Skip to main content

Analyzing Markets with PolyOracle AI

Learn how to use PolyOracle AI to perform comprehensive market analysis on Polymarket prediction markets.

Overview​

PolyOracle AI's market analysis combines:

  • Deep Research: Web-based research using Exa search
  • Technical Analysis: Orderbook, volume, and price analysis
  • Sentiment Analysis: News and social sentiment evaluation
  • Risk Assessment: Liquidity and market efficiency evaluation

Starting Your First Analysis​

Using the Web Interface​

  1. Access the Interface

    cd frontend
    pnpm dev

    Visit http://localhost:3000

  2. Connect Your Wallet

    • Click "Connect Wallet"
    • Select your Web3 wallet (MetaMask, etc.)
    • Switch to Polygon network if prompted

  3. Select a Market

    • Browse available markets or search by keyword
    • Click on a market you want to analyze
    • The market question and current odds will display

  4. Start Analysis

    • Click "Analyze Market" to begin the AI analysis
    • The agent will run through its research and analysis workflow
    • View real-time progress in the interface

Using the Backend API​

You can also analyze markets directly through the LangGraph API:

curl -X POST http://localhost:2024/analyze \
  -H "Content-Type: application/json" \
  -d '{
    "market_id": "0x123...",
    "research_depth": 3
  }'

Understanding the Analysis Process​

Phase 1: Market Data Collection​

The agent first collects comprehensive market data:

Market Information:

  • Question and description
  • Current YES/NO prices
  • Trading volume and liquidity
  • Market expiration date
  • Historical price trends

Orderbook Analysis:

  • Bid/ask spreads
  • Market depth at different price levels
  • Order size distribution
  • Execution price estimates

Example Output:

Market: "Will Bitcoin reach $100K by December 2024?"
Current YES Price: $0.65 (65% implied probability)
24h Volume: $45,230
Total Liquidity: $234,500
Bid-Ask Spread: 2.3%

Phase 2: Deep Research​

The agent performs recursive web research using Exa search:

Research Areas:

  • Recent news and developments
  • Expert opinions and analysis
  • Historical precedents
  • Market-moving events
  • Social sentiment indicators

Research Depth: Configurable (1-5 levels)

  • Level 1: Basic keyword search
  • Level 2: Follow-up questions based on initial findings
  • Level 3: Deep dive into specific aspects
  • Level 4: Contrarian research and risk factors
  • Level 5: Comprehensive multi-angle analysis

Example Research Output:

Research Summary:
- 15 sources analyzed across 3 research iterations
- Key finding: Recent institutional adoption increasing
- Risk factor: Regulatory uncertainty in key markets
- Sentiment: 70% positive, 20% neutral, 10% negative
- Confidence: High (based on source quality and consensus)

Phase 3: Technical Analysis​

Quantitative analysis of market conditions:

Price Analysis:

  • Price momentum and trends
  • Support and resistance levels
  • Volume-weighted average price (VWAP)
  • Price volatility metrics

Liquidity Analysis:

  • Available liquidity at different price points
  • Market impact estimation
  • Slippage calculations
  • Best execution estimates

Market Microstructure:

  • Order flow patterns
  • Time-weighted spreads
  • Market maker presence
  • Trading frequency analysis

Phase 4: Synthesis and Decision​

The agent combines all analysis into a trading decision:

Decision Framework:

  1. Fundamental Assessment: Based on research findings
  2. Technical Assessment: Based on market data analysis
  3. Risk Assessment: Considering all risk factors
  4. Confidence Scoring: 0-1 confidence in the decision
  5. Final Decision: BUY/SELL/NO_TRADE with rationale

Interpreting Analysis Results​

Research Quality Indicators​

Source Quality:

  • ✅ High: Major news outlets, expert analysis, official sources
  • ⚠️ Medium: Industry publications, blog posts, social media
  • ❌ Low: Unverified sources, outdated information

Research Confidence:

  • High (0.8-1.0): Strong consensus across quality sources
  • Medium (0.5-0.7): Mixed signals or limited information
  • Low (0.0-0.4): Conflicting information or poor source quality

Market Health Indicators​

Liquidity Assessment:

  • Excellent: Over $100K total liquidity, under 1% spread
  • Good: $50K-100K liquidity, 1-3% spread
  • Fair: $10K-50K liquidity, 3-5% spread
  • Poor: Under $10K liquidity, over 5% spread

Market Efficiency:

  • Efficient: Prices reflect available information well
  • Inefficient: Potential arbitrage or mispricing opportunities
  • Distorted: Unusual patterns suggesting manipulation

Trading Recommendations​

BUY Signal Criteria:

  • Research supports higher probability than market price
  • Sufficient liquidity for position size
  • Acceptable risk/reward ratio
  • High confidence in analysis (over 0.7)

SELL Signal Criteria (if holding position):

  • Research suggests lower probability than market price
  • Position sizing or risk management triggers
  • Market conditions have changed significantly

NO_TRADE Criteria:

  • Insufficient information or low confidence
  • Poor liquidity or excessive spreads
  • Market fairly priced based on available information

Advanced Analysis Features​

Custom Research Parameters​

Configure research depth and focus:

research_config = {
    "max_depth": 3,
    "sources_per_query": 10,
    "focus_areas": ["technical_analysis", "sentiment", "fundamentals"],
    "exclude_sources": ["reddit", "twitter"],  # Optional
    "time_filter": "1w"  # Last week only
}

Market Comparison​

Compare similar markets for relative value:

comparative_analysis = {
    "base_market": "0x123...",
    "comparison_markets": ["0x456...", "0x789..."],
    "comparison_metrics": ["implied_probability", "volume", "liquidity"]
}

Historical Backtesting​

Test analysis accuracy on historical markets:

backtest_config = {
    "start_date": "2024-01-01",
    "end_date": "2024-06-01", 
    "market_categories": ["politics", "economics"],
    "min_liquidity": 10000
}

Best Practices​

Research Optimization​

  1. Balance Depth vs Speed

    • Use deeper research for larger positions
    • Quick analysis for time-sensitive opportunities
    • Consider market volatility in timing

  2. Source Diversification

    • Include multiple perspectives
    • Verify information across sources
    • Consider potential biases

  3. Regular Updates

    • Re-analyze before major trades
    • Monitor for new information
    • Update analysis for position management

Risk Management​

  1. Position Sizing

    • Start small with new markets
    • Scale based on confidence and liquidity
    • Consider correlation across positions

  2. Market Selection

    • Focus on liquid markets (over $10K)
    • Avoid markets close to expiration
    • Prefer clear, objective questions

  3. Information Quality

    • Prioritize recent, relevant information
    • Cross-reference claims across sources
    • Be skeptical of outlier predictions

Troubleshooting Analysis Issues​

Common Problems​

"Insufficient research data"

  • Market topic may be too niche
  • Try broader search terms
  • Increase research depth setting

"Analysis taking too long"

  • Reduce research depth temporarily
  • Check internet connection
  • Verify API rate limits

"Conflicting signals"

  • Normal for complex markets
  • Focus on source quality
  • Consider waiting for more information

Performance Optimization​

Speed Improvements:

  • Use parallel research queries
  • Cache frequently accessed data
  • Optimize search query specificity

Quality Improvements:

  • Adjust source filtering
  • Tune confidence thresholds
  • Validate against known outcomes

Next Steps​

After mastering market analysis:

  1. Trading Guide - Learn how to execute trades
  2. API Reference - Explore advanced API usage for portfolio management
  3. FAQ - Find answers to common questions about performance

Ready to start trading? Check out the Trading Guide to learn how to act on your market analysis.