Skip to content

Reputation System

⚠️ STUB DOCUMENTATION: This file contains placeholder content and needs to be expanded with detailed reputation system documentation.

How the oracle reputation and selection system works in Verdikta Dispatcher.

Overview

The reputation system ensures high-quality dispute resolution by tracking oracle performance and selecting the best oracles for each request.

Reputation Scoring

Performance Metrics

  • Response Time: How quickly oracles respond to requests
  • Accuracy: Quality of decisions compared to consensus
  • Availability: Uptime and reliability metrics
  • Stake: Amount of tokens staked by the oracle

Score Calculation

Reputation Score = (Accuracy × 40%) + (Speed × 30%) + (Availability × 20%) + (Stake × 10%)

Oracle Selection

Weighted Selection

Oracles are selected using weighted random selection based on: - Reputation score - Availability status - Oracle class matching - Fee competitiveness

Alpha Parameter

The alpha parameter (0-1000) controls the balance between: - High Alpha (700-1000): Favor highly reputed oracles - Medium Alpha (300-700): Balanced selection - Low Alpha (0-300): More random selection, lower fees

Implementation Details

For technical implementation details, see: - ReputationKeeper Contract - ReputationAggregator Contract - Integration Examples