Baldur’s Gate 3 Was So Huge, Larian Chose to Triple in Size Rather Than Shrink the Game

“It codified your creativity, the ability to come up with an adventure, to play in it, to come up with all kinds of crazy shit, overcome it,” Vincke says. “I’ll give you an example of the most powerful thing I did with it. I have four children, and so when we’d take long drives I’d play D&D with them. I’m sitting in front of the car and I’m saying, ‘Okay, you see a witch in the woods. What are you going to do?’ And so in my head I rolled the dice, and I cheat, and so we go on an adventure and we can pass hours this way…It makes them drop their iPads so that they’re not locked into a screen the entire time. And they have a lot of fun and it’s really cool…We do that just by the power of story and the power of dice.”

In Larian Studios’ two and a half decades of existence, he’s had plenty of opportunities to translate that tabletop storytelling into video game format through the Divinity series and other projects. But now, one of the most recognizable tabletop-based RPGs has fallen into Larian’s lap, and the studio’s desire to bring true tabletop storytelling to video games is proving to be a profound challenge.

Vincke describes Baldur’s Gate 3 as a “dream project” for Larian, understandably so given its history. The first two Baldur’s Gate games were BioWare-developed RPGs set in the familiar D&D campaign setting of the Forgotten Realms, and while there have been various expansions and enhanced editions since Baldur’s Gate 2 in 2000, there hasn’t been a true third game until now. Larian, with its Divinity track record, seems a perfect fit for Baldur’s Gate 3, and came with the built-in advantage of early access experience. Both Original Sin 1 and 2 had gone through extensive, successful early access periods where Larian worked closely with the community to iron out each game’s wrinkles.

As Vincke explains in his talk, Larian prepared itself for its biggest game ever. The studio spent over a year in pre-production, hired for many new positions, and made the best projections it could for how many words the game would have, how much recording time would be needed, what would have to happen to make all the cinematics, and so much more.

But ultimately it wouldn’t be enough on its own, in no small part because of the enormity of trying to convert the complexities of a Dungeons & Dragons game — where players can do anything they can think of, effectively — into a video game system.

“You will often hear, ‘Oh, you have to give the illusion of choice,’” Vincke tells me. “You know you really have to give actual choice too, otherwise there’s no consequences, it doesn’t matter. And we have this dice mechanic, which is really deeply ingrained, you’ll discover when you play. And so failure on the dice, success on the dice have to matter. If they don’t then you don’t have a feeling that actually all this rolling and all these character abilities that you are leveling up have any impact on the game.”

But creating that kind of system was even more work than Larian expected. Baldur’s Gate 3 needed all 12 D&D classes, all of them very different from one another, and all the spells and abilities from 5th Edition D&D had to not only function but also feel impactful within the world.

One example Vincke shares is that of a druid: for a while to make the game’s cinematics work, druids couldn’t talk to, say, a cow, if the druid was currently transformed into a cow. They had to transform back into a human to talk to the cow. That didn’t make any sense. Why couldn’t a cow speak to a cow? But fixing it required a ton of cinematics work to ensure all cinematic conversations with animals were still functional if a druid was transformed.

On top of problems like that, there was multiplayer, multiple languages for the enormous volume of dialogue in the game, and a need for approachability amidst it all: a player who had never touched D&D before should be able to understand Baldur’s Gate 3.

With all this at stake, it became apparent to Larian that they could not ship the game they wanted to make with the team they had. It was either scale Baldur’s Gate 3 down, or grow the studio massively to meet the challenge. So Larian grew, jumping from 150 people to 400 in a short span of time. Larian had the resources to do it thanks to how well Original Sin 2 sold, but the upscale came with its own problems. Larian didn’t have processes in place for so many people. The sheer volume of work, teams, and individuals quickly revealed weaknesses in how they worked together that hadn’t been apparent before. And the COVID-19 pandemic certainly didn’t help matters either, with Vincke saying that the “fragmentation” of the company necessitated by lockdowns was “not necessarily the best thing in the world for boosting creativity.”

Between all these obstacles, there came a point ahead of early access where production nearly ground to a halt.

“Every single week we woke up and said, ‘Okay, this is the week we’re going to finish this [feature],’” Vincke says. “And then at the end of the week nothing had happened.”

At one point, Vincke wondered if Larian should even put Baldur’s Gate 3 into early access. For both Original Sin games the advantages had been very clear: player feedback had made the games better, and they had built considerable good will within the community. Larian wanted that for Baldur’s Gate 3 given its complexity, but were worried it was too big.

“You put it into the early access community, you go back the next day, you see plenty of things that went well, you see all your bad design decisions exposed, your good design decisions you see no one talks about them, so you’re good,” Vincke says of early access. “It allows you to rapidly try things out, you see what resonates, what doesn’t resonate, and when you have to put so many rules as we had to convert and figure out ways people would understand them or not understand them, it is a very, very useful tool to have.”

But Larian put Baldur’s Gate 3 into early access anyway. There were still some hiccups. The initial early access launch was rough and Larian has had to make extensive changes based on player feedback, though Vincke feels that ultimately made the game much better. He admits it was disheartening to see critics publish scored reviews of a game that was very much still in progress, though he acknowledges they had every right to do so — Baldur’s Gate 3 was out in the wild, being sold for money after all.

But with better perspective on what he ought to have done differently in the lead-up to early access, Vincke affirms he wouldn’t change his decision if he had to do it all over again.

“I would certainly organize ourselves better than what we did,” he says. “There were things that I did not expect were going to happen. But the benefits are so clear…You can literally trace the paths of why things were done based back to community discussions, reactivity that you saw, analytics that you saw. That’s the beautiful thing about it. And you can’t do that — I wouldn’t actually know how to do it without having a community. You have a team of thousands and thousands of beta testers I guess. But even then it’s not going to be the same thing.”

Because of the early access community Vincke says Baldur’s Gate 3 is much better able to support the storytelling fantasies Larian wanted to implement from the beginning. It’s because of player feedback that they can answer strange questions that never came up in design, like, can a bear climb up a ladder? (Yes.) What about a deep rothé? (It can, says Vincke, and has a very funny animation.)

And more than just the goofy campaign stories people share about D&D, Vincke says that success for Baldur’s Gate 3 means hearing more stories like that of me and my partner playing long distance, or of people who felt their lives were better for having experienced the emergent storytelling Larian made for them.

“Walking around PAX and having people come to you and tell you, ‘Hey, this is what happened to us and this made a difference in our life. I was going through a really shitty time and I played your game.’ …Games matter, and they can make a difference. And if you make a good game, it can matter more than if you made a bad game. So success looks like your game matters to people and it makes a difference.”

Rebekah Valentine is a news reporter for IGN. You can find her on Twitter @duckvalentine.