| name | game-analysis-criteria |
| description | This skill provides templates, analysis frameworks, and quality criteria for game reverse-engineering and competitive analysis. Used when analyzing external games for mechanics, technology, marketing, retention, and market positioning. Loaded by game-researcher agent and game-analyze command. |
Game Analysis Criteria
Purpose
Framework for systematic reverse-engineering and analysis of external games to extract actionable insights for player retention, market positioning, and game design decisions.
Templates
Mechanics (per-mechanic reverse engineering)
Analysis Reports
- technology-report-template.md - Engine, frameworks, networking, performance
- marketing-report-template.md - Marketing strategies, community engagement, influencer relations
- retention-report-template.md - Retention mechanics, monetization, live ops
- market-position-report-template.md - Competitive landscape, market share, audience
- ui-ux-report-template.md - UI/UX design, onboarding, accessibility
- game-feel-report-template.md - Polish, juice, feedback systems, audio design
Synthesis
Storage Convention
All game research is stored per game under the user's project:
docs/game-research/
{game-slug}/
raw-research.md # Raw data from game-researcher
mechanics/ # Per-mechanic reverse engineering
overview.md # sr-game-designer: all mechanics map, core loop
{mechanic-slug}.md # mechanics-developer: detailed per-mechanic analysis
technology-analysis.md # Engine, frameworks, performance
marketing-analysis.md # Marketing strategies, community
retention-analysis.md # Retention mechanics, monetization
market-position.md # Competitive landscape
ui-ux-analysis.md # UI/UX design analysis
game-feel-analysis.md # Polish, juice, feedback
executive-summary.md # Final synthesis with recommendations
source-code/ # Source code notes (if available)
architecture.md
patterns.md
Naming: {game-slug} = lowercase, hyphenated game name (e.g., hollow-knight, stardew-valley, clash-royale)
Analysis Framework
Seven Pillars of Game Analysis
Each game should be analyzed across these dimensions:
| Pillar | Key Questions | Primary Agent |
|---|
| 1. Core Loop | What is the fundamental gameplay cycle? How long is one loop? What drives repetition? | sr-game-designer |
| 2. Progression | How does the player advance? What are the unlock gates? What creates the "one more turn" feeling? | sr-game-designer |
| 3. Retention Mechanics | What brings players back? Daily hooks? Social obligations? FOMO? Loss aversion? | data-scientist |
| 4. Monetization | What is the revenue model? What are the conversion triggers? How does spending enhance experience? | market-analyst |
| 5. Technology | What engine/framework? How do they handle performance? What are technical differentiators? | mechanics-developer |
| 6. Marketing & Community | How was the game marketed? What communities exist? How is content distributed? | market-analyst |
| 7. Game Feel | What makes the game "feel good"? Audio design, visual feedback, input responsiveness? | game-feel-developer |
Retention Analysis Deep Dive
Special focus area — what keeps players engaged:
Short-Term Retention (D1-D7)
- First-time user experience (FTUE) quality
- Tutorial effectiveness
- Time-to-fun measurement
- Initial content variety
- Early reward pacing
Mid-Term Retention (D7-D30)
- Content depth discovery
- Social features activation
- Habit loop formation
- Difficulty curve management
- Second-session hooks
Long-Term Retention (D30+)
- Endgame content systems
- Community integration
- Competitive/cooperative hooks
- Content update cadence
- Sunk cost and identity investment
Source Code Analysis (when available)
For open-source games or games with accessible source code:
| Analysis Area | What to Extract |
|---|
| Architecture | Module structure, dependency graph, entry points |
| State Management | How game state is stored, updated, persisted |
| Event System | Communication patterns, decoupling strategy |
| Networking | Client-server model, sync strategy, lag compensation |
| Performance | Object pooling, rendering pipeline, memory management |
| AI Systems | Decision trees, state machines, behavior patterns |
| Content Pipeline | How content is loaded, cached, hot-swapped |
Quality Criteria for Research
Data Quality Tiers
| Tier | Source Type | Confidence | Usage |
|---|
| Verified | Official developer data, store APIs, public financials | 90%+ | Direct citation |
| Inferred | Review aggregation, community data, tool analysis | 60-89% | Cite with confidence level |
| Estimated | Industry benchmarks, comparable titles, analyst reports | 30-59% | Mark as estimate |
| Speculative | Educated guesses based on patterns | <30% | Mark as hypothesis |
Minimum Research Requirements
| Depth Level | Sources Required | Reports Generated | Estimated Duration |
|---|
| Quick | Store page + 3 reviews + 1 article | Executive Summary only | ~10 min |
| Standard | Store page + 10 reviews + 3 articles + community scan | All 8 reports | ~30 min |
| Deep | Standard + source code + gameplay analysis + developer interviews | All 8 reports + source code analysis | ~60 min |
Report Quality Gates
Before a report is considered complete:
Cross-Game Comparison
When analyzing multiple games, generate a comparison matrix:
## Comparison Matrix: [Game A] vs [Game B] vs [Game C]
| Dimension | Game A | Game B | Game C | Our Opportunity |
|-----------|--------|--------|--------|----------------|
| Core Loop Duration | X min | Y min | Z min | Target |
| D1 Retention | X% | Y% | Z% | Target |
| D30 Retention | X% | Y% | Z% | Target |
| ARPU | $X | $Y | $Z | Target |
| Session Length | X min | Y min | Z min | Target |
| Metacritic | X | Y | Z | Target |
| Community Size | X | Y | Z | Target |
