Playtesting Approach

Our Approach: Playtest-Driven Development

  • We prioritize incremental improvements that directly enhance the player experience.
  • Each weekly playtest sets a clear, focused goal, ensuring we improve a small piece of the core loop every cycle.
  • This approach allows us to reprioritize engineering tasks based on what we learn, keeping development aligned with player experience.

Primary Objectives for Nov 3rd

  • Full E2E Loop: Player can create a creature, see it in an inventory, and use it in a playable battle.
  • Complete Creature Assets: Each creature must have a functional card and a v1-generated image.
  • Functional Battle: Combat is playable against a random AI, with the majority of card effects working.
  • Visual Clarity: Battle resolution is clearly communicated to the player via a 2D dice/probability UI.

Execution Plan & Tasks

Playtest 1: Core Play Baseline

Battle Animations & Creature Art

Oct 9

Goal: Create and play your creature against an AI with low visual fidelity.

Scope:
  • Heal, Defense, and Attack cards are functional.
  • Status Effects are not functional.
  • No local models are used.

Playtest 2: Battle Animations & Creature Art

Battle Animations & Creature Art

Oct 16

Goal: Make combat tactically readable through UI animations with creature art.

Scope:
  • UI animations for combat events are implemented.
  • Creature art is integrated into the battle screen.
  • Visual fidelity of the UI will match 90% of the Figma designs for the battle screen.

Playtest Tasks

Status Effects Goal

JP

Focus: UI animations for combat events are implemented.

Integrate Status Effects Into Simulation
  • Implement 90% of known Status Effects within simulation.
  • Unit test validating correctness.
Animate Status Effects in UI
  • Sim must emit events when statuses are applied or expire.
  • UI must play a visual indicator for status application and expiration events.

Creature Image Generation Loop

Patrick

Focus: Creature art is integrated into the battle screen.

Set Up Creature Image Generation API with Public URLs
  • Build an API that takes a creature definition, generates an image, stores it in R2, and returns a public URL.
Call API During Creature Creation and Store URL
  • Godot client must call the image gen API on creature creation and save the returned URL.
Load and Display Image from URL in Inventory and Battle
  • Inventory and battle UI must load and display creature images from their public URL.
  • If an image fails to load, show a fallback.

Battle UI Polish & Dice Odds Panel

Paul

Focus: Visual fidelity of the UI will match 90% of the Figma designs for the battle screen.

Integrate Simulation Events into Gameplay
  • Simulation must emit events for all core state changes (HP, statuses).
Consume Simulation Events in Battle UI
  • Battle UI must listen for sim events and trigger corresponding animations.
Polish Battle UI to Match Figma Design (90%)
  • Refine UI timing, spacing, and layering to achieve 90% Figma fidelity.
(Stretch) Add Dice Odds Panel
  • Implement a UI panel to display dice faces and their probabilities.

Playtest 3: Full Integration

Full Integration

Oct 23

Goal: Achieve full integration of creature generation and multiplayer battles.

Scope:
  • Creature generation (cloud) with custom image gen models is operational.
  • Multiplayer & single-player battles are fully functional with animations.
  • Visual clarity will match 90% of the Figma designs across all screens.

Playtest Tasks

Full Core Loop Playable

JP

Focus: Multiplayer & single-player battles are fully functional with animations.

Implement Database Storage and Retrieval for Creature Data
  • Set up a database to store creature data (including avatar URLs and schema).
  • The inventory screen must query from and display this database data.
Paul
Multiplayer Support for Full Core Loop
  • The full loop (creation, inventory, battle) must be functional in multiplayer.
Complete Effect Support and Safe No-Op for Unsupported Effects
  • Finalize implementation for all required card effects.
  • Unsupported effects must fail gracefully (no-op) without crashing.
Prompt Engineering for Balanced Decks
  • Refine prompt engineering to produce balanced, MP-ready decks.
Ennio

Visual Fidelity

Paul

Focus: Visual clarity will match 90% of the Figma designs across all screens.

Implement Figma Components in Godot
  • Create reusable Godot components for all key Figma UI elements.
Apply Components Across All Relevant Screens
  • Integrate new components across all key screens to match Figma layouts.
Enhance Dice Roll Clarity and Other Visual Fine-Tuning
  • Final polish pass on dice roll animations and other key UI details.

Image Generation Fine-Tuned Model

Patrick

Focus: Creature generation (cloud) with custom image gen models is operational.

Integrate Custom Fine-Tuned Model for Image Gen
  • Image generation pipeline must use our fine-tuned LoRA model.
  • (Stretch) The model version should be a configurable parameter.
Improve Visual Clarity and Player Feedback During Generation
  • Add in-app feedback (e.g., loading animations) for the image generation process.