Bet Edge Analyzer

AI-powered edge analysis with natural language queries and conversational refinement.

What It Does

Provides enhanced edge detection workflow with:

• Natural Language Queries: "Find edges for Chiefs in week 7"
• Conversational Refinement: Iterative parameter adjustment
• Personalized Recommendations: Based on user preferences
• Dashboard Integration: Seamless UI integration
• Advanced Analysis: Multi-model comparison and insights

When to Use

• "Analyze edges for Chiefs in week 7"
• "Find the best QB TD props for this week"
• "Show me edges with at least 15% advantage"
• "Compare v1 vs v2 model results"
• "What's the optimal bet size for my bankroll?"

How It Works

1. Parse Query: Extract parameters from natural language
2. Run Analysis: Execute edge detection with parsed parameters
3. Present Results: Show edges with recommendations
4. Refine Parameters: Allow user to adjust criteria
5. Re-analyze: Run analysis with new parameters
6. Generate Insights: Provide actionable recommendations

Natural Language Examples

Basic Queries

• "Find edges for week 7"
• "Show me STRONG edges only"
• "Analyze Chiefs vs Giants"
• "What are the best QB TD props?"

Advanced Queries

• "Find edges for week 7 with at least 15% advantage using v2 model"
• "Show me edges for teams playing at home with good weather"
• "Analyze edges for QBs with recent injury concerns"
• "Find edges for games with high total points"

Refinement Queries

• "Increase the minimum edge to 20%"
• "Switch to v1 model"
• "Only show edges for AFC teams"
• "Filter out games with bad weather"

Parameter Parsing

Supported Parameters

• Week: "week 7", "this week", "next week"
• Model: "v1 model", "v2 model", "advanced model"
• Edge Threshold: "at least 15%", "minimum 10% edge"
• Teams: "Chiefs", "Bengals", "AFC teams"
• Bankroll: "with $1000 bankroll", "5% of bankroll"
• Confidence: "high confidence", "conservative"

Example Parsing

Input: "Find edges for Chiefs in week 7 with at least 15% advantage using v2 model"

Parsed Parameters:
{
  "week": 7,
  "teams": ["Chiefs"],
  "min_edge": 0.15,
  "model": "v2",
  "bankroll": 1000,
  "confidence": "medium"
}

Integration

• utils/edge_calculator.py - Edge detection engine
• utils/query_tools.py - Database queries
• dashboard/app.py - Dashboard integration
• find_edges.py - CLI edge detection

Usage Examples

Basic Analysis

# Natural language query
python scripts/conversational_analyzer.py "Find edges for week 7"

# With specific parameters
python scripts/conversational_analyzer.py "Show me STRONG edges for Chiefs"

Conversational Refinement

from scripts.conversational_analyzer import ConversationalAnalyzer

analyzer = ConversationalAnalyzer()

# Initial query
results = analyzer.analyze("Find edges for week 7")
print(results.summary)

# Refinement
refined_results = analyzer.refine("Increase minimum edge to 20%")
print(refined_results.summary)

# Further refinement
final_results = analyzer.refine("Only show AFC teams")
print(final_results.recommendations)

Dashboard Integration

// Frontend integration
async function analyzeEdges(query) {
    const response = await fetch('/api/analyze-edges', {
        method: 'POST',
        headers: {
            'Content-Type': 'application/json',
        },
        body: JSON.stringify({ query: query })
    });
    
    const results = await response.json();
    displayResults(results);
}

// Usage
analyzeEdges("Find edges for Chiefs in week 7");

Output Format

Standard Output

🎯 EDGE ANALYSIS RESULTS

Query: "Find edges for Chiefs in week 7 with at least 15% advantage"

Parameters Used:
- Week: 7
- Teams: Chiefs
- Minimum Edge: 15%
- Model: v2
- Bankroll: $1000

📊 EDGES FOUND (2 total)

1. Patrick Mahomes Over 0.5 TD vs Giants
   Edge: +18.2% | Tier: STRONG EDGE
   True Probability: 92% | Implied Odds: 77.3%
   Recommendation: $48.00 bet (4.8% of bankroll)
   Kelly Fraction: 19.2% Kelly (conservative)

2. Travis Kelce Over 0.5 TD vs Giants
   Edge: +12.5% | Tier: GOOD EDGE
   True Probability: 78% | Implied Odds: 69.2%
   Recommendation: $25.00 bet (2.5% of bankroll)
   Kelly Fraction: 12.5% Kelly (conservative)

💡 INSIGHTS
- Chiefs have favorable matchup against Giants defense
- Home field advantage adds +10% to probabilities
- Weather conditions are optimal for passing game
- Recent injury reports show no concerns

🔄 REFINEMENT OPTIONS
- "Increase minimum edge to 20%"
- "Switch to v1 model"
- "Show only STRONG edges"
- "Analyze other teams"

Conversational Flow

User: "Find edges for week 7"
System: "Found 15 edges for week 7. 3 STRONG, 8 GOOD, 4 WEAK. Would you like to refine?"

User: "Only show STRONG edges"
System: "Found 3 STRONG edges. All involve QB TD props. Would you like to see details?"

User: "Show me the best one"
System: "Patrick Mahomes Over 0.5 TD vs Giants has 18.2% edge. Recommend $48 bet. Want to analyze further?"

User: "What about the weather?"
System: "Weather is optimal - 72°F, clear skies, 5mph winds. No impact on passing game."

Configuration

See resources/model_configs.json:

{
  "models": {
    "v1": {
      "description": "Simple probability model",
      "parameters": {
        "league_average_tds": 1.5,
        "home_advantage": 0.1
      }
    },
    "v2": {
      "description": "Advanced probability model",
      "parameters": {
        "league_average_tds": 1.5,
        "home_advantage": 0.1,
        "weather_factor": 0.05,
        "injury_factor": 0.15,
        "recent_form_weight": 0.3
      }
    }
  },
  "defaults": {
    "week": "current",
    "model": "v2",
    "min_edge": 0.1,
    "bankroll": 1000,
    "confidence": "medium"
  }
}

Error Handling

Common Issues

• Invalid Week: Check if week is within NFL season
• Team Not Found: Verify team name spelling
• No Data: Check if data exists for specified week
• Model Errors: Verify model parameters

Troubleshooting

# Test parameter parsing
python scripts/conversational_analyzer.py --test-parse "Find edges for week 7"

# Check available data
python scripts/conversational_analyzer.py --list-weeks

# Verify model configuration
python scripts/conversational_analyzer.py --check-models

Performance

• Response Time: <30 seconds per analysis
• Accuracy: Matches CLI results 100%
• Usability: >90% natural language query success
• Token Usage: ~10,100 tokens per analysis

Success Metrics

• Usability: Natural language queries work >90% of time
• Accuracy: Matches CLI results 100%
• Latency: Responses in <30 seconds
• User Satisfaction: Positive qualitative feedback

Future Enhancements

• Machine Learning: Improved parameter parsing
• Context Awareness: Remember previous queries
• Multi-language Support: Spanish, French queries
• Voice Interface: Speech-to-text integration
• Advanced Insights: Weather, injury, trend analysis