Simulate Parrondo's paradox: two losing games that produce a winning strategy when combined. Adjust probabilities and see trajectory results.
Parrondo's Paradox demonstrates one of the most counterintuitive results in probability theory: two individually losing games can produce a winning outcome when played in alternation. This calculator lets you configure both games, set a mixing pattern, and run Monte Carlo simulations to observe the paradox in action.
Game A is a simple biased coin flip with a probability slightly below 50%. Game B is capital-dependent — the win probability changes based on whether your current capital is divisible by a modulus (typically 3). Both games are calibrated to lose money on average when played alone, but their alternating combination exploits the modular structure to produce net positive returns.
The simulation runs thousands of trials to estimate average profits, win rates, and capital trajectories for each strategy. The trajectory visualization and comparison table make it easy to see exactly where the paradox emerges and how strongly the mixed strategy outperforms the individual games.
Parrondo's paradox is a fascinating demonstration of how complex interactions between simple systems can produce unexpected emergent behavior. This calculator makes the paradox tangible by letting you adjust parameters and immediately see results through simulation, rather than working through abstract mathematical proofs.
It's an excellent educational tool for probability courses, game theory classes, and anyone interested in counterintuitive mathematics. The simulation approach also teaches Monte Carlo methods — a widely used technique in finance, physics, and engineering.
Game A: P(win) = p_A. Game B: P(win) = p_bad if capital mod M = 0, else p_good. Expected value per round: E[A] = 2p_A - 1. E[B] ≈ (1/M)×(2p_bad - 1) + ((M-1)/M)×(2p_good - 1). Paradox: E[A] < 0 and E[B] < 0, but E[Mixed] > 0.
Result: Game A loses ~-1.0, Game B loses ~-1.5, Mixed AABB wins ~+2.5
With the classic parameters, Game A (p=0.495) and Game B (mod 3, good=0.745, bad=0.095) both lose money individually. But the AABB alternation produces positive expected returns because Game A disrupts the modular pattern that Game B exploits against the player.
Parrondo's paradox, discovered by Spanish physicist Juan Parrondo in 1996, shows that combining two losing strategies can create a winning one. Game A is a simple coin flip with probability p < 0.5 (losing). Game B has two states: when capital is divisible by M, the win probability is very low (p_bad); otherwise it's high (p_good). Both games are calibrated so their long-run expected value is negative.
The paradox arises because Game A disrupts the modular pattern that makes Game B lose. When played alone, Game B's capital tends to land on multiples of M (the bad state) more often than the 1/M stationary probability would suggest. Interspersing Game A rounds shifts capital away from these trap states, allowing Game B to access its high probability more often.
Parrondo's paradox has deep connections to Brownian ratchets in physics — mechanisms that extract directed motion from random fluctuations by periodically switching between asymmetric potential energy landscapes. Just as two potentials that individually don't produce net flow can drive directed transport when alternated, two losing games can produce net profit when combined.
In evolutionary biology, the paradox models how organisms switching between two individually unfavorable environments can thrive. In portfolio theory, it helps explain why periodic rebalancing between losing assets can sometimes improve returns. In engineering, it informs the design of micro-scale transport devices and genetic algorithms. Understanding when and why mixing strategies beats individual execution is a profound insight with wide-ranging implications.
The key is that Game B's probability depends on capital state (modular arithmetic). Playing Game A between rounds of Game B shifts the capital away from the "bad" states that Game B exploits. This interaction creates a ratchet-like effect that overall produces positive expected value.
The modulus determines when Game B uses its bad (low) probability. If set to 3, Game B uses the bad probability whenever capital is divisible by 3. The modular structure is what creates the paradox — without it, mixing wouldn't help.
Parrondo's paradox is a mathematical curiosity that works with very specific probability structures. Real casino games don't have the capital-dependent probability structure needed. The paradox is more relevant to physics, biology, and finance than to gambling.
The classic pattern is AABB (two A rounds, then two B rounds). Different patterns produce different results. Try AB, AABB, ABBB, or other combinations to see which maximizes profit for your parameters.
The simulation uses random number generation (Monte Carlo method). With more runs (e.g., 5000), the results converge and become more stable. The variance decreases as the square root of the number of runs.
Analogous effects appear in evolutionary biology (flashing ratchets), investment portfolio rebalancing, gene regulation, population genetics, and Brownian motor physics. The mathematical structure has applications far beyond gambling.