Welcome to the DiceTrader, a Python game that combines dice rolling, betting, and advanced risk management strategies. This project is designed to be both fun and educational, with a focus on simulating market-like dynamics in an interactive way.
 |
 |
 |
|
- Overview
- Key Features
- Game Rules
- Probability and Payouts
- Betting Strategies
- AI Integration
- Advanced Features
- How to Play
- Installation
- Project Structure
- Contributing
- License
The DiceTrader is a Python-based game where players bet on the outcome of rolling two six-sided dice. The game includes:
- Betting Mechanics: Players can bet on the sum of two dice (2 to 12)
- Probability Calculations: The game displays the probability of each sum
- Payouts: Higher payouts for less likely sums, lower payouts for more likely sums
- Risk Management: Multiple betting strategies to manage bankroll
- Market Simulation: Dynamic trends and volatility that affect probabilities
- AI Integration: Real-time strategy recommendations from AI advisor
- Bet on dice sums (2-12) with dynamic payouts
- Multiple rounds with bankroll management
- Real-time probability calculations
- Dynamic market trends (bull/bear)
- Market volatility affecting outcomes
- Simulated financial news events
- Personalized betting recommendations
- Strategy suggestions based on market conditions
- Reinforcement learning for optimal decisions
- Powered by Gemini 2.0 via OpenRouter API
- Risk analysis and expected returns
- Portfolio diversification
- Performance tracking with key metrics
- Visual representations of betting history
- Sum profitability analysis
- Starting Money: The player starts with $100
- Betting: Choose a sum (2-12) and wager amount
- Dice Roll: The game rolls two dice and calculates the sum
- Winning: If your bet matches the dice sum, win bet ร payout
- Losing: If your bet doesn't match, lose your wager
- Game Over: Continue until you run out of money or quit
| Sum | Ways to Roll | Probability | Payout |
|---|---|---|---|
| 2 | 1 | 2.78% | 36x |
| 3 | 2 | 5.56% | 18x |
| 4 | 3 | 8.33% | 12x |
| 5 | 4 | 11.11% | 9x |
| 6 | 5 | 13.89% | 7x |
| 7 | 6 | 16.67% | 6x |
| 8 | 5 | 13.89% | 7x |
| 9 | 4 | 11.11% | 9x |
| 10 | 3 | 8.33% | 12x |
| 11 | 2 | 5.56% | 18x |
| 12 | 1 | 2.78% | 36x |
- Risk Level: Moderate
- Approach: Adjusts stake based on bankroll and previous results
- Best For: Balanced play with moderate risk
- Risk Level: High
- Approach: Doubles stake after each loss to recover previous losses
- Best For: Players with large bankrolls willing to take risks
- Risk Level: High
- Approach: Uses Fibonacci sequence to determine stake after losses
- Best For: Progressive recovery with somewhat controlled risk
- Risk Level: Moderate
- Approach: Increases stake by 1 unit after loss, decreases by 1 after win
- Best For: More conservative alternative to Martingale
- Risk Level: Low
- Approach: Bets a fixed percentage of current bankroll
- Best For: Bankroll preservation and steady play
- Risk Level: Moderate
- Approach: Calculates optimal bet size based on odds and probability
- Best For: Mathematically optimal long-term growth
- Risk Level: Low
- Approach: Always bets the same amount
- Best For: Beginners and conservative players
- Personalized Recommendations: Suggests optimal sums to bet on
- Strategy Selection: Recommends the most appropriate betting strategy
- Reasoning Explanation: Provides clear rationale for its recommendations
- Adaptive Learning: Improves over time through reinforcement learning
- Bankroll Level: Categorized as low/medium/high
- Market Trend: Current trend (bull/bear)
- Streak Analysis: Winning/losing streak tracking
- Win Rate: Recent win percentage
- Q-Learning: Updates Q-values after each bet
- Reward System: Positive/negative rewards tied to actual profit/loss
- Exploration: Epsilon-greedy strategy for balanced learning
- Persistence: Saves learned knowledge between sessions
- Learning Rate (ฮฑ): 0.1 - Balances new vs existing knowledge
- Discount Factor (ฮณ): 0.95 - Values future rewards
- Exploration Rate (ฮต): 1.0 - Ensures sufficient exploration
- API Integration: Seamless connection to OpenRouter for Gemini 2.0 access
- Fallback Mechanism: Local prediction algorithms when API is unavailable
- State Representation: Sophisticated state encoding for RL model
- Performance Tracking: Evaluation of AI recommendation accuracy
- Data Persistence: Automatic saving/loading of Q-values
- Statistical Tracking: Win/loss counts, total profit, prediction accuracy
- Analyzes betting history and performance
- Provides customized strategy recommendations
- Uses reinforcement learning models
- Dynamic probability adjustments
- Time-series models for market trends
- Bayesian probability models
- Real-time performance tracking
- Risk-adjusted return calculations
- Interactive visualizations
-
Run the Game:
python game.py
-
Place Your Bet:
- Choose a sum (2-12)
- Enter your wager amount
-
Roll the Dice: The game will roll and display the result
-
Win or Lose:
- Match: Win payout ร bet
- No match: Lose your bet
-
Continue or Quit: Decide whether to play again or quit
Here's an example of what a game session might look like:
python game.py
Welcome to the DiceTrader!
You start with $100. Place your bets on the sum of two dice!
Current market trend: BULL
๐ฐ Market optimism rises as investors flock to higher sums!
Type 'help' at any time to see available commands.
AI model loaded from /Users/dicetrader/data/q_values.json
Available strategies:
1. masaniello - Adjusts stake based on bankroll
2. martingale - Doubles stake after each loss
3. fibonacci - Uses Fibonacci sequence for stakes
4. d'alembert - Increases stake by 1 unit after loss, decreases by 1 after win
5. percentage - Bets a fixed percentage of your bankroll
6. kelly - Uses Kelly Criterion for optimal bet sizing
7. fixed - Uses a fixed stake amount
Choose a strategy (masaniello, martingale, fibonacci, dalembert, percentage, kelly, fixed): kelly
Using KELLY strategy with AI advisor
Current balance: $100.00
Current market trend: BULL
Here are the possible sums and their payouts:
Sum 2: Payout 36x (Probability: 2.15%)
Sum 3: Payout 18x (Probability: 4.30%)
Sum 4: Payout 12x (Probability: 6.45%)
Sum 5: Payout 9x (Probability: 8.60%)
Sum 6: Payout 7x (Probability: 10.75%)
Sum 7: Payout 6x (Probability: 19.35%)
Sum 8: Payout 7x (Probability: 16.13%)
Sum 9: Payout 9x (Probability: 12.90%)
Sum 10: Payout 12x (Probability: 9.68%)
Sum 11: Payout 18x (Probability: 6.45%)
Sum 12: Payout 36x (Probability: 3.23%)
Commands:
- Type a number (2-12) to bet on a single sum
- Type 'portfolio' to manage your bet portfolio
- Type 'stats' to view your performance analytics
- Type 'strategy' to change your betting strategy
- Type 'help' to see all available commands
Attempting to use DeepSeek API...
DeepSeek API response received successfully
Successfully used DeepSeek API
๐ง AI Strategy Advisor:
Recommended bet: Sum 7
Current KELLY strategy stake: $1.61
AI recommends KELLY strategy with stake: $1.61
Reasoning: The sum 7 has the highest probability of occurring at 19.35%, making it the most statistically favorable bet. The Kelly Criterion is recommended because it optimizes bet sizing based on the probability of winning and the current bankroll, ensuring long-term growth while minimizing risk. This is particularly suitable given the BULL market trend, which suggests a favorable environment for growth-oriented strategies.
Enter your choice (2-12 for bet, or command): 7
Betting options for sum 7:
1. Use your KELLY strategy stake: $1.61
2. Use AI recommended stake: $1.61 (AI also recommends this sum)
3. Enter a custom stake amount
Choose your stake option (1-3): 2
Using AI recommended stake: $1.61
The dice rolled a 7!
Congratulations! You won $9.68!
Winning bets:
Sum 7: $9.68
AI model saved to /Users/user/dicetrader/data/q_values.json
Do you want to play again? (yes/no): yes
Current balance: $109.68
Current market trend: BULL
Here are the possible sums and their payouts:
Sum 2: Payout 36x (Probability: 2.15%)
Sum 3: Payout 18x (Probability: 4.30%)
Sum 4: Payout 12x (Probability: 6.45%)
Sum 5: Payout 9x (Probability: 8.60%)
Sum 6: Payout 7x (Probability: 10.75%)
Sum 7: Payout 6x (Probability: 19.35%)
Sum 8: Payout 7x (Probability: 16.13%)
Sum 9: Payout 9x (Probability: 12.90%)
Sum 10: Payout 12x (Probability: 9.68%)
Sum 11: Payout 18x (Probability: 6.45%)
Sum 12: Payout 36x (Probability: 3.23%)
Commands:
- Type a number (2-12) to bet on a single sum
- Type 'portfolio' to manage your bet portfolio
- Type 'stats' to view your performance analytics
- Type 'strategy' to change your betting strategy
- Type 'help' to see all available commands
Attempting to use DeepSeek API...
DeepSeek API response received successfully
Successfully used DeepSeek API
๐ง AI Strategy Advisor:
Recommended bet: Sum 7
Current KELLY strategy stake: $1.77
AI recommends KELLY strategy with stake: $1.77
Reasoning: The sum 7 has the highest probability (19.35%) of occurring, making it the most statistically favorable bet. The Kelly Criterion is recommended because it optimizes bet sizing based on the probability of winning and the current bankroll, ensuring long-term growth while minimizing risk. Given the 100% win rate and BULL market trend, the Kelly Criterion will help capitalize on the favorable conditions while managing risk effectively.
Enter your choice (2-12 for bet, or command): stats
๐ PERFORMANCE REPORT ๐
==============================
Total Rounds: 1
Win Rate: 100.00%
Average Win: $1.61
Average Loss: $0.00
Starting Bankroll: $100.00
Current Bankroll: $109.68
Profit/Loss: $9.68
Sharpe Ratio: 0.10
Maximum Drawdown: 0.00%
Most Profitable Sums:
Sum 7: $1.61
Least Profitable Sums:
Sum 7: $1.61
Analytics data and plots have been saved:
Data: /Users/user/Documents/dicetrader/analytics/analytics_data_20250318_214648.json
Plot: /Users/user/Documents/dicetrader/analytics/bankroll_20250318_214647.png
Plot: /Users/user/Documents/dicetrader/analytics/win_loss_20250318_214647.png
Plot: /Users/user/Documents/dicetrader/analytics/trend_returns_20250318_214647.png
Press Enter to continue...-
Clone the repository:
git clone https://github.com/FaustoS88/dicetrader.git
-
Navigate to the project directory:
cd dicetrader -
Install dependencies:
pip install -r requirements.txt
-
Set up API access (optional):
- Create a .env file in the project root
- Add your OpenRouter API key: OPENROUTER_API_KEY=your_api_key_here
- Add yout Deepseek API key: DEEPSEEK_KEY=your_deepseek_api_key
dicetrader/
โ
โโโ game.py # Main game script
โโโ README.md # Project documentation
โโโ LICENSE # License file
โโโ requirements.txt # Dependencies
โ
โโโ strategies/ # Betting strategy implementations
โ โโโ masaniello.py # Masaniello system
โ โโโ martingale.py # Martingale system
โ โโโ fibonacci.py # Fibonacci system
โ โโโ dalembert.py # D'Alembert system
โ โโโ percentage.py # Percentage betting
โ โโโ kelly.py # Kelly Criterion
โ โโโ fixed.py # Fixed stake
โ โโโ ai_advisor.py # AI strategy advisor
โ
โโโ ai_services/ # AI integration
โ โโโ openrouter_client.py # OpenRouter API client
โ
โโโ RL_Agent/ # Reinforcement learning
โ โโโ RL_agent.py # RL implementation
โ
โโโ analytics/ # Analytics data storage
โ โโโ *.json # Saved game analytics
โ
โโโ analytics_dashboard.py # Analytics visualization
Contributions are welcome! Follow these steps:
- Fork the repository
- Create a new branch for your feature/bug fix
- Commit your changes
- Submit a pull request
This project is licensed under the MIT License. See the LICENSE file for details.
Have fun!