Obsidian's Josh Sawyer on Pillars of Eternity II, Publishers and Dwarven Discrimination

By Ian Dransfield on at

Josh Sawyer is the design director at Obsidian Entertainment on Pillars of Eternity II: Deadfire. He is a man with chops when it comes to your hardcore CRPGs, having been in the industry since the days of Black Isle Studios and the original Icewind Dale. He is handsome, has cool tattoos, and can grow a proper beard. What I’m saying is, I think I might fancy Josh Sawyer.

What a hunk. So please excuse any fawning noises, mere animal attraction shall not waylay my quest for TRUTH about Deadfire – TRUTH about double-dipping into crowdfunding, TRUTH about designing around D&D when lots of your players might not be D&D superheroes, and most importantly the TRUTH about how most players are, frankly, racist against dwarves.

I guess crowdfunding the sequel to a highly successful crowdfunded game is what we in the business call a 'no-brainer.'

Sawyer: The nice thing about crowdfunding in general is that when the publisher is not paying the bills, their input is still valuable – but it's not as integral to how you shape the game. And so it's nice to be able to just think about, ‘okay we got all all these people going nuts on the beta, what are their issues, what are they having fun with, what are they really not having fun with?’ And trying to really shape the experience around them.

From that perspective I'd say, it's largely the same as the first game. I would say that because this is our second time around, we really have to nail a lot of stuff that on the first game was maybe a little uneven and sloppy. For example, the stronghold [from the original Pillars] was something that people were like, "This doesn't really feel like it's integrated that well." And, yeah, it was a stretch goal – but it just doesn't feel like there's a lot of it there.

So with this ship/crew system [Pillars II’s version of a stronghold, essentially], we actually have gone through three revisions of this. There's a chance that people won't like it, that would be really rough, but at least they're not gonna say that there's no effort that went into it.

So from that perspective, I feel like there were certain things where, because it was our first time back at making this type of game with Pillars I, for a lot of the developers it was their first time making this type of game at all, there were things that were a little uneven that in this game we just really want to hit and nail perfectly. Because I think our backers have high expectations, which they should have, for the quality of the game that we're making.

No publishers saw the success of the original and tried to grab a piece?

I wouldn't say we had a lot of publishers kicking our door down because again, especially to the big publishers, this is still a niche game. This is still a game that is considered... well it is primarily PC focused, PC, Mac, and Linux focused. Paradox did make a console port of the original, but really it is a PC role-playing game, with a hardcore fan audience. And it sold very well for the sort of game that it is, but you're probably not gonna find Electronic Arts saying like, "Oh man, if only we could sell one million units of a game." They don't care.

Are there times you ever miss a publisher behind you cracking the whip, in terms of budgets and feature creep?

Our initial goal in terms of the actual area map size was to be a little bit smaller than Pillars, actually. Whoops. It's bigger, it's not maybe tremendously bigger, but it is bigger and the world map itself and all of this exploration adds a tonne of gameplay to it.

So crowdfunding did allow us to expand in terms of, we redid all of the rendering, we redid all of AI from the ground up. And [we didn’t just re-do] the visual effects, but the gameplay effects, like the game logic was rewritten entirely. I think without the additional funding it would have been really hard to justify some of that stuff, and I think the fans would have been bummed out if we hadn't improved on those things that were maybe okay in the first game, but could have been a lot better.

Obviously we have to be cognisant of there is a real lump of money that is going down as we make this thing. But I try to separate out what I do a lot from what the money concerns are. And obviously there's a point where people come and kick in my door and say like, "Money, it's running out. What are you going to do?"

But all the choices that I make are really driven by the player experience, and the range of tastes that our players have, and trying to make our really hardcore fans happy, and make the people who don't have as much time, who just are interested in the story, also have a really good time with the game.

You’ve been working on this series a long time now – since at least 2012 – how do you find that?

I will admit, it can be exhausting working on the same IP for a long period of time. I mean, anyone who's worked on an MMO can attest to how draining that can be. But because there is so much new stuff in Deadfire, it is exciting to see how people are interacting with new systems, all the new features and stuff like that.

You have a lot of data from the original showing what players actually do as opposed to what they say they do. What’s been the most interesting thing for you there?

I think one of the most notable ones is race preference. So in Pillars I, we started doing something, which we continue in Pillars II, which when you hit New Game, your character defaults to human, but it randomises the gender of the character, and it randomises the ethnicity of the character.

So that has already been that way since Pillars I, but we still find that the majority of players make white, human, I think typically men, which is fine, but it's also interesting to see what they don't make. Almost no one makes dwarves.

We hear people say that they like dwarves, they like the dwarves in the world, everyone seems to be like, "Yeah, fantasy dwarves." But they don't make them. They want a companion who's a dwarf, but they don't want to make a dwarf. So it's an interesting sort of dichotomy where you see like, well they like priests, but they don't necessarily like making priests. They want to have a healer, they want to have a support character – but they want to make the two-handed sword-swinging psychotic.

We find this overwhelming tendency towards making devoted fighters who are the single weapon crazy people, and assassins. It's like the concept is very appealing, they do a tonne of damage, and people are just like, "Yeah, assassin. White dude assassin."

They're playing all these characters, and they're really doing an analysis of all these different classes, but when it comes to the characters they make, they like making this certain type of character. Which is fine, it doesn't mean that we should emphasise things differently, but it does make us view their feedback in a different way.

Personally, I usually make savannah folk characters, who are more like Mesoamerican, like humans. Or I make Boreal dwarves, like Sagani, from the first game, a female. They look cool, and they're cool characters, and nobody else plays them. But there are people that like Sagani, from the first game, they really like her as a companion, they're just not going to make them their main character.



How much does accessibility factor into making something that is so unashamedly for the hardcore players?

It's a difficult balance because, I mean, I grew up playing D&D, I still play tabletop role playing games, I like hardcore RPGs, but I do know that there's a tonne of stuff about the games that I make that can be very off-putting to people. So for me, I don't think we will succeed if we try to make the game fundamentally something different.

For example in Pillars I, we had an affliction system. There were, no shit, like 30 different types of afflictions that could applied to you by different characters. Keeping all those memorised, and what countered what, that was really... Even people who were experienced players, were like, "What pre-spell counters this versus that?" And it's like ‘oh my god’.

So for Pillars II, we said, "Okay, let's rethink how we structure afflictions." Now each attribute has three afflictions. So there are three afflictions each [for dexterity, resolve and constitution]. And they each tier up from one another, and you can counter them by applying an inspiration from the same attribute on the other side. That's it.

We still have a tonne of afflictions – there are some that fall outside of this framework – but for the most part those core 18 afflictions get used a tonne. And they're all progressive from each other, so you always know that the one above it has everything effect below it. Easier to understand, easier to learn. The basics of countering are much more consistent, and so that's something where it's like, "You know what? This game can still be very tactical, have great depth, but let's not just make it a mountain to climb for players who are trying to understand how the system works."

You guys have been working on games like this for a long time – is it hard to remember to keep things streamlined, or accessible?

Yes. It's a game that we know how to make. I guess the thing is though, going all the way back to Icewind Dale ... I worked on Icewind Dale, and then I was the lead on Icewind Dale II, and ... In QA there was such, at Black Isle, there was an enormous gulf in player capability. And a lot of it depended on system mastery. And I was someone who, I grew up playing D&D, I played a tonne of it at university, too much at university, and I kind of just assumed that everyone understood how to play D&D, and so I designed combat and counters that were just ruthless, and brutal, and psychotic.

I remember one experience in particular, where we had a QA tester come up to my office and he said, "This is impossible. This fight is ridiculous, I've been trying to get through this for two and a half hours. What in the world were you thinking?" And I'm like, "What fight?" And he told me the fight, and I turned to my office mate who was also really good at these sort of games, I was like, "How many tries did it take you [to beat that fight]?" He's like, "I think I got through it on the first try." I'm like, "Yeah, I got through it on the first try too."

And so this tester says, "You are fucking lying. You're so full of shit." He's like, "Show me – show me how you did it." So I load it in, and I started pre-buffing. And I had three casters going for five rounds pre-buffing, and people drinking potions. He's asks, "What are you doing?" "I'm pre-buffing." He's like, "What do you mean?" So I explained all the different bonuses that I got, and how they stacked with each other, and how I cast the longest duration spells first, so that by the time I got to my shortest duration ones, that they were at the end of sequence and all this stuff. And then I transitioned, and I went in, and I fucking just wiped out the whole map. And he was like, "That's how you're supposed to do that?" I'm like, "Yeah, that's how I do every fight."

There are people who want pre-buffing actually in Pillars II. I'm like, "No, dude." Because it's these things that create gulfs. We have buffs, and they're powerful buffs, but you can't pre-buff. You have to do it in combat. So there's an opportunity cost. The smart character can still time the things out and be crafty and clever, but it doesn't create this enormous gap between players.

That's the stuff that I want to avoid. I want there to be tactical complexity and depth, I don't want it to be something where there's no-brainer choices that if you don't make you’ll handicap yourself, and if you don't play the game in a very specific way, you're just going to set up for failure.