Inside Klivii: Designing Chomp Circuit, Dot Glide, and Ichi

Why I Started With Three Very Different Games

I built Klivii around a simple question: how quickly can a browser game earn a player's attention without wasting it? Chomp Circuit, Dot Glide, and Ichi answer that question in three different ways. One begins with motion, one with a board you can study, and one with a familiar hand of cards. Developing them side by side made the differences useful. A control decision that feels natural in a maze can feel noisy in a puzzle, while feedback that is charming at a card table can interrupt an action game.

This is a practical design journal, not a claim that any of the three games is finished forever. I still play them on a phone and a desktop, watch for hesitation, and change the places where the interface asks too much. The current versions record those decisions in their rules, artwork, sound, timing, and reset behavior. If you play along, the links below let you see each choice in the game itself rather than taking my word for it.

Chomp Circuit: Making a Busy Maze Readable

The core of Chomp Circuit is intentionally immediate: clear every circuit dot while avoiding the BioMu Crew. The hard part was not inventing that goal. It was making the next useful move visible while the board, enemies, score, power-ups, and player character all compete for attention. I used a restricted neon palette and gave the collectible path a steady visual rhythm. The player should be able to glance ahead and read a route instead of searching the whole screen for the objective.

The five enemy characters share the same threat but need to feel like a crew, not anonymous obstacles. Their presence adds pressure, yet the board still has to explain cause and effect: a collision should never look like a random failure. That is why pause, start, status, and score controls stay close to the play area and why the character art is more expressive than the maze furniture. On smaller screens I repeatedly checked whether the character remained readable under a thumb and whether the control area competed with the maze.

Chomp Circuit: Pressure Without a Long Lecture

A maze game can become difficult in two unhelpful ways: movement feels unreliable, or the player cannot tell why danger changed. I wanted the pressure to come from routing. The opening screen therefore carries only the controls needed to begin, and the game moves into a focused play state rather than placing a manual over the board. The score and high-score loop are there for players who want another run, but the first success condition remains plain: finish the circuit.

Audio also has to serve that clarity. Start, collection, danger, and completion sounds are short signals, not a second soundtrack of instructions. I test with sound muted because every essential state must still be understandable. I also test with sound on because a browser game can feel surprisingly flat when a correct action has no weight. The design target is a response that confirms the action, then gets out of the way before the next decision.

Dot Glide: Letting the Board Be the Tutorial

In Dot Glide, each pair of colored endpoints asks for a continuous route, and routes cannot overlap. I wanted the first board to teach that with the pieces themselves. Matching color, endpoint size, line glow, and immediate connection feedback do most of the explaining. The puzzle becomes interesting when an obvious short path blocks a later connection, so the interface must make undoing and trying again feel normal rather than punitive.

The green-and-black presentation is deliberately quieter than Chomp Circuit. A route puzzle needs empty space because empty cells are part of the information. If every tile carried decoration, it would be harder to plan two or three moves ahead. The glow belongs to active endpoints and completed paths; it is feedback with a job. I adjusted endpoint sizes for touch screens so they are easy to select without making the grid feel cramped on a laptop.

Dot Glide: Failure Should Invite Another Attempt

Puzzle friction is not the same as interface friction. Taking a wrong route is the puzzle. Struggling to select a node is the interface. I treat those as separate problems during testing. The game saves a lightweight checkpoint, uses a clear wrong-action response, and provides board and route completion cues so the player knows whether the system understood the gesture. None of those features solves the board, but each keeps the conversation between player and puzzle honest.

I also avoided turning every attempt into a public performance. A player can work through a board without creating an account or protecting a leaderboard rank. Completion is its own reward, while progression supplies a reason to continue. That fits the kind of daily browser session I want Klivii to support: a few concentrated minutes that feel complete, not a funnel that withholds the satisfying part.

Ichi: Computer Opponents Need Table Presence

The familiar part of Ichi is matching a card by color or number and trying to empty your hand. The original design work sits in the table around that rule. Milo, Nova, and Zara have names, expressions, voice reactions, and different seats because a four-player card game feels hollow when the other hands behave like database rows. Their reactions are brief and situational: drawing cards, waiting on a turn, being skipped, winning, or watching someone call Ichi.

Character does not excuse delay. Early builds made the table feel lively but sometimes stacked voice lines until the actual turn was less important than the performance. I added cooldowns, queues, and interruption rules so a useful game event wins over a joke. The opponents can complain if a player waits, but they cannot become a punishment for reading the hand. That balance—personality without obstruction—is one of the details I still revisit most often.

Ichi: Showing the State of a Complicated Turn

Action cards create edge cases quickly. A skip changes whose turn matters. Reverse changes direction. Draw penalties can stack. A wild card asks for another choice before play can continue. The interface needs to show playable cards and the active player while keeping hidden hands hidden. I use highlights, card motion, counts, turn indicators, and short voice cues as overlapping evidence. If one signal is missed, another should still explain what happened.

Dealing and result screens are paced as transitions rather than instant data swaps. That small amount of motion gives the table weight, but it also creates a risk: animation can leave input available at the wrong moment. Much of the unglamorous work in Ichi is state control—locking a hand during the deal, preventing duplicate input, revealing the result at the correct time, and resetting every temporary effect for a clean rematch.

What All Three Games Taught Me About Browser Play

The browser is not a lesser game platform; it is a different promise. The player arrives through a link and should be able to understand the page before deciding how much time to give it. That makes loading, responsive layout, keyboard and touch behavior, sound permissions, and visible reset controls part of game design. A beautiful opening does not help if the play area jumps when it loads or if the main action only works with a mouse.

I now review each featured game with the same questions. Can a first-time player state the goal after looking at the opening screen? Is the primary action reachable on a phone? Does every failure have a legible cause? Can the game be played with sound off? Does replay remove stale state? Is the result honest about what was scored? These checks are less dramatic than inventing a character or board, but they are where a browser game earns trust.

Why I Am Publishing the Process

Klivii's Journal should add something that the game cards cannot. For original games, the most useful thing I can add is the reasoning behind what players see and a record of what changed. That is why this article names specific controls, states, characters, and tradeoffs. It also gives me a standard to be accountable to. If the live game contradicts the principle described here, the right response is to improve the game or update the journal—not pretend the mismatch is invisible.

My next passes will keep focusing on first-play clarity and accessibility across the featured catalog. You can start with Chomp Circuit for movement, Dot Glide for planning, or Ichi for a more theatrical table. Each one is free to play in the browser, and each will keep evolving as I test it in the same place players do: on the live Klivii site.