the player narrative, marking the player advancing in their relationship with the character. The assigning of the nickname would then be a story point in and of itself. After all, even George W. Bush assigned nicknames because of shared experience on the basis of building rapport with the characters in his (somewhat disastrous) presidential sim game.

Emotionally Representative Systems

These are systems whose mechanics are set up to induce an emotional state representative of the narrative context. Consider the spatial movement of Passage or even the move vs. shoot controls of Resident Evil 5 – the dynamics created by these mechanics are meant to convey the feeling of the passage of a lifetime of experience or the constrictions of choice that fear places upon you in life or death situations. Respectively.

These of course are elements of the story world the game developer provides, but they provide an emotional framing for the player. Without imposing specific plot constraints, they create the feel of the memento mori or the horror game. They give the player narrative footing to move around, while the emotional tone they provide is something the player can build and improvise upon.

Symbolism & Metaphor in System Mechanics

The best immediate example I can think of here is from God of War, Krato’s climatic battle to save his family from many versions of himself. He can sacrifice his own health and transfer it to his family to save them. Leaving a wake of death in his path up to that point, this is the one positive act Kratos can perform in the game. These mechanics are not emotionally representative because going up and hitting a button doesn’t make you feel like you’re hugging your family to save their lives… But it’s a metaphor for the destruction Kratos has brought upon his own family, and his coming to grips with it by sacrificing himself for them.

The problem with applying this example to storymaking is there’s not much actual player choice in the system. If you want to keep going, you have to complete the fight, and depending on your skill level you’ll have to sacrifice some of your health for them regardless. Now if these mechanics were presented to the player as a choice, the player could drive their narrative one of two ways, the way God of War goes, or to an even darker place (where Kratos refuses to deal with the harm he’s brought his family). But that’s only possible by using these mechanics combined with that context to create the metaphor for the change in his character.

Event Driven Systems

It’s important to look at not just the mechanics of a given system, but what events those mechanics imply. A high density of events that have an emotional context ties into the natural memory formation that comes from experiencing emotionally impactful events. By managing this, making sure a system includes events that have an inherent emotional perspective (like success or loss), it’ll be easier to remember & create stories about those events during play.

The melee combat system of a beat-em up provides has a simple palette of events to start – a successful hit on an NPC or the player, and the knock down or death of either. If you add in grappling, that gives you the dramatic events of struggling with an enemy, being pinned down (when they temporarily win the grapple), and defeating an enemy via a grapple.

If you throw ally characters into that system, you have even more events. The injury or loss of an ally might become an important element of the player’s story. It’s even better if you can support these events by tying them into the larger narrative. Take the buddies of Far Cry 2 – playing the game, if you accept missions from your buddy, or if you die and your buddy saves you, you advance your relationship, potentially unlocking more missions. Tying these inflection points of the system (like death in combat) back into the narrative gives the player more control over both their moment-to-moment story and larger plot direction.

Systemic Pacing

What Left 4 Dead achieved with it’s AI Director is nothing new (yeesh that was in 2003), it does mark the highest profile, and probably most polished, success of such work. Systemic pacing really can’t be looked at as a solution in and of itself to this problem. But it can be a key layer on top of all these all these other elements.

The thing to watch out for is the hope that it will somehow solve other problems with lower level systems and somehow tie them together narratively for the player – it won’t. It can take them to the next level, integrating them in more varied ways to increase the storymaking possibility space, even if it’s as simple as modulating when, how many, and which kind of, enemies spawn.

Layering Systems & Training Storytelling

With all these types of systems, the amount of choice can be overwhelming. For example, Krystian Majewski at Game Design Scrapbook describes PlayFirst’s user Testing of Emerald City Confidential (via Emily Short), which found that casual gamers freeze when presented with even just multiple choice dialog – being accustomed to one “correct” choice, and lacking feedback as to the consequences of their choice.

Besides using what I’ll call the Optometrist’s Gambit (only ever present two choices, as clearly different as possible – the ol’ Is A better? Or B?), this is kind of a thorny problem. Choice is what drives both character and plot – if players are paralyzed by fear of choosing incorrectly, neither can really progress. 

We have various tools at our disposal, from slowly introducing systems to throttle the changes in complexity, to creating audio/visual/mechanical language for the player to understand the consequences of their narrative choices, to varying the impact of any individual choice (potentially requiring many choices to add up to anything drastic). But whatever the strategy is, better have one!


Anyway, that’s all (he says as if somehow diminishes the massive text block above) from me for a while. With GDC, the inevitable post-GDC cold, and a few other things going on, look for more interesting stuff here late April – and have a happy March!

One Response to You got some narrative in my system.