My Role:

Game Designer, Level Designer

On PatchWord, I owned the game's design from initial concept through prototyping, including writing the MVP spec, iterating on the core loop mid-development, and driving the FTUE design iterations that took the game from confusing and frustrating to where players quickly grasped its unorthodox mechanics.

PatchWord went through more meaningful design evolution than most prototypes I've worked on. The version that made it to user testing looked quite different from what I first specced out — and that gap is where most of the interesting work happened.

My key contributions across the project were:

• Write the initial MVP spec and establish the first playable build based on a Sudoku and Woodoku-inspired concept, then rapidly redesign the core loop after the first internal playtest revealed that, while there was something interesting in the concept, the Sudoku rules and structure were limiting the fun.

• Redesign the board system from uniform 3x3 grids to color-coded, uniquely shaped sections. This was a change that turned every level into a visually and strategically distinct experience and helped give the game its identity.

• Define the level goal and win condition, shifting from a board-clearing objective to collecting tiles of a specific color, which gave the game a cleaner, more satisfying loop and a natural lean into match-3 hybrid territory.

• Diagnose and solve a significant FTUE problem, using iterative user testing to identify exactly where players were losing the thread — then redesigning the tutorial flow, UX, and even the game's vocabulary to fix it.

Project Process & Breakdown

Early look and feel:

PatchWord began as a fusion of the spatial pressure of Woodoku and the sectioned-grid logic of Sudoku. The initial concept put players on a board of random letters divided into 3x3 segments. Swiping letters to spell words would highlight the tiles used, and clearing an entire row, column, or segment would turn the letters into solid tiles. The goal was to turn every letter on the board into a solid tile to clear the level.

While interesting, this first concept introduced some very tricky challenges to solve, and cases that made levels nearly impossible to beat as you neared the end. The Woodoku aspect was quickly scrapped, and instead of turning letters into blocks, they were cleared from the board and replaced by new letter tiles. After we played the first prototype build, we knew there was something interesting here, but we wanted to make several changes for the next iteration.

Finding the real game:

The Sudoku structure that felt clever in the spec felt rigid and predictable in play. Every board looked essentially the same, and the clearing conditions didn't create the kind of satisfying moments we were hoping for. Rather than iterate on the surface, I went back to the fundamentals and made a few structural changes that ended up redefining what PatchWord actually was.

• The row and column clearing was dropped entirely, as it was pulling the game toward Sudoku when the more interesting part was the ever-changing game board, as sections were filled in. The uniform 3x3 grid was replaced with irregular, color-coded sections that varied in shape and size from level to level. Suddenly, each level had its own unique layout and strategic approach.

• From there, the level goal evolved naturally: Instead of clearing the whole board, players were given a target: collect tiles of a specific color by spelling words through that section. This small shift gave the game a clear win condition, a more focused objective per level, and an unexpected, but satisfying lean into the match-3 genre. Internally, people kept replaying the same levels, watching the cascades of letters fall down as one or even multiple sections were cleared off at the same time. This was a really strong signal that we were on to something.

Each level felt unique with its bespoke shape and section layout, and requiring players to collect specific colored tiles meant they were not only encouraged, but required to spell words all across the game board. The same colors were often represented in different areas of the level, so if the player had a hard time filling in letters in one area of the board, they could move to another part of the board instead. The shape of levels and the sections within played a much bigger role in this iteration: A level with long, narrow sections often proved to be difficult to fill in, and sometimes the player had to clear other sections until the right letters landed in the section they actually wanted cleared. Levels with smaller sections allowed for quick clearing, but yielded less progress towards the goal, while large sections could clear a goal in one quick swoop, but it took far more effort to clear. This gave players meaningful strategic decisions to make, depending on their needs in the moment.

Challenges:

The most demanding part of this project was making the onboarding and teaching of the game itself to brand-new players. This was due to all the conditions required of the player to make progress in the game: Spell valid words, then with those words fill in a section of the same color, THEN you score. In classic match-3 or word games, your payout comes immediately after you make the right move, but here it was delayed. It required players to often make multiple moves for a payout. While this feels satisfying and has a nice setup into payoff, it meant the game felt far less intuitive than most word or match-3 games.

In the first user test we ran, most players didn't intuitively grasp the rules at all, and were visibly and audibly frustrated as they did not understand what they were meant to do. And in the second user test, we had tutorials in place, but nearly half of the players still didn't fully understand the game loop. What made this particularly tricky was the nature of the confusion: players weren't completely lost. They almost grasped the rules, but often drew the wrong conclusion as to why they won the level. So they would clear a level with ease, but misunderstand exactly what they did to win.

The cause-and-effect chain of; spell a word>highlight tiles>fill a section to collect it needed to be broken into explicit steps and taught one at a time, then reinforced through early levels. I changed the vocabulary in tutorials by referring to each section as “Patches” rather than “Sections”, tying the vocabulary directly to the game's title and giving players a consistent word to anchor the concept.

In addition to this, our tech artist made UX adjustments to make the board's sections read more clearly at a glance and changed the in-game language.

The difference in subsequent tests was significant, as nearly every player understood the rules within the first three levels, with all testers confident by level five. The frustration and hesitation we'd seen early on were replaced with curiosity instead, where players were forming their understanding and eagerly testing it to confirm they understood the rules.

Takeaways:

An engaging game loop is not the same as an understandable one, and on mobile especially, you can't afford to assume players will figure it out if they play long enough. If onboarding is fighting the game's complexity, you'll lose players before they ever get to the part that would have hooked them.

Keeping the game loop clear and intuitive will do a lot of heavy lifting in the onboarding phase of your game, and you can spend more time thinking about how to build on top of that loop rather than spending resources teaching it.