Designing Radically Non-Linear Single Player Levels

Abrey Sir Reviewer Hi everyone, I'm Abrey Sir. I've been doing level design for 10 years.

These days I run a small indie game company called 12 Beast Games where I still do level design.

But I'm better known for my previous work at Wolfire, where we organized the first Humble Bundle and made games like Overgrowth.

I also collaborate with other indies making art in levels. My focus is on a specific type of level, type of level, which is why I'm here to talk to you at GDC.

Today I'm going to share with you my favorite kind of levels to design, which I call radically nonlinear levels.

My talk has three sections. First, I'll define exactly what I mean by radically nonlinear levels, how they're based around the idea of free movement and sandbox play, and activating the space.

Second, I'll describe the basics on how to build them, like layout options and typical dynamics.

And finally, I'll explain some techniques I use to fix common issues that may come up in radically non-linear maps, like the map feeling empty or not having enough structure.

Then I'll wrap it up with a recap and final example.

But before that, I want to mention an idea I think we as level designers probably agree with.

That idea is that all conventional levels feature some openness and some linearity.

In other words, all levels have places you can go and places that are blocked.

These are some of our most fundamental level design tools, paths and walls. Linear levels start with making paths.

Often, even when we talk about nonlinear levels, what's described are levels with branching paths.

These kinds of levels can be great, but an alternative kind of level is based on the opposite idea.

Instead of building paths, you start with an area of movable space and place walls to block the player.

The team behind Hitman calls these types of levels nonlinear sandboxes.

A radically nonlinear level is a sandbox level that is designed to encourage unpredictable player movement and exploration of the space.

All sorts of single player levels fit this paradigm, which I'll give some examples of in a moment, and there's also a lot of stuff in between.

Even highly linear levels have open areas in them, and there's an obvious overlap between this kind of level and most multiplayer arenas.

So I think these techniques apply to any space the player is free to move in.

So I hope some of these ideas can help you in whatever kind of level you might be working on.

Despite that overlap, I think this type of level is distinct and interesting. I didn't always feel that way, though.

Let me tell you about my first commercial level design job. It was for a game called Darkest of Days.

I was hired as an artist, but was immediately put to work on level design.

Outside of a few Excitebike tracks I made when I was eight, I had almost no level design experience.

In fact, this was the first major game project for most of the team.

We had our own custom engine and tools, but they were lacking some basic functions. For example, the map editor had no undo function.

So in that way, it was a lot like this talk. I had to try to do everything perfectly the first try.

On the one hand, it's nice to get paid to learn. On the other hand, it's a lot of pressure to care about your work and not know how to do it.

But don't worry, it turned out okay.

As we were making these maps, our lead engineer Jeff kept mentioning that he thought his test maps were more fun than what we were making.

We were constantly running up against a problem we couldn't reconcile.

We wanted to make a cinematic game like Call of Duty Modern Warfare, but we also wanted large dynamic battles in open spaces.

Trying to solve this problem is how I became interested in the idea of radically nonlinear maps, worked on a lot of games with them, and ultimately how they became my favorite kind of map.

With that out of the way, here's part one. I'll define exactly what I mean by radically nonlinear levels.

When I first started researching this question, I wanted to find a really fundamental example of what I meant that everybody would be familiar with.

And the oldest, clearest example I found was Pac-Man, but we're going to talk about Ms. Pac-Man because it's a better game.

So why Ms. Pac-Man? The map is open enough to offer free-form tactical choices,

like which path to take to avoid ghosts, and large enough for strategic choices if you're going to eat a power pellet or go for the bonus fruit.

The designer of Ms. Pac-Man mentioned that each ghost has a different behavior, and in this way, each ghost is a unique risk factor which the player needs to consider.

And there's some other interesting things about Ms. Pac-Man maps. It's small enough so that the whole map is in play. And I'm going to touch on why this is important later.

All the systems are legible in some sense, complex enough to be interesting, but transparent enough to be predictable. There's no exit.

This is a common thing for this kind of map. Even if a radically nonlinear map has an exit, the player often ignores it until other objectives are met.