The Martingale, doubling after losses, is the betting system everyone can name, and our piece on the Martingale in blackjack takes it apart in detail. But the Martingale is only the loudest member of a large family. D'Alembert, Labouchere, Fibonacci, Oscar's Grind and the Paroli all promise roughly the same thing: a schedule of bet sizes that turns a chaotic game into an orderly one.
This article covers the rest of the family, and it will save you the suspense: they differ in how the ride feels and agree in where it ends. The definitions below follow Wizard of Odds, the gambling math site, and so do the conclusions. Nothing here is staking advice, and every unit mentioned is a virtual chip.
The family, in one table
| System | How it moves the bet |
|---|---|
| D'Alembert | Raise the bet one unit after a loss, lower it one unit after a win |
| Labouchere | Write a line of numbers; each bet is the sum of the first and last, a win crosses both off, a loss adds the amount lost to the end; the session ends when the line is gone |
| Fibonacci | Bets follow the sequence 1, 1, 2, 3, 5, 8 and so on, starting on the second 1; a loss moves one place up, a win moves back two places, and falling off the start ends the session |
| Oscar's Grind | Flat one unit bets, repeated after a loss or a tie, raised by one unit after a win, aiming for a small fixed profit per series |
| Anti-Martingale, including the Paroli | Double after each win until a goal is reached, and quit on a loss |
Wizard of Odds describes D'Alembert as less aggressive than the Martingale, Labouchere and Fibonacci, giving more time at the table and less volatility. Oscar's Grind aims low, grinding toward a small fixed profit per series; Wizard of Odds' simulations show it reaching that small goal in most sessions, while the expected result per session stays negative. The Paroli sits at the other pole, and our Martingale article describes its shape exactly: raising bets after wins instead of losses produces many small losses and rare large wins.
Notice what all five have in common. Every rule in every system looks at the results of past hands to size the next bet. Hold that thought, because it is the heart of the case against the family.
A short Labouchere run, in units
It is easier to see what a progression actually does in a short worked example than in a definition. Here is a Labouchere session in virtual units, using a starting line of 1 1 1 1. Each bet is the sum of the first and last numbers in the line.
| Line before the bet | Bet | Result | Running total |
|---|---|---|---|
| 1 1 1 1 | 1 + 1 = 2 | Loss | -2 |
| 1 1 1 1 2 | 1 + 2 = 3 | Loss | -5 |
| 1 1 1 1 2 3 | 1 + 3 = 4 | Win | -1 |
| 1 1 1 2 | 1 + 2 = 3 | Win | +2 |
| 1 1 | 1 + 1 = 2 | Win | +4 |
Check the arithmetic: minus 2, minus 3, plus 4, plus 3, plus 2 is plus 4, the line is gone, and the session ends four units ahead on two losses and three wins. Two features are worth noticing. The two early losses forced the bets upward, so the wins landed at larger sizes, which is the entire mechanism. And the session wagered 14 units in total (2 plus 3 plus 4 plus 3 plus 2) to finish plus 4. Had the losses kept coming instead, each one would have joined the end of the line and pushed the next bet higher, and the line would have grown instead of shrinking. The schedule did not change the game; it chose which hands got the big bets.
Why none of them changes the expectation
The core fact is that each blackjack hand is independent of the results of the hands before it. The cards have no memory of your losses, a point our article on streaks and what they prove develops in full: past results, the premise of both the due-for-a-win fallacy and the ride-the-hot-streak version, have no causal reach into the next deal. Because the next hand does not know what you wagered, the size of the next bet cannot change that hand's expectation per unit wagered.
What a progression changes is how much is wagered, and when. And expected loss equals the total amount wagered multiplied by the house edge, the identity our guides to the house edge and expected value unpack. A system that raises the total wagered therefore raises the expected loss. Wizard of Odds states the general conclusion plainly: no betting system can turn a game with a house edge into a profitable one. Systems can raise the chance of a winning session, but only at the cost of larger losses in losing sessions, and over the long run the ratio of money lost to money bet approaches the house edge. Every system in the table above is a different way of rearranging that sentence.
Where blackjack breaks the bookkeeping
All of these systems assume simple even money bets, and blackjack is not one. Commonly published figures put the push rate at roughly 8 to 9 percent of hands in standard six deck games where the dealer stands on soft 17, so a D'Alembert or Labouchere sequence repeatedly stalls on a tie, a result its win and loss rules never mention. Doubles and splits add money in the middle of a hand, so a single round can move several times the nominal bet and wreck a careful ladder. The 3:2 payout on a natural is a windfall the sequence never budgeted for. Our Martingale article covers these wrinkles for the Martingale; they apply across the whole family tree.
The honest summary is that blackjack makes progressions messier without making them better. The mess is real: any system whose appeal rests on tidy bookkeeping will not survive contact with a game that pushes roughly one hand in twelve and asks you to double an 11 now and then.
The bet variation that does change expectation
There is exactly one way to make bet size carry information in blackjack, and it has nothing to do with past wins and losses. If the remaining cards are rich in tens and aces, the next hand's expectation itself is different, and a bet sized to that composition, the bet spread a counter uses, puts more money on the hands whose expectation is better. That is card counting, and the difference is that the shoe has a memory while the scoreboard does not. Three caveats belong here: counting with your head is legal in the United States, casinos are private businesses that may refuse service, and no technique removes the risk from the game.
Watching a progression safely
None of the above makes progressions uninteresting. As objects of study they are genuinely instructive: each one is a different answer to the question of how to distribute bet sizes across independent negative-expectation hands, and watching a system's session results pile up can teach more about variance than a description. The Pro upgrade's bankroll simulator runs on virtual chips, so you can bet a progression hand by hand across several sessions and watch its session results spread out. It is a way to meet the arithmetic of the family that risks nothing.
Frequently asked questions
Do betting systems work in blackjack?
No. Each hand is independent of the results of the hands before it, so the size of the next bet cannot change that hand's expectation per unit wagered. A progression changes how much is wagered and when, which reshapes the distribution of session results without changing the expected loss per unit wagered. Over the long run, the ratio of money lost to money bet approaches the house edge.
What is the difference between the Martingale and the Paroli?
The Martingale raises the bet after a loss, chasing a recovery; the Paroli, part of the anti-Martingale family, doubles after a win and quits on a loss. They produce opposite shapes of results: many small wins and rare large losses for the Martingale, many small losses and rare large wins for the Paroli. Neither shape changes the expectation.
Is the D'Alembert system safer than the Martingale?
It is gentler, but no better at beating the house. Raising the bet one unit after a loss instead of doubling keeps the ladder shallow, so the system gives more time at the table and less volatility, as Wizard of Odds describes. The same source's conclusion still applies: no betting system can overcome the house edge, or even dent it.
How does the Labouchere betting system work?
Also called the cancellation system, it starts with a written line of numbers. Each bet is the sum of the first and last numbers in the line: a win crosses both off, and a loss adds the amount lost to the end. The session ends when the line is gone. A short run can finish with a small profit, but losing streaks grow the line, and the bets along with it.
Can you try betting systems without risking money?
Yes. The Pro upgrade's bankroll simulator runs on virtual chips, so you can bet a progression hand by hand and watch its session results spread out with nothing at stake. Over enough sessions, it shows the shape of a system's results, many modest sessions one way and occasional large ones the other, without any real money being risked.