PRD: NFT Rarity Analyzer
Summary
One-liner: Analyze NFT collections to calculate trait rarity scores and rank tokens by uniqueness.
Domain: Cryptocurrency / NFT / Digital Collectibles
Users: NFT Collectors, Traders, Analysts, Developers
Problem Statement
NFT collectors need to quickly assess rarity to make informed buying decisions. Manual trait analysis is time-consuming and error-prone. Existing tools are often expensive, require accounts, or lack transparency in their scoring methods.
User Stories
As an NFT collector, I want to check rarity scores for tokens in a collection, so that I can identify undervalued pieces.
As an NFT trader, I want to compare multiple tokens by rarity rank, so that I can make data-driven purchase decisions.
As a collection creator, I want to analyze my collection's trait distribution, so that I can ensure balanced rarity across items.
As an analyst, I want to export rarity data for further analysis, so that I can create custom rarity models.
Functional Requirements
- REQ-1: Fetch NFT collection metadata from OpenSea, Alchemy, or direct contract
- REQ-2: Parse and normalize trait attributes across collection
- REQ-3: Calculate trait rarity using statistical frequency analysis
- REQ-4: Compute composite rarity scores using multiple algorithms
- REQ-5: Rank tokens within collection by rarity
- REQ-6: Display trait breakdown with rarity percentages
- REQ-7: Support ERC-721 and ERC-1155 token standards
- REQ-8: Cache collection data for performance
- REQ-9: Export rarity data in JSON and CSV formats
Rarity Algorithms
Rarity Score / Statistical Rarity
Score = sum(1 / trait_frequency) for each trait
Both statistical and rarity_score use this formula - they are equivalent in the implementation.
Average Rarity
Score = sum(trait_rarities) / trait_count
Information Content
Score = sum(-log2(trait_frequency)) (higher = rarer)
Score Normalization (Post-processing)
After calculating scores, they can be normalized to 0-100 scale:
normalized = (score - min_score) / (max_score - min_score) * 100
API Integrations
- OpenSea API: Collection metadata, token traits
- Alchemy NFT API: Fast metadata fetching
- Ethereum RPC: Direct tokenURI reads
- IPFS Gateway: Metadata from decentralized storage
Success Metrics
- Activation triggers on NFT rarity queries
- Accurate rarity rankings match established tools
- Fast analysis (<5s for cached collections)
- Support for collections up to 20,000 items
Non-Goals
- Price predictions based on rarity
- Automated trading or sniping
- Collection indexing from scratch
- Real-time floor price tracking