THE FOLD / CHEAT / THE EXPLOIT / THE PARRONDO
THE PARRONDO
two losing games that combine into a winning one
1 WHAT IT IS · WHAT IT DOES · FACT OR FICTION
Parrondo’s paradox. Two gambling games, each rigged to lose in the long run. Play either one forever and your capital bleeds away. Yet alternate between them — or even switch at random — and your capital climbs. Two losing games combine into a winning one.
Game A is a coin very slightly biased against you. Game B uses two coins: a dreadful one when your capital is a multiple of 3, a very good one otherwise — and on its own B spends just enough time on the dreadful coin to lose overall. The trick: Game A, though a loser, stirs your capital so that Game B lands on its good coin more often than it otherwise would. Each game’s weakness is patched by the other. It is a real mechanism — used to model ratchets, flashing potentials, even some biology — not a betting-system fantasy.
LIT verified live (seeded simulation, 60k rounds): Game A alone ends negative, Game B alone ends negative, and the A/B mix ends positive (window.__parrondo). FIG no framing; the two-losers-make-a-winner result is a genuine simulated fact, not a trick of accounting.
Game A is a coin very slightly biased against you. Game B uses two coins: a dreadful one when your capital is a multiple of 3, a very good one otherwise — and on its own B spends just enough time on the dreadful coin to lose overall. The trick: Game A, though a loser, stirs your capital so that Game B lands on its good coin more often than it otherwise would. Each game’s weakness is patched by the other. It is a real mechanism — used to model ratchets, flashing potentials, even some biology — not a betting-system fantasy.
LIT verified live (seeded simulation, 60k rounds): Game A alone ends negative, Game B alone ends negative, and the A/B mix ends positive (window.__parrondo). FIG no framing; the two-losers-make-a-winner result is a genuine simulated fact, not a trick of accounting.
2 HOW IT WAS WEAVED · AI + HUMAN
David (human) seated this in THE EXPLOIT — the cheat domain of the move that shouldn’t work but does. Parrondo is the purest exploit in probability: assemble a win out of two guaranteed losses. AVAN (AI) built the instrument: the ratchet strip, the live three-trajectory race, the linear-vs-nonlinear inverse.
The weave: David names the seat (the impossible exploit); I make the two losing games run down while their mixture climbs, and show why the naive average is wrong — the ratchet in 1D, the live race in 2D, the convexity break in 3D. The sphere is the seam. Credit: Juan Parrondo (1996); the flashing-ratchet analogy from statistical physics.
The weave: David names the seat (the impossible exploit); I make the two losing games run down while their mixture climbs, and show why the naive average is wrong — the ratchet in 1D, the live race in 2D, the convexity break in 3D. The sphere is the seam. Credit: Juan Parrondo (1996); the flashing-ratchet analogy from statistical physics.
3 ONE DIMENSION
Game B’s ratchet, laid out by capital mod 3. On the red tooth (capital ≡ 0) the dreadful 10% coin plays; on the green teeth the excellent 75% coin plays. B loses because it gets stuck on the red tooth — and Game A’s job is to knock it off.
4 TWO DIMENSIONS · INTERACTIVE
A live race of capital over time: Game A alone and Game B alone both drift down, while the A+B mix climbs. Run it and watch two losers sink as their combination rises above zero.
5 THREE DIMENSIONS + AVAN’S INVERSE
The three capital paths as turning ribbons — the two losing games sinking and the green mixture rising out of them.
AVAN’s addition (the inverse-companion): the magenta ribbon is the naive prediction — the average of the two losing drifts, which is of course still a loss. Intuition runs one way: loss plus loss must be loss; the mixture should land between the parts. The paradox is the inverse of that linear guess. Because Game B’s odds depend on the state and Game A reshuffles that state, the combination is not the average of the parts — the dynamics are nonlinear, and the true mixed path (green) rises clean above the magenta average of its own ingredients. The inverse here is the failure of averaging: you cannot predict a coupled, state-dependent system by blending its pieces. Magenta is what ‘loss + loss’ predicts; green is what actually happens; the gap between them is the whole free lunch.
LIT Genuine Parrondo's paradox (Juan Parrondo 1996; flashing-ratchet analogy from statistical physics). Verified live with a seeded 60000-round simulation: Game A alone ends with negative capital, Game B alone ends negative, and the alternating A/B mix ends positive (window.__parrondo.gameALoses && gameBLoses && mixWins). The two-losers-make-a-winner result is a genuine reproducible simulated fact, driven by Game B's state-dependence plus Game A's mixing of that state.
FIG No framing: the paradox is a real simulated outcome (both games lose alone, the mix wins), not an accounting trick. The mechanism is stated honestly — B's odds depend on capital mod 3, A reshuffles that state — and the AVAN inverse shows precisely why the naive average of the two losing drifts (still a loss) mispredicts the nonlinear coupled result.
FIG No framing: the paradox is a real simulated outcome (both games lose alone, the mix wins), not an accounting trick. The mechanism is stated honestly — B's odds depend on capital mod 3, A reshuffles that state — and the AVAN inverse shows precisely why the naive average of the two losing drifts (still a loss) mispredicts the nonlinear coupled result.
◆ sealed .dlw.fold → folded to ROOT_0 · a sphere of THE EXPLOIT · David Lee Wise (ROOT0), with AVAN