Parrondo's Paradox
Simulate Parrondo's paradox — discover how two losing coin-tossing games combine to produce a winning outcome. Interactive Monte Carlo simulation with 5 strategies and charts.
About This Calculator
Parrondo's paradox is a counterintuitive phenomenon in probability and game theory: two games that are each guaranteed to lose money in the long run can be combined into a winning game by alternating between them. Discovered by Spanish physicist Juan Parrondo in 1996 while studying the Brownian ratchet, this paradox challenges our intuition about probability and has found applications in economics, biology, and population dynamics.
The classic formulation involves two coin-tossing games. Game A uses a slightly biased coin — you win $1 with probability 0.495 (with epsilon = 0.005) and lose $1 with probability 0.505. Game B is more complex: if your current capital is a multiple of 3, you play a coin with P(win) = 0.095 (heavily biased against you); otherwise you play a coin with P(win) = 0.745 (favorable). Individually, both games have negative expected value. However, when alternated in the right sequence (e.g., A-B-A-B or A-A-B-B), the combined game can have positive expected value.
This calculator runs a Monte Carlo simulation across thousands of trials, each playing up to hundreds of games. You can choose from five strategies: Only A, Only B, alternating ABAB, AABB, or Random selection. The simulation computes the average final capital, win rate, and expected value per game, and charts the capital trajectory over time. A Strategy Comparison chart lets you see how each approach performs side-by-side. Adjust epsilon to explore how the bias affects the paradox — smaller values make the effect more subtle, while larger values eventually break the paradox entirely.
Frequently Asked Questions
What is Parrondo's paradox?
Parrondo's paradox is a counterintuitive phenomenon in game theory where two individually losing games can be combined in alternating sequence to produce a winning outcome. Discovered by Spanish physicist Juan Parrondo, the paradox demonstrates how randomness and state-dependent rules can create unexpected results. It has applications in economics, biology, and population dynamics.
How does the coin-tossing simulation work?
The simulation runs thousands of Monte Carlo trials. In each trial, you start at $0 and play games based on your chosen strategy. Game A uses a coin with P(win) = 0.5 minus epsilon. Game B uses two coins: if your capital is a multiple of 3, P(win) = 0.1 minus epsilon; otherwise P(win) = 0.75 minus epsilon. With epsilon = 0.005, both games are individually losing, but alternating them produces a winning strategy.
What do the results mean?
Average Final Capital shows the mean profit or loss across all trials. Win Rate is the fraction of trials that ended with positive capital. Average per Game shows the expected value per coin flip. The Capital Trajectory chart plots the average capital after each game. The Strategy Comparison chart lets you compare all five strategies side by side.
What is the role of epsilon?
Epsilon (ε) controls the bias of the coin flips. Standard values are around 0.005. Game A win probability is 0.5 - ε, Game B1 win probability is 0.1 - ε, and Game B2 win probability is 0.75 - ε. A smaller epsilon makes all games closer to fair, while a larger epsilon increases the losing bias. The paradox typically holds for epsilon values up to about 0.01.
Can Parrondo's paradox be applied to real gambling?
No. Parrondo's paradox requires a specific mathematical relationship between the games — the outcome of one game affects the state (modulo 3 of capital) that determines the odds of the other. In real casino games, such coupling does not exist, and no strategy can turn two negative-expectation games into a positive one. The paradox is a mathematical curiosity, not a gambling system.
What strategies can I try in this simulator?
Five strategies are available: Only Game A, Only Game B, alternating A-B-A-B pattern, A-A-B-B pattern, and Random selection. Only A and Only B should both lose in the long run. The alternating patterns (ABAB and AABB) and Random strategies can produce a winning outcome, demonstrating the paradox. Try them all to see which performs best.
Who discovered Parrondo's paradox?
Parrondo's paradox was discovered by Spanish physicist Juan Parrondo in 1996 while studying the Brownian ratchet, a thought experiment in thermodynamics. The paradox was later popularized by Derek Abbott and colleagues at the University of Adelaide, who published the coin-tossing formulation that this calculator simulates.
Can the paradox be explained using Markov chains?
Yes. Each game can be modeled as a Markov chain with three states (capital modulo 3). The steady-state distribution shows how often the system is in each capital state. For Game B alone, the probability of being in state 0 (where you play the losing B1 coin) is about 38.4%. By alternating with Game A, the system spends more time in favorable states where B2's high win probability dominates, making the combined game profitable.