The process of bringing a spark of inspiration into a successful game feature is often fraught with danger, not to mention plenty of terrible ideas. Speaking in panel today at the 2013 Game Developers Conference, Maciej Szczesnik, lead gameplay designer at CD Projekt Red, outlined a comprehensive eight-step process for dreaming up and developing game features.
Peppered with stories from the development of The Witcher series, Szczesnik gave the impression of both a tight ship and a highly collaborative environment at CD Projekt Red. Starting with an examination of imagination and the importance of having constraints in the creative process, he immediately went into the importance of asking questions and finding inspiration in new and personal experiences.
Calling the first step the "sheriff’s star" method, Szczesnik and team ask as many questions as possible when considering a new feature. They consider the why’s and how’s, addressing "why add the feature, who will enjoy it most, when will the player use it, how will it make the game fun, where will it be most spectacular?"
"The more questions you ask, the more precise your design will be at the end."
The next step is proper research. "This is what I call obvious research," he said, pointing to a slide with pictures of familiar books, movies and other games. "We’re all game developers," he said, with similar roots in geek culture. Leaning too heavily on these, he argues, makes for boring uniformity.
Instead, Szczesnik advocated for finding inspiration in unique experiences. "Look where no one else is looking," he said, pointing towards travel and even old family photos as great sources for rich, interesting ideas.
He also advocated going out and seeing things yourself, instead of experiencing things second-hand. For The Witcher series, attending battle reenactments and using medieval swordfighting texts as a point of reference helped to ground the look and feel of the series.
The talk wasn’t just focused on creativity and inspiration. Szczesnik's next steps highlighted the studio’s rigorous process for assessing feature ideas, encouraging buy-in from all team members and mapping out development. The process involves measurable tools such as a voting system for separating the proverbial wheat from the chaff early on, a features matrix to measure the impact of interdependencies among systems, a consequences graph for mapping out unexpected possibilities and a game roadmap that allows the team to decide what order to produce features.
Szczesnik made the case for the studio’s methodology, arguing that without such a system, bad ideas would make it too far into development. One prime example was a proposed "stinker" skill that would have allowed monsters to have a fart attack in the game. It was canned early in preproduction, since it didn’t pass muster in the internal voting system.
He also admitted that the romance system from the first game would have been more thoughtfully implemented if these internal tools had been in place at the time. "Remember the sex cards from the first game?" he asked, recalling a feature that caused controversy in The Witcher. Szczesnik explained that the intention was to breathe life into a deep romance system – but without considering the consequences systematically, what evolved was essentially a sexed up collectable card system.