Ken Levine has outlined the strategies he hopes to pursue in the projects he'll be working on at his new studio, talking about a system of narrative blocks which could make a story infinitely replayable.
Levine began by discussing his history as a writer of firmly linear narratives in games like BioShock, expressing pride in that work alongside his desire to move towards a new model which allows players to experience a game in a different way each time.
"We were known for doing a lot of linear narrative games," Levine said of Irrational. "I'm pretty proud of what we did with in our narrative, we took it pretty far, So why change? Linear narrative puts a boundary between the developer and the audience."
Citing the example of BioShock again, Levine spoke about how his recent model of narrative twists make for great reveals, but can't surprise a player twice. "Would you kindly only works once," he explained.
In opposition to that model, systemic games like Civilization allow almost infinite potential experiences. Those games, favourites of Levine's, sacrifice authorial intent and narrative drive for the ability to continually surprise. What the writer wants to do is combine those two models, resulting in a strongly story-driven game without restricting that experience to a limited number of outcomes.
"What happened to Booker at the beginning of Infinite talked to what happened to him at the end in the player's mind, but there's nothing in the game to really affect that. In Civilization the choices you make at the beginning really affect the way the end in a meaningful way - what types of cities you've built, what cultural trees you've taken. They don't really do that in a linear narrative. Branching narratives do exist, although not really in our games, but they still exist in X number of states. What I want to talk about today is narrative that exists in X to the Y number of states."
Multiple paths and endings may mean you don't get exactly the same experience as your friends, but you'll have a shared narrative with at least some other people. You're still taking one of a set of pre-defined paths, rather than creating your own. Most importantly, said Levine, it's not player-driven. His plan is to create something which is.
"It's my way of contributing to a conversation that I already think a lot of smart people are talking about," Levine admitted. "It's my way of scratching at making player-driven narrative."
Levine described the nuts and bolts of the system he hopes will facilitate that as a broadly faction-based system, where groups of NPCs are motivated by various and often conflicting 'passions'. Those passions, when met or confounded, will affect an NPC's attitude towards players, granting advantages or causing difficulties for the player. Whilst those factors may be broadly aligned within groups, individuals within that group may have conflicting motivations. A village of orcs, for example, may have a broadly held animosity against a nearby elven settlement, but an individual orc could carry a secret torch for an individual elf. Slaughtering elves wholesale will broadly raise your profile with orcs generally, but lower that individual's opinion of you.
The complexity of that interdependent narrative web offers a significant variation of perspectives within a story. By appeasing different passions and affecting your relationships, a player experiences different journeys.
So far, so Fallout, you may think - or Elder Scrolls, or Fable or STALKER. What makes Levine's plans more interesting is his idea of shuffling those passions around each time, meaning characters change each time you play, forcing players to renegotiate that network of social interactions differently in each playthrough. NPC stars could have a list of 10 or 20 potential passions, a few of which can be randomly assigned at the beginning of each playthrough. Those passions will be related to the point of not being mutually exclusive, but could vary wildly in each iteration.
These passions relate to quests, which affect the passion 'sliders' of everyone who is invested in them, either positively or negatively. In Levine's system, goodwill is a limited resource, making appeasing passions a zero-sum game. You can't please everybody all of the time.
Levine went on to to ponder the ability to add content 'in' to a game, rather than 'on' to it. DLC for BioShock was an add-on, he says, because it's in addition to the original content. Adding new technologies or social systems to Civilization would be adding 'in', because it affects the path of the play through by creating new branches and paths. In the system Levine describes, this could take the form of new characters, passions for existing characters, entire new factions or races, or the imposition of external forces which unite otherwise warring factions, overriding passions or giving focus to certain motivations like survival.
By the time Levine had wrapped up, what his audience was left with sounded a lot like a loose plan for semi-procedurally generated narrative systems which could provide a basis for a game operated as a service. Whilst he didn't explain that explicitly, the outline here would clearly fit that mould: an iterative, easily updated game which would be massively monetisable, with obvious potential for user-generated content, scenario generation or modding. As he was very clear to point out, Levine wasn't revealing any specific game plans or making definite announcements, but this marks a considerable shift in process for a man who is famed for his authorial intent and grand narratives.