AI-assisted content
Drafted with AI assistance and reviewed by Lei Abelai. The worked examples are illustrative calculations demonstrating how each system behaves, not predictions of any particular session. Full AI disclosure.
Betting systems are the most persistent idea in gambling, and the most thoroughly disproven. They survive because they contain a genuine insight — bet sizing does matter — attached to a false conclusion about what it can achieve.
This guide walks through the popular ones honestly, including what each does well, before explaining the wall they all hit.
1. The one sentence that kills every system
The expected value of a series of bets is the sum of the expected values of each individual bet.
That's it. That's the whole argument. If each bet you place carries a 2.70% house edge, then ten of them carry a 2.70% edge, and a hundred of them carry a 2.70% edge — regardless of what order you place them in or how you vary the size.
Expected value is additive. There is no arrangement of negative-expectation bets that sums to a positive expectation. This isn't a claim about any particular system; it's a property of arithmetic that every system must somehow evade, and none can.
A system can move where your losses occur — making them rarer but larger, or more frequent but smaller. It cannot move the total.
2. Martingale, worked through in pesos
The most famous system. Bet on an even-money proposition. After every loss, double your stake. When you eventually win, you recover all losses plus one unit of profit. Then start again.
It genuinely works most of the time. That's what makes it dangerous.
| Loss streak | Next bet | Total staked so far | Profit if this one wins |
|---|---|---|---|
| 0 | ₱100 | ₱0 | ₱100 |
| 1 | ₱200 | ₱100 | ₱100 |
| 2 | ₱400 | ₱300 | ₱100 |
| 3 | ₱800 | ₱700 | ₱100 |
| 4 | ₱1,600 | ₱1,500 | ₱100 |
| 5 | ₱3,200 | ₱3,100 | ₱100 |
| 6 | ₱6,400 | ₱6,300 | ₱100 |
| 7 | ₱12,800 | ₱12,700 | ₱100 |
| 8 | ₱25,600 | ₱25,500 | ₱100 |
Look at the last row carefully. You are risking ₱25,600 to win ₱100. You have ₱25,500 already committed. And this is after only eight consecutive losses — an event that occurs roughly once every 400 sequences on a red/black bet, which is to say, regularly.
Two walls stop you here:
- Your bankroll. To survive eight losses from ₱100 you need over ₱25,000 available, to protect ₱100 of profit per cycle.
- The table maximum. This is the decisive one. Every table has a limit, and it exists precisely to cap doubling sequences. Hit it and you cannot place the recovery bet. The sequence is broken and the accumulated loss is locked in.
Martingale converts many small wins into one large loss
The distribution is what changes: you win ₱100 many times, then lose ₱25,500 once. Sum it up over enough cycles and you land exactly where the house edge predicted — but you get there via a single catastrophic session rather than a gradual drift. That's psychologically much worse, and financially identical.
3. Fibonacci, d'Alembert and Labouchère
| System | Mechanic | Honest assessment |
|---|---|---|
| Fibonacci | Stake follows 1, 1, 2, 3, 5, 8, 13… advancing on a loss, retreating two steps on a win. | A gentler Martingale. Grows more slowly, so you survive longer streaks — but hits the same table-limit wall eventually, and recovery is slower when it comes. Expected value unchanged. |
| d'Alembert | Increase stake by one unit after a loss, decrease by one after a win. | The most conservative progression. Genuinely reduces the risk of catastrophic loss compared to Martingale. Rests on the false premise that wins and losses "even out". Expected value unchanged. |
| Labouchère | Write a sequence of numbers; bet the sum of first and last; cross both off on a win, append the loss on a loss. | Elaborate bookkeeping that feels like a method. The list grows during losing runs, so stakes escalate. Expected value unchanged. |
| Oscar's Grind | Increase stake by one unit after a win only, targeting one unit of profit per series. | Low volatility, very slow. Long series can run indefinitely. Expected value unchanged. |
Notice the last cell in every row. That's not laziness — it's the point. These systems differ meaningfully in risk profile, and it's fair to say d'Alembert is safer than Martingale. None of them differ at all in expectation.
4. Positive progressions
Systems like Paroli (double after a win) and 1-3-2-6 increase stakes on winning streaks rather than losing ones. They deserve slightly more credit than the negative progressions, for one honest reason: they cannot bankrupt you the way Martingale can. You're escalating with money you just won rather than money you're trying to recover.
What they don't do is generate profit. A winning streak is only identifiable after it has ended. The system captures streaks that happen to occur and gives back the escalated stake on the streaks that don't — which nets to exactly the house edge.
If you find a positive progression enjoyable, it's a reasonable way to structure a session. Just don't mistake "won't ruin me quickly" for "will make money."
5. Why systems feel like they work
This is the interesting part, because the feeling is real and worth explaining.
- They win most sessions. Martingale produces a small profit in the large majority of short sessions. The rare catastrophic session is what balances the books, and you may not experience it for weeks.
- Small frequent wins are memorable; one big loss gets externalised. The wins feel like the system working. The blowup feels like bad luck, or a mistake in execution, or the table limit's fault.
- They provide structure. Having a rule to follow genuinely reduces impulsive betting, which is a real benefit that gets misattributed to the system's maths.
- Survivorship bias online. People post winning system sessions. Nobody posts the night the sequence broke.
- The gambler's fallacy underpins all of them. Negative progressions implicitly assume a win becomes "due" after losses. It doesn't. Each spin, hand and deal is independent.
6. What bet sizing actually controls
Here's the genuine insight buried inside the systems, stated properly:
Bet size controls variance and survival time. It does not control expectation.
| Bet sizing choice | Effect on expected loss | Effect on your session |
|---|---|---|
| Flat small bets | Same % of turnover | Longer play, smaller swings, lower total turnover per hour |
| Flat large bets | Same % of turnover | Shorter play, bigger swings, higher total turnover per hour |
| Negative progression | Same % of turnover | Frequent small wins, rare large losses |
| Positive progression | Same % of turnover | Frequent small losses, occasional larger wins |
The only two levers that change your total expected cost are the ones nobody sells a system for: how much you stake in total, and which bets you choose. Playing baccarat banker at 1.06% instead of the tie at 14.36% is worth more than every betting system ever devised, combined.
The genuinely useful version of "having a system"
Flat stakes at roughly 1% of your session budget, on the lowest-edge bet you enjoy, with a pre-set stop-loss and stop-win. That's structure, discipline and cost control — everything a betting system pretends to offer, minus the false promise.
7. Anyone selling you a system
A short and reliable test. If a system genuinely produced a positive expectation against a casino, the rational action for its inventor would be to use it quietly and indefinitely, not to sell it for ₱2,000 on social media.
Treat these as settled:
- Paid "winning formulas", predictor apps and hack tools do not work.
- Screenshots of winning balances prove nothing — they're trivially fabricated, and even genuine ones are a filtered sample.
- Anyone offering to place bets on your behalf, or asking for account access, is running a different scam entirely.
- Groups charging for "signals" or "patterns" are selling the gambler's fallacy at a subscription price.
The mathematics of casino games has been public and settled for a very long time. Nobody is holding back the answer.
What to do instead
Bankroll structure is the real version of the thing betting systems pretend to be.
21+ only · Gambling involves risk of loss · If it stops being fun, stop.