- name
- analytics-balance
- description
- Extract analytics and game balance entities from narrative text. Use when analyzing player metrics, difficulty curves, drop rates, loot tables, heatmaps, and balance tuning.
# analytics-balance
Domain skill for Analytics Balance Specialist. Specific extraction rules and expertise.
## Domain Expertise
- **Player analytics**: Session length, completion rates, engagement metrics
- **Game balance**: Difficulty curves, loot tables, progression speed
- **Heatmaps**: Player movement, hot zones, unused areas
- **Drop rates**: Loot probability, item scarcity, RNG balancing
- **Conversion rates**: Player retention, monetization, engagement
- **Difficulty scaling**: Early game vs late game balance
## Entity Types (8 total)
- **player_metric** - Player metrics
- **session_data** - Session data
- **heatmap** - Heatmaps
- **drop_rate** - Drop rates
- **conversion_rate** - Conversion rates
- **difficulty_curve** - Difficulty curves
- **loot_table_weight** - Loot table weights
- **balance_entity** - Balance entities
## Processing Guidelines
When extracting analytics and balance entities from chapter text:
1. **Identify analytics/balance elements**:
- Difficulty or scaling mentioned
- Loot drops or rewards
- Player progress or metrics
- Game balance references
- Session data or tracking
2. **Extract analytics/balance details**:
- Difficulty levels and progression
- Loot tables and probabilities
- Player engagement metrics
- Heatmap data patterns
- Balance issues or adjustments
3. **Analyze analytics/balance context**:
- Is progression too fast or slow?
- Are rewards fair for difficulty?
- Are players engaging properly?
- Are there balance exploits?
4. **Create schema-compliant entities** with proper JSON structure
## Key Considerations
- **Fairness**: Loot and difficulty should feel fair
- **Progression speed**: Neither too fast nor too slow
- **Retention**: Players should want to keep playing
- **Data-driven**: Balance changes based on metrics, not guesses
- **Player agency**: Difficulty options for different playstyles
GitHubで見る