Bart Bonte’s Factory Balls is an old, and I’m sure familiar, flash game, that is now available on browser, iOS and Android. I have no doubt you have played it before actually, I hadn’t even realized I had until I opened the browser and was hit with the sense of familiarity. I say this because I recall playing this game at 8 years old, and today at 21 once I got to puzzle 10 I was stumped. Clearly this game is for a wide range of audiences.
Factory Balls is deceptively simple at first glance, a conveyor belt, an empty ball, and a set of tools. The goal, also deceptively simple, is to replicate a specific pattern on the ball using the available equipment.
Despite its simplistic nature that one would assume they would get tired of quickly, the game’s success lies in how elegantly its mechanics shape the player’s experience: teaching by doing, thinking through doing, and gradually stretching the player’s mental model of how each tool works in combination with others. Its simplistic nature the becomes what makes the game so special. Its excellence lies in its limited but effective set of interactions. Players dip a ball into paint, apply protective gear like hats and belts, and use tools to create pattern with the paint. These simple mechanics become powerful when layered and can truly stretch the mind in unique ways. Each action has consequences, and these consequences are irreversible, an early mistake means restarting the puzzle. This in turn causes players to think ahead, planning not just what to do but in what order.
Compared to other puzzle games like Portal or The Witness, Factory Balls lacks a narrative or immersive world. However, I would argue this is not an inherently bad thing. Its challenge emerges purely from the design of its puzzles. Each level introduces a new complexity, adding goggles, or grass, or even light-reactive elements, requiring players to adjust their strategy and stretch the mind. While Portal leans on spatial logic and physics, and The Witness on observation and environmental clues, Factory Balls focuses entirely on transformation through sequential operations, something which I believe makes it stand out.
Through the lens of the MDA framework, Factory Balls excels at aligning all three. Its mechanics (tools and materials) drive the dynamics (planning, trial-and-error, order sensitivity), which in turn evoke aesthetics of satisfaction, clarity, and curiosity. The minimal visuals and lack of time pressure enhance its meditative rhythm. Players experience flow not through speed, but through precision and logical sequencing. This perfect balance allows the user to enjoy the game in a way that still promotes learning through an incredibly effective flow state, very much situated in the Zone of Proximal Development I would say.
A central strength (or some would say weakeness) of the game is how it scaffolds learning. There are no tutorials, not at the beginning and not when a new tool is introduced. Players must infer functionality via experimentation. This aligns with constructivist theories of learning, where knowledge is built through active problem-solving. However, this strength also reveals a flaw: the game makes quite explicit assumptions about the kind of logic its players possess. Specifically, Factory Balls assumes a comfort with abstract thinking, and a familiarity with procedural reasoning often taught in Western school systems or STEM-related fields. A player unfamiliar with cause-effect modeling or spatial sequencing may find the game opaque.
This leads to an important ethical consideration: While the game appears universal, there are no words, minimal symbols, its puzzles rely on an assumed literacy in puzzle logic, particularly the ability to visualize transformations and reverse steps. This may unintentionally exclude players from educational backgrounds without access to formal logic training or exposure to games with similar mechanics. Additionally, because tools like “spray paint” or “belt” may be culturally specific, the metaphors of masking and layering might not be intuitive to all players.
One way to address this would be to include some kind of hinting system, perhaps optional visual walkthroughs that demonstrate one correct solution after several failed attempts, or tool tips that visually preview a tool’s effect on the ball. These wouldn’t dilute the learning-by-doing model, but would open the game to broader audiences by supporting different learning styles and different contextual backgrounds.
If you were curious which one I got stuck on, it was this one.
