Magicube (available now on iOS!) began as a simple idea and nearly ended up as an unsolvable puzzle. To this day, some of my friends—avid board gamers and people who lock themselves in rooms with logic puzzles for fun—look at it and say, “Oh, I’m really bad at this kind of game.”
Magicube takes a 2D image divided into a grid of square pieces and jumbles them into a new configuration. The player must shift the pieces back into their correct positions, re-forming the original image. The columns and rows “wrap” around the grid: When you slide a row to the left, the piece on the far left slides back into that row from the right. So, you can’t just move one piece, you need to plan a sequence of shifts that gets all the pieces where they need to go. Otherwise, you end up chasing one piece around the grid:
Essentially, it’s a sliding block puzzle…or a 2D puzzle cube…or maybe a 2D puzzle doughnut? I’m very proud of it, and I spent the bulk of development discovering all the ways it would fail as anything but a simple game.
The obvious way it would fail
Magicube relies on the player knowing exactly where a piece is supposed to go. This fundamental constraint drives the game’s design: There is only one destination for a given piece, so if a player can’t immediately understand where a piece is supposed to go, the game breaks.
Once scrambled, complex graphics turn into a noisy, indecipherable mess. But the graphics need enough complexity to make every piece recognizably unique. We ultimately landed on a solution that uses simple, geometric patterns with bold color palettes that nod to the game’s 3D counterpart, the Rubik’s Cube.
Failing by being too hard
Despite the deceptively simple premise, I almost abandoned development several times because the game itself was too hard for me to solve. After years of (admittedly sporadic) work on the game, I still had barely managed to solve the 4x4 and 5x5 grids. For such a straightforward concept, solving the puzzle was seemingly impossible.
At this point, I wasn’t sure who the audience of this game would be. I knew it was a great puzzle, but if it was too hard for me and my hardcore puzzle friends, who would want to buy it?
That’s when I ran into a problem I didn’t expect.
Failing by being too easy?
Perspective changes everything. As my wife—the art & marketing side of Stonegate Games—evolved my prototype graphics into their final designs, she started playing the game to test the interaction of the pieces. She looked at the grid differently and formed a strategy that was simple, straightforward, and surprisingly easy. While I took nearly 10 minutes to solve the 4x4 grid, she could do it in under a minute. Suddenly, my niche, impenetrable game was not just winnable, it was solved.
Failing by being too hard (again)
Once the base game was solved, I tried to find ways to make it harder. What if one row flowed into the next row like a snake or conveyor belt? Could it go a different direction? Could a row shift into a column?
These ideas transformed a visually straightforward puzzle into sheer chaos. Messing with the geometry of rows and columns turned the grid into a madness of shifting pieces. I had managed to increase the difficulty by making the mental model arbitrarily difficult to project and plan. If a player can’t easily see and understand how their moves affect the pieces, they will never be able to solve the puzzle.
A simple challenge
I still wasn’t happy that the original game could be solved so easily, but drastically changing how the pieces shifted increased the difficulty in ways that weren’t fun. So I added a mode that “froze” certain pieces in place.
Not only did this provide strategic differences from the original game, the core mechanics of the game remained recognizable.
Solved is not failed
I realized that having an algorithm to solve a Rubik’s Cube hasn’t negatively affected its popularity. The Rubik’s Cube community shows there’s a compelling difference between being able to do something and being the best at it.
Magicube is a game of two parts: first discovery, then mastery. On launch, it comes with the “Classic” mode and a “Puzzle” mode with locked pieces. Both are simple, and that’s ok; they are the best versions of the game.
Achievements set milestones for players on their path to mastery, and daily leaderboards let players compete against each other. Every day brings a new universal starting configuration across all devices, so the daily leaderboards are synced against each other.
Released at last
Magicube is now available for all iOS devices. Please try it out!
You can find me on Twitter @StonegateGames