Wolfeye Studios' Weird West is being made by a team of veterans who've long worked within the immersive simulation genre. It's a type of game infamous for enabling player choice, sometimes to the detriment of designers trying to make sure they still have a fun, coherent experience.
At GDC 2022, Weird West narrative designer Lucas Loredo discussed this particular problem, and offered some useful techniques for immersive sim developers to manage the multiple narrative ends that can emerge in this game genre.
Loredo's fundamental principle was that immersive sim narrative designers are trying to wrangle with a genre where the "verbs" that define gameplay can "hit" story elements. In Weird West, a player can befriend a local sheriff, they can also shoot the sheriff, they can also lure the sheriff into a fire tornado that they started, but in such a way that they are not directly responsible for the sheriff's death.
If the sheriff is a key path forward for Weird West's story, that presents a number of design challenges and branching ends to account for. Here's how Loredo broke down the tools that Wolfeye uses to wrangle and imagine all that content
Loredo broke the different tools for grappling with emergent story possibilities into four categories, ranking them by "difficulty," which he measured as how much design budget was needed to account for these possibilities.
Skipping over the techniques that linear games use to advance the story (cutscenes or invincible non-playable characters), Loredo advanced into the "easiest" solutions designers can use to ensure that key story points are delivered to the players.
The first of these solutions was to embed narrative story information in in-game objects. If your game doesn't let players destroy inventory items—just move them from the world into inventory. t become "invincible objects" that you can anchor storytelling possibilities to. Anything from a mythical artifact, to a note, to a pocketwatch with a secret on the back, fits this category.
The next tool he described was the "character in a box." This is a design tool when a character delivering important information isn't physically in the game world, and communicates with the player from an inaccessible location. In many games, this is the "friend in the chair" who's supporting the player through an earpiece.
Weird West is in a magical wild west world, so Loredo said the team originally choose to not use this conceit. But if he went back now, he said he would gladly create such a magic system that lets NPCs communicate with players over great distances.
Finally, the next easiest solution deals with organically preventing players from accessing locations too early. In a game where you want players to explore locations out of order, why would you ever deny them access to a place if they have premature knowledge that there's something interesting there?
The answer: Create locations that can be safely gatekept until there's narrative context to let the player in. This also applies to destructible locations. A tornado might tear a house down in Weird West, but the basement would be safe from destruction in such an event. Said basement can also be locked until players find the right key.
These are simple solutions that re-frame "invincibility" to fit the player experience. The next techniques are more advanced.
More elegant solutions
In Weird West, Wolfeye Games wanted it to be possible for players to kill every NPC. So how do players get to the game's climax if everyone who could point them there is dead?
Well you could start by giving players reasons to not murder everyone they come across. Loredo called these "soft influences." If enough work is put into an NPC to make them memorable, that can shift what percentage of your players will commit to killing them.
Designers might do this by giving them funny dialogue, using a funny speech affect (HK-47, of Knights of the Old Republic, uses descriptive adjectives to this effect). The NPC might regularly dispense rewards for the player, or information that leads to secrets. If players value the character for abstract reasons—they have incentive to keep them alive.
Next, designers can consider multiple solutions to quest objectives. If a player shoots the sheriff, but needed their help to enter an area, designers can leave in-world solutions that persist after the sheriff's death. Maybe the sheriff left a journal containing the needed information. Maybe another NPC likes that you killed the sheriff and offers to help.
"The key here is that each option is narratively rich," Loredo explained. "When the player breaks [the objective] and sees they can continue, it creates a feeling of there being more solutions than we've actually designed."
Some of these options should be labeled as "failsafes." They need to be indestructible tools from the "easy" solutions that can't break no matter how hard the player tries. That way the player can feel agency over the world without bringing it to a grinding halt.
Loredo took a deep breath before continuing here. These solutions are based on the notion of building non-linear quests that can be completed with solutions that the player isn't explicitly directed to. Some players will follow every instruction they're given, others will be rewarded for taking shortcuts or brute-forcing other solutions.
If you don't want a player to use these solutions on certain locations or NPCs, you can use the cheaper solutions to wall them off. But Loredo encouraged developers to create narrative justifications for these moments. If a notable boss can't die until the player has done X task, you can fill their hideout with notes explaining why they aren't present, for instance.
The more involved solution involves "diligently documenting" every case of what the player can do, and what state the world can be when they arrive to an area.
Wolfeye Studios maintains a robust tag system to support these possibilities. Let's return to the aforementioned sheriff. If the player needs the sheriff's help to kill bandits, and the player arrives to this point in the story, what is the current state of the sheriff? Did they die? How did they die? Who reacts to their death?
From there, you can create alternate dialogue or game states that respond to that data. NPCs can reference how the sheriff died. The bandits can take different actions depending on if the sheriff is alive or dead. "This is really hard," Loredo admitted, and said that he and his colleagues spent ages "banging their heads against the wall" to make these states possible.
But Loredo said these efforts are worth it because when it works, players feel "really smart." "We have as much storytelling material for if the sheriff dies as if he lives," he explained.
It's one hell of a task, but when developers do it right, "something magical happens that only happens in games," he said. He called it "the feeling of co-authorship" between players and developers.
Loredo ended his talk with a call for the audience to make use of these tools in their own games "There is a magic when you invite players to co-author with you," he said with a smile.
It's not quite the same magic that lets players summon fire tornadoes, but it's a special magic nevertheless.