Paroli

The Paroli betting system is a positive progression: the player increases the next wager after a win and returns to a base stake after a loss or after reaching a predetermined win cap.

It is often described as a reverse Martingale because a Martingale increases stakes after losses, while Paroli increases stakes after wins.

The usual attraction is behavioral rather than mathematical. A capped Paroli sequence can prevent the runaway loss escalation of a loss-chasing system. It can also turn a short winning streak into a larger session gain. What it cannot do is change the probability, payout, house edge, or expected value of the underlying wager.

A basic Paroli sequence

A common version uses a three-win cap and doubles after each win.

With a $10 base unit:

If all three wagers win at even money, the sequence earns:

$10 + $20 + $40 = $70

If the sequence loses at any step, its net result is -$10 under this exact structure.

Why? Consider the second-step loss:

Step 1 win  = +$10
Step 2 loss = -$20
Net         = -$10

Or a third-step loss:

Step 1 win  = +$10
Step 2 win  = +$20
Step 3 loss = -$40
Net         = -$10

The progression therefore creates a distinctive payoff pattern: many failed cycles lose one base unit, while a completed three-win cycle wins seven base units.

The three-step cycle in probability form

Let:

• (b) = base stake;
• (p) = probability that one qualifying wager wins;
• the sequence require three consecutive wins;
• winning wagers pay 1:1;
• outcomes be independent for the purpose of this example.

The probability of completing the cycle is:

[ P(\text{three wins})=p^3 ]

The cycle earns (+7b) if all three bets win and loses (-b) otherwise.

So expected value is:

[ EV=(p^3)(7b)+(1-p^3)(-b) ]

which simplifies to:

[ EV=b(8p^3-1) ]

On a single-zero roulette even-money wager:

[ p=\frac{18}{37} ]

With a $10 base unit:

[ EV=10\left[8\left(\frac{18}{37}\right)^3-1\right]\approx-$0.79 ]

That does not mean every Paroli cycle loses 79 cents. A completed cycle can finish at +$70, while any failed cycle finishes at -$10. The -$0.79 is the long-run average per cycle under the stated assumptions.

The negative expectation remains because roulette pays even money even though 19 of 37 outcomes lose an ordinary red/black, odd/even, or high/low wager.

Paroli changes the path, not the price

This is the most important distinction.

A betting progression can change:

• how quickly stakes rise;
• how much of a winning run is re-risked;
• how often a cycle ends with a small loss;
• how large a successful cycle looks;
• the distribution of session outcomes;
• the emotional experience of play.

It does not, by itself, change:

• the probability of the next roulette spin;
• the casino payout schedule;
• the house edge;
• the expected value per dollar wagered.

If a game has a negative expected value, changing the order of bet sizes does not convert it into positive expectation.

For the wheel-specific version, including roulette zero rules and sequence examples, see Reverse Martingale / Paroli System. This glossary entry owns the definition and general math; the roulette page owns the game-specific strategy discussion.

“House money” is still the player’s money

Paroli often feels comfortable because the larger wagers follow wins. Players may say that the $20 or $40 pressed stake is “casino money.”

That is mental accounting, not financial accounting.

Once a winning payout belongs to the player, risking it again is a new decision. A $40 chip won five minutes ago has the same cash value as a $40 chip brought from home.

This distinction matters because pressing can quietly increase total action. Suppose two players each start with a $10 bet:

• Flat bettor: three $10 wagers = $30 total action.
• Paroli bettor who wins the first two: $10 + $20 + $40 = $70 total action.

If the same house-edge rate applies to every dollar, the second path exposes more money to that edge during the winning run.

Expected loss can be expressed as:

[ \text{Expected Loss}=\text{Total Action}\times\text{House Edge} ]

Paroli does not necessarily produce more action in every session because a first-step loss immediately ends a cycle. The point is narrower: when stakes are pressed, the additional money is real exposure even if it came from earlier wins.

Research on the “house-money effect” helps explain why prior gains can make later risk feel easier to accept. The classic real-money experiments reported that people sometimes became more willing to take risk after a prior gain. That behavioral result explains the appeal of pressing wins; it does not create a mathematical advantage.

Paroli is not the same as chasing losses

This is where Paroli deserves a fair distinction from the Martingale system.

A Martingale typically doubles after losses. The required stake can grow exponentially during a cold run:

$10, $20, $40, $80, $160, $320 ...

Paroli normally resets after a loss. With a strict three-step cap, one failed cycle loses only the base unit.

That can make Paroli easier to control than an uncapped negative progression. But “easier to control” is not the same as “profitable.”

The comparison is about risk shape, not about which system beats the game.

Independence needs one qualification

Paroli explanations often say, “Every casino result is independent.” That is too broad.

Roulette spins on a properly operating wheel are modeled as independent for ordinary betting analysis. RNG game outcomes are designed according to their approved mathematical model. But card games dealt from a finite shoe are not perfectly independent because removing cards changes the composition of what remains.

That does not rescue Paroli.

A stake progression based only on the fact that the player just won does not automatically exploit card composition. Advantage play requires information that changes expected value—not merely a sequence of previous wins.

For the probability distinction, OpenStax explains how independent events are analyzed and why conditional information matters when dependence is present. The key question is whether the underlying probability changed, not whether the previous bet won.

Define the rules before using the name

Players often call any “press the winner” behavior Paroli. A testable Paroli plan should specify:

• base unit;
• exact multiplier after a win;
• maximum number of steps;
• what happens after a push;
• what happens after a partial-pay result;
• when a successful cycle resets;
• whether the session has a separate stop condition.

Without those rules, the system can become improvised bet escalation. A player who planned three steps may keep pressing after the third win because the table “feels hot.” At that point, the original cap no longer controls risk.

What Paroli can legitimately be used for

Paroli can be used as a predefined staking pattern by someone who prefers to:

• keep losses at the base-unit level during failed cycles;
• press a limited part of a winning run;
• use a clear reset rule;
• avoid doubling after losses.

It should not be described as a method that:

• predicts streaks;
• makes a recent win increase the next win probability;
• converts a negative-edge game into positive EV;
• protects winnings from becoming real money at risk;
• guarantees that a session ends ahead.

The broader category is positive progression. For the alternative of keeping wagers constant, compare flat betting.

The definition in one sentence

Paroli is a capped positive progression that increases bets after wins, usually resets after a loss, and changes the pattern of bankroll swings without changing the underlying game’s mathematical edge.