Skip to content

RELIABILITY_INDEX_WHITEPAPER.md

Latest commit

 

History

History
211 lines (158 loc) · 7 KB

RELIABILITY_INDEX_WHITEPAPER.md

File metadata and controls

211 lines (158 loc) · 7 KB
Raw

SoroSusu Reliability Index (RI) Technical Whitepaper

Overview

The Reliability Index (RI) is SoroSusu's proprietary "Social Credit" system that quantifies user trustworthiness and reliability within the decentralized savings circle ecosystem. The RI serves as a decentralized reputation score that influences protocol fees, access to advanced features, and integration with third-party DeFi services.

Core Components

The RI is calculated as the arithmetic mean of two primary sub-scores:

RI = (Reliability Score + Social Capital Score) / 2

Both sub-scores are expressed in basis points (0-10000, representing 0-100%).

1. Reliability Score

The Reliability Score measures a user's historical performance in meeting their financial obligations within savings circles.

Calculation Formula:

Reliability Score = min(10000, On-Time Rate + Volume Bonus)

Where:

  • On-Time Rate = (Total On-Time Contributions × 10000) / Total Expected Contributions
  • Volume Bonus = min(100, Total Volume Saved / 1000000) × 50

Parameters:

  • On-Time Contributions: Contributions made before or on the circle's deadline
  • Total Expected Contributions: All contributions the user was scheduled to make
  • Total Volume Saved: Cumulative value of payouts received (in stroops)

Scoring Examples:

  • Perfect record (100% on-time, high volume): 10000 bps
  • 95% on-time rate, moderate volume: ~9750 bps
  • 85% on-time rate, low volume: ~8500 bps
  • New user (no history): 5000 bps (baseline)

2. Social Capital Score

The Social Capital Score measures a user's participation in community governance and social support mechanisms.

Calculation Formula:

Social Capital Score = min(10000, Baseline + Leniency Bonus + Voting Bonus + Decay Penalty)

Where:

  • Baseline: 5000 bps (neutral starting point)
  • Leniency Bonus: +50 bps per leniency vote given, +25 bps per leniency received
  • Voting Bonus: +10 bps per governance vote cast
  • Decay Penalty: -5% monthly if inactive for >6 months

Parameters:

  • Leniency Given: Number of times user voted to grant extensions to struggling members
  • Leniency Received: Number of times user received leniency from the community
  • Governance Votes: Total quadratic votes cast in proposals
  • Activity Timestamp: Last recorded participation in any circle activity

RI Update Triggers

The RI is recalculated and stored whenever a user performs any of the following actions:

  1. Deposit Contribution: Updates reliability score based on timeliness
  2. Cast Governance Vote: Increases social capital score
  3. Request/Grant Leniency: Updates social capital for both parties
  4. Complete Circle Cycle: Updates volume metrics
  5. Monthly Heartbeat Check: Applies inactivity decay

RI Applications

1. Fee Discount Logic

Protocol fees are discounted based on RI tiers:

  • RI ≥ 9000 (Diamond): 75% fee discount
  • RI ≥ 8000 (Platinum): 50% fee discount
  • RI ≥ 7000 (Gold): 25% fee discount
  • RI ≥ 6000 (Silver): 10% fee discount
  • RI < 6000 (Bronze): Standard fees

2. Access Control

Higher RI unlocks advanced features:

  • RI ≥ 8000: Early access to liquidity advances
  • RI ≥ 7000: Priority in randomized payout queues
  • RI ≥ 6000: Eligibility for governance voting
  • RI ≥ 5000: Basic circle participation

3. Third-Party Integrations

RI scores are exposed via the get_reputation() function for:

  • DeFi lending protocols (credit scoring)
  • Insurance providers (risk assessment)
  • NFT marketplaces (soulbound token minting)
  • Cross-chain bridges (identity verification)

Sybil Resistance Mechanisms

To prevent reputation manipulation through multiple accounts:

1. Proof of Personhood Integration

  • increase_reputation() functions require verified unique identity
  • Integration with decentralized identity providers
  • On-chain verification of personhood proofs

2. Activity-Based Validation

  • RI increases only through provable on-chain actions
  • No direct reputation transfers between accounts
  • Decay mechanisms prevent reputation hoarding

3. Economic Incentives

  • High RI provides tangible benefits (fee discounts)
  • Sybil attacks become economically unviable
  • Community reporting mechanisms for suspicious activity

Inactivity Decay

To ensure RI reflects current reliability:

Decay Formula:

New RI = Current RI × (1 - 0.05)  // 5% monthly decay

Trigger Conditions:

  • No circle participation for >6 months
  • Automatic monthly check via "heartbeat" mechanism
  • Decay applies to both sub-scores equally

Recovery:

  • Decay can be reversed through renewed participation
  • No minimum RI floor (can decay to 0)
  • Historical performance provides foundation for recovery

Technical Implementation

Data Structures

#[contracttype]
pub struct UserReputationMetrics {
    pub reliability_score: u32,     // 0-10000 bps
    pub social_capital_score: u32, // 0-10000 bps
    pub total_cycles: u32,
    pub perfect_cycles: u32,
    pub total_volume_saved: i128,
    pub last_activity: u64,
    pub last_decay: u64,
}

#[contracttype]
pub struct ReputationData {
    pub user_address: Address,
    pub susu_score: u32,        // RI (0-10000 bps)
    pub reliability_score: u32, // 0-10000 bps
    pub total_contributions: u32,
    pub on_time_rate: u32,      // 0-10000 bps
    pub volume_saved: i128,
    pub social_capital: u32,    // 0-10000 bps
    pub last_updated: u64,
    pub is_active: bool,
}

Storage Keys

  • URep:{user_address}: UserReputationMetrics
  • LastDecay:{user_address}: Last decay timestamp

Public Functions

  • get_reputation(user: Address) -> ReputationData
  • calculate_reliability_score(user: Address) -> u32
  • apply_inactivity_decay(user: Address)
  • update_reputation_on_deposit(user: Address, was_on_time: bool)

Security Considerations

Manipulation Prevention

  • All reputation updates require on-chain transaction proof
  • No admin override capabilities for reputation scores
  • Cryptographic verification of contribution timestamps

Privacy Protection

  • Reputation data is publicly readable but not linkable without user consent
  • No personal identifying information stored
  • Soulbound token mechanism for optional public verification

Economic Security

  • Fee discounts create natural demand for high RI
  • Decay prevents long-term reputation squatting
  • Community governance over reputation parameters

Future Enhancements

Advanced Metrics

  • Cross-circle reputation aggregation
  • Temporal weighting (recent activity > old activity)
  • Predictive scoring using machine learning

Integration APIs

  • Standardized reputation oracle interface
  • Cross-protocol reputation portability
  • Decentralized identity verification bridges

Governance

  • Community voting on reputation parameters
  • Adjustable decay rates and thresholds
  • Emergency reputation reset mechanisms

This document provides the technical foundation for SoroSusu's Reliability Index system. Implementation details may evolve through community governance and security audits.