Infinite Possibilities: Ken Levine

What’s next for a guy whose last game sold 6 million copies and nabbed more than 85 awards? Starting over from scratch.


These days, Ken Levine is getting back to the proverbial drawing board. During his periodic five-hour walks with Andres Gonzalez, the lead level designer on BioShock Infinite: Burial at Sea, the pair ponder gaming’s more nebulous questions. Do game narratives need to be linear? Can designers give players the power to rearrange and customize storylines? Would players respond to that at all? Seven months after the shutdown of Irrational Games, the powerhouse studio best known for developing the BioShock franchise, the company’s co-founder is focusing on the fundamentals—a pared-down staff, an upcoming game in extremely early development phase, and a reassessment of the core principles of narrative design. At his to-be-named new studio, Levine is on a mission to expand plot-driven games by breaking the story into “Narrative Lego” the player can impact. From there, the bite-sized narrative bits will systemically rearrange themselves to create a larger unique narrative. It’s a dangerous experiment, one that could very well prove commercially unsuccessful despite Levine’s hefty industry accolades, but it’s also an endeavor that could put the player into the creator’s chair and change the way we interact with games.

Get In Media: What challenges does implementing a Narrative Lego framework present?

Ken Levine: If you break down a game like BioShock Infinite to its most basic level, story happens when you cross a certain trigger point and we run essentially a narrative sequence of some kind. That can be in engine or out of engine, or a player can be locked in place or not be, but still you’re waiting for a known story element to play out. You have to change how you think about it: “OK, the story I have in my head, I can’t really tell a specific story because I’m not going to force the player down a specific path.”

What you’re really doing is breaking down the narrative. … We’re thinking of story more as scenario, a situation where a bunch of things can come out of it. It’s probably better to think of a TV show than a movie. Say you start at the beginning of Breaking Bad. The scenario is you have a world, a city that’s highly involved in drug trafficking. You have a character who has found out he’s dying of cancer and wants to provide for his family. He has a brother-in-law who’s a DEA [agent]. He has a marriage that’s sort of losing its spark. He has a kid with some physical challenges who is going to need probably extra money to make sure he’s OK. He has a ride-along seeing the drug world and he realizes … maybe he can make money improving on the product that the drug lords are making using his chemistry abilities. That’s a scenario in which a million different potential stories can roll out of. There are dozens and dozens of stories that rolled out from the actual show, and if you sat down you could probably make up hundreds more stories that could roll out of that exact scenario, that exact set up.

[In our upcoming game] we started building out the potential, like who are the characters in this scenario? What are the conflicts in this scenario? What can the player do to push and pull on the characters in this scenario? … It’s really about a broad range of characters, a broad range of conflicting passions. We think of zero-sum games between passions, like obviously the passions of Tuco, the drug lord in Breaking Bad, and the passions of Hank, the DEA agent, are very, very different. They’re competing and there’s a zero-sum game there where the passions of Walter, some tie in with Tuco’s passions and some tie in with his brother-in-law’s passions. …

Finding those interesting sets of characters, that interesting set of conflicting and overlapping passions, and setting the player loose as a person who can affect those in the world is something we find very interesting. I don’t think it’s going to work if you just either approach it as a linear narrative or you approach it as, OK, here’s a bunch of branches. You have to break down the narrative impulses into their smallest coherent chunk and build those and then have those, like any other systemic game, be able to interact with each other in lots of interesting ways. …

Photo Credit: Ben Leuner/AMCPhoto Credit: Ben Leuner/AMC

GIM: Does having that open-endedness and giving the player that power create problems in terms of establishing the size of the game?

KL: Yeah. … We are heavily invested in thinking about that when we make certain decisions on the design. We have a fair amount of thinking already about the math that will yield. How many potential narrative chunks, a Lego as it were, how many chunks is this decision going to yield? We’re sort of thinking of the game as … different towns that are essentially different factions. There’s lots of mixed allegiances within each of these towns, but you’re interacting with them. We’re sort of thinking it’s three towns, three major characters in each town or village, [and] every time you add in a character, guess what happens to the amount of content you have to create? It goes substantially up. We actually have guys talking about things like factorials and things like that, math I don’t really understand, but fortunately it yields a number for me of how many lines I’d have to write. … That said, you have to have enough content so the experience is really meaningful and really rich and really deep, but you also can’t put yourself in a situation where you can literally spin out of control and set yourself to write an unwritable amount of content. …

Knowing that from the beginning, knowing that you’re heading into something very dangerous, is important, and on a smaller team we already set up a bunch of what we call pillars for the game. … Pillars are the core principals of the game, and we have about four pillars for the game. Any time you make a decision, you have to look back at those pillars and say, “How supportive of one of those pillars is this design decision?” If it’s not supportive of the decision and it’s expensive to make, it’s going to take a lot of time and resources to make, it’s probably the wrong decision for the game. … It’s a way to keep you from creeping in features by establishing these pillars and sort of always asking yourself how to tie into the pillars. The number 2 pillar of this game is making a game that has a meaningful, re-playable narrative. …

GIM: Do you have any way of estimating what adding a narrative chunk does to the size of the game?

KL: Fortunately, my lead designer and my lead tech guy, they both come from an engineering background, and one of the reasons that’s valuable is that they’re already very much thinking, as we make design decisions, how we model what the impact on the amount of narrative we’re going to have to create. We think about how many characters [and] how many different passions. … What are these things that the characters care about? For instance, Walter White’s passion is taking care of his family. That’s a big passion. If you help him with that, he’s going to like you. If you don’t help with that, he’s going to like you less. Every passion you add is more content, and then if you have a character with a zero-sum game with another character, say Tuco and Hank [with conflicting passions] … if you can keep those zero sums aligned, meaning if it makes Hank like you one unit more, it makes Tuco like you one unit less, that’s what you call zero-sum game alignment. You know that if Hank is on this point of liking you on this passion, Tuco is always at the inverse point numerically. That gives you a smaller amount of content you have to create because it limits the number of states along that passion.

We were having this discussion about the communication of information in the game. When you take an action that’s going to make someone like you more or dislike you more, is there a way for you to intercept the message? Think about Game of Thrones. Remember how they’re always shooting down the [ravens] in Game of Thrones to keep information from getting to their enemies or to their friends? Do you have a system where there’s some way where you can impact the information world to intercept messages so you can do things that people may not like but they’ll never hear about it? … If I do an action that Hank is going to like and Tuco is going to not like and the information gets to Tuco and not Hank, if you have a way of stopping information then the problem is that they’re going to then be unaligned on their feelings about it. They’re going to be not necessarily at the inverse … therefore your potential state, your narrative state related to those two characters in this one passion, is out of sync and that essentially creates another state that you have to support. …

GIM: Finances are a huge challenge for young developers. When you are working on a blockbuster franchise, are finances still a hurdle?

KL: Part of the reason I’m doing something different in looking into the future, I do have some questions. … BioShock Infinite and BioShock were fortunate to be linear single-player narrative games that still sold really well and made money, [but] they don’t really have re-playability. … Each second of linear narrative is very, very expensive to produce. I won’t be lying when I say that looking into the future, one of my motivations for thinking about narrative and thinking about re-playability in narrative is, as a narrative guy, as a person who loves narrative, I was worried that to some degree we’re going to price ourselves out of existence. That could happen.

The middle tier game is disappearing. …People tend to buy, especially on the console side, the biggest super big titles: the GTAs and the Call of Dutys. They’re not really buying a lot of the midrange stuff. … The amount of numbers you have to move to be successful in the midrange is getting higher and higher because the costs are going up. Where people are having success is on the more independent side, making smaller passion projects that rely more on innovation and freshness of vision than they do on scale and scope and raw muscle of, “Here’s a ton of stuff. Great, cut scenes.” That stuff is all great. I’m not criticizing any of it, but narrative, unless you have a huge, huge 25 million seller, I think is becoming a more challenging proposition to make economical sense. That’s one of the reasons we’re doing this, is to find a way that narrative can be re-playable, hopefully having hundreds of hours of gameplay without having to spend that minute-per-minute cost that you traditionally see in making a narrative. …

GIM: For a student who really has no portfolio and no contacts in the industry, what do they need to build their resume?

KL: They need to work on things and ship them. I think the way they do that first is what can you do now that can help you demonstrate your capabilities? What mods are out there? What can you make on your own? Anything you can make that can demonstrate I can make something, even if it’s crude, even if it’s not super polished or doesn’t have great artwork. There are millions of people out there now … who want to make stuff. There are tools that exist to help you, whether it’s Unity or Unreal or whatever, that will help you get so far down the field.

REALTED: Breaking into narrative design with Obsidian Entertainment creative director Chris Avellone.

If you’re not a technical person, you need to align yourself with a technical person. There are a million people out there who want to make stuff, a million people, and you’re going to deal with frustration, especially if you’re working with a bunch of people casually and they’re not being paid. People are going to flake out and they’re not going to perform, but it’s no different than a young filmmaker who wants to get started and [is] making small things with his friends. Let me tell you, I’ve made those things. People don’t show up. People disappear. Your lead actor gets arrested. Terrible, annoying, horrible things will happen and generally the prize goes to the guy or girl who sticks it out, who says, “OK, I’m going to make this thing and I don’t care what happens. I’m going to make something and I’m going to demonstrate my quality to people.”

The way people will be convinced to hire you is they’ll see something of yours and they’ll be like, “Oh my god, how can I not have this person at the company?” But they’re not going to probably find that beauty hidden within you. …

GIM: Do you have any trepidation about trying to change how story is delivered in games?

KL: There’s always a reasonable chance of, we’ll try to put it up on its feet and what if it’s terrible? What if it absolutely is not interesting at all? That’s less of an issue if you’re making a BioShock game. Having done a System Shock game, System Shock 2, [BioShock] Infinite, you feel that there are things you can fall back on. At least you won’t completely fall on your face when you’re starting out the project, where here we don’t really have that security.

…You may look back at this article and say, “Wow, I should have warned that dude, because he was about to really fail horribly. I should have said something.” It’s always a possibility when you’re working in a new space, but we never made a game on console before BioShock and we were worried about that. We had done a game before that actually never came out, that we pulled from the market. … We were fortunate that we were able to pull it and not ask the audience to play our failed experiment. There were a lot of good things about it, but in the end it was our failure. Nobody wants to go through that, but if you don’t risk things you just end up in stasis and you end up being the guy creating the same number 12 of the series. You’re not challenging yourself any more. Listen, I totally get why people do that. It must be very comfortable and there’s often a lot of money in it. I don’t really want a life where I’m super comfortable going to work every day knowing exactly what I’m making.

Related Content

Have some feedback for our editors? Contact Us