Design Bits Analysis: Limbo's Level Design - Mostly Perfect

Prologue: A Flawed Masterpiece

Limbo helped break down the door for indie developers back in 2010. It wasn't the first puzzle platformer with 'atmosphere', but it was the first with capital 'A' Atmosphere. It spawned a slew of 'Limbo-likes' after it, and it is no wonder. Imitation is the sincerest form of flattery, and Limbo tackles its genre's inherent challenges like a master.

But, even the master makes mistakes. So, as much as I will laud Limbo’s accomplishments - and I will - I will also point out its flaws. Understanding the depth of Limbo's successes and failures requires some game design knowledge, SO, I will break it down step by step hopefully without boring the boogers out of you seasoned veterans.

Chapter 1: Affordances - How a Game Talks to a Player

Affordances implicitly explain rules by having something sound, look, or act a certain way. The more intuitive the affordance is, the easier it is for the player to pick up the implicit rule. For example: ‘handles’ in Limbo. Handles can be grasped and pulled. Or, put another way, handles afford grasping and pulling. So, they are a perfect affordance for “pullables” like carts. Once the player has found one pullable thing with a handle, the player can intuit other things with handles as pullable. In this way, a game not only teaches rules - it teaches an audio/visual language that makes its rules intuitive. The more a game teaches through affordances instead of text, the more heavy-lifting your affordances have to do. Limbo tells you nothing up-front. It relies solely on audio/visual storytelling to communicate its rules. Everything must be intuited. It is a risky choice, and, most of the time, it works brilliantly.

Chapter 2: A Puzzle is a Tutorial

An affordance alone does not make a rule intuitive to learn. A rule’s tutorial plays a big part, and for this, Limbo follows some classic rules of game design: (1) Make it easy to try again after they fail to learn, and (2) Remove distractions from teaching the new rule. Or, shortened up: (1) Safe to fail, (2) No distractions. This is good, but what I love about Limbo is how it teaches more with less. In a word: elegance. In the player’s first encounter with bear traps, they learn (1) bear traps kill, (2) how to move bear traps without them killing you, and (3) to do a running start before they jump to get over bigger obstacles. My favorite example is the machine gun tutorial. The player naturally jumps into the machine gun’s line of sight and pulls themselves up before the machine gun fires. First time players likely won’t notice the machine gun until it fires, but by that point, they will be safe up on the ledge. It’s intuitive! It’s efficient! It’s Limbo! *chef’s kiss*.

However, Limbo isn’t perfect.

Chapter 3: When Communication Breaks Down 

For the elevator panel tutorial, I spent a good few minutes trying random things until I realized I needed to stand at a specific point by the panels to interact with them. The only thing to suggest when you can interact with the panel is when the nameless boy’s hand goes up, but that will only happen if you are already standing in the right spot for half a second. The panels lack appropriate affordances, and I think Playdead (the studio that made Limbo) kept this in mind when developing their next game, Inside. All of Inside’s panels have physical buttons and switches that afford interaction, and it is clear when the player can and cannot interact with them. Many designers intentionally fall into the same trap Limbo did. The thinking is: “by making things vague, you make your world more mysterious.” But having vague rules or affordances can be a one way ticket to bad, bad, bad station. They may initially invoke mystery, but vague rules are more likely to invoke arbitrary frustration - especially in a puzzle game where the fun comes from understanding rules and discovering their consequences. After all, trying random things is not a real challenge. The player is not making any meaningful decisions, and hunting for a solution without clues is not very fun at all. It is an arbitrary “pixel hunt” masquerading behind 'being mysterious'. A good puzzle requires the player to explore what they know in order to discover an unknown and surprising solution - as is often the case with Limbo. Unless there is no other way to achieve a game’s goals, run away from vague rules. They cripple trust between the player and the game's audo-visual language and, in doing so, make it harder to build a common language of affordances.

Chapter 4: Death in the Puzzle - a Tradeoff

Usually, Limbo is great with pacing. It knows when to throw a low reflex challenge after a high mental challenge or vice versa. It knows when to mix it up with a crazy set-piece (i.e. spider, “girl by the grave” mid-game, when the world starts spinning, etc.). I want to focus on pacing problems that come out of poor puzzle design.

Example 1: The player solves the puzzle in their head. Now, the player needs to go through all the puzzle’s steps. Going through all those steps might take a little bit, but the player does not mind because they are still riding an “epiphany-high” from solving the puzzle in their head. They are riding that "I got it!" feeling. Now, imagine they are about to finish the puzzle's steps when *WHAM*, they die because they failed a reflex challenge - nothing to do with solving th puzzle. The player respawns and the puzzle is reset. Bored, frustrated, and kicked off their “epiphany-high”, the player repeats going step-by-step through a puzzle they have already solved. This is a worst-case scenario you risk having when you add costly reflex challenges into puzzles. Alas, even Limbo falls into this trap.

At this point in this puzzle, I had solved it after a few minutes and was just finishing the last steps when *PLONK* - the box that gets thrown in the air fell on my head. I respawned and the puzzle was the reset. Finishing all the puzzle’s steps just to make sure I didn’t mess up a reflex challenge at the end felt bad - not in a “well, I’ll do better next-time” way but in an arbitrary way. In a reflex-based game, sudden arbitrary death can happen too, but at least the reflex challenges are still challenging after I respawn. Once a player solves a puzzle in their head, the mental challenge is gone. Unless you forget the solution, redoing a puzzle has no mental challenge, and it kills pacing. Any challenge left is reflex-based and, admittedly, Limbo probably leans on reflex a little too much for its longer puzzles. In an effort to make puzzles more interesting if the player dies, you can add more reflex challenge, but, the more you add, the harder it will be for a player to explore a puzzle’s components without dying and resetting the puzzle. This is why games with bigger, longer puzzles tend to avoid combining their puzzles with reflex challenges, or they lessen the cost of failing reflex challenges (i.e. Braid, The Talos Principle, Portal 2, The Witness). Beyond small timing challenges here and there, the pacing is purely determined by a puzzle’s epiphany loop. That being said, having reflex challenges during puzzles can work well, but one should be aware of the tradeoffs. 

Chapter 5: Reflex Not Always Bad Thing

However, when you introduce real-time bosses like the Spider in Limbo or the long-armed Janitor from Little Nightmares, reflex challenges become essential. Limbo’s Spider boss battle is nearly all reflex-based, and this is actually great. It is better for something to be mostly reflex-based and a little mental-based than vice versa (i.e. think about any good Legend of Zelda boss battle or Dark Souls enemy). Once a puzzle is solved, there is no challenge, but a reflex challenge is a challenge every time - no matter how many times you respawn. It might not be as interesting as the first time, but it is still a challenge, and the pacing won’t die. 

Chapter 6: Atmosphere through Cinematography

Limbo’s finely tuned camera work and composition not only achieve its gameplay goals but also its thematic goals. I could go on about the cinematography for HOURS, however, this is a Design 'Bits' blog, not a Design 'Slog' blog so I'll forget about the last couple ending shots, and focus on some bits like: the use of blurred foreground elements to seamlessly transition from camera zooms, frame important scene elements, and make the world feel a little unsettling - like something far away is looking over your shoulder.

Buuuuuut here's one specific shot essential to setting up Limbo's atmosphere. The boat scene is genius classic level design: transition from a small claustrophobic space to a big open space. It condenses the emotional impact of a smooth transition into one moment. Open-world games use transitions like this to invoke wanderlust (i.e. Breath of the Wild, Fallout 3), but that is not how Limbo uses it. Limbo is about surviving a strange, desolate world that is indifferent to your desires. Here, a nameless boy crosses into a dangerous, unforgiving land. And so, the camera zooms out to reveal - emptiness. You and your little ship in a big empty world. Phenomenal. Intentional cinematography like this injects another layer of theme and emotion like a great big dollop of vanilla ice cream on an already-delicious portion of game design cake. It makes me happy.

Chapter 7: In Conclusion

A decade is an eternity in game design, but Limbo's design challenges remain relevant. Affordances, Pacing, Cinematography: essential components of any puzzle platformer (and any video game for that matter). They are easy to screw up. Limbo does screw up, but, 98% of the time, it performs brilliantly.

I could have focused on other design bits related to Game Feel, Pacing Between Areas, or how its careful selection of mechanics helps achieve its themes, but, alas, those bits may be for another blog.

'Til next time