My friend Lucien Soulban is a Lead Narrative Designer at Obsidian Entertainment. He's been nominated for a BAFTA for Best Story in 2013 for Far Cry 3 and in 2015 for Far Cry 4. He has a slew of credits on AAA titles, including assorted Far Cries, Rainbox Sixes, Deus Ex: Human Revolution and Warhammer: Dawn of War. He is also extremely well loved in the writing community in Montreal; we'll be sorry to see him go once it's important to go into an office again.
Alex: So I gather you just moved to Obsidian, and, let's get technical right away, that means learning a flock of new software tools.
Lucien: Obsidian has its own tool set because it has a lot of branching. The tools have been around for a while now. It's a lot of learning on the fly, because I don't have the benefit of being at the company to sit down with people. I handed in my two weeks notice at Ubisoft, and literally the next day we went into quarantine. And so I left Ubisoft remotely... my last two weeks were remote. I began at Obsidian in April, and it's also been remote. So there's a lot of assumed knowledge that everybody else possesses and they've been doing RPGs for a while.
Alex: Well, learning any toolset is harder when you can't go over to somebody and say, Will you take a look at my screen and tell me what the hell's wrong? What did I break here?
Alex: How would you say that the style of storytelling -- you're talking about branching narrative tools. How is the storytelling different at Obsidian?
Lucien: Well, Ubisoft was only beginning to look at RPG in greater detail over the last few years; there was very little in the way of branching dialogs initially. There was no need to give the character multiple options for dialog. There was no need for UI to call up dialog options or anything of the sort. It was a lot more linear. Obsidian has been doing more of the sort of traditional RPG approach where a lot of it was, “This is the node where you ask more questions. Here is the node where you progress the quest. So there's a lot more compartmentalizing what you write in order to fit the stubs as opposed to a linear path. You've got to be a lot more careful with how information is transmitted, because if you pass information out along the tree, the player may not see it. So you have to be really surgical about what the nodes say along the critical path in order to feed that information. There's a lot more troubleshooting on the fly.
Alex: So are there any debugging flags that you can set in your software to say, over multiple playtests, does this bit always play, does this bit ever play?
Lucien: I’m still learning because the system is seriously robust. But yeah, you've got to go into the toolset to do it. It’s built with all kinds of conditionals and pathing options. It’s crazy how much love and finessing went into it.
Alex: Does the software let you visualize the tree easily?
Lucien: Oh yeah, absolutely. Basically, the best way I can put it? It almost looks like a mind map sort of thing, you know, where it can expand outward into a variety of different responses. But you can collapse responses or look at the tree in its entirety. You can highlight specific nodes through menu selection.
Alex: I hope you can do, like, Minority Report. (Waves hands as if moving holograms around.)
Lucien: I wish. I wish. It basically looks like a poster board with cards on it and branching lines. But you can also put conditionals in there. There's a bit of programing. So, for example, you'll put a card there, you'll say, okay, this node is going to trigger three options for dialog, for responses from the player. And then you can put conditionals saying, well, this note is only triggered if the player has inadvertently gone forward and finished the quest before ever talking to the NPC. If that condition isn't met, it then gives you just the other two options. Or this node feeds is shared with multiple conversations. Coming from linear games, that’s a new trick I’m learning to master, like going from Chess to Star Trek 3D Chess.
Alex: A lot of what we do is just linear, and so voicing your story is like making a movie. You've got a character with a name and a backstory and relationships with people. Even if you were to give the player dialog options, they would have to be within a narrow realm of what's reasonable for this character and the situation. Say you're playing Hamlet, he's not just going to go off and say, "All y'all go fuck yourselves." That's just not Hamlet.
Alex: So if you suddenly start defining the main character-- What does low IQ Hamlet sound like? Wow, that's a can of worms.
Lucien: It's up to us to limit the parameters. In general you want a character who sounds at least reasonably intelligent, but not really make any sort of moral choices or decisions until the player is specifically asked to make a moral choice or decision. It's going require testing. There's gonna be times where we step over the line and we give too much characterization or go too flat. I think it's just a matter of, like, range finding. When you pick up a bow, you fire an arrow. The first arrow is too high. The second arrow is too low.
Alex: And the statistician yells, "We hit the target! On average."
So when you're writing for a blank character versus a character with a voice and the past -- how do you think the player experiences those differently?
Lucien: I think that the genre of the game -- whether it's an FPS or over-the-shoulder third-person, whether you can see your character or not -- I think that creates a player expectation of who their character is. The minute you're in the character's head, I think a lot of players are used to having the characterization pulled back more into stereotype or wish fulfillment. But when you can see your character, when the camera shows you your character performing, then there's a lot more of an expectation of characterization. In Ubisoft FPS's, we made a conscious decision never to pull the camera out of the character's head, or at least only in very rare moments. The character is a tabula rasa, because [Ubisoft HQ in] Paris and Editorial wanted you to feel like the real you could be dropped into the game world. Accessible to your personal reality was a part of the discussion.
Alex: In an RPG, I am the main character, so whatever I want to do is whatever the character wants to do. In a game with a defined main character, the character wants to do things that the player doesn't automatically care about. For example, in our last game, Arthur wants to escape the town of Wellington Wells to find his long-lost brother. As a story-teller, my job is to get you to care about the brother, which we did through a series of audio flashbacks. Is that enough, though? Or do you have to give the player a player-focused motivation for pursuing the levels that make up the game? E.g. "I want to explore a weird world" or "I came here to chew gum and bash heads, and I'm all out of gum."
Lucien: The player needs to empathize or sympathize with the characters. When you're talking about a tabula rasa character in an open world game, you force them to act because the situation itself is going haywire. Their life is in danger. It's more immediate of a concern... it’s survival.
Alex: You're talking about Maslow's hierarchy of needs.
Lucien: Yeah. I've been using Maslow's hierarchy of needs so much I was trying to avoid invoking it. But talking about the very base of the pyramid, if you're in a first person perspective, you've got these immediate threats that rob you of security, that rob you of your immediate survival. I think if you have a character who is better realized with motivations and sense of priority and everything, there's more opportunity to push them higher on the hierarchy. Even, by the end of the game, allowing them to arrive at that self-actualization top of the pyramid. But you really have to work to get the player on board with those motivations. You take a look at a game like Watchdogs 1. The issue, I think, with Watchdogs 1 is that they had a revenge storyline. The main character was seeking vengeance. But we almost never saw the niece, we never got to know her, interact with her in a meaningful way, or say, fully saw what happened to her. That character was never a part of our experience. As a result, there was a disconnect. The motivation of the character was only that of the character's. The players were more interested in exploring the open world than they were in pursuing his revenge quest. I wonder if it mattered to them beyond their own desire to stay invested.
Alex: I had the same problem with Last of Us 2. Ellie is so hell bent on getting revenge for something that I don't think she needs revenge for, I don't approve of what she's doing to get it, right, "I'm out. I can't do this anymore." First of all, you don't need revenge. Second, it's not going to help anybody-- To the creative director, sure, it's a metaphor for the pointlessness of the endless Israeli-Palestinian conflict, but if you want to make that point score, first you have to bring me in, make me feel, "Yes, I must have my revenge, no matter what the cost!!!" *before* you show me that it's pointless. Otherwise it's just pointlessly pointless.
Lucien: Revenge has become a shorthand in a lot of games, because revenge implies a call for action that can involve violence. "Oh, people want revenge. It'll be fine. Somebody wronged you." And you can empathize that if it happened to your family, you'd want to do that as well. But it's a tricky line to walk. The player needs to feel that they've been brought into the situation, or it's just the character going off on their own, and the player feeling, “Well, OK, I guess I'm along for the ride and I'll just use this as an excuse to perpetuate violence."