The Ultimate Prepared Environment

When I was nine, my parents gave my sister and me a Nintendo for Christmas. They were late to the party. By that time, we had already logged dozens (hundreds?) of game hours at friends’ houses, so we immediately booted up the console to play the game we had spent most of those hours on: Super Mario Bros.

Revisiting Super Mario Bros. thirty years later as a game industry veteran and student of Montessori, I find it impressive that most people over the age of five can learn the basics of Super Mario Bros. simply by picking up a controller and playing it. This is true even though the game doesn’t rely on traditional instructional techniques or even those used in other games: you don’t need direct instruction or coaching; you don’t have to read the manual; and the game doesn’t include any embedded tutorials.

How does SMB accomplish this? In game-making terms, it’s through the skillfulness of its level design: the way its virtual environments are laid out.

Many game levels aren’t just about creating a fun, challenging experience for players: they’re also prepared learning environments just like Montessori classrooms. While a Montessori classroom is designed so that children develop the capacities to succeed in life, an effective early game level is designed so that players develop the capacities to succeed at the game. Seen through this lens, the overlap between great game design¹ and Montessori pedagogy starts to emerge.

Super Mario Bros.’ first level, Level 1-1, is frequently cited as a prime example of great introductory level design. You’ll find hundreds of YouTube videos like this one analyzing the design choices. If you’ve never played Super Mario Bros., it’s worth checking out that example to get a lay of the land and to see how the myriad concrete design elements add up to such an effective learning environment. How exactly does the design of Level 1-1 do this? I think there are three key principles at play, and all of them have parallels in Montessori pedagogy:

Inviting engagement: Games are an interactive form of entertainment – one where the player must do something to make the experience unfold. Good game design invites the player to engage with and discover the game’s world. In SMB, blocks with blinking question marks beg the player to interact with them. The vast empty space to Mario’s right stimulates the player to move in that direction and explore (leading her to discover a fundamental principle of how the game works: go right!).

Montessori recognized that true learning is an active process. To build effective knowledge, the child must engage willingly with the world around them. Learning can be facilitated by creating the right conditions for this engagement, but can’t be forced. This is why Montessori classrooms and materials are carefully designed to be enticing: the beauty of the materials and the manner by which they are arranged invite the child to interact, to touch, to explore.

Isolation of difficulty: Succeeding in a video game like Super Mario Bros. requires the player to perform complex, fluid sequences of movements. But first, the player must understand and be able to perform the individual component skills that make up those sequences.

Level 1-1 presents component skills as well as any game I know, establishing what has turned out to be the most effective pattern for doing so: introducing one skill at a time in isolation. By my reckoning, the following crucial, difficult-to-master skills are all introduced, one at a time, within the first two screens of SMB: lateral movement, precisely stopping lateral movement, jumping in place, timing a jump to avoid or stomp on an enemy, hitting blocks from underneath, controlling the height of a jump, and jumping while moving laterally.

We see this same design principle at work in the Montessori notion of isolation of difficulty. Consider the Montessori approach to introducing scissors. Cutting out an irregular shape requires the coordination of several individual skills: holding the scissors properly, performing the snipping motion, stringing multiple snips together to make a continuous cut, moving the scissors along a desired path, etc. Each step is difficult for a child who has never done it before. In the Montessori approach, each is introduced in isolation with an appropriate presentation and activity, for example cutting strips of paper in a single snip without simultaneously having to follow a desired path.

The purpose of isolating difficulty – both in video game design and in Montessori classrooms – is so that different parts of a complex skill can be introduced in a specific order, all culminating in the child eventually mastering the full skill and becoming independent. A key to successfully designing this sequencing, in any context, is ensuring that the learner has attained prerequisite and component skills before asking them to tackle challenges that depend on those earlier skills.²

Control of error: Matt wrote about control of error in a Montessori context a few weeks ago, and the concept is just as important to video games. Game designers tend to think in terms of providing the player with feedback, both positive (e.g. points or other rewards) and negative (e.g. dying and having to restart a level, or losing a power-up). This feedback allows and encourages the player to adjust their perceptions and then, critically, their behavior.

The techniques at work are the same as the ones you’d see in a Montessori classroom. When a SMB player first encounters a goomba, they might assume that they can ‘collect’ it like a coin or power-up. But if they try, it kills them, and the level must be restarted: although this only sets them back a short distance. Think of the child who tries stacking the big pink tower blocks on top of the smaller ones, or the student who counts his spindles incorrectly: the environment immediately provides direct feedback that prompts the child to rethink their approach. Allowing players and children to experience controlled failures with fair consequences encourages them to explore, experiment, and self-correct without fear of suffering a catastrophe.

I’m fascinated by cases like this where game designers and Montessorians are ‘sniffing around’ similar goals and principles, but where the unique attributes of digital games have the potential to amplify or transform those principles beyond the constraints imposed by physical environments (despite their many virtues). Even more fascinating to me are cases where Montessori learning principles seem to be at odds with those of great game design. For instance, Montessori stresses the role of the prepared adult and presentations in connecting the child with materials. Yet we praise games like Super Mario Bros. for being effectively self-presenting and criticize games when they over-rely on presentational techniques like tutorials.

Ultimately, I think apparent conflicts like this are mostly false or superficial. There’s more to learn from the deeper, more fundamental similarities. For example, consider what Shigeru Miyamoto, the great designer behind Super Mario Bros., Donkey Kong and many other classic Nintendo titles, has to say about his approach to a learning-focused level in his own commentary on Level 1-1 (emphasis mine):

“We wanted the player to gradually and naturally understand what they’re doing. [Level 1-1] was designed for that purpose: so they can learn what the game is all about. But then, after that, we want them to play more freely… Once the player realizes what they need to do, it becomes their game.”

Consider how Miyamoto’s quote parallels this sentiment from Maria Montessori in The Absorbent Mind:

“Thus man develops gradually and by means of these successive steps of independence, he becomes free. It is not a question of will, it is, a phenomenon of independence. Really, it is nature that is giving to the child the opportunity of growing, gives him independence and at the same time leads him to freedom.”

I love what these quotes capture: the idea that as a learner naturally progresses, he sees both his range of freedom to act with efficacy and his ownership over his domain of action increase. This is the essence of agency, and I suspect Montessori and Miyamoto are so aligned in their approaches because they’ve ‘tapped into’ the same fundamental truth: that independence is both a goal and requirement of learning and of life.

1. A lot of discourse is devoted to games that are explicitly designed to be educational, or entertainment games that are designed to encourage exploration and creativity (for example “open world” games like Minecraft or games with embedded creation tools like Roblox). But I think it’s instructive to look at a game that, in spite of its linear structure and (by today’s standards) limited ‘palette,’ manages to be so good at teaching its players how to play.

2. SMB deserves praise for its approach to ensuring the player has mastered foundational skills before taking on more complex challenges. Level 1-1 makes heavy use of gating where the player literally can’t progress to complex challenges without first having mastered the prerequisites. For example, the first enemy encounter is designed so that the player has to successfully time a jump to avoid or defeat the goomba enemy. Only after he succeeds will he be able to move past the enemy and on to challenges that require him to combine timed jumping with other skills. One of the implications of this – and one that I think is incredibly powerful when it comes to the unique potential of games for learning – is that well-designed games are inherently self-assessing.