Before this class, I thought I understood game design mostly because I’d already made things that “worked” in the loose, artistic sense: people felt something, laughed at the right lines, or told me a moment stuck with them. My mental model was basically: have a strong concept, build the content, add interactivity, polish. Mechanics were either a garnish (nice, optional) or a burden (fine, I’ll do it). I also underestimated how physical game design is. Not “physical” as in cardboard, but as in: the player has a nervous system, limited working memory, and a very finite amount of patience for confusion before it turns into disengagement.
Then I spent a quarter building three very different games—Heartland, Battle of the Bands, and Giant Steps—and the class kind of bullied me (politely, constructively) into upgrading my whole understanding of what “the work” is.
In the class, I did these things that were particularly helpful
1) I learned to treat playtesting as the core creative act, not a chore.
Watching people play your game is brutal in the best way. It’s the fastest way to find the difference between what you meant and what you made. The class forced me into a rhythm where I had to observe confusion, revise, and test again. That loop made everything else sharper.
In Giant Steps v4, this was especially intense because I was trying to deliver a bodily feeling: the experience of “thinking while you’re playing music” When playtesters replayed the solo, it told me the system had become legible enough that a player could form hypotheses, test them, and care about the result.
In Battle of the Bands, playtesting did something different: it exposed cognitive load as the real boss fight. Early tests weren’t failing because the math was wrong—they were failing because the learning curve was wrong. That made me start treating onboarding, scaffolding, and visual translation as first-class mechanics.
2) I started thinking in systems instead of scenes.
This was a huge shift. I’ve always loved story and tone, so my default is to build games as sequences of moments. The class kept pushing us to map the underlying system: objects, attributes, relationships, loops, arcs, feedback. Once I started doing that, my games stopped feeling like collections of parts and started feeling like machines with intent.
You can see this across the three projects:
-
Heartland taught me how a system can create social dynamics—player-to-player interaction, tension, collaboration—without needing narrative exposition. Your teammate roleplaying an almond farmer watching their trees burn from a wildfire is a system-generated story that lands because the mechanics support it.
-
Battle of the Bands forced me into explicit system design: drafting -> shop -> battle -> scoring, plus the deeper “why” of the multiplication formula (Technique × Style × Audience) and the anti-snowballing decisions (snake draft, rotating city styles, catch-up rewards).
-
Giant Steps made system design emotional: anxiety as a number, risk as probability, breathing as regulation, and thresholds that turn stress into consequences.
3) I learned how much “onboarding” is actually design, not instructions.
This was probably the most practical and transferable lesson. In Heartland, one quote from the process writeup stuck with me because it’s painfully true: if players are busy thinking about the rules and where they can play stuff, they won’t think about strategy. Our later iterations leaned hard into cognitive offloading—making the physical layout and rulebook intuitive enough that players could stop wrestling the interface and start engaging the actual game.
What surprised me in Heartland was that onboarding wasn’t just about making things easier; it was about making the learning feel good. The pre-test / post-test dynamic (players “claiming” they didn’t learn much, but casually recognizing cards like “Bill of Lading” and “Red Flag Warning” later) made me realize that stealth learning works when the game is genuinely fun and the interface is frictionless enough that players have spare attention for meaning.
In Battle of the Bands, onboarding became the main design fight. The project log basically documents us learning—over and over—that giving players more explanation can increase confusion if it’s not layered correctly. Flowcharts helped because they built a mental model of sequence. A full system map sometimes hurt because it asked players to digest logic before they had felt the loop in their hands. The final direction—visual scaffolding distributed across components—feels like the “grown-up” version of onboarding: not “read the manual,” but “the game teaches you while you play.”
I learned this
Mechanics are expressive. Numbers tell stories.
I don’t mean that in a poetic way; I mean it literally. Point values, probability curves, thresholds, resource economies, turn order—these things write the emotional experience.
-
In Giant Steps v4, tuning the solo system taught me that balance is meaning. If breathing erases anxiety too quickly, the nervous system stops feeling real. If risky play is too punishing, it becomes a coin-flip misery machine. If safe play is too strong, the tension disappears and the theme gets muted.
-
In Battle of the Bands, the multiplication model makes a philosophical claim: success is contextual and multidimensional. You can’t max one stat and coast. The system pushes players to build a band the way real bands function—balancing musicianship, fit, and audience reaction—because the math makes single-minded min-maxing collapse.
-
In Heartland, the mechanics shaped both fun and learning: rivalry, collaboration, and roleplay emerged because the system supported it, and educational content stuck because players weren’t drowning in rules.
Iteration is not “polish.” It’s design.
The class trained me out of a trap I love: trying to perfect an early idea instead of letting the work evolve. Across all three projects, the best versions were not more “finished.” They were more honest about what players actually experience. This is clearest in Battle of the Bands, where “kill your darlings” was survival. When a maximalist version created cognitive overload, the solution was subtraction and translation: simplify phases, restructure economy, make the core loop obvious, let strategy emerge from the base system instead of from piles of add-ons.
A thing I’m taking with me: players often start confused, and that’s okay if the game helps them convert confusion into agency quickly. The goal is a fast ramp into “ohhh, I get it,” where the player feels smarter, not punished. That principle connects directly from Heartland’s onboarding wins, to Battle of the Bands’ scaffolding lessons, to Giant Steps’ mirror tutorial approach (feedback without an over-explained lecture).
When I go to make games in the future, I will…
-
Prototype the core loop earlier and test it sooner—even in narrative-heavy work. If the loop doesn’t create the feeling, no amount of writing will rescue it.
-
Design onboarding as part of the experience, using visual scaffolding and in-world teaching whenever possible. I want learning to feel like play, not like paperwork.
-
Use state intentionally to build memory and consequence. Giant Steps got dramatically more alive once the world could remember what you’d done and reflect it back.
-
Treat tuning and balance as expressive tools, not finishing touches. The numbers are the story’s physics.
-
Keep building systems that make empathy happen through interaction. I care about players feeling something specific—stress, tenderness, competitiveness, solidarity—and this class gave me a much clearer path to constructing that feeling instead of just hoping it arrives.
