🎰 Python Slot Machine

A fully custom terminal-based slot machine built in Python featuring weighted reels, hold mechanics, payout logic, and statistical simulation tools.

---

Features

• 🎲 Weighted Reel System
  Customisable reel weights allow realistic probability balancing and RTP tuning.

• 🔒 Hold Mechanic
  Players can hold reels strategically between spins, with anti-abuse logic to prevent infinite hold exploitation.

• 💰 Dynamic Payout System
  Symbol-based payout table supporting:

  • Low-tier fruit payouts for 2+ matches
  • High-tier jackpot symbols requiring 3 matches

• 📊 Monte Carlo Analysis Tool
  Includes simulation engine to analyse:

  • RTP (Return To Player)
  • Win Rate / Hit Rate
  • Symbol Frequency Distribution
  • RTP Contribution by Win Type

• 🌐 Flask API Backend Experimental backend endpoints for integrating slot spins into web applications.

---

Project Structure

slot-machine/
│
├── slots.py        # Main slot machine gameplay logic
├── settings.py     # Configurable weights, symbols and payout tables
├── analysis.py     # Statistical simulation / RTP analysis tool
├── app.py          # Flask API backend prototype
└── screenshots/    # Gameplay + analysis screenshots

---

Recent Updates

Codebase Improvements

• Refactored slots.py to improve readability and separate gameplay logic more cleanly.
• Improved hold-state management and win-condition tracking.
• Added better structured helper methods for payout / win checking.

Probability Balancing

• Reworked reel weighting values in settings.py for more realistic slot odds and smoother RTP balancing.

Statistical Validation & Expected Value Analysis

• In addition to Monte Carlo simulation, this project now includes a formal statistical validation layer to evaluate whether the slot machine behaves consistently with its intended probability model.

• 🎯 Expected Value (EV) Analysis

• The Expected Value represents the theoretical average profit or loss per spin.

• This allows us to measure whether the slot machine is profitable or loss-making in the long run.

• EV < 0 → House edge (expected in real casino systems)

• EV = 0 → Fair game (theoretical balance point)

• EV > 0 → Player advantage (model imbalance or exploit risk)

• 📊 RTP vs EV Relationship

• Return To Player (RTP) and Expected Value are directly linked:

• EV=RTP−1

• This means:

• RTP expresses return as a ratio

• EV expresses return as net profit per spin

• Both are used to validate whether the simulation matches the intended payout design.

📉 Z-Test RTP Validation

• A one-sample z-test is used to compare simulated RTP against the expected theoretical RTP.

• This checks whether observed differences are statistically significant or just random variation.

• Null hypothesis: observed RTP = expected RTP

• Alternative hypothesis: RTP deviates significantly

• Output includes:

• Z-score

• 95% confidence interval

• Statistical conclusion (pass/fail consistency check)

---

Screenshots

Gameplay Example

Statistical Analysis Output

---

🛠️ Future Improvements

• Link in to my more solid random number generator project
• 🔊 Sound effects
• 🎰 Animated/timed reel-strip implementation
• Add endpoints for holding reels to the api version ( /hold )
• Add player sessions (currently only a single shared game state)
• Build a frontend (web UI) to interact with the API

---

📜 License

This project is open source and available under the MIT License.

---

🙌 Acknowledgements

Built as a learning project to explore:

• Python OOP design
• Randomness and probability
• Game logic implementation
• Z-Test and T-Test pracice