UNIR REIS
Let's Talk
Back to Projects
Project 02 / 13

Project 02 · Game

Bottle Flip Fever

Physics-based puzzle game with 50+ handcrafted levels.

  • Puzzle
  • Physics
  • Mobile
  • Games
Genre:
Physics Puzzle
Scope:
Level Design · Systems Design · Release & Validation
Development Time:
Ongoing project
Team Size:
2
Tools:
Unity (C#) · Blender · Photoshop
View Gallery

PlaceholderBottle Flip Fever — key art16:10

Project Resume

01

Overview

Every level is a arrangement of surfaces, not a set of new rules. Difficulty comes from the geometry the player has to read before committing to a flip — landing angle, surface size, and what happens on a near miss. Four environments give the 50 levels visual variety without changing what the player does.

  • Platform Mobile
  • Team Size 2
  • Duration Ongoing project
  • Engine Unity

My Role

Game Designer & Level Designer

  • Level Design
  • Systems
  • Physics Tuning
  • Live Validation
Full responsibilities
Read chapter 01
02

Problem 1

PlaceholderProblem 116:10

Players felt that identical flips produced different outcomes.

  • Players stopped attributing failure to the game
  • Retry rate went up instead of drop-off — a failed attempt read as learnable
Read chapter 02
03

Problem 2

PlaceholderProblem 216:10

Adding interest to later levels kept pushing toward new mechanics.

  • The level count grew without the rule set growing
  • Later levels stayed readable to a new player who skipped ahead
Read chapter 03

Media

The work, up close

Project gallery · 2

PlaceholderBottle Flip Fever — key art16:10

Bottle Flip Fever — key art

Physics-based puzzle with 50+ handcrafted levels.

Roles

Roles & Responsibilities

What I did

  • Designed and handcrafted 50+ levels across four environments
  • Tuned the physics model until difficulty came from reading the shot, not from luck
  • Built the difficulty curve and the level ordering
  • Ran the beta release and read the retention data back into level design

Full Case Study

4 chapters

Overview

Design Brief

Design Problem

Bottle flipping is a one-second action with a binary outcome. The design problem was to build fifty levels of interest out of a single verb, without adding mechanics that would dilute it.

Constraints

  • One verb: flip
  • Physics must stay deterministic enough to feel fair
  • Mobile, one-handed, no tutorial
  • Levels had to be hand-built, not generated

Overview

Design Solution

Every level is a arrangement of surfaces, not a set of new rules. Difficulty comes from the geometry the player has to read before committing to a flip — landing angle, surface size, and what happens on a near miss. Four environments give the 50 levels visual variety without changing what the player does.

Key Design Choices

  • Difficulty lives in level geometry, never in randomness
  • A near miss must be visibly a near miss, so failure teaches
  • No mechanic added after launch that the first level could not explain
  • Environments group levels into difficulty bands the player can feel

PlaceholderLevel set4:3

Problem 1

Problem

Fairness in a Physics Game

Problem

Players felt that identical flips produced different outcomes.

Cause

The physics allowed small variance the player could neither see nor predict.

Solution

Tightened the physics response and made landing surfaces read their own tolerance visually. The shot became predictable enough that a miss was clearly a misjudgement.

Result

  • Players stopped attributing failure to the game
  • Retry rate went up instead of drop-off — a failed attempt read as learnable

Problem 2

Problem

Fifty Levels Without Fifty Mechanics

Problem

Adding interest to later levels kept pushing toward new mechanics.

Cause

Treating difficulty as content rather than as composition.

Solution

Built a vocabulary of surface types and combined them instead of extending the rules. Every hard level is a new sentence in a language the first ten levels already taught.

Result

  • The level count grew without the rule set growing
  • Later levels stayed readable to a new player who skipped ahead

Results & Learnings

Learnings

Takeaways

Five thousand beta downloads were worth more than any amount of internal review. Organic acquisition told me which levels people actually stopped at, and that list did not match the one I would have guessed.

A single verb is not a limitation — it is the thing that makes fifty levels legible. Every time I considered adding a mechanic, the better answer was a surface I had not tried yet.

“Games are a way to turn ideas into feelings.”— Unir Reis

Next project

Farm RPG