Back to Journal
ConceptFizzics

Fizzics: From Suika Clone to Fluid Physics Puzzler

How a casual puzzle game concept evolved from 'Suika with liquids' into a visually differentiated physics puzzler with a neon mad scientist aesthetic.

March 15, 20265 min read

The idea started simple: what if Suika Game used liquid instead of fruit?

That's it. That was the whole pitch. Five seconds of thought that turned into months of work. Classic.

Suika's merge mechanic is endlessly satisfying. Drop objects, match pairs, watch them combine into something bigger. But every Suika clone looks the same. Circles with fruit textures. What if the visuals were the actual differentiator instead of just a skin on the same mechanics?

Finding the Gap

I looked at what's out there and the pattern was obvious. Casual puzzle games have proven mechanics but they all look... fine. Just fine.

  • Block Blast: 300M monthly active users, generic blocks, zero visual identity
  • Suika Game: 11M+ downloads, satisfying merge loop, fruit textures from 2001
  • Water Sort Puzzle: 100M+ downloads, colored liquids in test tubes, flat UI

Every one of these games is mechanically sound. And visually forgettable. Nobody's ever screenshot a Block Blast clear and thought "I need to share this."

The gap: no casual puzzle game uses procedural visual effects as the hook. Every clear looks the same. Every chain reaction is predictable. What if every detonation was unique, genuinely procedural, never the same twice?

The Core Mechanic

Fizzics takes Suika's pair-merge concept and wraps it in real-time fluid physics:

  1. Drop: slide to position a glowing blob above a beaker, release to drop
  2. Merge: same-colored, same-sized blobs combine into the next size tier (a metaball shader makes them flow together like actual liquid)
  3. Grow: merged blobs progress through 4 tiers: Droplet → Globule → Cluster → Volatile
  4. Detonate: max-size blobs explode in procedural neon particle bursts
  5. Cascade: explosions push nearby blobs into new merges, triggering chain reactions with exponential scoring

The strategic depth comes from placement. Different-colored blobs don't mix. They layer like oil and water. Where you drop matters because you're managing a container of immiscible liquids, not just stacking shapes. It's Suika's "think ahead" gameplay with a physics layer that makes every game state genuinely different.

The North Star

I landed on a single phrase to guide every design decision: "Satisfying hands, mesmerizing eyes."

The mechanics should feel clean: tight input, responsive physics, instant feedback. The visual layer goes all out with procedural neon explosions, glow effects, and haptic feedback. A mad scientist lab theme ties it together. Beaker, pipette, catalyst vial. Every UI element reinforces the fiction without a single line of dialogue or tutorial text.

If someone watches over your shoulder and asks "what is that?", the game is working.

The Conditional Go

Phase 0 ended with a conditional Go. The market gap is real, the mechanic is validated through existing games, and costs are negligible ($99/year for an Apple developer account).

But one critical question remained: can I actually render fluid physics at 60fps on an iPhone?

If the answer is no, none of this matters. Cool concept, impossible execution, back to the drawing board. So that became the first spike in Pre-Production.

// The concept crystallized into a simple config
public enum BlobTier { Droplet, Globule, Cluster, Volatile }
public enum BlobColor { Cyan, Magenta, Lime, Amber, Violet, HotPink }
 
// Same color + same size → merge up one tier
// Max tier → detonate → chain reactions → spectacle
// Simple rules, emergent chaos. The best kind.