Republique: What Happens When Almost Everything Goes Wrong?
Posted on Tuesday, March 25 @ 13:00:00 Eastern by blake_peterson
One of my favorite games released last year was Camouflaj's Republique: Episode I: Exordium for iOS, a game that merged tropes of survival horror and tactical stealth to generate what creative lead Ryan Payton calls a "stealth, survival action game." In the game you play a hacker using a cell phone to help a young girl named Hope escape a fascist society. The game, which was funded by Kickstarter and featured an A-list voice cast with the likes of David Hayter and Jennifer Hale, had a strong media push and positive Kickstarter updates throughout development.
So it was fascinating to hear Payton and fellow Camouflaj developer Greg Raab's talk at GDC, titled "500 Days of Failure: Designing Republique's 'One Touch' Gameplay," since from their positive media push I never knew they were in any trouble. Payton began almost immediately with a brief exposé on how for 500 of the 600 days of development the primary mechanic they had been trying to sell the game on just didn't work. How the game failed to come together, and the solutions they discovered, is a great example of creative problem solving in game development.
A big part of Republique's one-touch design philosophy was born out of frustration. He would take the triple-A games he had worked on home (MGS4 and Halo 4) and wanted to share them with his friends and family. However, it was very difficult for them, as non-gamers, to navigate a seventeen-button controller. He wanted to provide the triple-A experience on a ubiquitous format and picked iOS due to its high degree of saturation and uniform distribution across multiple cultures. He used Chair's Infinity Blade as an example of what could be done with the format.
In the game you play as an as-yet-unnamed individual who moves from security cam to security cam using Hope's cellphone to propel you from one hacking objective to another. You help Hope by hacking doors, computers, phones, etc, and giving her directions on where and when to move. However, Hope is designed to be her own character, with her own decisions on how she moves.
In the prototyping phase against pre-rendered backgrounds, they were able to get a look that they wanted, which turned out similar to Resident Evil in terms of style, though this quickly fell into a stealth gameplay mechanic that was centered heavily about avoiding detection. After the Kickstarter's success, the complications were almost immediate. The pre-rendered backgrounds would be too costly in terms of space and resources for high-resolution devices they had promised them for, like the iPad, prompting them to adopt 3D architecture in Unity. Still, there were successes using single touches to move the player's view from one security camera to another or automatically jump to the next camera when Hope left the area. There were other issues, though.
Combat didn't work. Payton wanted to create a timing mini-game where Hope would have to wait for the guard, then attack when the player directed at just the right moment. However, this was at odds with another aspect of their design philosophy, that the player doesn't have direct control over Hope, just elements of her environment and giving her prompts on where to move. Hope, as an AI, picks the path that is the least dangerous to get there.
Since Hope was supposed to be an autonomous character, she was designed to avoid detection by guards automatically. Here Payton threw up some images of the Influence Map and Waypoints vs. NavMesh for Hope's AI navigation in a room; the green blocked areas are safe, yellow is where Hope will be spotted if she stands up, and red is where she will definitely be spotted by a guard (Prizrak). In a short video, these areas dynamically changed with Hope hugging the shifting green areas behind cover as a guard moved on the opposite side of the cover surrounded by red blocks. Since Hope was designed to avoid detection, having her stay in one place when the guards attacked was directly against the basic character programming.
One fix for this that they tried was having the player click on Hope when they wanted her to stop for combat. However, this created new problems. With Hope being clickable, if there was something behind her they wanted her to interact with, it was negated by the hitbox they had set up for the player's finger: Instead of clicking behind Hope, they clicked on her.
But these problems were simple compared to the most basic issue that the one-touch gameplay didn't work. Some of this was due to elements they didn't expect; particularly a small subset of two-player types, what Payton and Raab called "Diablo Clickers" and "Lazy Tappers." The Diablo Clickers would spam-click the screen in order to try to get Hope to move faster, and Lazy Tappers would tap and then slightly drag their finger across the screen—dragging a finger across the screen, deliberately, moves Republique's camera—so for a small set of players they were constantly moving the camera when they simply meant to click.
FOLLOW ME ON TWITTER. YOU KNOW YOU WANT TO.