In the past, I’d design nonlinear games with a certain series of events in mind. I’d try to put myself in the player’s shoes and ponder, “What will they probably do here?” I’d place enemies, traps and items in ways to encourage a player to think and do certain things — but whenever I had the chance to see someone playing my level, they would not do as I had anticipated. It was kind of frustrating, but also a lesson. Short of non-interactive cutscenes or linear level design, you can’t control the way a player is going to play your game. They struggle on parts you thought were easy, will attempt to explore areas you don’t anticipate, and indeed have a will of their own.
In my current project, whose WIP title is Super Team Fortress 2, the world is nonlinear, modeled after games like Super Metroid and Castlevania: SOTN. I’m putting forth real effort to simply design an interesting and engaging world rather than attempting to fit sections of a level to a script. The player will be using 8 different characters to navigate the world, each with their own unique puzzle-solving and navigation tools. It feels not unlike accepting criticism for my written work — it’s kind of a bittersweet feeling, but the right thing to do.
Ultimately, certain checkpoints will require that the player have explored most of the world by the time they reach the end of the game. Other than that, it’s a free range. Of course, I’m not done yet, and am open to having certain parts (like boss battles) linear — I had a lot of fun making linear, gimmicky levels in my last project.
Just looking at this slightly out of date picture gives me a sense of satisfaction, though — it looks like a place you can explore, and that’s what I want.