Slaying the Course

Dyscourse by Owlchemy Labs

Published: October 15. 2014


Owlchemy Labs’ Dyscourse takes a newfangled approach to the point-and-click adventure game

The origin of Dyscourse happened in a cabin full of a bunch of indie developers. “We had been playing a lot of the party game Werewolf, so Devin [Reimer] and Carrie [Witt] and myself were thinking about group dynamics, trust, betrayal and hard choices,” says Alex Schwartz, chief scientist at Owlchemy Labs. “We got thinking about how to manifest these game attributes in a digital game—especially one without zombies, which we felt was done to death.” So the group began talking about what they wanted to see done in the survival game space; Lord of the Flies, along with certain episodes of The Twilight Zone were brought up. “Rod Serling’s views on the human condition were interesting, especially his portrayals of mobs of people quickly devolving to the ‘witch hunt’ mentality,” says Schwartz.


From there, once they realized they were getting moving towards the idea of a branching narrative, the developers decided to focus on the context of group interaction. “Building relationships and having them tested is something great that just isn’t seen in games very often,” says Schwartz. He notes that they kept reminding themselves of their goal to not let the game take itself too seriously; as they’ve tried to do with their past portfolio of games, Owlchemy let its sense of humor shine through.


The initial impetus with Dyscourse was to build a 3D game without the overhead of typical 3D game budgets. Implementing a unique visual style, they have been able to play to their strengths in 2D art and create a visually impressive 3D game quickly and cost effectively. “Unity has helped immensely, in the sense that we're able to operate with far less risk than a typical studio creating 3D games,” says Reimer. The initial semi-2D, semi-3D layout was the result of experimentation within Unity when sitting down on the first day of prototyping for Dyscourse. “We started messing with a hybrid alpha-testing / alpha-blending shader, which would allow simple 2D imagery to appear as cutouts and cast shadows into the world and developed the rest of the style from there,” says Reimer.


Creating a compelling game based around a branching narrative comes with its share of challenges, of course, and the team at Owlchemy Labs has embraced them. “[There’s often] the nagging feeling that the game designer is trying to sway you one way or another in a tough choice, and there’s [also the problem] of labeling certain choices as having a specific allegiance,” says Reimer. One of the key tenets of the team’s design is to never label a specific path as completely right or completely wrong; as with real life, every choice has a ripple effect on the world, potentially strengthening bonds with one person and possibly weakening bonds with others. “There is no Paragon and there is no Renegade,” he says. “Depending on your mood, motives, and past choices that led up to your current predicament, the decisions you make certainly have a lot of dependencies. Showing the gravity of your choices in an easy-to-understand way is one of the challenges of the design, and something we've been working on throughout development.”

Those choices available to the player are also never static. For example, if you’ve learned a specific bit of information about a character, the option to bring that up or use that information against said character will be “unlocked” in the choice tree, whereas players who didn't come across that info due to going down a different branch won’t be able to see that option. So if you and 10 friends play one scenario, it’s likely that you’ll each end up a vastly different course of events. “Your buddy might be telling you how his group crashed and burned by day five, while you managed to get three survivors safely to rescue by day 10,” says Reimer. “Your choices directly determine who does or does not make it out alive in the end, which is the overall mantra we’ve been following throughout development.”


“Early on we made sure that we would be able to create content for the game extremely quickly and at a very high quality without needing complicated and expensive 3D rigged models,” says Schwartz. “That means a simple tree asset painted by Carrie in Photoshop can be brought into the engine, set up, animated, and placed in the world quickly without having to pass off art content between an illustrator, a technical artist, an animator, and finally a level designer. One person can do it all. The asset can then cast a complex shadow and, using the built-in animation system, we can ensure each item in the world feels alive, with subtle hinge-based animations to leaves and branches, creating a dynamic feel instead of a drab, static world.” Each and every item in the world is affected by the lighting system, changing the length and angle of shadows as well as creating that distinct feel of early sunlight or deep red sunset.


The art style for the game is based off of lead artist Carrie Witt’s own personal 2D art style, with some tweaks and changes to make it work within the semi-2D, semi-3D world. Environment art is made to look 3D in place of actual 3D assets, while the characters are more simplified and flat in order to animate in a hinge-like fashion. “Carrie’s love of interesting shapes and colors is exemplified in the style, and it’s been fun to push and explore those angles for the art in this game,” says Schwartz, who points to Witt’s inspiration from artists like Mary Blair, Bruce Tim, Mike Mignola, Leyendecker and Matt Lyon. “Her work is definitely unique and charming in a way that’s extremely hard to communicate in most 3D games.”

Schwartz says that the art style took roughly six months to get to a point where it felt really solid, at which point all of the art was re-done. “We’re using deferred rendering to achieve wonderful effects like radial shadows casting from a point-lit object like the campfire or a torch,” says Schwartz. “Unity gives us the power to do this right out of the box.”


Indeed, Owlchemy has been using Unity since 2009, and has built up a library of reusable development bits that it keeps handy. “We can create games super fast with this toolset, it only makes sense to continue using Unity,” says Reimer. “I honestly think that this game would never get finished—or we’d run out of money before getting to the point of getting to a playable demo, let alone shipping the game—if we had to use a different engine that didn’t place a focus on quick iteration while developing.”

“Aside from Unity making it dead-simple to create an awesome and unique visual style, the ease of editor scripting within the engine has been a huge boon to our workflow through the development of Dyscourse,” Reimer continues. “We were able to reuse editor tools from previous projects, and were able to quickly create tools that helped with designer-driven workflows and content creation.” The game’s initial art test was up and running within a day, and the team had locked in a general style within a week.

Still, even with efficient software, creating something unique takes work: “The systems for branching narrative and our dialog-specific workflow were very, very time-consuming, and while we had a first workable solution for narrative within a few months, it took over a year of iteration to come up with a workflow that allowed for content to be created that’s being used in our current playable,” says Reimer. Aside from the art import pipeline, scripting engine, scene creation tools, and other basic Unity functions, Owlchemy Labs is utilizing the built-in NavMesh solution within Unity to a great extent. It’s also making use of Unity’s new deferred rendering pipeline, shadow system, and custom shaders to achieve Dyscourse’s distinct look.


“Some of the toughest moments in the game’s design, as with many of our prior designs, has been the moments where you feel as if you’re walking through the dark in an unexplored area of design, questioning your motives, your direction, and the feasibility of seeing the light at the end of the tunnel,” says Schwartz. “There are so many decisions that go into creating a cohesive design, and sometimes it can be overwhelming to step back and think about what’s best for the overall game. I’m proud of our team’s ability to consistently tackle these big decisions, get into long discussions about the validity of one path over another, and come out stronger and more cohesive than when we went into the discussion. It’s hard to find a team that knows each other so well that we can consistently come up against such difficult challenges and come out the other side net positive.”

Dyscourse is slated for release in 2015.

Unity Asset Store

“On many of our projects, we use a myriad of third-party tech solutions from the Unity Asset Store to aid in development,” says Devin Reimer, Owlchemy Labs’ Chief Technical Owl (CTO). “With Dyscourse specifically, we’re using 2D Toolkit, InControl, MultiPlatform Toolkit, Ludosity's Steamworks Wrapper and likely tons of Prime31 integrations down the line.”


“We wanted to be as realistic as possible with ourselves and our community,” says Devin Reimer, Owlchemy Labs’ Chief Technical Owl (CTO), of his studio’s recent successful Kickstarter campaign to fund the development of Dyscourse. “We knew that Dyscourse would end up costing 3-4x as much as we’d likely be able to raise from the community, but we went forward with the Kickstarter anyway, asking for a conservative amount that would assist with development but certainly not fund the entire thing.” This had the rather wonderful and unexpected side effect of allowing for the community to share in the team’s excitement for the game before it was completed. “We analyzed the current market and made some guesses as to what target we could likely hit, settled on 40k and hoped for the best,” says Reimer. “Kickstarter was a completely new experience for us and in the end, after an absurd amount of hours and painstaking effort, we came out alive and with a successfully funded Kickstarter. We’re still trying to figure out how many gray hairs that Kickstarter campaign ended up causing!”

Course Corrected

“Initially, our concept for Dyscourse was extremely AI-driven,” says Alex Schwartz, chief scientist at Owlchemy Labs. “We were aiming to create a simulation in which intelligent simulations would work together, or betray/deceive one another in a group situation, while being heavily narratively weighted.” As they started to delve deeper into this concept, they realized they were setting out to build something akin to “The Sims x10”. The team wasn’t quite sure it was onto something special until Devin Reimer, the studio’s chief technology owl, commented on the fact that, “if he were able to pull off the system he was currently implementing, he’d likely receive some kind of award in the AI field.”

Read more about Dyscourse and Owlchemy Labs

« Back to overview