Game Narratives vs Game Systems

Whenever we make a video game, we make some mixture of static content and systems, and this mixture produces the game that players will someday experience.

Traditional AAA game production tends to see “systems” as things you use for AI design and loot pickups, whereas other things like game levels, boss fights, and game stories… are all hand-crafted content.

Typically: producing static content that provides high quality player experience is easier than making systemic experiences, and it’s easier to estimate how long the work will take and what resources you will need.

However: while it’s easier to make content than systems, when your production focuses mostly on content… you can’t just pull a lever and get ‘more player experience’.

I've argued for the merits of a systemic approach to making player experiences before, and one of the challenges people throw back at me is “yeah, but what about games with really great stories?”

It’s a good question. Most really great video game stories are handcrafted experiences. For example: Last of Us had such a great story that it’s the primary player reward. I didn't grab loot to upgrade my weapon so I could score more headshots; I gathered loot to keep Ellie alive and advance the story; I had to know what would happen next.

But I can imagine how a systemic approach could be used to also tell a great story. Maybe even make a great story.

Traditional Games: Players Can Do Anything Except Change the Story

The problem with the completely handcrafted story, is that somebody has written it before the player arrives. This means certain events just have to happen, and the player can’t stop the story from happening.

That condescending guy you suspect may be the hooded wizard who destroyed your village? You can’t use your powers to take him out before the end of the cutscene. He’s essential to the plot.

In terms of game production, we have reasons for this lack of agency. That cutscene cost the studio a lot of money. In fact, everybody who worked on that scene was probably trained in the film industry, as were the actors who played it out in mo-cap. So because we spent money: the player MUST see it as-intended.

If we really believed that pre-rendered video is the only way to tell a good story in a video game, we could stop trying to innovate now. (Nah.)

Use Systems to Evaluate Who the Hero Is

Unless your story only ever goes in one direction, then “who the hero is as a person” isn't a Narrative Director decision; it’s a result of how the player plays the game.

There are great examples in “choose your own adventure”, but really this is about making a design commitment to support cause and effect in your story.

Step one: design systems that keep track of how the player behaves.

Step two: have your story events and characters use the information systems from step one to modify the story.

Let’s do an easy example: many games play on player choice. Let’s say we’re making a cat burglar game, and the game mechanics support players doing heists in stealth, or using violence and fear to steal the goods. You’re the player, and you decide how each mission goes down.

The game character who buys the stolen goods from you at the end of each mission could react differently to you depending on how you completed the mission. “You didn't need to kill all those people!” or “like a ghost” are the obvious choices, but you could try harder.

Maybe the mission giver character stops trusting you because you’re clearly a dangerous sociopath, and begins equipping extra security… Or maybe she asks you to look after her brother, because she thinks you’re somebody who cares about people even though you pay your bills by stealing.

The number of ways you track ‘who the hero is’ and the depth of reaction that the story supports will determine how much money this design choice costs your production.

For many producers the very idea of having branching dialogues and narrative scenes, possibly producing expensive content that some players may not even see, is pure money-burning madness. And they've got a point.

Reducing Story-Branch Costs With Systems

One of my best early moments in playing Mass Effect was realizing that in every cutscene, “my” hero was really the hero I had built. I was wearing the right helmet, holding the right weapon, and I had the right hair.

This is smart production, calling the scripted events on the player’s avatar in a way that feels contiguous with my player’s degree of agency. Let’s look at other ways of using systems to support variation in game stories.

Bad Guy X : what happens if the player kills the bad guy, or befriends the bad guy? Then the story needs a new bad guy. What if we had systems that would allow us to write a mission sequence in which somebody is the bad guy? We’d need maybe five bad guy voices, possibly more than one body type doing the bad-guy sequence. But then we could let the game events decide who is the bad guy for Mission 17, and that’d be pretty cool.

Side Mission Generator: We already do this in some games, and every nonlinear game experience would benefit from using this idea. Making a decent generator for side missions means you get X number of permutations (remember all of these systems can listen to “who the hero is”, above) and once you have the system working, the mission generator can create quests for any new section of the game world. Like that new planet/island/continent you have planned for DLC in the first quarter post-launch.

If you set yourselves the goal of designing systems and data structures that will create a wide variety of player experience instead of linear sequences, solutions will present themselves.

Game stories can be designed as a system of things that can happen, instead of a linear sequence that must happen.