Playtonic Games’ Yooka-Laylee has broken $2.5 million with 29 days left on Kickstarter, making it far and away one of the most successful campaigns to date. With a development team made up of the masters of the 3D platformer genre in its heyday, this does not come as too much of a surprise. The good folk at Gamasutra reached out to Playtonic for their sage advice on making a 3D collection-themed platformer like Yooka-Laylee or its predecessor Banjo-Kazooie, and they got back six bits of wisdom from all over the field, from the character artist to the technical director and beyond.
Character artist Steve Mayles started by saying that appeal rules in character design. Appeal comes from things like a recognizable shape, expressiveness, standout features—in Yooka’s case he notes the colorful crest—and can be “something of an X-factor.” He states that if a player cannot make a connection to a character after one look, the character is weak. He also notes that a lot of a character’s character comes from out-of-game, such as posing in promotional art or box art.
Chris Sutherland, Yooka-Laylee‘s project director, says that before a character should look good, they should feel good. Control of the character is the most important part of a game’s design, as a strong game feels good even if you are just moving around the map. He says he starts out with a box, and once navigating an area with no animations or flavor feels right, that is when you can add all the color and aesthetic touches. He also notes that most characters are controlled with “a few digital buttons and maybe an analog stick, that’s actually a tiny bit of information, so you need to extract everything you can from those inputs and use them to bring the character to life in a responsive and meaningful way.” He offers an example of having a character lean in the direction of the physical controller input if the direction of movement and input direction do not match up exactly for whatever reason (cameras complicate these sorts of things).
The creative lead, Gavin Price, says that world design must be surprising. “Everything the game does,” from bosses to dialog to tiny details, should surprise the player and be memorable in some way or another.
Steven Hurst, the environment artist, adds that the world itself should provoke exploration without being convoluted. A well constructed world encourages players to “to explore every nook and cranny trying to find those subtle clues to those secret areas that you just know are there somewhere.” It should also however, make it difficult to get lost, which can sometimes be as easy as putting a sign in the right place or having plenty of easily recognizable landmarks.
Technical artist Mark Stevenson notes that the collectibles that fill this world should be “smart.” Collectibles are versatile, and aside from being a McGuffin, can also be used for anything from guiding the player (by literally placing a trail of them down), to encouraging exploration (by placing them somewhere they might not have thought to go), to challenging them (by placing them in hard-to-reach places). Good collectibles should also be easy to recognize, even from a distance against the background through any number of visual obstacles—both physical or mechanical (like depth-of-field blurring distant objects).
Finally, Jens Restemeier, the technical director, cautions against trying to build everything from the ground up. Efficiency is king in technical design, and “Off-the-shelf engines and middleware are worth it.” Rather than trying to reinvent the wheel with a custom, in-house engine, don’t be afraid to build on what decades of others have built before you. Back in the day this was not so much an option, and maintaining an environment to build the game in was as difficult if not more so than making the game itself. Polishing the game and improving workflow can take a front seat when you don’t need to have engineers working on the tools for it.
It is all great advice—for any genre of game, really. Strong character and world design is not exclusive to 3D platformers, and through collectibles in the literal sense may be, there are parallels in almost every game (treasure chests in adventures, ammo or weapon caches in shooters…). I can personally attest to Restemeier’s advice. Unless you have a team of master engineers and several years on your hands, it is going to be difficult to make an engine worth the effort with all the pre-existing options out there. You can see the exact quotes from the members at the source.