game-vision

What this skill does

# Game Vision

**Purpose:** Bridge the gap between creative intuition and structural rigor. A game concept that lives only in your head is untestable. This skill turns "I have an idea" into a validated, layered design that can be handed to specialized skills for detailed work.

**Position in workflow:** This is the UPSTREAM skill. Define vision here first, then fan out to **game-design** (mechanic evaluation), **experience-design** (engagement loops), **systems-design** (system interactions), and others. If you skip this step, downstream skills lack a target to evaluate against.

---

## When to Activate

Use this skill when:
- Starting a new game project from scratch
- Pivoting or rethinking a game concept that isn't working
- Evaluating whether an idea is worth building at all
- Prototyping has gone on too long without a clear hypothesis to test
- The team (or your own thinking) can't agree on what the game IS
- Scope keeps expanding and there's no north star for cutting decisions
- You need to communicate a game concept to collaborators or stakeholders
- Returning to a shelved concept and need to re-evaluate it

---

## Core Framework: The Vision Stack

A game vision is a stack of dependent layers. Each layer constrains and informs the layers below it. If you change a layer, everything below may need to change.

```
Layer 1: Core Fantasy          ← One sentence. The emotional promise.
Layer 2: Experience Pillars    ← 3-5 pillars that define the target experience
Layer 3: Core Loop             ← The irreducible verb sequence
Layer 4: Supporting Systems    ← Derived from pillars, each must serve one
Layer 5: Content Strategy      ← What fills the systems (levels, items, enemies)
```

### Layer 1: Core Fantasy

The answer to: **"What does the player FEEL like they're doing?"**

One sentence. Not what the game IS, but what the player FEELS. This is the emotional promise that every other design decision must serve.

| Weak Core Fantasy | Strong Core Fantasy |
|-------------------|---------------------|
| "A platformer with combat" | "You're a scrappy underdog who outsmarts enemies three times your size" |
| "An open-world RPG" | "You're a wanderer piecing together a forgotten civilization's final days" |
| "A city builder" | "You're a mayor making impossible trade-offs to keep a fragile city alive" |

**Test:** If two different games could share your core fantasy statement, it's too generic. If it doesn't imply an emotion, it's a genre label, not a fantasy.

### Layer 2: Experience Pillars

3-5 pillars that define the target experience. These are the qualities the game MUST deliver. Everything else is negotiable.

See [Experience Pillars Method](#experience-pillars-method) below for how to derive these.

### Layer 3: Core Loop

The irreducible verb sequence that IS the game. The smallest cycle of player action that produces engagement.

See [Core Loop Crystallization](#core-loop-crystallization) below.

### Layer 4: Supporting Systems

Systems derived from the pillars. Each system must serve at least one pillar, and each pillar must be served by at least one system.

See [System Derivation](#system-derivation) below.

### Layer 5: Content Strategy

What fills the systems: levels, enemies, items, dialogue, environments. Content is the fuel; systems are the engine. Once you reach this layer, fan out to specialized skills (**game-balance** for stats, **progression-systems** for pacing, **narrative-design** for story, **encounter-design** for level population).

---

## Experience Pillars Method

Pillars are the experience qualities the game must deliver. Deriving them is a five-step process.

### Step 1: Brainstorm

List everything exciting about the concept. No filtering, no judgment. Include feelings, verbs, moments, aesthetics, references to other games, half-formed ideas. Quantity over quality.

### Step 2: Cluster

Group related ideas. Look for themes. "Tense combat," "split-second decisions," and "barely surviving" might cluster into one group. "Building a base," "watching your creation grow," and "expressing yourself through design" might form another.

### Step 3: Prioritize

Force-rank clusters to exactly 3-5 pillars. This is the hard part. The discipline of cutting is the entire point.

**The pillar rule:** If it's not a pillar, it's a feature. Features can be cut. Pillars cannot. If everything is a pillar, nothing is.

### Step 4: Validate

Each pillar must pass two tests:
- **Testable:** Can you build a prototype that specifically tests this pillar?
- **Actionable:** Does this pillar imply concrete design decisions? ("Fun" is not actionable. "Tense survival under resource pressure" is.)

### Step 5: Conflict-Check

Do any pillars tension against each other? This is not necessarily bad — creative tension can define a game's identity — but the tension must be intentional, not accidental.

| Pillar A | Pillar B | Tension | Resolution |
|----------|----------|---------|------------|
| Strategic depth | Fast-paced action | Thinking vs. reacting | Alternate between planning phases and execution phases |
| Creative expression | Competitive balance | Freedom vs. fairness | Cosmetic creativity is free; mechanical creativity is constrained |
| Narrative immersion | Player agency | Authored story vs. player choice | Branching narrative with authored branches |

If pillars conflict, establish priority. When a design decision must favor one pillar over another, which one wins?

---

## Target Experience Statement

A single statement that serves as the north star for every design decision.

**Template:**

> In **[game title]**, players experience **[primary emotion sequence]** by **[core verbs]** in a world where **[fiction/setting premise]**.

Every design decision should be testable against this statement: "Does this serve the target experience?"

### Worked Examples

**Action-survival:**
> In *Ashfall*, players experience desperate resourcefulness by scavenging, crafting, and fortifying in a world where the volcanic winter is closing in and every supply run could be your last.

**Puzzle-exploration:**
> In *Hollow Light*, players experience wonder and dawning comprehension by navigating, observing, and experimenting in a world where ancient machines still run but nobody remembers what they were for.

**Strategy:**
> In *Tidelock*, players experience the weight of impossible trade-offs by allocating, negotiating, and sacrificing in a world where three island nations share one shrinking freshwater source.

**Roguelike:**
> In *Thornwild*, players experience improvisational mastery by combining, adapting, and risking in a world where every run through the overgrown ruins deals you a different hand of abilities.

---

## Core Loop Crystallization

The core loop is the smallest cycle of player action that is independently satisfying. Finding it requires distillation, not invention.

### Process

1. **List all player actions** the concept suggests. Everything the player might do.

2. **Draw dependency chains.** What must happen before what? What feeds into what?

3. **Find the smallest loop.** The tightest cycle: `Action → Feedback → Evaluation → Decision → Action ...`

4. **Test for independence.** Does this loop produce the target emotions WITHOUT supporting systems (progression, economy, narrative)? If the loop only becomes interesting with progression bolted on, the loop isn't strong enough.

5. **Test on paper.** The core loop should be compelling as a description before it's compelling as code. If you can't make someone excited about it in conversation, implementation won't save it.

### Core Loop Examples

| Game Type | Core Loop |
|-----------|-----------|
| Action-platformer | Move → Encounter obstacle → Execute skill → Survive/fail → Adapt approach |
| Deckbuilder | Draw hand → Evaluate options → Play cards → See outcome → Adjust strategy |
| City builder | Observe needs → Zone/build → Simulate → See consequences → Rebalance |
| Stealth | Scout →