Adam Brennecke didn’t know whether to feel optimistic or frustrated. He settled for both.
As the executive producer and lead programmer on Pillars of Eternity, he was proud of what Obsidian had built. It was also buggy—a somewhat unavoidable situation due to the small size of the team and their budget as well as the massive scope of the game, but one that underlined the company’s “Bugsidian” reputation.
“We had a reputation for buggy games, and I knew how many open bugs we still had, so I was fairly disappointed with the state of the game when we launched,” Brennecke admitted. “It reflected poorly on the programming team. It's not our fault, per se, it's just something you have to deal with as a developer: Having a game go out, and still having open bugs.”
Smaller bugs, such as glitches in the user interface, were relatively harmless. Unfortunately, users were reporting instances of much more serious bugs. In one instance, double-clicking on an item to equip it to the active character deleted many of the character’s benefits—permanently. Brennecke responded quickly after the bug was reported on Obsidian’s forums to reassure players that a fix was coming.
Another bug flipped increased defensive bonuses on a character’s armor when players reloaded their saved progress, making the game too easy even on higher difficulty levels. Yet another error increased the amount of time the game needed to load areas every time players saved their progress.
Fortunately, Obsidian was prepared to put out fires. A skeleton crew monitored forums and news sites for word of major bugs and rolled out patches to address them. “Usually it's a few programmers and a producer, just to manage the build process and uploading builds to various distribution platforms like Steam, GOG, and Origin. Occasionally we'll pull people from other teams to look at bugs and help fix issues,” Brennecke said.
Not every patch was aimed at putting out fires, referred to as “big ticket” items. In addition to bugs, version 1.03, released in early April, implemented balance changes such as raising or lowering damage for certain characters, attacks, and weapons, adjusting prices on items and spells, and modified other values across the board such as bonus experience gained during combat and the duration of penalties and other effects.
By late April, the need for fixes diminished from code-red status to changes centered on cosmetics and game balance—important, but not so dire as errors that crashed the game or ruined a player’s progress. Developers trickled out of the office for some much-needed time off.
“The schedule toward the end of a project is often fairly hectic,” said Carrie Patel. “You've got people who are working a fair bit of overtime. It's always nice to get a little bit of creative and personal downtime.”
At Obsidian, crunch is an exception, not a rule. “There have been some short periods of crunch at Obsidian. Some departments have asked to put in some time over a month or six weeks. That's happened a few times, just to be clear,” said Feargus Urquhart, CEO and co-founder. “That was for some very specific reasons, things that needed to get handled. But from the start, even back at Black Isle, I never wanted to force someone to work extra because I want them to have a life.”
Some studios have become so notorious for crunch that their reputation for grinding employees into dust overshadows their products. In 2004, Electronic Arts developer Erin Hoffman wrote a blog post under the pseudonym “ea_spouse” that censured Electronic Arts for work hours that took emotional, physical, and mental tolls on employees. At Blizzard North, new hires were given sleeping bags along with their on-boarding paperwork during development of Diablo 2, a project on which many developers crunched for eighteen months.
In 2017, games writer and narrative designer Walt Williams wrote Significant Zero, a memoir that tackled his career in the industry—highlighted by the toll crunching had taken on him over years of working at 2K Games. Writing with candor and self-awareness, Williams confessed that crunch was a poor practice, but that many creative types were drawn to it. “Crunch is destructive. It is not necessary (with good planning), and should not be forced on people. It is also seductive to certain types,” he wrote in an author’s note that preceded an excerpt of his book. “For whatever reason, I'm broken in certain ways. Destroying myself to make something fills an emptiness that I can't shake. That's bad. The excerpt is from a moment in my life when I was at my lowest and giving in to my most self-destructive tendencies.”
Many developers believe crunch is intrinsic to developing games, and to most creative industries. Careers in such industries attract artistic types like Williams who voluntarily work themselves ragged in pursuit of their art. Several Blizzard North developers admitted that without their crunch, many of the features that made Diablo 2 a global phenomenon may not have taken shape.
“It is a constant fight for me,” Williams continued, stating that his choices led to a mental and physical breakdown, but that he still craved crunch. “As an industry, we need to talk about crunch—how we define it, and especially how exploitative it can be. I didn't go into that, because I didn't want it to seem like I was forced to work this way. I did this to myself. Still do, to be honest. And, if I'm being just really open about it, I wasn't sure I could do that discussion justice because I have a hard time seeing it clearly.”
Williams’ situation remains common. Many developers enter the industry when they’re young and unattached, able and willing to push themselves hard to achieve career goals. “We all worked a lot, and it wasn't because we were forced to,” Brennecke said, recalling his schedule on early projects at Obsidian. “It was because we loved the projects and wanted to make the best games possible. Some of my fondest memories are working until one in the morning and seeing the entire team [beside me].”
Long hours making games are when bonds are formed and tightened, and memorable anecdotes made. At three o’clock one morning in 2004, several Obsidian developers including Brennecke went to Denny’s to celebrate. They had just sent off their final release candidate for Star Wars: Knights of the Old Republic 2. “There was some crazy guy harassing us at Denny's, so the police department had to come,” Brennecke remembered, laughing.
Josh Sawyer remembered feeling right at home when he joined Obsidian. He recognized friendly faces from Black Isle and Interplay. Other faces were new, and made him realize he had come full circle. “There was a wave of young guys, basically, which was strange for me. I came to Black Isle when I was twenty-three, and I would say most of the people there were older than me. Then I'm coming to Obsidian several years later and thinking, Whoa, who are these kids? Adam Brennecke came to Obsidian right out of DigiPen. Eric Fenstermaker, I think, had just graduated from college.”
Developers like Fenstermaker and Brennecke reminded Sawyer of himself during Black Isle projects with tight deadlines such as Icewind Dale II. Moreover, he recognized habits that had been dictated by circumstance. “They were so passionate that they were driving themselves into the ground. It did feel like, when we were at Black Isle, we were part of Interplay, but at Obsidian, we were on our own. We sink or swim based on how we do. If we don't execute, it's going to be really rough for us—which, to be honest, for the first few projects, it was pretty rough.”
Memoirs like Williams’ and the backlash against Electronic Arts have led to more conversations about work-life balance in the games industry. Almost nine years after her first blog post as “ea_spouse,” Erin Hoffman said that EA radically overhauled many of its labor policies and practices, an act for which the juggernaut publishing/development studio had received little credit.
As far back as the founding of Obsidian Entertainment in 2003 through the present day, Feargus Urquhart, Chris Parker, and Darren Monahan have talked about the realities of crunch and what part it should play in their company’s culture. Urquhart hires employees because he believes they have something to offer in the long term. If he runs them through meat grinders, forced to watch their personal lives disintegrate because the demands of their job never let up, they’ll either burn out or quit, and he’ll lose a developer who has or may have taken Obsidian to new heights.
With the exception of special circumstances such as post-launch support, overtime is an employee’s choice. “I'm on Pillars of Eternity II now, and I've never worked this much overtime at Obsidian, ever, but it was purely my choice,” said Dimitri Berman. “A lot of people here really love what they're doing and want to make it better. I think our games show that. People will stay extra and put in a lot more effort just to put stuff in that's not necessary for the game, but it makes the experience a tiny bit better for players. People who work really late end up coming in later so they can sleep in, or if you come in early, you can leave early. It's up to us.”
There are two core reasons why companies—not just game studios—provide amenities like gyms, showers, theaters, and catered meals. To make work more comfortable for employees, and to give employees the benefits of home so they’re encouraged to stay at work. In most cases, those reasons are not insidious. Happier developers lead to more productive developers. Obsidian’s managers strike a balance between keeping their team productive and healthy, and encouraging them to maintain their personal lives.
On Thursdays, several developers—Adam Brennecke, Chris Parker, Darren Monahan, and Bobby Null are regulars—go out for drinks. Some employees play on the company softball team. Others stick around to play board games—not because they have to, but because they want to, hearkening back to life at Interplay where Tim Cain hosted GURPS sessions.
As passionate, creative-minded types get older and start families, their endurance for crunch can wane. “I can put in a month of crunch, and I actually am energized and get a lot of stuff done, but then I've got to stop,” Urquhart admitted. “A short crunch can see great results, but after that, crunching is almost like alcohol. It's awesome for a few hours, but then you've got a day-long hangover.”
Age and lifestyle are secondary influencers at Obsidian. Many division leads set positive examples for their teams to follow. “If you're staying late all the time, some people will feel pressure to do the same,” Patel said. “If you're making sure to go home on time and set some boundaries with your work, and speak up when things are snowballing, other people are going to feel emboldened to do the same.”
“As Obsidian has gotten older over the years, I think we're trying to do things better: more processes, taking fewer risks in our projects, making sure we're not all being cowboys and cowgirls,” Brennecke added. “A lot of it back then was just, 'Let's figure it out and work as hard as possible. We'll get the game out the door.' Now it's, 'Let's take a step back and not kill ourselves. Let's see if we can make games with proper processes' and stuff like that.”
The freedom to have a personal life is one pillar of Obsidian’s culture. Transparency is another. “I've been around since the very beginning, and one thing I really like about Obsidian that I don't think I could find anywhere else is really open communication with the leadership,” Brennecke said. “We have four owners at Obsidian, and they're always here working every day. Just this morning I had a discussion with the majority of the ownership about something. Their doors are always open.”
Every three weeks, management holds a company meeting where they discuss projects and other company-specific situations. Urquhart, as the company’s CEO, tends to set tone and direction. His policy on transparency is appropriately transparent: If employees have a question, they can ask, and if he is in a position to answer, he will.
“I think that's an easy thing for a lot of companies to say, but we try to live it every day,” he said. “If you ask me a question, I'll answer practically everything. If I don't want to answer it, I'll just say I don't want to answer it.”
(Author’s note: One situation Urquhart opts to remain tight-lipped about is the circumstances surrounding the departure of Obsidian co-founder and creative director Chris Avellone, who announced that he would be moving on to new ventures in June 2015. Both parties preferred that I focus on the history of Infinity Engine games and the development of the Pillars of Eternity franchise for Beneath a Starless Sky. After investigating the allegations made by Avellone and discussing them with both parties, I agreed for two reasons: Obsidian developers, specifically Urquhart as CEO, do not comment on these matters publically; and because Obsidian, Avellone, and I agreed that the issues had little bearing on the story I set out to tell. I informed all parties that I would mention the issue so as to avoid giving the impression that I purposely ignored or negligently overlooked it in my research.)
Pillars of Eternity’s developers took vacations in shifts so the game would continue to receive support. Before development had concluded, work was underway on other projects such as Tyranny, an isometric RPG built on Pillars of Eternity’s underlying technology, and The White March, a two-part expansion to continue the first game’s story.
Designing the The White March was a comparatively smoother process: Developers knew the technology so they could focus on creating content, while production managers such as Justin Britch could concentrate on improving production processes for The White March and future titles. “In some ways, The White March was sort of a nice transition,” said Carrie Patel. “It was obviously smaller than Pillars, but it also gave us a chance to put into use a lot of the lessons we'd learned, and to take a few more risks with some of the content and the storytelling we were doing, to make something a little bit more unique and different. That was nice, to transition onto something that was going to be a slight change of pace from what we'd just made.”
The White March: Part 1 premiered on August 25, 2015, less than six months after Pillars of Eternity’s release. The White March: Part 2 followed in mid-February of 2016. Changes and new features included a higher level cap for characters, more abilities and companions, and Story Time, a difficulty mode that simplified gameplay systems such as combat so that players more interested in experiencing a story could appreciate the game without getting too frustrated at tough spots.
The White March also provided opportunities for team leads to iron out wrinkles. One was the sense of ownership that area and narrative designers wanted to claim on assets involved in story progression. “We worked out what I think was a much stronger compromise,” Eric Fenstermaker explained. “Narrative designers would review area design documents and attend the area design review meetings of the areas they were responsible for, so there would be a collaboration from the start. That all went very well. It got investment from everybody, and team chemistry at that point was such that I don’t think people felt much of a loss of ownership.”
By the late summer of 2015, Adam Brennecke was growing restless. He had taken a week off to visit with his parents in the Bay Area and then returned to assist in getting The White March underway. All the while he’d been stockpiling ideas.
That October, Brennecke went into Sawyer’s office. Their last wager, worth much more than a penny, had paid off. It was time to bet on their collaboration again and break ground on a proper sequel. Many of their ideas centered on refining development pipelines. “It's important for us, in anything we do with Pillars of Eternity, that the IP is received in such a way that when someone thinks of it, they think in terms of quality; they know it's going to be a good roleplaying experience,” Brennecke explained. “It's important to us that because we own this thing, the brand and the IP stay strong. We don't want to tarnish it with a bad product.”
Everyone at Obsidian had followed reception to Pillars of Eternity: the good, such as praise for companions and the scripted interactions that forced players to make choices; the bad, such as criticisms that environments had been too static, a lack of complexity compared to the Dungeons & Dragons-driven rules of Infinity Engine games, the lackluster stronghold area, and dense expository text that occasionally hindered pacing; and the ugly, specifically the game-breaking bugs that had hobbled Pillars for many players until patches had mitigated or solved them.
Sawyer had likewise spent weeks and months ruminating on Pillars of Eternity. He had published a colorful, presentation-style postmortem on what had gone right and where Obsidian could do better. Brennecke suggested a name for the sequel, and Sawyer took to it at once. “I really liked the name Deadfire,” Brennecke recalled. “It was a region Josh defined early on in the worldbuilding of Pillars 1. I thought that would be a cool name for a sequel.”
Throughout Pillars of Eternity, players find books that describe areas, customs, cultures, and flora and fauna throughout Eora. Reading those books is an optional but fun way to learn more about the world and its inhabitants. One tome, Monsters of the Deadfire Archipelago, described a swath of islands between the Great Eastern Ocean and the Restless Ocean where storms and monsters terrorize captains brave enough to sail the region’s waters.
“I thought that would be a good way of setting the Pillars of Eternity world in a more exotic and different-looking setting right off the bat: Island environments, jungles, a desert, different types of architecture, maybe integrating Polynesian architecture and costumes,” Brennecke continued.
Developers grew excited by the opportunity to build on the foundation they had set—not only in terms of story and setting, but in technical composition.
“On the first title, there was so little time to put extra care into all the stuff we were cranking out,” explained Kaz Aruga, concept and UI artist. “We knew we wanted to polish assets, pay more attention, give more love and care to the renders. We also figured it was going to be a huge game—again—and that we'd have more resources, but not a night-and-day [difference in resources]. So, we had to figure out ways to make the art team more efficient so that we can spend all of our time on art and less time resolving technical issues.”
“With Deadfire, we had a little more time to iterate on those ideas, which was nice,” Sawyer said. “We had a lot more time to iterate on outfits and clothing. We had more time to iterate on the look and feel of the individual companions. All of that really made a big difference.”
“You'd think that, on your second [franchise installment], you'd be past all that,” added art director Rob Nesler. “But the truth is that you did one, and now the consumer expectation is that the next product has got to be better visually, and in terms of the experience, and in terms of features, and be more polished and improved in ways that satisfy existing [fans] and attract new ones. The truth is we are not always 100 percent certain how we're going to make what we're going to make.”
Leaving behind the first game’s European wilderness and setting sail for tropical islands beset by storms and pirates represented a major departure from Dungeons & Dragons- and Infinity Engine-inspired milieus. That was intentional. “I thought it was more important for us to say, ‘These are things that worked for Pillars 1, and we think they can be better in Pillars II if we take a step away from [convention].’ I think now we're iterating on Pillars as opposed to iterating on the Infinity Engine games,” Sawyer continued.
Swapping settings represented a significant but appealing challenge. “Take Pillars 1, which is a very traditional, western-European fantasy setting,” Patel said. “On the one hand, everybody has a fairly common understanding of what that means. That can make generating content and flavor easier, but it's also familiar, which can be a good thing and a bad thing. With a place like Deadfire, where you're having to define something you haven't seen before, or haven't seen as much of before, everybody has a lot of room to play around and say, ‘This would be cool. Let's do this.’”
“Not that I didn’t find the setting of the first game interesting, but Deadfire’s setting felt a lot fresher to me,” Fenstermaker added. “There just aren’t many pirate-y RPGs out there.”
One thing the team agreed on was that while Pillars of Eternity II would usher in a new setting, it should not deviate far from the deep gameplay systems that had made the original game an instant success on Kickstarter. Combat, tactics, progression, parties, exploration—all of those systems should be refined, rather than replaced.
“If you make it less hardcore, does that alienate some hardcore fans? I'm sure,” said Bobby Null, senior designer. “How big of a market does this type of game have? Could it sell ten million units? Probably not. You want to be careful about going after too big a market for a product that's not there, and then alienating the core audience you have. We're always looking at feedback, but I want to make sure that fans of this type of game are happy with the content we're making.”
“In one of the easiest decisions I ever had to make, we brought in the tooltip system that had been built for explaining gameplay systems and then later used in the dialogues in Tyranny to help remind people about key parts of the lore they might be fuzzy on,” said Fenstermaker. Obsidian’s Tyranny RPG utilized a tooltip system that functioned as an encyclopedia, letting players easily look up information on characters, places, items, and game systems. “That eased a big burden,” he continued.
“A lot of the verbiage and name-dropping is less daunting for newcomers,” Brennecke agreed of the tooltip system, “just because they can read about, ‘Oh, who's Eothas? What's the Dyrwood? What's an animancer?’ They can read that stuff in the game and on their own time.”
Developers riffed on ideas. Players should be able to buy, sail, and customize ships. And crews for their ships. And weapons. And the monsters described in Pillars of Eternity’s Deadfire lore book should be only one danger they would face. “The dynamic weather system is one thing that Josh really pushed for,” Brennecke said.
“The game was taking place in a largely tropical area, I felt weather should be more volatile and present in everybody's lives,” Sawyer said. “Also, one of the things people criticized in the first game was that environments felt very static. I believed the addition of rolling clouds on a landscape, rain, lightning, wind that affects foliage—all of those things would make levels much less static.”
Like everyone else, Fenstermaker had a wish list of changes he wanted to see implemented. “I wanted a smaller game. Or at least, one that was no larger than Pillars. Mostly I’d been hoping to squirrel away some resources for polish time and added depth. We did manage to get the companion count down, however, which was at least one small part of my dream. That let us do a lot more with intra-party interactions and relationships and world reactivity.”
Fenstermaker and Carrie Patel developed several treatments of the game’s story. In one, a magical apocalypse laid waste to a society. That germ of an idea morphed into the game’s inciting event: Eothas, god of renewal and light, would come to life and inhabit a gigantic statue of adra. Players would assume the role of their Watcher character—importing their decisions thanks to an option that let them carry over their progress from the first game—and set off after the god.
“Eric and I worked over the months of preproduction to flesh out specific story beats,” Patel said, “how the player encounters factions, how the player encounters Eothas, how we keep this off-screen antagonist—or is he not an antagonist?—present in the player's story, but also tie in things the player is doing to the factions of Deadfire and things that are happening locally.”
Patel and Fenstermaker looked for openings to incorporate elements unique to Deadfire’s setting. “I remember, for example, that for one plot point we needed a big event in the midgame that would involve gaining faction support and using your boat, and we needed a way for the player to force a connection with Eothas and obtain the next clue as to his intentions,” he said. So we kicked that around for a while and settled on having a big Moby-Dick-style whaling chase where you’d try to lance Eothas with a soul-draining harpoon to force a connection with him and gain the next piece of the puzzle. In retrospect, I don’t think it would have lined up well with where the rest of the story ended up, but it was a fun idea.”
Fenstermaker’s favorite part of preproduction was his research into Deadfire’s time period. Just as Sawyer had drawn inspiration from the Renaissance period for the original title, he and Fenstermaker drew from the seventeenth century for the sequel: the advent of corporations, and a push by Europeans to build their empire by controlling trade. Strapped for resources, the Europeans had established trading posts that granted them absurd levels of power. The outposts used that power to claim territory for their countries.
“Imagine if the US government told Comcast they could build fortresses in distant lands and annex territory on their sovereign’s behalf,” Fenstermaker said. “As much as they wanted. And government would also look the other way in terms of any atrocities and war crimes. You know Comcast would be down.”
Fenstermaker read Merchant Kings, a book that describes leaders of the major trading companies of the time in gory detail. He decided that most of them must have been sociopaths: They had been bent on attaining power at any cost. Companies waged war against one another, murdered or enslaved natives to usurp their land, and severely punished anyone who opposed them. “If you violated a company law, they might blast your hand off with gunpowder. I’d never seen that setting in a game, at least not examined in such a way that it didn’t gloss over all of the insanity. It was something I thought would be great to explore and drop the player in the middle of.”
By late 2016, Fenstermaker and Patel—with plenty of input from Sawyer—had outlined Deadfire’s major story beats. Then Fenstermaker, a new father, resigned. His parting was amicable. He had realized dreams by working at Obsidian for so many years. Now, he wanted to raise his child.
“White March benefited from many such lessons about writing style, lore distribution, design and writing process, collaboration, et cetera. I was excited to see what we could do with a full-sized game after all that. And then of course I totally bailed on my team. Sorry, dudes,” Fenstermaker said. “By the time I handed off my duties to Josh, I think we'd laid out the places you'd visit, the basic sequence of events, and the role the major factions would play. Call it a first draft. It had holes in it, and a lot of the details had yet to be added or continued to evolve afterward.”
There was no question as to how Obsidian would bankroll Deadfire. Crowdfunding had worked once. All the enthusiasm surrounding Pillars of Eternity practically guaranteed it would work again.
“We were able to go and make Eternity II whatever we wanted it to be,” said Urquhart. “That's what that success of Eternity 1 did: I don't have to be on the road right now selling games, and selling games that, while we're proud of and enjoy every game we've made, a lot of times, we're a business, and if we want to keep people employed, we have to sell a game to a publisher so they will hire us and so we can pay people.”
Although Pillars of Eternity II: Deadfire would be crowdfunded, the campaign shared few similarities with its predecessor. Instead of going through Kickstarter, Obsidian’s management chose Fig, a crowdfunding platform devoted to games on which Feargus Urquhart remains an advisor. Fig allows users to fund projects by pledging money to various tiers in exchange for rewards, but it also enables investors to buy equity in a project.
As he had done in 2012, Adam Brennecke spearheaded the effort to build Deadfire’s campaign. Working with a small team of developers and marketers from Obsidian’s then-newly minted marketing department, he outlined reward tiers and the Deadfire project, and helped direct a trailer that invited players to return to Eora.
Urquhart pressed Brennecke to launch Deadfire’s campaign as soon as was feasibly possible. Brennecke resisted, preferring to wait until each discipline had made a certain amount of progress refining pipelines and generating content. “The thing that was important about the Fig campaign for me was showing off something that looked brand-new. The first time you saw Pillars II, I wanted you to say, ‘Wow, that looks really good.’ That definitely was our focus for the Fig campaign.”
After eight months of preproduction, the Fig campaign for Pillars of Eternity II: Deadfire went live in January 2017. Players got their first look at the game’s dynamic weather system, navigation by ship, and the eponymous archipelago’s tropical paradise.
While different in terms of crowdfunding platform and the option to buy equity, Deadfire’s campaign shared one thing in common with the first game’s Kickstarter. Both games set a funding goal of $1.1 million. “We decided that going with the same amount as the first campaign would be fine,” said Brennecke. “People would have questions if we had tried to raise more or less, so we went with the same amount.”
Deadfire broke the original game’s funding record, hitting its goal in twenty-three hours. Obsidian’s developers held another viewing party, then kept working.
“The things we tried to address were things that could be better,” Brennecke remembered. “Like, if this was our second chance at Pillars, how could we improve combat? How could we offer more player choice to improve customization? How could we improve issues with making the narrative offer more player choice?”