Betting systems — Martingale, Fibonacci, 1-3-2-6, D’Alembert, Paroli — are among the oldest ideas in gambling. Every one promises a structured path to profit. And every one fails to change the fundamental mathematics of the games they are applied to.
This guide explains how each popular system works, shows the real math behind it, and demonstrates why no staking sequence can overcome a negative-edge game. We also explain what systems can do — which is reshape your risk profile — and when that might actually be useful.
The Core Principle: Why No System Beats the House
Before diving into individual systems, this mathematical fact needs to be stated clearly:
If every individual bet has negative expected value, no combination of those bets can produce positive expected value.
This is not an opinion or a simplification. It is a provable mathematical theorem. A roulette spin has a 2.7% house edge whether you bet $1 or $1,000, whether it is your first bet or your hundredth, whether you won the last five spins or lost them. Betting systems change when and how much you bet, but not the expected value of each dollar wagered.
What systems can do is redistribute your outcomes: many small wins with rare large losses (Martingale), or many small losses with rare large wins (reverse Martingale). The total expected loss over time is the same — only the shape of the journey changes.
System 1: Martingale (Double After Loss)
How it works: Start with 1 unit. After each loss, double the bet. After a win, reset to 1 unit. The logic: one win recovers all previous losses plus 1 unit profit.
Example sequence (starting at $10, even-money bet):
- Bet 1: $10 → Loss (down $10)
- Bet 2: $20 → Loss (down $30)
- Bet 3: $40 → Loss (down $70)
- Bet 4: $80 → Win (up $10 net)
Why it fails:
- Exponential growth: After 7 losses, a $10 base bet requires $1,280 on bet 8. After 10 losses: $10,240. The required bankroll grows astronomically.
- Table limits: Most tables cap at 100-500× the minimum. At a $10 min / $5,000 max table, you can only double 8 times before hitting the limit.
- Risk/reward imbalance: You risk thousands to win $10. The expected profit per resolved cycle is +1 unit, but the expected loss from a busted cycle wipes out dozens of wins.
- Probability of ruin: At European roulette (48.6% win rate on even money), the chance of 10 consecutive losses is 0.7%. That sounds small — but over 1,000 betting cycles, you will likely encounter 7+ streaks multiple times.
What it actually does: Converts a steady drip of losses into a pattern of frequent small wins punctuated by rare catastrophic losses. Your short-term winning probability increases, but your average loss over time is unchanged.
Model the risk: Martingale Risk of Ruin Calculator
System 2: Fibonacci (Sequence After Losses)
How it works: Follow the Fibonacci sequence (1, 1, 2, 3, 5, 8, 13, 21, 34…) for bet sizing. After a loss, move one step forward. After a win, move two steps back. Reset to the beginning after a full recovery.
Example ($10 unit):
- Bet $10 → Loss. Next: $10
- Bet $10 → Loss. Next: $20
- Bet $20 → Loss. Next: $30
- Bet $30 → Win. Move back two steps → Next: $10
- Bet $10 → Win. Sequence complete.
Why it fails: The same fundamental issue as Martingale — bets escalate during losing streaks — but more slowly. After 10 consecutive losses, your Fibonacci bet is 89 units (vs 1,024 for Martingale). Gentler, but still unsustainable over long periods.
What it actually does: A slower-burning version of Martingale. The risk profile is less extreme (smaller max bets, longer to reach table limits), but the expected loss per dollar wagered is identical to flat betting.
System 3: D’Alembert (Add One Unit After Loss)
How it works: Start at a chosen bet size. After a loss, increase by 1 unit. After a win, decrease by 1 unit. The assumption: wins and losses should “balance out,” so increasing after a loss positions you for a larger recovery.
Example ($10 unit, start at 5 units = $50):
- Bet $50 → Loss. Next: $60
- Bet $60 → Loss. Next: $70
- Bet $70 → Win. Next: $60
- Bet $60 → Win. Next: $50 (back to start)
Why it fails: The “balancing” assumption is the gambler’s fallacy. Outcomes are independent — a loss does not make a win more likely. D’Alembert grows bets more slowly than Martingale or Fibonacci, which makes it less volatile, but the expected loss per unit wagered is still determined by the house edge, not the staking pattern.
What it actually does: The most conservative negative progression. Lower peak bets, lower variance, but also no mathematical advantage. It is essentially a structured way to increase bets during losing streaks — which means you wager more money when losing, accelerating the house edge’s effect on your bankroll.
System 4: 1-3-2-6 (Positive Progression)
How it works: After each win, progress through the sequence: bet 1, then 3, then 2, then 6 units. After any loss or after completing all four bets, reset to 1 unit.
Cycle outcomes (1 unit = $10, even-money bets):
- Lose on bet 1: −$10 (−1 unit)
- Win bet 1, lose bet 2: Win $10, lose $30 = −$20 (−2 units)
- Win bets 1-2, lose bet 3: Win $10+$30, lose $20 = +$20 (+2 units)
- Win bets 1-3, lose bet 4: Win $10+$30+$20, lose $60 = $0 (break even)
- Win all four: $10+$30+$20+$60 = +$120 (+12 units)
Why it is popular: Maximum exposure per cycle is just 2 units (lose on bet 2), while the full cycle pays 12 units. The risk/reward ratio feels attractive, and you only increase bets with “house money.”
Why it does not beat the house: The probability of winning four consecutive even-money bets at European roulette is approximately 5.6%. The frequency-weighted average of all cycle outcomes still equals the flat-bet expected loss. The system feels better because losses are capped and wins are concentrated, but the total amount lost per dollar wagered is unchanged.
System 5: 1324 (Variant of 1-3-2-6)
How it works: Same concept as 1-3-2-6 but the sequence is 1, 3, 2, 4 units. Maximum bet is 4 units instead of 6.
Cycle outcomes:
- Lose on bet 1: −1 unit
- Win bet 1, lose bet 2: −2 units
- Win bets 1-2, lose bet 3: +2 units
- Win bets 1-3, lose bet 4: +2 units
- Win all four: +10 units
Compared to 1-3-2-6, the 1324 system has a lower ceiling (10 units vs 12 on a full win) but a better outcome when you win 3 of 4 bets (+2 vs 0). It is slightly more conservative. Same fundamental limitation: the house edge on each bet is unchanged.
System 6: Paroli / Reverse Martingale (Double After Win)
How it works: Start with 1 unit. After a win, double the bet. After a loss (or after 3 consecutive wins), reset to 1 unit.
What it actually does: The opposite risk profile from Martingale. You have frequent small losses (every time your first bet loses) with occasional large wins (when you catch a 3+ win streak). Maximum exposure is always just 1 unit — you never escalate during losses.
Why it does not beat the house: Same principle. The expected value per dollar bet is determined by the game’s edge, not your staking sequence. But Paroli is arguably the safest progression system because your losses are always capped at 1 unit and you only escalate with winnings.
Side-by-Side Comparison
| System | Type | Max Bet Growth | Risk Profile | Beats House Edge? |
|---|---|---|---|---|
| Martingale | Negative | Exponential (2×) | Many small wins, rare huge loss | No |
| Fibonacci | Negative | Slower exponential | Gentler Martingale | No |
| D’Alembert | Negative | Linear (+1) | Low variance | No |
| 1-3-2-6 | Positive | Capped (6 units) | Capped loss, big win on streak | No |
| 1324 | Positive | Capped (4 units) | Conservative 1-3-2-6 | No |
| Paroli | Positive | Capped (3 doubles) | Small losses, occasional big win | No |
| Flat Betting | None | Constant | Steady, predictable | No |
The last column is the same for every row. That is the point.
What Systems Can Legitimately Do
Systems do not create edge, but they do shape your experience. This is worth understanding:
- Martingale maximizes your probability of ending a short session ahead — at the cost of occasional devastating losses. If your goal is “win $50 tonight and leave,” Martingale gives you the highest chance of achieving it. The downside is the 5-10% chance of losing your entire bankroll.
- Positive progressions (1-3-2-6, Paroli) cap your downside while giving you a shot at a big win during hot streaks. They are the least harmful systems because you never escalate during losses.
- Flat betting gives you the most predictable outcome and the simplest bankroll management. It is the baseline that all systems should be compared against.
If you understand that no system changes your expected loss, but you prefer the experience of one system over another, that is a valid personal choice. The danger is believing the system creates an edge that does not exist.
The Only Real Edge: Positive Expected Value
The only way to have a long-term mathematical advantage is to find bets where the expected value is positive — where the true probability of winning exceeds what the odds imply. This exists in a few narrow contexts: card counting in blackjack, advantage play on promotions, sharp sports betting at mispriced lines, and certain video poker games with optimal strategy plus comps.
In these contexts, staking systems like Kelly Criterion become genuinely useful — not because they create edge, but because they optimize how much to bet when you already have one.
Related tools: Kelly Criterion Calculator | Progression Builder | Martingale Risk Calculator | Roulette Progression Simulator
Frequently Asked Questions
What is the 1-3-2-6 strategy?
A positive progression: bet 1, 3, 2, 6 units after consecutive wins. Reset after any loss or completing the cycle. Maximum exposure is 2 units per cycle. Does not change the house edge.
Is the Martingale system illegal?
No. Completely legal everywhere. Casinos do not ban it because it does not give you an edge. Table limits (not legality) are what prevent indefinite doubling.
Can the Martingale beat the casino?
No. It produces frequent small wins and rare catastrophic losses. The expected total loss over time equals flat betting at the same house edge.
What is the Fibonacci method in gambling?
A negative progression following the Fibonacci sequence (1,1,2,3,5,8,13…). Advance one step after a loss, retreat two after a win. Gentler than Martingale but same fundamental limitation.
What is the 1324 strategy?
Variant of 1-3-2-6 with sequence 1,3,2,4. Lower maximum bet (4 vs 6 units), slightly better outcome on 3-of-4 wins. Same mathematical limitation.
Is there a gambling system that never loses?
No. Mathematical theorem: if each bet has negative expected value, no combination of bets can be positive. Systems change the distribution of outcomes, not the total expected loss.
What is the safest betting system?
Flat betting (constant bet size) is the safest in terms of variance. Among progression systems, Paroli (double after wins, not losses) has the lowest downside risk because you never escalate during losing streaks.
Do professional gamblers use betting systems?
Professional advantage players use Kelly Criterion staking — but only because they already have a positive edge from card counting, sharp lines, or exploitation plays. They do not use Martingale or Fibonacci on negative-edge games.