🎲 Talent vs. Luck — A Simulation-Based Graduation Thesis

Author: Kento Morita
Supervisor: Takafumi Koshinaka
Department: Data Science
Year: 2023 (Reiwa 5)

---

📘 Overview

This research explores a timeless question: Is success driven more by talent, effort, or sheer luck?

Using multi-agent simulation and reinforcement learning (Q-learning), this project builds upon the Talent vs. Luck (TvL) model proposed by Pluchino et al., extending it with agents capable of learning from their environment to test whether effort can consistently overcome randomness.

---

🔬 Methodology

• Multi-Agent Simulation: Agents move through a grid-based environment encountering lucky and unlucky events.
• Reinforcement Learning: Q-learning is used to simulate agent learning and adaptive behavior.
• Metrics: Success is measured by accumulated virtual assets over simulated lifetimes.
• Variables: Agents are initialized with varied talents and learning abilities (effort proxy).

---

💡 Key Findings

• Luck remains the dominant factor in determining extreme success—even when learning is introduced.
• Neither high talent nor strong learning rate alone guarantees wealth or top performance.
• Learning (effort) can act as a multiplier but cannot substitute for opportunity.
• Emergent wealth distributions closely mimic real-world inequality (Pareto distributions).

---

🧠 Implications

This study challenges the meritocratic assumption that success is purely earned. It shows how randomness, often overlooked, plays a powerful role in shaping careers and outcomes—and suggests the need for more nuanced views of fairness and reward.

---

🛠️ Tech Stack

• Python (OOP)
• NumPy, Matplotlib
• Q-Learning (Reinforcement Learning)
• Simulation & Modeling

---

📈 Future Work

• Introduce agent-to-agent interactions (multi-agent RL)
• Apply network-based environments (e.g., social graphs)
• Explore causality between effort, access, and opportunity
• Investigate policy implications in education and economics

---

📚 References

• Pluchino et al. Talent vs Luck: The Role of Randomness in Success and Failure
• N.N. Taleb, Fooled by Randomness
• Additional references included in the thesis

---

“Success is never as fair as we think. But what if we could simulate fairness itself?”