How Larian Hopes To Transcend Baldur's Gate 3's Legacy
Turns out: making RPGs is incredibly difficult.
During a panel at PAX East, the first
gameplay from the much-anticipated Baldur's Gate 3 was revealed by developer Larian Studios, which is best known for the acclaimed
Divinity: Original Sin and its
sequel. Prior to this showcase at PAX, we were able to sit down with Larian creative director Swen Vincke to discuss the challenges of creating the game, working with Wizards of the Coast, and much more. Read on for our full chat and check out our
breakdown of the new intro cinematic.
Mike Mahardy: I know Larian has been pursuing the Baldur's Gate license for a while now. Is it true that Divinity: Original Sin 2 was kind of like a tryout for Wizards of the Coast? A way for you to say, "Hey, we can handle this license?"
Swen Vincke, Creative Director: I approached them after Divinity: Original Sin (DOS 1). I said, "You should let us make Baldur's Gate 3. We have a lot of Dungeons and Dragons fans in the office. It's something that would really motivate them." They said no, then. Then we were working on DOS 2. In between, I kept on bumping into Nate Stewart at Dungeons and Dragons. He asked me, "What game would you make actually?" I said, "This is what I would make." He said, "Okay, all right, that's kind of cool." But yeah, at the time, the stars were not aligned.
I then got a phone call from him in December before DOS 2 released - so that was in 2016. He said, "Do you still want to do this?" I said, "Yes, I am actually really interested in still doing this." He says, "Fly over to Seattle." I fly over Seattle and went to an obscure bar with him. There, he showed me three PowerPoints, which contained everything I'd talked about all this time. He said, "I'll present this to the board of Hasbro next week. Do you agree?" I said, "Yes." Then six months later we signed the agreement. We were still working on DOS 2 back then.
Mike: What had changed, with the green light happening then? Was it just the ongoing conversation you were having?
Swen: I think there were a couple of things happening. Management had changed. Dungeons and Dragons (DnD) was doing really well, and DOS 2 happened also. It was in early access at that time, and so people could see what was happening with DOS 2. I think all those things conversed with one another. We took it from there.
Mike: In my mind, as a DOS 2 fan, there was a sense of, "Do they need to do Baldur's Gate 3?" Because you've already done Divinity, which was about as close as video games have gotten to the spirit of DnD. I guess I'm asking: how are you setting Baldur's Gate apart from Divinity?
Swen: There's a lot of stuff. In BG3, I think the very first thing that's already apparent is that everything is cinematic. One of the reasons why we went cinematic for everything—every single dialogue is cinematic—is that we want it to have the party front and center this time, which was not necessarily the case in DOS 2. It wasn't really as strong as what we're doing now. What the two have to do with each other, well if you want to have intimacy, want to show relationship building, you can do it really well when you go cinematic.
Then you have the rule set. It's class based. It's more limiting in a certain way, so you have to be more creative in how you do their things. You have less things that you can do per turn, because you only have a couple of spells while they build up of course, if you're playing wizards or a warlock or a sorcerer. What I'm trying to say is, just look at the level-up curve, for instance. The leveling is something that is very slow in Baldur's Gate. It's very fast in Divinity.
I mucked it up a few times during the demo today, but we're also focusing on the idea of "light and dark." The entire way that it works with "Heavily Obscured" and "Lightly Obscured," that's straight from the [DnD] books. That ports very well. It works really well actually. When you compliment it with the environmental interactions that we introduced in DOS - I can douse fire with water, which makes it dark, which is beneficial to me. That's cool systemic gameplay. That's the kind of stuff that we're looking for.
Mike: Can you delve more into the combat here? What were the biggest challenges in adapting the Fifth Edition rule-set into a video game?
Swen: We were very worried about things like the Fighter and the Rogue, especially at the lower levels, because they have very limited actions to them. That's why I picked a rogue to demonstrate today, because it actually shows how much stuff you can do, because you start thinking, "What can I do in this world?" You start playing to the strengths of the Rogue. As we were developing, we became more relaxed about that. We were quite worried because we thought we were going to have to invent a whole bunch of extra stuff so that every single turn you would have a choice.
There's also always that interaction that you can do between the characters. You're always thinking about, "That guy is going to do that. Then this guy's going to do that." And you make combinations like that. That gives you a lot of agency in trying things out as you progress through the game. That was a challenge that worked out well.
Mike: Going from a classless system to one that's class-based, what are the clear advantages? Was it more of a restraint, or did it actually free you up creatively?
Swen: Classes communicate really well to the player. So you are thinking "My cleric will do this. My fighter is going to go into the front line." It wasn't necessarily as clear in Original Sin, where it was really like, "What cool things did I put together?"
Mike: Divinity was very much your own thing. No other group had its hands in it. Now, with Baldur's Gate, you come in, and it's kind of the exact opposite. With Baldur's Gate, you have that added weight of nostalgia. What's that been like?
Swen: There's multiple ways that you can deal with that. You can let yourself be paralyzed by that pressure, because everybody has a different game in their heads that they want. But we're making a game that we think is going to be fun. We generally tend to make games that we like to play, and then hopefully there are people that will want to play it. I think that in BG3 they will find a lot that they will enjoy. There will be people that will be disappointed that it's not real-time and fast. We just chose to go for turn-based because we think it's closer to the source material, and allows us to do more things. It is what it is. We're just making a game, and we try to make it really cool.
Mike: There's been somewhat of a CRPG renaissance lately. Divinity rests comfortably within that. Why are they having such a moment now? What is it about the audience's taste, or the goals of developers, that led this to happen right now?
Swen: I think that actually is because you have this widespread distribution, so there is more access to broader markets and specific initiatives and markets that in the past were not necessarily identified by publishers, because they didn't think that was going to work. I explicitly was told back in the day that that turn-based [games never sell].
Mike: I feel like games in the early 2000s were very much about accessibility and, in some cases, handholding. But in the past several years, it seems like players have gotten used to learning super complex systems. Larian's games seem to be benefiting from that.
Swen: That's always been the trick, right? "Easy to learn, hard to master." That's literally what it is. You've seen me do very advanced things [in today's demo]. That's not how you start playing as a normal player. You see a tutorial that will ease you in, but then you start discovering things. The magic of these games is that moment where you say, "I tried something and it works," and you don't expect it to work in general, because in the more guided games you don't have it.
It's like, "This is your corridor. Walk in your corridor." You can't get out of your corridor. You see plenty of objects. You can't interact with a single one of them except if it highlights. Then you play this game like this and you said, "Huh. Here's an object. I wonder what I could do with it. Oh, if nothing else, I can throw it at somebody and do some damage."
That experimentation leads you to progress that you probably weren't supposed to have. Now I say, "Hey, now I'm on the real adventure. I'm in over my head. I'll have to keep on doing these things if I want to survive," which is basically what happened to me today.
Mike: That's such an opposite idea of what so many games today seem to be about: min-maxing and optimizing everything to death. But then you watch a demo like today, and I thought it was extremely entertaining because of how often things went wrong, and you had to adapt to that failure.
Swen: It's super, super important in this game, because when we put the dice roll in there, we knew that the dice roll was going to fail also. We had to make sure that failing in the dice roll, be it narratively or be it systemically, needed to lead you to a place where you actually were having fun. Because then you start to really care, because you know stuff is going to happen. You know you can continue to play.
You know you're not going to reload because hopefully, you'll trust that the game is going to guide you to a fun place, just like an experienced dungeon master would. We do also give you abilities, very limited ones, where we give you the ability to manipulate the dice. We give you inspiration points, right? Then you can say, "When will I use my inspirations?" Maybe you'll just hoard it until the end of the game. But you have things like Lucky points that allow you to reroll. These things are going to make that roll an important event, which is what it is on the tabletop.
Mike: Another thing during combat that you seemed to stress today was verticality. The earlier Baldur's Gates had these flat, pre-rendered backgrounds. What's it been like to rework that? What does that allow for? Was that a big focus from the start or was that something that just unfolds when you were saying, "How do we adapt these rules into combat?"
Swen: The latter actually, because we were starting to think about advantage and disadvantage. DnD 5th Edition streamlined a lot of things into advantage and disadvantage. One of the big drives that we had is how do we bring that in an actual way to the player? Height is a very easy one. "I have a height advantage, so, okay, let's start thinking about this. Oh, look at that, you can shove people. Well, that's interesting. Let's start shoving people. Okay, this is rather fun."
That's one of them, but there are others. Light and dark are also really good sources of advantage. Sneaking up on somebody from the darkness, surprising them so you get the benefits. These are things that sound super mechanical when you read them in the books, but they're not mechanical though. When you see them and you play them, they're just as logical.
Mike: Whether it be combat or lore or the storyline, have you worked closely with Wizards of the Coast, or do they just kind of let you do whatever you want?
Swen: Both. When we were setting up a storyline, we worked together with Adam Lee, who is the man who was partially responsible for Descent Into Avernus. He was one of their writers. He came over to Ghent to work with us and we went through storyline permutations and so forth. Then we spent a lot of time at Wizards' offices also. We traveled a lot there as we were building up the entire concept. Over time, trust started increasing in what we're doing. At this point, they're just like, "Surprise us." They're very big fans.
Mike: In terms of lore: coming into this project, did you know what parts of the DnD universe you wanted to pull from?
Swen: Well we actually started out differently than where we ended up. That is partially because of the work that Adam was doing. We knew that the city was going to be central. You eventually get to Baldur's Gate, and the things that you will encounter in Baldur's Gate will depend on how you deal with events before then. There's quite a lot of reactivity in there.
Then we also knew that we wanted to have something like what the previous Baldur's Gates did, where you had lots of diversity in terms of where are you going to go. Baldur's Gate was almost like a carousel of iconic creatures that you had, or locations that you wanted to visit, which means it's very DnD. It's no different here. You're going to see lots of things that you like about Realms reappear.
Mike: Do you see Larian always kind of sticking to fantasy as far as-
Swen: No.
Mike: No? Do you have genres you have talked about wanting to go into?
Swen: No. We always wanted to make very big RPGs. The idea is that if we can manage to be successful with this, we essentially free ourselves, so that we can actually start exploring other things at the scale that we want to do it at.
Mike: You mentioned an early access for Baldur's Gate 3. Have you talked about how you're going to structure that in terms of what that will include as opposed to the full game?
Swen: We'll do it similar to DOS. We'll come in with content, and then we'll add pieces of content, like extra classes, extra origin characters, extra systems, extra spells. We'll add some regions over time also. We want to limit it to first act content, if you want. Players that are joining us during early access, when they play the game on release, they still have plenty of content to go through. But at the same time they do have sufficient content to actually have a really cool adventure during early access.
Mike: Do you feel as if that kind of a release structure could hinder the overall experience? As opposed to just having the game and jumping in and being able to play for 80 hours straight?
Swen: You can! There are two types of people that go into early access. There are the people that are fans and that want to support us early and just... try it out. They generally spend five or ten minutes and that's it. They don't touch it anymore. Then you have the people who want to be part of development, and that send us plenty of feedback, which we listen to, because otherwise we wouldn't be doing it. Then we adapt the game. If you look at DOS2, the game that went in and the game that came out of early access is a very different game. Same thing goes for DOS1.
Mike: Were there challenges with DOS1 and 2, specific you could point out that were really valuable lessons for making Baldur's Gate 3?
Swen: There's plenty, yeah. This is our own engine that we're building. There's a whole bunch of stuff that the team wanted to change on the engine. Now that we have more resources, obviously, we could. That's been a big one. New problems appear, of course. We set up an entire cinematics pipeline to be able to record all this dialogue for every single scene. That really is a big chunk of new work and new problems that appear, because we're doing multiplayer also..
I don't think you've seen that before at this scale. It's quite the challenge. In other ways, there's so much stuff that's written about DOS2 and there are so many opinions. We just tried to make a game that we liked, because that's the only guideline that we can follow anyway.
Mike: Yeah, I imagine that must be tough. I work for a site that gave it a 10 out of 10. It's like, "Of course, people really like this thing. So how do we do something different?" It's always hard to ignore criticism, but it must be even harder in some ways to ignore praise. Are you trying to just block that out while you make this game you want to make?
Swen: Yes. Yeah, otherwise it wouldn't work.
Mike: How's that going?
Swen: Well you've seen the results. I think it's shaping up to be a really good game. We still have a lot of work. We're certainly not done yet. There was so much demand. We also think that we're at the stage where we're ready to start showing it. In a couple of months, we will probably be ready to throw it into early access. Then the real fine-tuning can begin. We have a lot of content. There's a lot of stuff that I bypassed today that you didn't notice it because you didn't know that it was there or I turned my camera at the right spot. There is a lot of stuff there.
Mike: I just imagine that the pressure of your success, coupled with the legacy of Baldur's Gate, must be its own challenge throughout development.
Swen: The thing we are telling people is that we are the dungeon master. And we're just going to play DnD together. That's literally what we're doing here. I think that's an okay approach. Every single person that plays DnD has their own version of it that they're playing.
When we talked to Wizards, we initially thought this is going to be impossible because they're going to want every single rule strictly implemented. On the contrary, they're very open to [change]. There are things that they want to have in there. Then there are things they say, "Well, try to make the best out of it." That's also how they approached 5th edition towards their own players. I think that's a large reason for their success.
Mike: 5th edition is pretty widely revered now. Is that mainly because of the systemic streamlining you mentioned?
Swen: I think first of all, they did a really, really good job designing it. I think the way that they're approaching storytelling, lore, freedom for their players, is exceptionally good. That's a very important part of it. The fact that it's really easy to learn now, but still you have all the depth and complexity that's in there. That's probably the main reason it's doing so well.
You take Dragon, you take the Mine of Phandelver and the starter kits, and you can walk into a room with people who've never played it and they'll come off fans. It's a powerful thing. We had the same thing with DOS. We gave it to people who never played it. We saw that at PAX inevitably, when people walked in there with a friend and they said, "Try this out." We walked out with somebody converted and said, "I can actually play this. I didn't think I was going to be able to play this. I thought it was too complex, but this is actually really a lot of fun." You have the same thing with DnD.
Mike: Larian was a respectable name coming out of DOS1. Now it's a household name in the world of CRPGs, or just RPGs in general. What's it like making that jump in a pretty short time?
Swen: It doesn't feel like that, because we are still bitching at each other continuously about how we can do this better and can do this better. We are very focused on the things we don't do well. Of the things that we do well, we then start taking them for granted because we want to improve our game. There's so much stuff that we still want to do and there's so little time.
You have to remember we rebooted ourself in 2010. Went completely independent. Stopped being dependent on publishers. Meaning that we didn't listen to anybody else anymore telling us how to do our stuff. We started building our engines from scratch. We decided to make multiplayer CRPGs, which is really complicated. It's much harder than single player CRPGs, I can tell you that, and they have to be consistent with the ambition of recreating a virtual version of what a tabletop is.
We're just going forward with that and now, we're adding the entire cinematic angle to it, which we always wanted to do. We have so much dialogue and so many permutations. When you see it cinematically like you see now, it's this [new] layer of polish. You can already see what the potential is of the destination, and how much more emotion you can bring into the scenes.
Then being able to play that with my friends, and all having the same experience in a narrative that we're all affecting, that's pretty awesome.
Mike: It feels like RPGs in video games have been chasing the magic of DnD for so long. Do you think that's something they'll ever catch? So much of a DnD lives in people's heads, whereas video games are very much, "Hey, we're putting this concrete thing in front of you that you're going to interact with." Do you think they'll ever really cross paths?
Swen: What is so cool about DnD is that your agency gets rewarded because you have a dungeon master that reacts to that agency and creates the result of your agency. That's what we're trying to do. We're trying to give you as much agency as we can, giving you an ever increasing toolset of things that you can do within a world that promises to react to you in a way that is going to be rewarding to you. I don't think you'll ever be able to replace a DM completely, but I think we can go very far with that.
Mike: You still have to play DM, it's just for millions of people at a time, not four.
Swen: (Laughs) Yeah, yeah, yeah, I know. But yeah, for each person, it's their own thing. A lot of how the story is going to evolve is going to be a function of their [individual] agency. For instance, you didn't have to decide to bite Shadow Heart in that scene.
(Editor's note: In today's demo, Vincke played as a vampire character and decided to suck his party member's blood while she was sleeping at camp) The fact that you bit her - if the die rolled wrong, that's your choice.
You created that in the game. From now on, she's gone. She's never going to be back. You will encounter situations that were dealing with her and they're going to react to the fact that she's gone. It's the same thing that you would have with a DM. It's like, "I did this. The world happens to evolve this way. I have to deal with this." That is a very powerful feeling for a player. That's also unique to video games.